Embed Size (px)
Citation preview
Plasmodium knowlesi malaria:
Epidemiology, clinical features, diagnosis and
pathogenesis
Bridget Barber
July 2013
A thesis submitted in fulfillment of the requirements for the degree
of Doctor of Philosophy of Charles Darwin University
i
Declaration
I hereby declare that the work herein, now submitted as a thesis for the degree of Doctor of
Philosophy of the Charles Darwin University, is a result of my own investigations, and all
references to ideas and work of other researchers have been specifically acknowledged. I
hereby certify that the work embodied in this thesis has not already been accepted in
substance for any degree, and is not being currently submitted in candidature for any other
degree.
I also declare that all individuals who can be identified from the photographs included in this
thesis provided consent for their photograph to be taken.
_________________
Bridget Barber Date
26th July 2013
ii
All of the papers included in this thesis in which I am first or last author were written by me.
This thesis is substantially my own work, and was conducted under the guidance of my
supervisors Professor Nick Anstey and Dr Tsin Yeo, and with the support of Dr Timothy
William. For the two retrospective studies at Kudat District Hospital (Chapters 5 and 6), and
the review of the Sabah Department of Health malaria notification data (Chapter 7), I was
primarily responsible for study design, collected and analysed and interpreted all data, and
wrote the first draft of the manuscripts. The prospective observational study at Queen
Elizabeth Hospital (Chapters 8, 10, 11 and 12) was implemented and conducted by me. I
analysed and interpreted all data for these chapters, and wrote the first draft of the
manuscripts. Enrolment and daily follow-up of patients was done predominantly by me, with
assistance from Dr Tsin Yeo, Dr Matthew Grigg, and my research assistants Rita Wong,
Beatrice Wong and Ann Wee. All the statistical analyses presented in the thesis were
performed by me, with advice from Tsin Yeo, Menzies statistician Mark Chatfield and Nick
Anstey.
I specifically acknowledge the following additional contributions of others:
Nick Anstey and Tsin Yeo developed the protocol for the prospective study at QEH
For the retrospective review of fatal cases of malaria in Sabah (Chapter 9), Dr Giri
Rajahram retrieved and reviewed the medical notes, with assistance from me. Giri
and I co-wrote the first draft of the paper, with Nick Anstey, Tsin Yeo, and Timothy
William contributing substantially to the final manuscript.
For the study investigating red cell deformability (Chapter 12), Dr Matthew Grigg was
primarily responsible for enrolling patients after October 2011
iii
Ferryanto Chalfein performed all the quantification of parasitemia by microscopy for
the QEH prospective study
Dr Jutta Marfurt and Dr Sarah Auburn performed the PCR assays for the QEH
prospective study
Beatrice Wong and Ann Wee made the malaria blood films, performed the rapid
diagnostic tests, and conducted the measurements of red cell deformability
iv
I would like to acknowledge a number of individuals who made substantial contributions to
this research project. Firstly, I would like to thank my supervisor Professor Nicholas Anstey
and co-supervisor Dr Tsin Yeo, who developed the protocol for the prospective study that
comprises the major component of this thesis, and who provided me this very exciting
opportunity to study an emerging disease of which we knew relatively little about. Nick has
been an inspiring mentor who has provided constant and enthusiastic support and advice,
and his attentive oversight ensured that the study was conducted to the highest possible
standard. I thank him also for all his insightful and constructive comments on the numerous
drafts of the manuscripts that make up this thesis. Tsin’s calm approach to any difficulties
encountered throughout the project was greatly appreciated, and I would also like to thank
Tsin for covering me in Sabah when I was on leave, for designing the database used in the
prospective study, and for statistical advice.
Very special thanks to Dr Timothy William, who, on top of an already daunting clinical
workload, took on the responsibility of implementing this project at Queen Elizabeth Hospital
(QEH), and assisted with all the associated logistical issues, from coordinating referrals from
district hospitals, to finding space for an office and laboratory, recruiting staff, assisting with
visas and medical registration, and even helping me to find a car. He ensured the
cooperation of all involved staff at the Health Department, the Sabah State Reference
Laboratory, the District Hospitals, and various departments of QEH, to facilitate the smooth
running of the study. He was always available to assist when problems arose, and always
made me feel most welcome in Sabah.
Thanks to Dr Matthew Grigg, who assisted with enrolments towards the end of my time in
Sabah, covered me while I was on leave, and provided welcome company. I am also very
grateful to Matt for taking over the coordination of the QEH study from October 2011,
v
allowing me to return home to write up this thesis. Thanks also to Matt, and to Dr Uma
Parameswaran, for answering my many emails requesting missing data.
Great thanks to my research assistants, Rita Wong, Beatrice Wong, and Ann Wee. Rita for
her efficiency, energy and good humour despite very long days, for chasing up so many
blood slides and results, and for bringing me food when there was no time for lunch breaks;
Beatrice for her hard work in the lab and working late into the nights without complaint; and
Ann for her flexibility with both laboratory and nursing work, and for her help with translation.
Thanks to Professor Arjen Dondorp for the loan of the Laser-assisted Optical Rotational Cell
Analyser (LORCA), and for helpful discussions regarding red cell deformability, and to Dr
Richard Maude and Forradee Nuchsongsin for their trips to Sabah and their training and
advice regarding operating the LORCA.
Thanks to all the staff at the infectious diseases clinic at QEH for accommodating us and
making us feel welcome, and to all the nurses and medical officers at QEH who cared for the
patients. Thanks also to the medical officers at the District Hospitals for keeping us informed
about malaria patients at these hospitals, and for all the referrals. Thanks to the
microscopists at QEH for notifying us of positive slide results, and for quantitating
parasitemia for us. Thanks also to the staff at Sabah State Reference Laboratory, in
particular Fread Anderios, for performing PCR on all our QEH malaria patients, and for
allowing us access to all the malaria slides referred to their laboratory.
Great thanks to Kim Piera for setting up the laboratory at QEH, for training staff, and for all
her logistical support throughout the study. Thanks to Dr Jutta Marfurt and Dr Sarah Auburn
for performing the PCR assays on all our malaria patients, and to Ferryanto Chalfein for
reading the malaria slides. Thanks also to Dr Vijaya Joshi, Dr Ella Curry and Melissa Gallop
for administrative assistance throughout the study.
vi
Finally, deepest thanks to my partner Mitch for tolerating my long absence, for his many trips
to Sabah, for the many long phone calls, and for reaching the top of Mt Kinabalu; and to my
darling daughter Jemma, who has tolerated my writing of this thesis during the first 9 months
of her life.
vii
Figure 1. The malaria research team outside Lingzhi Infectious Diseases Ward
From left: Beatrice Wong, Rita Wong, Nicholas Anstey, Bridget Barber, Kim Piera, and Timothy William
Figure 2. Hard at work in Kota Kinabalu
From left: Matt Grigg, Bridget Barber, and Tsin Yeo.
viii
Figure 3. Photo taken for the Daily Express, prior to the 1st Borneo Scientific Meeting on Tropical Diseases, Kota Kinabalu, Aug 29th-30th
ix
The simian parasite Plasmodium knowlesi is a common cause of human malaria in
Malaysian Borneo, and can cause severe and fatal disease. Substantial knowledge gaps
exist in regards to the epidemiology, clinical features, diagnosis, and pathophysiology of
knowlesi malaria, and the overall aim of this thesis was to enhance understanding in these
areas.
First, two retrospective studies were conducted at Kudat District Hospital, northeast Sabah,
with these studies confirming that P. knowlesi was the most common cause of malaria
among adults and children at this hospital. A wide age-distribution among patients with
knowlesi malaria was described, and two family clusters were identified. Second, Sabah
Department of Health malaria notification data were reviewed, with this study demonstrating
that, while notifications of P. falciparum and P. vivax had decreased markedly over the past
20 years, over the past decade notifications of “P. malariae/P. knowlesi” increased
significantly. Third, a prospective observational study was conducted, involving all malaria
patients admitted to Queen Elizabeth Hospital (QEH), a tertiary-referral hospital in Sabah,
from September 2010 to October 2011. This study found that P. knowlesi was the most
common cause of severe malaria at QEH, and was associated with a 3-fold increased risk of
severity compared to P. falciparum. Parasite count was the major independent risk factor for
severe knowlesi malaria, with risk increasing 11-fold with parasitemia >20,000/μL and 28-
fold with parasitemia >100,000/μL. Artesunate therapy was highly effective for severe
malaria, and no deaths from any species occurred in this study. This study also evaluated
the accuracy of microscopy, and the sensitivity of two rapid diagnostic tests, for the
diagnosis of knowlesi malaria. Finally, in the first of a series of pathophysiological studies
x
conducted at QEH, red cell deformability among patients with knowlesi malaria was
investigated.
These studies have confirmed the importance of knowlesi malaria as a public health problem
in Sabah, and have extended the existing information regarding the epidemiological and
clinical features, as well as diagnosis and treatment, of knowlesi malaria. Future research
priorities are identified in order to enhance our understanding of this emerging disease.
xi
AMA-1 Apical Membrane Antigen-1
CSP Circumsporozoite Protein
DBP Duffy Binding Protein
DNA Deoxyribonucleic Acid
Fy Duffy Antigen
ICAM-1 Intercellular Adhesion Molecule-1
IL Interleukin
KDH Kudat District Hospital
LAMP Loop Mediated Isothermal Amplification
LDH Lactate Dehydrogenase
LORCA Laser Assisted Rotational Optical Cell Analyzer
MCP-1 Monocyte Chemotactic Protein-1
MIP-1b Macrophage Inhibitory Protein-1b
mtDNA mitochondrial DNA
PCR Polymerase Chain Reaction
PfEMP-1 Plasmodium falciparum Erythrocyte Membrane Protein-1
PfHRP2 Plasmodium falciparum Histidine Rich Protein 2
PkNBP Plasmodium knowlesi Normocyte Binding Protein
QEH Queen Elizabeth Hospital
RBC Red Blood Cell
RBC-D Red Blood Cell Deformability
RBL Reticulocyte Binding-Like
RDT Rapid Diagnostic Test
ROC Receiver Operator Curve
SSUrRNA Small Subunit Ribosomal RNA
xii
TNF Tumour Necrosis Factor
VCAM Vascular Cellular Adhesion Molecule
WHO World Health Organization
xiii
Author’s Statement ..............................................................................................................ii
Acknowledgements ............................................................................................................ iv
Abstract ............................................................................................................................... ix
Glossary .............................................................................................................................. xi
1. Aims ................................................................................................................................. 1
1.1 Background ................................................................................................................. 1
1.2 Aims ............................................................................................................................ 3
1.3 Overview of thesis ....................................................................................................... 5
2. Introduction I: Historical background and epidemiology ............................................. 8
2.1 Historical Background .................................................................................................. 8
2.2. Epidemiology ............................................................................................................ 13
2.2.1 Geographical distribution and emergence of P. knowlesi in Malaysia .................. 13
2.2.2 Geographical distribution of P. knowlesi outside of Malaysia ............................... 16
2.2.3 Reservoir Hosts................................................................................................... 19
2.2.4 Mosquito Vectors ................................................................................................ 20
2.2.5 Transmission of P. knowlesi ................................................................................ 22
2.3. Knowledge gaps identified from the literature search ................................................ 24
3. Introduction II: Diagnosis, Clinical Disease and Treatment ....................................... 26
3.1. Diagnosis .................................................................................................................. 26
3.1.1 Microscopy .......................................................................................................... 26
3.1.2 Rapid Diagnostic Tests ....................................................................................... 28
3.1.3. Molecular methods of diagnosis ......................................................................... 29
3.2. Clinical disease ......................................................................................................... 30
3.2.1 Demographics ..................................................................................................... 31
3.2.2. Incubation and pre-patent periods ...................................................................... 32
3.2.3. Clinical features on presentation ........................................................................ 32
3.2.4 Laboratory findings .............................................................................................. 33
3.2.5 Severe disease ................................................................................................... 33
3.3. Treatment ................................................................................................................. 37
3.3.1 Treatment of uncomplicated Plasmodium knowlesi malaria ................................ 37
xiv
3.3.2 Treatment of severe knowlesi malaria ................................................................. 38
3.4. Knowledge gaps identified from the literature search ................................................ 39
3.4.1 Diagnosis of knowlesi malaria ............................................................................. 39
3.4.2. Demographics and clinical features of knowlesi malaria ..................................... 39
4. Introduction III: Pathophysiology ................................................................................ 42
4.1 Histopathology of severe P. falciparum malaria ......................................................... 42
4.2. Histopathology of P. knowlesi in macaques and humans .......................................... 43
4.3. Abnormalities of the microcirculation in P. knowlesi and P. falciparum malaria ......... 45
4.3.1 Reduced red blood cell deformability in severe falciparum malaria ..................... 45
4.3.2 Abnormalities of the microcirculation in rhesus macaques infected with P. knowlesi .................................................................................................................................... 47
4.4. The role of platelets in the pathogenesis of severe knowlesi and falciparum malaria 49
4.5. Erythrocyte invasion by P. knowlesi merozoites ........................................................ 51
4.5.1. Duffy antigen and receptor ligands ..................................................................... 51
4.5.2. Influence of the age of red blood cells on invasion of erythrocytes by P. knowlesi merozoites ................................................................................................................... 52
4.6. Inflammatory response to P. knowlesi malaria .......................................................... 52
4.7. Knowledge gaps identified from the literature search ................................................ 53
5. Plasmodium knowlesi malaria in children ................................................................... 55
6. Epidemiology of Plasmodium knowlesi malaria in Kudat, Sabah: Family clusters and wide age distribution ................................................................................................. 59
7. Apparent emergence of P. knowlesi malaria in Sabah, 1992 - 2011 .......................... 62
8. Epidemiological and clinical features of P. falciparum, P. knowlesi and P. vivax malaria in adults: a prospective comparative study ....................................................... 65
8.1 Background ............................................................................................................... 65
8.2 Study Site .................................................................................................................. 67
8.3 Patient enrolment and study procedures ................................................................... 69
8.4 Significance of the paper ........................................................................................... 71
9. Fatal cases of Plasmodium knowlesi malaria in Sabah ............................................. 77
10. Evaluation of microscopy for the diagnosis of knowlesi malaria ............................ 79
11. Evaluation of Rapid Diagnostic Tests for the diagnosis of severe and uncomplicated P. falciparum, P. knowlesi and P. vivax malaria .................................... 84
12. Evaluation of red blood cell deformability among patients with knowlesi malaria 87
xv
13. Conclusions and future directions ............................................................................ 90
13.1 Epidemiology ........................................................................................................... 90
13.2 Diagnosis of Plasmodium knowlesi .......................................................................... 92
13.3 Clinical features ....................................................................................................... 93
13.4 Treatment ................................................................................................................ 95
13.4 Pathogenesis ........................................................................................................... 96
13.5 Summary ................................................................................................................. 99
14. References ................................................................................................................. 100
List of Tables
These do not include tables in each of the enclosed papers.
Table 1. Case reports of Rapid Diagnostic Tests used to diagnose P. knowlesi ...................... 29
List of Figures
These do not include figures contained within each of the enclosed papers.
Figure 1. The malaria research team outside Lingzhi Infectious Diseases Ward ................. vii
Figure 2. Hard at work in Kota Kinabalu .............................................................................. vii
Figure 3. Photo taken for the Daily Express, prior to the 1st Borneo Scientific Meeting on
Tropical Diseases, Kota Kinabalu, Aug 29th-30th ............................................................... viii
Figure 4. Map showing countries where human cases of knowlesi have been reported,
together with the natural distribution of long-tailed and pig-tailed macaques and the
Anopholes leucosphyrus group of mosquitoes. Reproduced from Singh and Daneshvar
(167). .................................................................................................................................. 23
Figure 5. Severity criteria for Plasmodium knowlesi malaria* .............................................. 36
Figure 6. Kudat District Hospital ......................................................................................... 56
Figure 7. Map of Sabah showing Districts and Divisions .................................................... 68
xvi
Figure 8. Queen Elizabeth Hospital Emergency Department .............................................. 70
Figure 9. Lingzhi Infectious Diseases Ward ........................................................................ 70
Figure 10. Beatrice Wong processing bloods at Lingzhi ..................................................... 70
Figure 11. Bridget Barber reviewing patients on the medical ward, Queen Elizabeth Hospital
........................................................................................................................................... 70
Published manuscripts forming the basis of this thesis
1. Barber BE, William T, Grigg MJ, Menon J, Auburn S, Marfurt J, Anstey NM, Yeo TW. A
prospective comparative study of knowlesi, falciparum, and vivax malaria in Sabah,
Malaysia: high proportion with severe disease from Plasmodium knowlesi and Plasmodium
vivax but no mortality with early referral and artesunate therapy. Clin Infect Dis 2013;
56(3):383-97
2. William T, Rahman HA, Jelip J, Ibrahim MY, Menon J, Grigg MJ, Yeo TW, Anstey NM,
Barber BE. Increasing incidence of Plasmodium knowlesi malaria following control of P.
falciparum and P. vivax malaria in Sabah, Malaysia. PLoS Negl Trop Dis 2013; 7(1):e2026.
3. Barber BE, William T, Grigg MJ, Yeo TW, Anstey NM. Evaluation of the sensitivity of a
pLDH-based and an aldolase-based Rapid Diagnostic Test for the diagnosis of
uncomplicated and severe malaria caused by PCR-confirmed Plasmodium knowlesi,
Plasmodium falciparum and Plasmodium vivax. J Clin Micro 2013; 51(4)1118-1123
4. Barber BE, William T, Grigg MJ, Yeo TW, Anstey NM. Limitations of microscopy to
differentiate Plasmodium species in a region co-endemic for Plasmodium falciparum,
Plasmodium vivax and Plasmodium knowlesi. Malar J 2013; 12:8
xvii
5. Barber BE, William T, Dhararaj P, Anderios F, Grigg MJ, Yeo TW, Anstey NM.
Epidemiology of Plasmodium knowlesi malaria in north-east Sabah, Malaysia: family clusters
and wide age distribution. Malar J 2012; 11:401
6. Rajahram GS, Barber BE, William T, Menon J, Anstey NM, Yeo TW Deaths due to
Plasmodium knowlesi malaria in Sabah, Malaysia: association with reporting as Plasmodium
malariae and delayed parenteral artesunate. Malar J 2012; 11:284
7. Barber BE, William T, Jikal M, Jilip J, Dhararaj P, Menon J, Yeo TW, Anstey NM.
Plasmodium knowlesi malaria in children. Emerg Infect Dis 2011; 17(5):814-20
1
1.1 Background
Plasmodium knowlesi was first identified in a long-tailed macaque in 1932 (1), and the first
naturally occurring human case was reported in 1965 (2). The microscopic similarity
between P. knowlesi and P. malariae however impeded the detection of any further cases of
human knowlesi malaria until 2004, when Singh et al. investigated a large number of
microscopically-diagnosed “P. malariae” cases in Sarawak, Malaysian Borneo, and found
them to be P. knowlesi by PCR (3). This seminal report led to a number of studies in
Malaysia and elsewhere involving molecular testing of archived samples from patients
diagnosed with “P. malariae”, and these studies confirmed the existence of human knowlesi
malaria in Malaysia and Thailand at least as early as 1996 (4, 5). Since these studies, cases
of human knowlesi malaria have been increasingly reported, with cases now described in
every Southeast Asian country except Laos and Timor Leste (6-11), and with P. knowlesi
also increasingly reported in returning travellers (12-22). The largest number of cases has
been reported from the eastern Malaysian states of Sabah and Sarawak, where in some
districts, P. knowlesi is the most common cause of malaria (23-26). Moreover, in these
states P. knowlesi has been reported to cause severe and fatal disease, with multi-organ
failure similar to that seen in severe falciparum malaria (23, 24, 27).
Significant gaps exist in our understanding of the epidemiology, clinical features, and
pathogenesis of knowlesi malaria. At the commencement of this thesis, the only information
regarding the geographic distribution of P. knowlesi in Sabah came from a study that
reported the detection of P. knowlesi by PCR in 41 of 49 archived “P. malariae” blood slides
obtained from multiple districts throughout Sabah (23). The number of “P. malariae” blood
films as a proportion of all malaria blood films in these districts was unknown. Information
2
regarding the recent trends of malaria notifications according to species was also limited.
Data regarding the clinical features of human knowlesi malaria, with the exception of the
early studies from the malariotherapy era, came from case reports (7, 9-11, 13, 14, 16, 18-
20, 22, 28, 29) and small case series (27, 30), a district-hospital retrospective study (3), a
tertiary-referral hospital retrospective study (31), and a single district-hospital prospective
study (24). Only 40 patients with severe disease had been described (23, 24, 27, 30, 31),
and risk factors for severity had not been determined by multivariate analysis. Furthermore,
no study had been able to directly compare in the same population the epidemiological and
clinical characteristics of knowlesi malaria to those of the other human malaria species.
Finally, only one study had reported on the pathogenesis of severe knowlesi malaria in
humans (evaluating the human cytokine response to P. knowlesi) (32), in addition to a single
autopsy report of fatal human knowlesi malaria (27).
The overall aim of this thesis was therefore to enhance our understanding of the
epidemiology, clinical features, diagnosis, and pathophysiology of knowlesi malaria. The
first aim was to investigate certain aspects of the epidemiology of P. knowlesi malaria in
Sabah, primarily through retrospective review of existing data. Second, by conducting a
prospective observational study of all malaria patients admitted to a tertiary-referral hospital
in Sabah, I aimed to describe the epidemiological and clinical features of knowlesi malaria
and to compare these to those of the other human malaria species. Third, as part of this
prospective study I aimed to evaluate two methods of diagnosis for knowlesi malaria,
namely, microscopy and rapid diagnostic tests. Finally, I aimed to investigate one aspect of
the pathogenesis of severe knowlesi malaria, that being red cell deformability. Additional
prospective pathophysiological studies have been conducted, or are in progress, but are
beyond the scope of this thesis.
3
1.2 Aims
Aim 1: Epidemiology of P. knowlesi malaria in Sabah
i. To describe the age and sex distribution of patients diagnosed with knowlesi malaria,
compared to patients diagnosed with falciparum and vivax malaria
ii. To investigate the presence of family clusters of patients diagnosed with knowlesi
malaria
iii. To describe the geographic distribution of notifications of “P. malariae/P. knowlesi”
cases throughout Sabah, compared to notifications of P. falciparum and P. vivax
iv. To describe the trends of malaria notifications over time, according to species and
according to district
Hypotheses 1:
i. That the age distribution of patients diagnosed with knowlesi malaria will differ from
that of patients diagnosed with falciparum or vivax malaria
ii. That family clusters of knowlesi malaria may be occurring
iii. That throughout Sabah there will be geographic variation in the notifications of “P.
malariae/P. knowlesi” as a proportion of all malaria notifications
This aim was achieved through retrospective review of existing data, including hospital
microscopy records and Sabah Department of Health malaria notification records. These
studies are discussed in Chapters 5 – 7.
4
Aim 2: Clinical features of knowlesi malaria
i. To describe in detail the clinical features of severe and non-severe knowlesi malaria,
and to compare these features with those of severe and non-severe falciparum and
vivax malaria
ii. To identify risk factor(s) for severe disease among patients with knowlesi malaria,
and to compare these with those of falciparum and vivax malaria
Hypotheses 2:
i. That the clinical and laboratory features of knowlesi malaria will differ from those of
falciparum and vivax malaria.
ii. That the risk factors for severe disease among patients with knowlesi malaria will
differ from those of falciparum and vivax malaria. In particular, that age will be a risk
factor for severe knowlesi malaria
Aim 3: Diagnosis of knowlesi malaria
i. To evaluate the accuracy of microscopy for the diagnosis of knowlesi malaria, in a
setting where P. knowlesi, P. falciparum and P. vivax all commonly occur
ii. To evaluation the sensitivity of two rapid diagnostic tests for the diagnosis of knowlesi
malaria
Aims 2 and 3 were achieved through a prospective, observational study of all malaria
patients admitted to Queen Elizabeth Hospital (QEH), tertiary-referral hospital in Sabah.
This study comprises the central component of this thesis, and is discussed in detail in
Chapter 8. The evaluations of microscopy and rapid diagnostic tests for the diagnosis of
knowlesi malaria are discussed in Chapters 10 and 11 respectively.
5
Aim 4: Pathogenesis of knowlesi malaria
i. To evaluate red blood cell deformability (RBC-D) among patients with severe and
non-severe knowlesi malaria, and to compare this to RBC-D among patients with
severe and non-severe falciparum malaria
Hypothesis 4:
i. That RBC-D will be reduced in knowlesi malaria in proportion to severity, and will be
comparable to patients with falciparum malaria
Aim 4 was achieved by measuring RBC-D (using a Laser-assisted Optical Rotational Cell
Analyzer) on patients admitted to QEH with knowlesi or falciparum malaria. This study is
presented in Chapter 12.
1.3 Overview of thesis
Chapters 2-4 review the existing literature on P. knowlesi. Chapter 2 reviews the history of
P. knowlesi, including its initial identification in 1932, its use in the treatment of neurosyphilis,
and the discovery of large numbers of human cases in Malaysia. The geographic
distribution of human knowlesi malaria cases is discussed, and information is provided
regarding the simian hosts and mosquito vectors of P. knowlesi. Chapter 3 provides an
introduction to the clinical features, diagnosis and treatment of knowlesi malaria, although
does not include data obtained from the prospective study conducted as part of this thesis
and discussed in detail in later chapters. Chapter 4 reviews the limited data on the
pathophysiology of knowlesi malaria in humans and monkeys, with comparative features of
P. falciparum pathophysiology also discussed.
6
Chapters 5 -7 discuss 3 retrospective studies that provide information on certain aspects of
the epidemiology of P. knowlesi malaria in Sabah. Chapter 5 presents a retrospective study
of the 2009 malaria microscopy records at Kudat District Hospital (KDH) in northeast Sabah.
This study revealed that in 2009 P. knowlesi was the most common cause of malaria
admissions to KDH, and that it was also the most common cause of paediatric malaria. The
clinical features of knowlesi malaria in children were also investigated in this retrospective
study, and are discussed in this chapter. Chapter 6 presents an extension of this study, with
microscopy records from 2009 – 2011 reviewed at KDH. This study confirms the
predominance of P. knowlesi as a cause of human malaria in northeast Sabah, and provides
a detailed description of the age-distribution of patients with knowlesi malaria. Chapter 7
presents a retrospective review of all available Sabah Department of Health malaria
notification data from 1992 – 2011, and provides information regarding the apparent recent
emergence of knowlesi malaria in Sabah.
Chapter 8 comprises the central component of this thesis: an observational prospective
study of all malaria patients admitted to Queen Elizabeth Hospital (QEH), a tertiary-referral
hospital in Sabah, from September 2010 to October 2011. This is the largest study of
knowlesi malaria cases to date, and the largest series of severe knowlesi malaria. It also
compares the clinical and epidemiological features of knowlesi, falciparum, and vivax
malaria. The chapter extends previous descriptions of the clinical features of non-severe
and severe knowlesi malaria. It also discusses risk factors for severe disease among
patients with knowlesi malaria, and provides information regarding the early therapeutic
efficacy of artemisinin derivatives.
Although no deaths occurred from any malaria species during the prospective study at QEH,
14 malaria deaths occurred in Sabah during 2010 and 2011, including 7 P. falciparum, 6 P.
7
knowlesi, and one P. vivax. To complement the findings from the prospective QEH study,
we reviewed the details of these 14 cases, in particular to review and compare the clinical
details and management of severe malaria caused by each species. This study is presented
in Chapter 9.
Chapters 10 and 11 present the results of two diagnostic studies conducted alongside the
QEH prospective study, evaluating microscopy and rapid diagnostic tests, respectively, for
this diagnosis of knowlesi malaria. Chapter 12 presents the first of a series of
pathophysiological studies that I have conducted at QEH. This study evaluates red blood
cell deformability among patients with knowlesi malaria, and compares this to patients with
falciparum malaria. The additional pathophysiological studies are outside the scope of this
thesis.
Chapter 13 summarises the findings of this thesis, and discusses the significant knowledge
gaps that still exist with regards to the epidemiology, clinical features, diagnosis, treatment,
and pathogenesis of knowlesi malaria. The research priorities needed to improve our
understanding of knowlesi malaria in all these areas are addressed in this final chapter.
8
2.1 Historical Background
Plasmodium knowlesi may have been first sighted by Giuseppe Franchini (33, 34), an Italian
physician, who in 1927 noted a parasite in the blood of a long-tailed macaque (Macaca
fascicularis) that appeared different morphologically to P. inui and P. cynomolgi, and
resembled P. malariae (35). It was several years later, in 1931 at the Calcutta School of
Tropical Medicine, India, that the parasite was observed again by Drs Campbell and Napier
who were undertaking research on leishmaniasis (1). Although their research was usually
conducted using the readily-available and cheap rhesus macaques (M. mullata, then known
as M. rhesus), a scarcity of these monkeys in Calcutta at the time had led to the fortuitous
use of a long-tailed macaque imported from Singapore, and it was in this monkey that the
unusual Plasmodium spp was noted (36, 37). Having never seen a Plasmodium infection in
a monkey before (36), Napier inoculated the infected blood into two other long-tailed
macaques and one rhesus macaque, and subsequently passaged the infection through both
macaque species (1). While the infection proved benign in the long-tailed macaques, the
rhesus macaques developed rapidly fatal disease, with haemoglobinuria noted as a terminal
event (1). Increased virulence of infection with passage through rhesus macaques was also
noted (1).
Around this time, in the protozoology department of the same institution, Dr Robert Knowles
and his assistant Biraj Mohan Das Gutpa had been hoping for many years to obtain a strain
of avian or simian malaria, to overcome the difficulties of studying malaria in patients who,
they note, “naturally have to be treated, and then become useless for clinical investigations”
(37). Remarking on the expense of canaries and the difficulties of keeping sparrows alive,
Knowles and Das Gupta felt that monkeys might offer a more hopeful option. It was for this
9
reason that Napier, knowing this interest of his colleagues, handed over his original infected
macaque for further study. With Knowles on leave at the time, Das Gupta passaged the
parasite through macaques of various species until Knowles’ return, at which time further
investigations were performed including the experimental infection of man (36, 37). In their
initial report of these studies Knowles and Das Gupta confirmed the findings of benign and
fulminant infection in long-tailed and rhesus macaques respectively, as well as increasing
virulence and haemoglobinuria with serial passage through the latter (37).
The ability of the parasite to infect humans was demonstrated through the inoculation of
three volunteer patients, the first with “paretic symptoms” of uncertain aetiology, the second
with a foot ulcer following a rat bite, and the third with lepromatous leprosy (37). Infection
was associated with daily fever spikes and an incubation period of 10 – 23 days, and, while
two patients experienced mild to moderate disease, the patient with the foot ulcer received
blood passaged through the first patient and became “very seriously ill” (37). Although
reportedly comatose, clinical details were limited, and the patient made a spontaneous
recovery (37).
Knowles and Das Gupta also described the morphological features of the parasite in both
man and macaques, highlighting the great differences in appearance of the parasite in the
different hosts – “a finding so amazing that it appeared to us to be incredible” (37). In long-
tailed macaques the parasite resembled P. vivax, with enlarged red cells and Schuffner’s
dots observed, while in rhesus macaques red cells were of normal size with no stippling
seen, and the appearance was closer to that of P. falciparum. In man, the parasite was
noted to have “a general resemblance towards P. malariae”. A difference in red cell internal
viscosity of the different hosts was postulated by Knowles and Das Gupta as a possible
explanation for this variation in morphological appearance.
10
Despite their extensive investigation into this apparently new species, Knowles and Das
Gupta had no desire to identify it as such, citing “worker after worker” who had “described
new species on the most slender grounds, or on no grounds at all”, and remarking that “the
chief duty which faces the honest medical protozoologist of today is to attempt to create
order where at present chaos reigns” (37). Knowles stated his regret that he had been
“responsible in his earlier and inexperienced days for adding to this muddle”, and concluded
by assuming that the newly observed simian Plasmodium species had probably already
been described. Later that year however the morphology of the parasite in macaques was
described in more detail by Sinton and Mulligan, who confirmed the parasite’s unique 24-
hour asexual replication cycle, and, convinced that this was indeed a new species, named
the parasite Plasmodium knowlesi (38, 39). Further descriptions of the life-cycle of P.
knowlesi were provided by Garnham et al. (40) and Coatney (34). In rhesus macaques it
was noted that the tissue forms of the parasite were first seen in the liver parenchyma at 92
hours after infection, with ring forms found in circulating blood by 120 hours (34, 40). No
hypnozoites were observed.
Following the discovery of P. knowlesi the parasite became widely used as a pyretic agent
for the treatment of neurosyphilis, with the first description provided by Rooyen and Pile in
1935 (41).1 In their report of twelve patients inoculated with P. knowlesi, the authors noted
the infection to be mild (“eminently suitable for the treatment of elderly debilitated patients”),
with an incubation period of 3 – 14 days. Marked variability in susceptibility to infection was
also reported, with infection particularly difficult to reproduce in three individuals who had
been previously exposed to P. vivax (41). This relative resistance to infection among those
previously exposed to malaria was also reported the same year by Nicol (42), who found that
among a total of 76 patients, only 1 of 16 with a previous history of malaria developed fever
following inoculation of P. knowlesi compared to 44% of non-immune patients. In 1937 1 It was following this 1935 report in the British Medical Journal that Knowles wrote to the journal, clarifying the “apparently overlooked” contribution of Napier and Campbell to the discovery of P. knowlesi, and even going so far as to mention that he had been on leave at the time that the parasite was discovered.
11
Ciuca and colleagues reported on their use of P. knowlesi to treat over 300 patients in
Romania (43, 44). Among these patients, 46% and 80% of immune and non-immune
patients respectively developed fever following exposure to P. knowlesi (43, 44). In contrast
to Nicol who reported loss of pathogenicity of P. knowlesi with repeated passage from
person-person (42), Ciuca et al. reported increased virulence of infection, leading to their
eventual discontinuation of this treatment in 1955 after 170 transfers (45). Increased
virulence with successive human transfers was also reported by Milam and Kusch in 1938,
who noted high parasite counts (>10%) in three patients infected after serial passage, in
contrast to the low parasitemia (<1%) noted in the majority of their patients, leading these
authors to speculate that the parasite may become better adapted to the human host with
successive transfers (46). This report also described relative resistance to P. knowlesi
infection in six African Americans, confirmed by another report the same year of decreased
susceptibility to infection in 12 African Americans compared to 30 Caucasian patients (47).
In both these reports several African American patients were described as having subclinical
infection, with no parasites seen on blood film and no fever experienced, but with infection
produced in monkeys inoculated with their blood.
In 1965 the first case of naturally acquired human infection with P. knowlesi was reported
(2). This occurred in a 37 year old American male, who became unwell in Bangkok after
having spent four weeks working as a surveyor for the U.S. Army in the forests of Pahang,
Peninsular Malaysia (2). After returning to his home in Maryland he was initially diagnosed
by microscopy with P. falciparum, before being referred to the Clinical Centre of the National
Institute of Health, where his diagnosis was amended to P. malariae. Prior to treatment, a
sample of his blood was forwarded to a research laboratory at the U.S Penitentiary, Atlanta,
Georgia, where malaria studies were being conducted, and a “P. malariae” strain had been
sought (34). Inoculation of the infected blood into seven human volunteers and
subsequently into rhesus monkeys led to infection of all humans and the death of the
12
monkeys, confirming the diagnosis of P. knowlesi (2). Several years later Chin et al.
demonstrated the experimental mosquito transmission of P. knowlesi from man to man,
monkey to man, and man to monkey (48). All five patients exposed to infected mosquitoes
became infected, with one acquiring infection after only a single bite.
These cases of human P. knowlesi infection, along with an earlier report of accidental
mosquito transmission to American malaria researchers of another simian malaria parasite,
P. cynomolgi (49), stimulated research efforts to determine if simian malaria posed a threat
to human populations. In a study involving the inoculation into rhesus monkeys of pooled
blood from 1117 humans living in forest-fringe villages in West Malaysia, no malaria
infections were produced, and it was assumed that human P. knowlesi infection was a rare
event (50).
In 2004 however Singh et al. reported a large focus of naturally acquired P. knowlesi
infections in humans in Sarawak, Malaysian Borneo (3). In the preceding years Singh and
colleagues had noticed that a higher than expected proportion of all malaria cases in the
Kapit district were being diagnosed by microscopy as P. malariae. Unusual features of
these infections were noted, including high parasitemia and increased severity of infection.
In a preliminary investigation PCR was performed on 5 of these isolates, and was positive
for Plasmodium species in each but negative for all 4 recognized human malaria species.
Subsequent DNA sequencing of the small-subunit (SSU) rRNA and the circumsporozoite
(csp) genes indicated that 8 patients diagnosed with “P. malariae” were infected with P.
knowlesi. P. knowlesi-specific primers were then developed (Pmk8 and Pmkr9), and in a
prospective study conducted in the Kapit division between 2000 and 2002, 120 (58%) of 208
malaria cases were found to be positive for P. knowlesi DNA, with no case of P. malariae
identified (3).
13
2.2. Epidemiology
2.2.1 Geographical distribution and emergence of P. knowlesi in Malaysia
Following the initial discovery of a large focus of P. knowlesi malaria in humans in Kapit,
Sarawak, Cox-Singh et al. reported detecting P. knowlesi DNA by PCR in 266 (28%) of 960
malaria blood slides taken from hospitalized patients in various districts in Sarawak between
2001 and 2006 (23). P. malariae DNA was detected in only four (0.4%) patients, all of whom
had recently returned from overseas, suggesting a lack of indigenous cases of P. malariae in
Sarawak. In Sabah, P. knowlesi DNA was detected in 41 (84%) of 49 “P. malariae” blood
films taken during 2003-2005, and in all of five “P. malariae” blood films from Pahang,
Peninsular Malaysia, during 2004-2005 (23). In another study published the same year,
Vythilingam et al. detected P. knowlesi DNA in 73 (78%) of 93 ”P. malariae” blood films
taken in Peninsular Malaysia during 2005-2008 (51). These studies confirmed widespread
distribution of P. knowlesi across Malaysia at least since the early 2000s. In an attempt to
confirm earlier cases of human P. knowlesi infection in Sarawak, PCR was performed on 36
archival “P. malariae” blood slides from 1996, the earliest available, with P. knowlesi
detected in 35 (97%), and only one being positive for P. malariae (with the origin of infection
unable to be determined) (4).
In 2009 Imwong et al. reported that the P. knowlesi Pmk8/Pmk9 PCR primers used in these
studies demonstrated cross-reactivity with a number of P. vivax isolates, potentially calling
into question the prevalence of P. knowlesi reported in the preceeding studies (52).
However, Imwong et al. made note of the fact that in all such studies the presence of P.
vivax in tested samples had been excluded and/or P. knowlesi had been confirmed through
amplification and sequencing of other P. knowlesi genes (52).
14
The existence of human cases of P. knowlesi in Malaysia prior to the PCR-confirmed cases
has been harder to determine. Evolutionary analyses of sequence data from samples
obtained from Sarawak indicate that P. knowlesi probably existed in macaques in Southeast
Asia more than 100,000 years ago, with infection in humans likely occurring from the time of
human arrival in the region (53). With a lack of archived blood samples on which to perform
PCR testing however, additional information is limited to descriptions of the earliest
microscopy-based malaria surveys conducted in Sabah and Sarawak.
The most detailed of these surveys was conducted in the Tambunan District, Sabah, during
1939 – 1942 (54). In a report of some of this work, McArthur describes a “peculiar and
interesting situation” on the Tambunan Plain, where villages in the middle of the plain were
“relatively healthy… but that malaria steadily increased on approaching the surrounding hills”
(55). This unusual distribution of malaria led McArthur and colleagues, after an extensive
two-year search, to the discovery that the jungle-breeding An. balabacencis (then known as
A. leucosphyrus, and now known to be capable of transmitting P. knowlesi) was the primary
vector of human malaria in Tambunan, and elsewhere throughout northern Borneo. In
another report, McArthur described “the presence in certain hill ravines of pockets of P.
malariae” (54), with this raising the possibility that these were actually cases of P. knowlesi.
Although McArthur notes a “peculiar age incidence” of malaria in certain kampongs in
Tambunan, with spleen rates peaking later than expected, the kampongs where this
occurred were not those in which the prevalence of P. malariae was high (54). Furthermore,
McArthur also reports that in the “malarious kampongs” (ie. those in the hill ravines) the
gametocytes of P. malariae were “by far the most commonly recorded”. A recent study
found that in an area where P. malariae and P. falciparum coexisted, P. malariae
gametocyte prevalence and density were higher than those of P. falciparum, despite P.
falciparum being the predominant species (56). In contrast, Lee et al. report that although
gametocytes were seen in 4 of 10 patients with knowlesi malaria, they comprised only 1.2 –
15
2.8% of infected erythrocytes (57). In our prospective study in Sabah (discussed in Chapter
8), gametocytes were reported by our research microscopist in <1% of patients with P.
knowlesi. The finding in McArthur’s survey of a high prevalence of “P. malariae”
gametocytes may therefore argue against these cases being P. knowlesi.
In the 2008 report by Cox-Singh et al. in which PCR was performed on 49 “P. malariae”
blood slides taken in Sabah during 2003 – 2005, 8 (16%) of these slides were confirmed to
be P. malariae by PCR (23). Six of these were from Kudat District in northeastern Sabah,
with 5 involving children aged 7 – 15 years, from 2 villages, and with no history of having
travelled outside Sabah. Although 70 years after the malaria surveys conducted on the
Tambunan Plain, one wonders if these 6 cases reflect a rare remaining “pocket” of P.
malariae. However, larger and more recent studies in Sabah have consistently found that
<1% of microscopy-diagnosed “P. malariae” cases are actually P. malariae by PCR, with the
large majority being P. knowlesi (25, 26, 58).
Recent microscopy data from the Sabah State Reference Laboratory, together with
notification data from the Sabah Department of Health, indicates a substantial recent
increase in cases of microscopy-diagnosed “P. malariae”. In 2001, 96 (1.6%) of 6050
malaria slides referred to the Sabah State Public Health Laboratory were diagnosed as P.
malariae monoinfection by microscopy, with the proportion increasing to 59 (2.2%) of 2741 in
2004 (26). In contrast, microscopy-diagnosed “P. malariae” accounted for 703 (35%) of
1936 malaria cases reported to the Sabah Department of Health in 2011 (59). This apparent
emergence of P. knowlesi in Sabah is discussed in detail in Chapter 7.
In Sarawak, the earliest documented malaria survey was conducted in 1952 (60). In this
report “P. malariae” accounted for 142 (33.7%) of 421 malaria cases detected during
community screening in six regions, and, as with Sabah, A. leucosphyrus was found to be
16
the predominant malaria vector (60). In contrast to Sabah, recent evidence from Sarawak
suggests a lack of indigenous cases of P. malariae when PCR methods are used (3, 23, 24),
and it has therefore been argued that at least some of these early cases of P. malariae may
have been P. knowlesi (4).
2.2.2 Geographical distribution of P. knowlesi outside of Malaysia
Since the recognition of P. knowlesi as a cause of human malaria in Malaysia, cases have
been reported among residents of nearly every other country in Southeast Asia, including
Thailand (5, 7, 61), Vietnam (10, 62), Singapore (11, 63, 64), Philippines (9), Indonesia (65),
Brunei (66), Myanmar (6, 67, 68), Cambodia (8), India (69), and southern China (67), with
cases also increasingly reported in travellers returning from these countries to non-endemic
regions (12-22). Among these countries, the largest surveys have been conducted in
Vietnam, Cambodia, Thailand, Myanmar, and India.
In the Ninh Thuan Province of south Vietnam, P. knowlesi was detected by PCR in 5 (5.2%)
of 95 “P. malariae” blood films collected during a cross-sectional survey in 2004, with all 5
taken from asymptomatic individuals (29). A higher prevalence of P. knowlesi infection was
reported in Khanh Hoa Province of south-central Vietnam, where P. knowlesi was detected
in 19 (51%) of 37 PCR-positive malaria slides identified from a cross-sectional survey, and in
13 (15%) of 88 PCR-positive malaria slides collected by targeted active case detection
during 2010 (10). Of these 32 cases, all were coinfected with another Plasmodium species,
and 6 (19%) reported fever. In Cambodia, P. knowlesi DNA was detected among 2 (0.3%)
of 754 PCR-positive malaria patients during 2007 – 2010 (8). In Thailand, P. knowlesi was
detected by PCR in 34 (0.65%) of 5,254 PCR-positive malaria blood films taken from
patients attending malaria clinics during 1996, 2006-2007, and 2008-2009 (5, 61). These
included 2 (0.9%) of 215 from Prachuab Khirikhan in the southwest, 16 (0.6%) of 2485 from
17
Yala and Narathiwat in the south, 8 (1.2%) of 662 blood films from Chantaburi in the east,
and 8 (0.4%) of 1897 blood films from Tak in the northwest (5, 61). Patients had a median
age of 34 years (range of 4 – 59 years), the majority lived in proximity to macaques, and all
had uncomplicated malaria (5). As with cases reported from Vietnam (10), most patients
(71%) in Thailand infected with P. knowlesi were also infected with other Plasmodium
species.
In southern Myanmar in 2008, P. knowlesi was detected in 32 (22%) of 146 patients with
uncomplicated malaria, with 28 (88%) occurring as coinfections (6). In a later study
conducted on the Thai-Myanmar border, P. knowlesi was detected by PCR in 2 (0.5%) of
419 patients attending a malaria clinic, with both patients having worked in the Koh Song
Province of southern Myanmar (68). In the Andaman and Nicobar islands, situated between
the Andaman Sea and the Bay of Bengal, P. knowlesi was detected by PCR in 53 (12%) of
445 malaria patients during 2004 – 2010, with 50 (94%) of these occurring as coinfections
(69).
While confirming widespread distribution of P. knowlesi across southeast Asia, these studies
are remarkable both for the high proportion of coinfections reported, and the lack of reports
of severe knowlesi malaria. These findings differ from studies conducted in Malaysian
Borneo, where coinfections are relatively rare (7/188 [4%] in the prospective Kapit study),
and severe disease is increasingly reported (24, 31). The high proportion of coinfections
reported in Thailand, Myanmar, Vietnam and India does not appear to be due to the
previously reported cross-reactivity between P. vivax, and P. knowlesi primers (Pmk8 and
Pmkr9) targeting the P. knowlesi ss rRNA gene (52). Although these primers were used in
the surveys conducted in Vietnam (10) and Myanmar (6), the authors of both papers
acknowledged the limitations of these primers, and consequently subjected their samples to
additional testing, including PCR targeting of the P. knowlesi csp gene in the Vietnam study
18
(51), and sequence analysis in the Myanmar study (6). Furthermore in the Myanmar study,
coinfections with P. falciparum were as common as coinfections with P. vivax. In the larger
Thailand survey (5) and in the Andaman/Nicobar Islands survey (69) a nested PCR assay
was used with Pk18SF and Pk18SRc primers, which the authors report (69) did not cross-
react with other Plasmodium species.
In Thailand, Vietnam, Myanmar and India the higher proportion of P. knowlesi/P. vivax and
P. knowlesi/P. falciparum coinfections occurs in the context of higher background
transmission of P. falciparum and P. vivax, compared to Malaysia where prevalence of these
species is now low (70). Jiang et al. speculate that the high proportion of coinfections in
their Myanmar study may indicate that humans infected with other malaria parasites may be
more vulnerable to P. knowlesi (6). Alternatively, infection with P. knowlesi may be
associated with P. vivax relapse or P. falciparum recrudescence, with similar interactions
reported to occur between these latter two species (71).
Despite these reports of P. knowlesi coinfections, the lack of reports of severe knowlesi
malaria in these countries may indicate a degree of cross-species immunity between P.
knowlesi and the other human malaria species. Although heterologous immunity does not
generally occur between human malaria species, it has been argued that a degree of cross-
resistance may be more likely to occur between species infecting different hosts (72). This
is supported by data from the neurosyphilis studies, where patients who had been previously
infected with P. vivax were less susceptible to infection with P. knowlesi (41). If occurring,
this cross-species immunity may explain the apparent higher prevalence of symptomatic
knowlesi malaria in Malaysia, where P. falciparum and P. vivax have largely been controlled,
and yet where environmental conditions remain suitable for the transmission of P. knowlesi
(70). Further discussion on possible interactions between P. knowlesi and other malaria
species, including density-dependent regulation, is discussed in Chapter 7.
19
While the above studies suggest widespread distribution of P. knowlesi across south-east
Asia, the true burden of disease remains unknown due to the inability to accurately diagnose
the species by microscopy. The population at risk of P. knowlesi infection includes all those
living within the overlapping distribution of the simian hosts and the Anopheles leucosphyrus
group of mosquitoes, which extends from eastern India and Bangladesh through to the
Philippine archipelago and Indonesia and includes ≈500 million people. Cross-sectional
surveys using molecular methods of diagnosis, in addition to enhanced clinical surveillance
throughout the region, are required to further determine the true burden of P. knowlesi
malaria.
2.2.3 Reservoir Hosts
The predominant natural hosts of P. knowlesi include the long tailed macaques (Macaca
fascicularis) and pig tailed macaques (M. nemistrina). These macaques share a similar
geographic distribution, extending from India, Bhutan and Bangladesh through to the
Philippines, and southwards to Indonesia (Figure 1). Long-tailed macaques have a diverse
habitat, including primary and secondary forests, and riverine, swamp, and coastal forests of
nipa palm and mangrove. They are found particularly in forest-fringe areas, and thrive near
human settlements, where they may approach or enter houses to forage for food, and are
sometimes kept as pets (73). Pig-tailed macaques, in contrast, prefer undisturbed primary
rainforest, although may be found in swamp and secondary forests.
The highest prevalence of P. knowlesi in macaques has been reported from Kapit, Sarawak,
where large numbers of human knowlesi malaria cases have been reported. In this region
P. knowlesi was detected in 71 (87%) of 82 long-tailed macaques and in 13 (50%) of 26 pig-
tailed macaques (53). Other Plasmodium species were also commonly detected in the
20
macaques, including P. inui (82%), P. coatneyi (66%), P. cynomolgi (56%), and P. fieldi
(4%), with mixed infections found in 84%. In Peninsular Malaysia, P. knowlesi was detected
in 10 (13%) of 75 long-tailed macaques from Kuala Lipis, Pahang, but was not identified in
any of 41 macaques from Selangor or in 2 macaques from Kuala Lumpur (51)
In southern Thailand where human cases of P. knowlesi have also been reported, P.
knowlesi was identified in 7 (1.2%) of 566 macaques in the Narathiwat Province, including 5
pig-tailed macaques, one long-tailed macaque and 1 dusky-leaf macaque (Semnopithecus
obscurus) (74). No infections however were identified from 70 macaques in the Yala
Province (74), or, in another survey, from 99 long-tailed macaques in the Ranong and
Prachuab Khirikhan Provinces (73). In Singapore, P. knowlesi was identified among 3 of 3
long-tailed macaques caught from a forested area where human knowlesi malaria had been
acquired, but not identified in any of 10 peri-domestic long-tailed macaques taken from a
nature reserve (64).
Earlier reports identified P. knowlesi among long-tailed macaques from the Philippines (75,
76), and in a banded-leaf monkey (Presbytis melalophus) from Peninsular Malaysia (77).
2.2.4 Mosquito Vectors
The mosquito vectors of P. knowlesi are members of the forest-dwelling Anopheles
leucosphyrus group. In 1961, in the first demonstration of a natural vector of simian malaria,
An. hackeri was found to be infected with P. knowlesi through the inoculation of sporozoites
into a rhesus monkey (78). The mosquito was found in the bases of nipa palm leaves along
the Selangor coast in Peninsular Malaysia, and was not attracted to humans (78). An
hackeri has also been found in the peninsular region of southern Thailand (79) and in the
Philippines (80), although its role as a vector in these countries has not been investigated.
21
More recently An. cracens was found to be the main vector of P. knowlesi in Pahang,
Peninsular Malaysia (51). The mosquito bites after dusk away from houses, and feeds on
monkeys at the canopy level and humans on the ground level (51). The mosquito has also
been found in Indonesia and Thailand (80).
In Kapit, Sarawak, An. latens was recently identified as the primary P. knowlesi vector, with
the mosquitoes biting both monkeys and humans, and with peak biting times of 19:00 –
20:00 in the farms and 01:00 – 02:00 in the forest (81). In this study although 126 (11.7%)
An. latens were found in longhouses, none were infected, suggesting transmission occurs
away from people’s homes (81). In Sabah, a P. knowlesi-infected An balabacensis was
recently found in Ranau (82). This mosquito readily bites monkeys at canopy level (83), and
has been found in laboratory studies to be a highly efficient vector of P. knowlesi, with >1000
sporozoites found in mosquito salivary glands and infection readily produced in monkeys
(84). More recently however, An. donaldi was shown to have replaced An. balabacensis as
the primary species in the Kinabatangan region of Sabah. The peak outdoor biting time for
this mosquito occurs from 18.00-19.00 hours, when humans are often outside their homes,
and indoor biting occurs throughout the night (85). Although sporozoites were detected in
An. donaldi in the Kinabatangan region, species identification was not performed and further
studies are required to determine the role of this species in the transmission of knowlesi
malaria.
In the Khanh Phu forest in southern Vietnam P. knowlesi sporozoites are carried by An.
dirus, mostly in combination with P. vivax (10). An dirus, the main vector of human malaria
in Thailand, Laos, Cambodia and Vietnam, and closely related to An balabacensis, has been
shown to maintain high levels of malaria endemicity at low population densities (86) and is
capable of adapting to deforestation (87).
22
2.2.5 Transmission of P. knowlesi
Although the experimental transmission of P. knowlesi from human to human by infected
mosquitoes was demonstrated in 1968, studies to date suggest that P. knowlesi remains
primarily a zoonosis. In Kapit, Sarawak, Lee et al. sequenced the circumsporozoite (csp)
gene and mtDNA of P. knowlesi isolates from 31 humans and 16 macaques, and found a
higher number of P. knowlesi csp alleles and mtDNA genome haplotyes per infection in
macaques compared to humans (53). Both hosts shared some diverse alleles of the P.
knowlesi csp gene and certain mtDNA haplotypes, and there were no mtDNA lineages
exclusive to either host. The authors therefore concluded that in the Kapit region P.
knowlesi remains a zoonosis with macaques as the reservoir hosts, a finding also supported
by the lack of any case-clustering in Kapit despite the presence of communal long houses in
the area (3, 24). Similar findings were reported from Singapore, where sequencing of the P.
knowlesi csp gene was conducted on samples from 3 infected macaques and 4 humans
(64). Again, genotypes were found to be shared between hosts, indicating macaque to
human transmission. It should however be noted that neither of these studies exclude the
possibility that human-human transmission of P. knowlesi does occur, with regular
transmission from humans back to macaques possibly accounting for the absence of any
gene sequences found to be exclusive to the human host.
In Thailand, Jongwutiwes et al. sequenced the complete P. knowlesi merozoite surface
protein 1 (Pkmsp-1) gene from 10 human and 5 macaque blood samples (5). Although they
found considerable genetic diversity among isolates, the sequence from one human was
identical to that from a pig-tailed macaque. An identical Pkmsp-1 gene sequence was also
found in 2 patients from the same area concurrently infected with P. knowlesi, with another
identical sequence seen in another 2 patients, from a different area, who developed fever a
few days apart. Although the authors report that these results support macaque-human
transmission, in a subsequent study involving further analysis of these and additional human
23
and macaque P. knowlesi isolates, they report that the number of haplotypes, haplotype
diversity, nucleotide diversity and recombination sites of human-derived Pkmsp-1 sequences
was greater than that of the monkey derived sequences, with these later results raising the
possibility of human-human transmission (88).
Figure 4. Map showing countries where human cases of knowlesi have been reported,
together with the natural distribution of long-tailed and pig-tailed macaques and the
Anopholes leucosphyrus group of mosquitoes. Reproduced from Singh and Daneshvar
(167).
24
2.3. Knowledge gaps identified from the literature search
The above review of the literature has identified significant gaps in our understanding of the
epidemiology of knowlesi malaria, with some of these addressed in this thesis. Firstly,
although Cox-Singh et al. (23) detected P. knowlesi by PCR in a significant proportion of “P.
malariae” blood slides in Sabah, information regarding the incidence of P. knowlesi in
comparison to the other human malaria species in Sabah was not available at the time of
commencement of this thesis. There was also no information regarding the distribution of P.
knowlesi throughout Sabah, in particular whether P. knowlesi was occurring throughout the
state or only in certain districts. In addition, there was no information on the trends of P.
knowlesi in Sabah over time, and hence it was not known whether the newly-recognised
human cases of P. knowlesi represented a true emergence of the species in humans.
This thesis seeks to address these knowledge gaps through the 3 retrospective studies
described in Chapters 5 – 7. Chapters 5 and 6 present retrospective studies of the 2009 –
2011 malaria microscopy records at a district hospital in northeast Sabah. These studies
aim to describe the number of “P. malariae/P. knowlesi” diagnoses in comparison to the
other malaria species, and to describe the trend of P. knowlesi, P. falciparum and P. vivax
malaria over the 3 year study period. Chapter 7 presents a retrospective review of all
available Sabah Department of Health malaria notification data from 1992 – 2011, in order to
obtain information regarding the distribution of knowlesi malaria throughout Sabah, and the
trend of P. knowlesi and the other human malaria species in Sabah over the past 2 decades.
Other substantial knowledge gaps are identified in the above literature search, and are
beyond the scope of this thesis. There is limited understanding of the interaction between P.
knowlesi and the other human malaria species, and it remains possible that this interaction
may account for the apparent differences in the epidemiology of P. knowlesi malaria
25
between Malaysian Borneo and other countries in Southeast Asia, including a greater
proportion of coinfections and a lower incidence of severe disease. There is also little
information available regarding the true burden of P. knowlesi malaria throughout Malaysian
Borneo and elsewhere in southeast Asia, and large cross-sectional surveys using molecular
methods will be required to address this question. Finally, this thesis does not address the
substantial knowledge gaps that exist regarding the occurrence of P. knowlesi and other
malaria species in macaques in Sabah, and moreover, the mosquito vector of P. knowlesi in
Sabah.
26
3.1. Diagnosis
3.1.1 Microscopy
The morphology of P. knowlesi in humans was first described by Knowles and Das Gupta,
who described the features of the parasite in man and macaques, and noted in particular the
difference in appearance of the parasite in the different hosts (37). In long-tailed macaques
the parasite resembled P. vivax, with enlarged red cells and Schuffner’s dots observed,
while in rhesus macaques the appearance was closer to that of P. falciparum, with red cells
of normal size and no stippling seen. In man, the parasite was noted to have “a general
resemblance towards P. malariae”, with little or no amoeboid activity, and red cells of normal
size. Stippling was not observed with Giemsa or Leishman’s stain in red cells infected by
rings and trophozoites, but was sometimes seen in schizont-infected red cells. Gametocytes
were “extremely rare findings.”
A more recent description has been provided by Lee et al., who described the appearance of
thick and thin blood films from 10 randomly selected patients with P. knowlesi (57). While
parasite counts were only moderate (median 5266 parasites/uL) one patient had a
parasitemia of 236,000 parasites/μL. Infections were generally asynchronous, with early
and late trophozoites and schizonts appearing in 7 patients. Gametocytes were seen in 4
patients, but comprised only 1.2 – 2.8% of infected red cells. Early trophozoites were
characterized by a ring-like cytoplasm and resembled those of P. falciparum, with double
chromatin dots, multiply-infected erythrocytes and applique forms all commonly seen (57).
Late trophozoites were almost indistinguishable from those of P. malariae, with dense
irregular cytoplasm and band-like forms commonly seen. The schizonts of P. knowlesi
27
contained up to 16 merozoites, more than the maximum of 12 observed in schizonts of P.
malariae, and P. knowlesi merozoites were more irregularly arranged (57). P. knowlesi-
infected erythrocytes were of normal size, and stippling was not observed, although in one
patient “irregular and sparse dots that were unevenly distributed” were seen in some
erythrocytes containing mature trophozoites. Stippling has reportedly been observed in a
case report from Thailand (7), and in an Australian man returning from Kalimantan,
Indonesian Borneo (16).
While microscopy remains the standard method of diagnosis for malaria in Malaysia and
most of the malaria-endemic world, the morphological similarities between P. knowlesi and
P. malariae render this method inadequate for the diagnosis of these species. The reliance
on microscopy for the diagnosis of Plasmodium species has allowed the existence of human
cases of knowlesi malaria to remain unrecognized for likely many decades, and continues to
make it impossible to estimate the true disease burden of P. knowlesi. The misdiagnosis of
P. knowlesi as the more-benign P. malariae is also likely to have clinical consequences, with
clinicians failing to recognize the potential for severe disease (discussed in more detail in
Chapter 9).
In addition to the difficulties with distinguishing P. knowlesi from P. malariae, P. knowlesi
may be confused with the other human Plasmodium species. In two previous studies in
Sarawak (3, 23), 15/141 (11%) and 33/312 (11%) patients diagnosed by microscopy as “P.
malariae” were actually P. falciparum by PCR, and 9/141 (6%) and 43/312 (14%) were P.
vivax. P. knowlesi DNA was detected in 4/25 (16%) and 11/216 (5%) patients diagnosed by
microscopy with P. falciparum, and 2/42 (5%) and 16/428 (4%) of those with microscopy-
diagnosed P. vivax. We have recently reported similar findings from Sabah, discussed in
Chapter 10. Alternative methods of diagnosis are therefore required in areas endemic for P.
knowlesi.
28
3.1.2 Rapid Diagnostic Tests
Rapid Diagnostic Tests (RDTs) provide an alternative method of malaria diagnosis. They
can be performed with minimal training, and are increasingly used in malaria-endemic
countries. The antigens most commonly detected by RDTs include P. falciparum-specific
histidine-rich protein 2 (PfHRP2), genus-specific aldolase (pan-aldolase), genus-specific
parasite lactate dehydrogenase (pan-pLDH), and P. vivax and P. falciparum-specific pLDH.
Prior to the commencement of this thesis, information regarding the utility of RDTs for the
diagnosis of P. knowlesi came primarily from case reports of returned travelers with knowlesi
malaria (Table 1). In some of these cases, RDTs detecting the pan-pLDH antigen were
positive (12, 20, 22, 89), as was an RDT detecting the pan-aldolase antigen (12, 22, 28).
However, the ability of RDTs to detect these antigens appeared to be poor at low parasite
densities (13, 15, 21, 22, 89). RDTs detecting P. vivax and P. falciparum-specific pLDH
were also found to cross-react with P. knowlesi (20, 22, 89, 90), suggesting that these RDTs
will not be able to differentiate P. knowlesi from these human malaria species.
A prospective evaluation of two RDTs for the diagnosis of knowlesi malaria is described in
Chapter 11.
29
Table 1. Case reports of Rapid Diagnostic Tests used to diagnose P. knowlesi
Rapid Diagnostic Test
SampleParasite count
BinaxNOW OptiMALCore Malarial Pan/Pv/Pf test RDT Palutop+4 Test
Entebe Malaria Cassette
RefPan-malaria aldolase
PfHRP2pan-malarial LDH
Pf-LDH
Pv-LDH
Pan-malarial LDH
PfHRP2 Pv-LDH
Pan-malarial LDH
PfHRP2 Pv-LDH
PfHRP2
Human 0.80% + ― + + - + + ― (12)Human 0.10% ― ― (13)Human 0.003% ― ― + + (20)
Human 2% + ― + + (22)
1587/μL ― ― + + 138/μL ― ― ― ―
Human 0.20% + + (28)Human 0.0005% ― ― (91)Human 0.5% ― ― (15)Human 0.20% ― ― (21)Monkey 28.30% + + + ― (89)
11.2% + ―0.04% ― ―
3.1.3. Molecular methods of diagnosis
Molecular methods are required to definitively diagnose P. knowlesi malaria, and several
assays have now been developed including nested PCR (3, 52), real-time PCR (91-94),
single-step PCR (95), and loop-mediated isothermal amplification (LAMP) (96, 97).
The first PCR assay developed for the detection of P. knowlesi, reported by Singh et al., was
a nested PCR assay which used Pmk8 and Pmkr9 primers to target a P. knowlesi small
subunit ribosomal RNA (ssrRNA) gene (3). This assay however was found to produce
occasional false positive amplification of P. vivax DNA, resulting in P. vivax monoinfections
being misdiagnosed as P. knowlesi/P. vivax mixed infections (52). Alternative primers
targeting the ssrRNA gene were subsequently developed by Imwong et al. (PkF1060 and
PkR1550) which did not demonstrate cross-reactivity with other plasmodium species, and
30
which detected P. knowlesi with a sensitivity of 1 - 10 parasite genomes/sample (52). In a
later study in Sarawak another primer set targeting the ssrRNA gene (Kn1f and Kn3r) was
reported by Lee et al. to have replaced the original Pmk8/Pmkr9 primers (53).
Several real-time PCR assays have also been developed (91-94), and one of these has
been validated with clinical samples from 40 patients (92). P. knowlesi DNA was detected
from all 40 samples, including one patient with a parasite count of only 3 parasites/μL, and
no cross-reactivity with other Plasmodium species was demonstrated.
More recently LAMP has been described as a potential low-cost and simple technique that
rapidly amplifies target DNA under isothermal conditions, and hence may be suitable for use
in field conditions. Two assays have been developed, with one targeting the β-tubulin gene
(96) and the other the apical membrane antigen-1 (AMA-1) gene (97). This latter assay
detected P. knowlesi in all of 13 patients with microscopy-diagnosed P. knowlesi, including a
sample with parasitemia of <0.01%, demonstrating a sensitivity greater than that of the
comparator nested PCR assay (97).
3.2. Clinical disease
At the time of writing, and with the exception of the studies undertaken for this thesis, the
more recent descriptions of the clinical features of knowlesi malaria come from case reports
(7, 9, 11-13, 15-22, 28, 66, 91, 98-100), two small case series in West Malaysia (30, 101), a
description of four fatal cases in Sarawak, Malaysia, a retrospective (3) and subsequent
prospective (24) study in Kapit, Sarawak, a case-control study in Sarikei and Sibu, Sarawak
(102), and a retrospective study at Queen Elizabeth Hospital (QEH), Sabah, Malaysia (31).
The results of these studies are discussed below in this introductory chapter. In addition,
31
further details of the clinical features of severe and non-severe knowlesi malaria are
provided in Chapter 8, which presents the results of a prospective comparative study of the
epidemiologic and clinical features of patients with knowlesi, falciparum and vivax malaria at
QEH, Sabah (103). This study represents the largest prospective study of knowlesi malaria
to date, and the largest series of severe knowlesi malaria.
3.2.1 Demographics
Only one previous study has compared the age distribution of patients with knowlesi malaria
to that of patients with falciparum or vivax malaria. In this prospective study of 107 adult
malaria patients admitted to Kapit District Hospital, patients with P. knowlesi had a mean age
of 45 years, compared to 39 years and 36 years for those infected with P. falciparum or P.
vivax respectively (p=0.006) (24). Confounding factors in this study however prevented
direct comparisons between species, with P. knowlesi infections being predominantly
acquired locally, while P. falciparum and P. vivax infections occurred in logging camp
workers returning from other countries. In the earlier retrospective study conducted at the
same site, the mean age of 106 patients with knowlesi malaria was 35 years (range 10 – 76
years), with 92% of these cases occurring in adults >15 years (3). Patients with severe
knowlesi malaria appear older than those with non-severe disease. In a retrospective study
at an adult tertiary referral hospital patients with severe knowlesi malaria had a mean age of
57 years compared to 37 years among those with non-severe disease (p<0.05) (31), while in
a recent case-control study, adults with severe and non-severe disease had a mean age of
50 and 43 years respectively (p=0.11) (102). Increased risk of severity with age is also
known to occur with P. falciparum (104).
Patients with knowlesi are predominantly male, with proportions of 56 – 80% reported from
the two previous prospective (24, 102) and 2 retrospective studies (3, 31). Among case
32
reports of knowlesi malaria in returned travelers, only one of 11 (9%) has involved a female
patient (14).
3.2.2. Incubation and pre-patent periods
Information regarding the incubation period of knowlesi malaria comes primarily from early
reports of experimental human infection and of patients infected with P. knowlesi as a
treatment of neurosyphilis. Unless otherwise stated the following reports refer to inoculation
of P. knowlesi-infected blood. In Das Gupta and Knowles’ first description of experimental
human infection, the incubation period of the three patients infected was 10, 20 and 23 days,
with parasites appearing in the blood up to 5 days after the initial temperature rise (37). Van
Rooyen and Pile described incubation periods of 3 – 14 days in their 12 patients treated for
neurosyphilis, with the majority experiencing the first temperature rise after the 8th day, and
with parasites appearing in the blood on about day 10 (41). Pre-patent periods of 4 - 27
days were reported in another series of 42 patients treated for neurosyphilis, with the longer
periods noted in African Americans (47). In a study involving 5 patients infected with P.
knowlesi through the bites of infected mosquitoes, pre-patent periods of 9 – 12 days were
reported (48).
3.2.3. Clinical features on presentation
The mean duration of illness on presentation has been reported as 4 – 5 days (3, 24, 31).
Clinical features are non-specific and similar to those of P. falciparum or P. vivax, including
fever and chills, rigors, headache, malaise and myalgias (24). Cough, nausea and
abdominal pain occur in about half, and diarrhea and vomiting occur in one third (24). Signs
of severe disease may be evident on admission, including jaundice, hypotension, or
33
respiratory distress (24). A palpable liver and spleen has been reported in 24% and 15%
respectively (24).
3.2.4 Laboratory findings
The most striking laboratory abnormality among patients with P.knowlesi is
thrombocytopenia, a universal finding in the prospective Kapit study, with a mean platelet
count of 72,000/μL (24). In the Sarikei/Sibu case-control study, 107 (97%) of 110 patients
had thrombocytopenia on enrolment, with a median platelet count of 38,000 and 69,000/μL
among patients with severe and non-severe malaria respectively (102), while in the QEH
retrospective study patients with severe and non-severe malaria had a mean platelet count
of 40,000 and 70,000/μL respectively (31). Platelet count is associated with parasite density
(24), and is lower in patients with severe malaria (31, 102). Anaemia may occur, although
appears less common in knowlesi than in falciparum malaria, and is rarely severe (24, 31).
Renal impairment is common, and is associated with parasite density (24). Other laboratory
abnormalities include hyponatremia and mild transaminitis (24).
3.2.5 Severe disease
In addition to the 44 cases described in this thesis (Chapters 8 and 9), descriptions of severe
knowlesi malaria (based on severity criteria for falciparum malaria, Figure 5) include 33
patients from Sarawak (23, 24, 102, 105), 4 patients from West Malaysia (30, 99), 24
patients from Sabah (27, 31, 100), and one patient from Brunei (66). These cases are
described in case reports and series (23, 27, 30, 66, 99, 100, 105), as well as a prospective
study at Kapit District Hospital, Sarawak (n=10) (24), a retrospective study at QEH, a tertiary
34
referral hospital in Sabah (n=22) (31), and a case-control study at two district hospitals
(Sarikei and Sibu) in Sarawak (n=17) (102).
At Kapit District Hospital, severe disease occurred in 10 (9.3%) of 107 patients (24), while at
QEH, 22 (39%) of 56 patients had severe disease, with the higher rates of severity likely
reflecting referral bias (31). In the QEH study, the Kapit study, and the Sarieki/Sibu case-
control study, the mean age of patients with severe disease was 57, 64 and 50 years
respectively, with age being a risk factor for severe disease in all of these. The Kapit study
and the QEH retrospective study reported female sex as a risk factor for severe disease;
however multivariate analysis adjusting for age was not performed.
Among the 62 patients with severe disease in the studies listed above, the most common
manifestations of severity included acute renal impairment (n=42, 68%), respiratory distress
(n=31, 50%), jaundice (n=28, 47%), hypotension (n=22, 35%) and hyperparasitemia (n=22,
35%), with metabolic acidosis, anaemia, and hypoglycaemia also reported. Gastrointestinal
bleeding has been reported in 2 patients with severe disease and hyperparasitemia: one
with gastric and duodenal ulcers (30); and one with invasive mucormycosis involving the
sigmoid colon (66). Perforated gastric ulcer was also reported in one fatal case (23). No
case of coma was reported among these 62 patients. In the Kapit, QEH and the
Sarikei/Sibu case-control study case fatality rates among patients with severe disease were
20% (2/10), 27% (6/22) and 24% (4/17) respectively. In the Kapit and QEH studies all but
one of the fatal cases were treated with intravenous quinine, while in the Sarikei/Sibu case-
control study all patients with severe malaria were reported to have been treated according
to WHO guidelines for severe malaria, although the drugs regimens were not specified.
Although parasite counts among patients with non-severe knowlesi malaria appear to be
lower than those of patients with non-severe falciparum malaria (24), high parasite counts
35
are not uncommon among patients with severe knowlesi malaria. Three (2.8%) patients in
the prospective Kapit study had parasite counts >100,000/μL (24), as did 16 patients in the
case-control study (102), another two fatal cases in Sarawak (23), all 4 patients with severe
knowlesi malaria in West Malaysia (30, 99), a fatal case in Sabah (100), and a case in
Brunei (66). In Kapit, parasite count was shown to be a predictor of complications, with an
area under the receiver operating curve (ROC) of 0.90 (24). Furthermore, parasitemia was
more strongly correlated with features of severity in knowlesi malaria than it was in either
falciparum or vivax malaria (32). In multivariate analysis in the case-control study in
Sarawak, a parasite count of >35,000/μL was associated with a 10-fold increased risk of
severity (102). Severe disease has however been reported with low parasite counts (31),
and the threshold parasitemia at which complications occur requires further evaluation.
While the multi-organ failure seen in knowlesi malaria is similar to that of severe falciparum
malaria, the proportion of patients with shock and acute respiratory distress syndrome
(~50%) was higher in the Kapit and QEH study than has been reported in severe P.
falciparum. Respiratory distress in these studies was characterised by low partial pressure
of oxygen, and was strongly associated with parasite density (24). In contrast to severe
falciparum malaria, severe anaemia does not appear common, and coma has not been
reported in the 21st century literature.
36
� Unrousable coma
� Severe anaemia (<7.1g/dL)
� Jaundice (bilirubin >42 μmol/L)
� Acute kidney injury (Cr>265 μmol/L)
� Hypotension (blood pressure ≤80 mmHg)
� Metabolic acidosis (lactate >6 mmol/L)
� Hypoglycaemia (serum glucose level <2.2 mmol/L)
� Respiratory distress (respiratory rate >30 breaths /
minute plus oxygens saturation <95% on room air,
and/or pulmonary infiltrates on chest radiograph
� Hyperparasitemia (parasite count >100,000/μL)
Figure 5. Severity criteria for Plasmodium knowlesi malaria*
*The studies described in this chapter use slightly variable severity criteria for P. knowlesi
malaria, but all are based on severity criteria for P. falciparum malaria (106). This figure lists
the criteria used in the 2009 prospective study by Daneshvar et al (24). The criteria used in
the prospective study described in Chapter 8 are listed in that chapter.
37
3.3. Treatment
3.3.1 Treatment of uncomplicated Plasmodium knowlesi malaria
P. knowlesi can be assumed to have no acquired drug resistance if wholly zoonotic, and
multiple antimalarial drugs have been used to successfully treat patients with uncomplicated
disease. Historically, chloroquine has been the most commonly used agent, primarily due to
the widespread misdiagnosis of P. knowlesi as P. malariae. Chloroquine was used to
successfully treat the first case of naturally acquired human knowlesi malaria (2), and in a
retrospective study was given in combination with primaquine to 92 patients at Kapit District
Hospital, Sarawak, Malaysia (3). In this latter report there was no evidence of early
treatment failures, and one patient with a parasitemia of 117,600 parasites/μL had a
negative blood film by day three. In a subsequent prospective study at the same site 73
patients with uncomplicated knowlesi malaria were treated with 3 days of chloroquine and 2
doses of primaquine (107). Median fever clearance time was 26.5 hours, and the mean time
to clearance of 50% and 90% of parasites was 3.1 and 10.3 hours respectively. No
recrudescences or re-infections were detected among the 60 patients followed up at day 28.
Chloroquine has also been used successfully to treat uncomplicated knowlesi malaria in
returned travellers (12, 15, 19, 22).
Despite this documented efficacy, a potential limitation of chloroquine may be an intrinsic
impairment in activity against certain life-cycle stages, as occurs with chloroquine against P.
vivax trophozoites (108). In contrast, artemisinins demonstrate activity against all stages of
other Plasmodium species, and have been shown to be more rapidly parasiticidal than
chloroquine (109). In keeping with this, in a recent retrospective study in Sabah, parasite
clearance times were faster among patients treated with artemether-lumefantrine (mean 1
day, range 0 – 3) than among patients treated with chloroquine (mean 2.5 days, range 1-3
days) or oral quinine (mean 2.5 days, range 1 – 3); p=0.01 (59). Returned travellers have
38
also been successfully treated with oral artemisinin therapies, including artemether-
lumefantrine (17) and mefloquine (13, 21). A randomized control trial of artemisinin
combination therapy (artesunate-mefloquine) versus chloroquine is currently underway in
Sabah, Malaysia (ClinicalTrials.gov Identifier: NCT01708876), and will provide important
information regarding the most effective treatment for uncomplicated knowlesi malaria.
Successful treatment of uncomplicated knowlesi malaria has also been reported with
atovaquone/proguanil (14, 16, 91).
3.3.2 Treatment of severe knowlesi malaria
Given that P. knowlesi has a 24-hour replication cycle that may be associated with rapidly
increasing parasite counts, and that parasite count is associated with the risk of
complications (24, 102), treatment with a rapidly parasiticidal agent is imperative for patients
with symptoms or signs of severe disease. Although intravenous quinine has been used to
successfully treat severe knowlesi malaria (24, 31), in a retrospective study at QEH, Sabah,
parasite clearance times were faster with intravenous artesunate (mean 2 days, range 1 – 3
days) than with intravenous quinine (mean 4 days, range 1 – 7; p=0.02), and fewer patients
treated with intravenous artesunate died (31). The WHO now recommends intravenous
artesunate for the treatment of all severe malaria from any species (110), and in a
prospective study conducted at QEH (see Chapter 8) we found that this treatment was
associated with zero mortality among 38 patients with severe knowlesi malaria.
39
3.4. Knowledge gaps identified from the literature search
3.4.1 Diagnosis of knowlesi malaria
The ability of RDTs for the diagnosis of P. knowlesi is important to establish, given the
difficulties with microscopic diagnosis of the species. However, at the time of
commencement of this thesis information regarding the use of RDTs for the diagnosis of P.
knowlesi came primarily from case reports of returned travellers, and no prospective
systematic evaluation had been performed. A prospective evaluation of the use of RDTs for
the diagnosis of knowlesi malaria was therefore performed as part of this thesis, and is
discussed in Chapter 11. The limitations of microscopy for the diagnosis of knowlesi
malaria are discussed further in Chapter 10, and add to existing data reported by Singh et al
(3) and Cox-Singh et al (23).
The limited data from the case reports discussed above suggest that RDTs may
demonstrate low sensitivity for the diagnosis of knowlesi malaria at low parasitemia (13, 15,
21, 22, 89). Hence, highly sensitive molecular methods will likely be required for diagnosis
of knowlesi malaria, and recently-developed LAMP assays may provide a low-cost
alternative to the existing PCR assays (96, 97). This thesis does not incorporate further
evaluation of molecular methods of diagnosis.
3.4.2. Demographics and clinical features of knowlesi malaria
Although the studies described in this chapter provide important information regarding the
demographics and clinical features of knowlesi malaria, substantial knowledge gaps remain.
Firstly, more information is required regarding the demographic features of patients with
knowlesi malaria, compared to those with falciparum and vivax malaria. Although the
40
prospective study conducted in Kapit (24) reported that patients with knowlesi malaria were
older than patients with the other malaria species, this study was confounded by the fact that
in this region knowlesi malaria was locally acquired whereas falciparum and vivax malaria
generally occurred in migrant logging workings who had recently returned from other
countries. Hence, no study had directly compared in the same population the demographic
and clinical features of patients with knowlesi, falciparum and vivax malaria. Moreover, the
Kapit study did not include children, and hence there was a need for further information
regarding the age distribution of patients presenting with knowlesi malaria, and in particular
the clinical feature of knowlesi malaria in children. Several of the studies included in this
thesis address these knowledge gaps. Chapter 5 presents a retrospective review of
knowlesi malaria in children; Chapters 6 and 7 provide further information regarding the age
distribution of patients presenting with knowlesi malaria at a district hospital, and Chapter 8
compares the demographic and clinical features of patients with knowlesi, falciparum and
vivax malaria.
The studies reviewed in this chapter also leave gaps in our knowledge of the clinical features
of severe knowlesi malaria. In particular, no study has directly compared the clinical
features of severe knowlesi malaria with those of severe falciparum and vivax malaria, and
risk factors for severity have also not been compared between the species. Furthermore,
while increasing age and increasing parasitemia both appear to be risk factors for severity
among patient with knowlesi malaria (24, 31, 102), these factors have not been evaluated
with multivariate analysis, and the association between age and parasitemia has also not
been investigated. These knowledge gaps have been addressed in the prospective study
presented in Chapter 8.
Finally, the effectiveness of intravenous artesunate and oral artemisinin combination
therapies (ACT) for the treatment of severe and non-severe knowlesi malaria respectively
41
had not been prospectively evaluated. While a controlled treatment trial does not form part
of this thesis, artemisinin therapy was standard treatment for all patients enrolled into the
prospective study described in Chapter 8, and hence this study allowed evaluation of the
effectiveness of this treatment.
42
Little is known about the pathophysiology of severe knowlesi malaria. In falciparum malaria,
the sequestration of parasitized red blood cells (RBCs) within vital organs is the hallmark of
severe disease, resulting from cytoadherence of parasitized RBCs to host endothelium,
increased aggregation of RBCs, and decreased RBC deformability, leading to microvascular
obstruction and impaired organ perfusion (111-113). These processes are exacerbated by a
dysregulated immune response, leading to increased production of proinflammatory
cytokines and activation of endothelial adhesion receptors (114, 115). In addition, severe
malarial anaemia results from intravascular haemolysis, splenic clearance, and destruction
of infected and uninfected RBCs (116, 117). In reported cases of severe knowlesi malaria
(23, 24, 27, 30, 31, 102) the multiorgan failure resembles that of severe falciparum malaria,
however the lack of coma, the greater degree of thrombocytopenia, and the few reports of
severe anaemia, suggest that mechanisms of disease are likely to differ between these two
species.
This chapter discusses the pathophysiology of P. knowlesi in macaques and other primates,
and reviews the few studies that provide information on the pathophysiology of P. knowlesi
infection in humans. Comparative features of P. falciparum malaria are also discussed.
4.1 Histopathology of severe P. falciparum malaria
Autopsy studies of patients who have died from falciparum malaria have provided much
information on the cellular and molecular processes underlying severe disease. The
histological features of severe falciparum malaria include sequestration of parasitized RBCs,
43
vascular congestion and pigment deposition within capillaries and post-capillary venules in
multiple organs including the brain, kidneys, heart, lung and muscle. In adults who have
died from cerebral malaria, electron microscopy studies have revealed extensive
sequestration of parasitized RBCs within cerebral vessels, without evidence of surrounding
inflammation or immune complex deposition (118, 119). The sequestration occurs primarily
as a result of adhesion of parasitized RBCs to host endothelium, which is mediated by
expression of the highly variable protein P. falciparum erythrocyte membrane protein 1
(PfEMP-1) on the surface of erythrocytes (120). PfEMP-1, visualized by electron microscopy
as knobs, binds to ligands on the surface of endothelial cells, including intracellular adhesion
molecule 1 (ICAM-1) and CD36 (119). Expression of ICAM-1 on cerebral blood vessels is
increased in patients with cerebral malaria, while CD36 is not usually detected in cerebral
vessels (119).
In addition to cytoadherence of parasitized RBCs to host endothelium, parasitized RBCs
also bind to other parasitized RBCs (autoagglutination) (121), to non-parasitized RBCs to
form rosettes (122-124), and to platelets to form platelet-mediated clumps (125), with these
processes further contributing to sequestration and microvascular obstruction.
4.2. Histopathology of P. knowlesi in macaques and humans
Several reports have described the histopathology of P. knowlesi infection in rhesus
macaques (Macaca mulatta) (126-128). In this highly susceptible species, infection with P.
knowlesi is characterized by widespread accumulation of parasitized RBCs, with malarial
pigment seen in multiple organs including the brain, and associated with widespread cellular
necrosis (126-128). Miller et al. described the occurrence of deep vascular schizogony (the
cyclic disappearance of infected RBCs from the peripheral circulation as the parasite
44
matures), with parasitized cells accumulating particularly in the liver at low parasitemia and
in multiple organs at higher parasitemia (128). In this report parasitized cells were
marginated along venules at low parasitemia, with capillaries containing only occasional
parasitized cells, while at higher parasitemia both venules and capillaries were packed with
infected cells. Distribution of parasites throughout cerebral capillaries was highly irregular,
with some capillaries completely occluded by parasitized cells while adjacent areas were
devoid of parasites. Deep vascular schizogony occurred to a lesser extent in P. knowlesi
than had been described in P. falciparum, and mechanisms appeared to be different, with an
occasional electron-dense invagination of the plasma membrane of P. knowlesi-infected
RBCs seen by electronmicroscopy, but no knob-like protrusions as seen with P. falciparum
(128). In another report involving rhesus macaques, intravascular fibrin deposits containing
parasitized cells “sometimes apparently were attached to the endothelium” (127).
Accumulation of parasitized cells in cerebral vasculature has also been noted in olive
baboons (Papio Anubis) infected with P. knowlesi, with more than 70% of cerebral venules
and capillaries filled with aggregates of infected and non-infected erythrocytes (129). In
contrast to the cerebral pathology of rhesus macaques infected with P. knowlesi however,
infiltration of lymphocytes and phagocytic cells between endothelial cells of cerebral blood
vessels was not observed.
There has been only one human autopsy report of a case of fatal knowlesi malaria (27). In
this report, there was widespread accumulation of pigmented infected RBCs in capillaries
and venules of the cerebrum, cerebellum, heart, kidney, spleen and liver, with multiple
petechial haemorrhages from small vessel rupture noted in the brain. No vasculitis or
perivascular inflammation was seen, and there were no platelet clumps or thrombi. In
contrast to severe falciparum malaria, cytoadherence of infected RBCs to host endothelial
cells did not appear to occur. Although electron microscopy was not performed in this
forensic autopsy, lack of cytoadherence was suggested by the interspersion of infected
45
RBCs with uninfected cells, and the lack of marginalization of parasitized cells. Furthermore,
ICAM-1, which mediates cytoadherence to brain endothelial cells in falciparum malaria (119,
130), was not detected, despite the finding from another study that P. knowlesi infected
RBCs have the ability to bind to ICAM-1 and VCAM (105). The report suggests that
important differences exist between the pathophysiology of severe knowlesi and severe
falciparum malaria, in particular the mechanisms by which parasitized cells accumulate in
the microvasculature.
4.3. Abnormalities of the microcirculation in P. knowlesi and P.
falciparum malaria
4.3.1 Reduced red blood cell deformability in severe falciparum malaria
RBCs have an average diameter of 7.5 μm, and hence must be able to elongate
considerably to pass through capillaries, which have an average luminal diameter of 3 – 7
μm. In falciparum malaria sequestration of parasitized cells reduces the luminal diameter of
capillaries and venules further, and in this context reduced deformability of RBCs may
critically impair normal microvascular flow, leading to further vascular congestion and tissue
hypoxia (113, 131). Several factors may contribute to a reduction in RBC deformability,
including reduced surface area to volume ratio, increased interval viscosity, and changes in
the cell membrane (113).
In severe falciparum malaria reduced deformability of both infected and uninfected RBCs
plays a central role in disease pathogenesis. Among infected cells, deformability reduces
with parasite maturation (113, 132). While RBCs infected with young ring forms
demonstrate increased rigidity due to increased sphericity of the RBC and consequent loss
46
of surface area to volume ratio (113), the decreased deformability of RBCs infected with P.
falciparum trophozoites and schizonts is due to increased membrane rigidity, as well as the
parasites themselves causing increased internal viscosity (113, 133). This reduction in
deformability of P. falciparum-infected RBCs is an important contributor to the microvascular
sequestration that characterizes severe falciparum malaria.
In addition to reduced deformability of parasitized RBCs, in severe falciparum malaria
uninfected RBCs also demonstrate increased rigidity. This phenomenon has been
investigated in detail through use of the Laser-assisted Optical Rotational Cell Analyzer
(LORCA) (134-136). With this method whole blood is added to a highly viscous medium and
the RBC suspension is sheared between two concentric rotating cylinders at a constant
temperature of 37oC. The increasing rotation of the outer cylinder leads to an increasing
shear stress that causes the RBCs to elongate and align themselves in the fluid layer. A
laser beam is directed through this fluid layer and forms a diffraction pattern on the screen
behind it. This diffraction pattern undergoes computer analysis to produce an elongation
index, which defines RBC deformability. Importantly, estimates of RBC deformability
obtained using this method reflect the deformability of the entire RBC population, and hence
are primarily accounted for by a reduction in deformability of non-parasitized cells. In studies
using the LORCA, RBC deformability has been shown to be reduced in patients with severe
falciparum malaria, and is a strong predictor of mortality among adults and children (135,
136). In addition, reduced deformability of uninfected RBCs contributes to increased splenic
clearance of these cells, leading to severe malarial anaemia (134).
The mechanisms of increased rigidity of non-parasitized cells in falciparum malaria have not
been fully determined. Reduced deformability has been found to be more marked at lower
shear stresses, such as those found in capillaries, where intracellular viscosity and
membrane deformability are important determinants of RBC deformability (135). Dondorp et
47
al. however reported that manipulation of internal viscosity did not restore deformability,
suggesting that membrane changes are likely responsible for the increased rigidity (137). In
one study Naumann et al. found that binding of a P. falciparum exoantigen to normal RBCs
led to a reduction in deformability (138), although this was not confirmed by a later study
(133). More recently Nuchsongsin et al. showed that hemin caused oxidative damage to the
membrane of RBCs and led to reduced deformability in a dose-dependent manner, with the
reduction in deformability prevented and reverted by the anti-oxidant N-acetylcysteine (139).
Determination of the mechanisms of reduced RBC deformability in severe falciparum malaria
is important as they may provide novel targets for intervention.
4.3.2 Abnormalities of the microcirculation in rhesus macaques infected
with P. knowlesi
Although abnormalities of the microcirculation have not yet been reported among humans
with knowlesi malaria, some insights are provided by early simian studies. In one such
report from 1945, Knisely et al. recorded a film documenting the circulatory changes that
occurred as result of P. knowlesi infection in rhesus macaques (140). In this report, Knisely
et al. described three stages of infection, with Stage I defined as the prepatent period. Stage
II was described as beginning with the appearance of parasites in the blood, and was
characterized by the formation of a thin “adhesive precipitate” that coated parasitized cells
following merozoite invasion. These coated parasitized cells reportedly stuck together to
form clumps, but at this stage did not adhere to non-parasitized cells or endothelium.
Unparasitized cells remained unaffected, and blood flow appeared unimpeded. Parasitized
cells coated with this precipitate were rapidly and selectively phaygocytosed within the liver,
spleen, and bone marrow. Knisely et al. then reported that with increasing parasitemia, a
new precipitate formed around all RBCs, binding parasitized and non-parasitized RBCs
together in “great tough wads and masses” and changing blood to a “thick muck-like sludge”.
48
This new precipitate was referred to as “Stage III sludge”, with Stage III lasting from sludge
formation until death. This sludge reportedly resisted flow through the microvasculature,
leading to “impaction of enormous numbers of small vessels.” Atabrine and quinine were
both reported to disrupt this sludge and restore normal blood flow, allowing treated
macaques to recover from infection.
In a later report, Knisely et al. demonstrated that heparin prevented the coating of
parasitized cells (141). This reportedly prevented the phaygocytosis of parasitized cells,
allowing the development of very high parasite counts (up to 1800 parasites per 1000 RBCs)
in rapidly flowing blood. In this same report, Knisely et al. provided detailed descriptions of
the behavior of rhesus macaques during “Stage III” infection, and contrasted this with the
behavior of a heparin-treated macaque who did not develop “Stage III sludge”. Untreated P.
knowlesi-infected macaques were noted to become increasingly somnolent following the
formation of “sludge”, with coma and death occurring within 12 hours of sludge formation.
Parasitemias reached a maximum of 300 parasites per 1000 RBCs prior to death. In
contrast, the heparin-treated macaque “pumped unsludged blood containing from 500 to
1000 parasites per 1000 RBCs for 12 hours, during which he was strong, vigorous and
active, showing almost no signs of clinical infection”. Death eventually occurred in this
macaque “when the parasites had consumed nearly all the haemoglobin.” Knisely et al.
concluded that this sludge was the primary pathogenic mechanism causing death, with no
animal who developed sludge dying until large numbers of small vessels were impacted, and
no animal surviving this process. Interestingly not all animals developed sludge, and in
these animals severe anaemia was reported to be the primary cause of death.
In another study involving rhesus macaques, Miller et al. investigated the deformability of
RBCs by measuring viscosity and filterability of P. knowlesi-infected red cell suspensions
(142). In this study the viscosity and resistance to flow of infected blood was greater than
49
that of controls, reflecting a decrease in RBC deformability. Resistance to flow increased
with increasing parasite count, and was greater with more mature trophozoites compared to
younger trophozoites. In addition Miller et al. demonstrated the exclusion of P.knowlesi
schizonts from rouleaux, suggesting increased rigidity of these parasitized cells.
Among humans with severe knowlesi malaria reduced deformability of RBCs may contribute
to microvascular obstruction and organ dysfunction, as occurs with severe falciparum
malaria. A prospective evaluation of RBC deformability among patients with severe and
non-severe knowlesi malaria is discussed in Chapter 12.
4.4. The role of platelets in the pathogenesis of severe knowlesi
and falciparum malaria
Studies to date have demonstrated that thrombocytopenia is near-universal in knowlesi
malaria (24, 102), is more profound than in falciparum malaria (24), and is associated with
severe disease (31, 102). However, the role of platelets in the pathogenesis of severe
disease from both species is unclear. In Malawian children with P. falciparum cerebral
malaria, accumulation of platelets was observed in cerebral vessels, with the degree of
accumulation greater than that seen among patients with severe malarial anaemia (143).
Recently, platelet-mediated clumping of infected erythrocytes has been shown to be an
important adhesive phenotype of P. falciparum, with the size of the clumps formed in vitro
sufficient to potentially cause significant disturbance to microvascular flow. Platelet-
mediated clumping has been associated in several studies with severe malaria (125, 144-
146), cerebral malaria (144, 145), prostration (146), and severe malaria anaemia (146).
Given these findings, in one of these studies Wassmer et al. investigated the possibility that
thrombocytopenia may have a protective effect in patients with cerebral malaria (145). In
50
this study plasma from thrombocytopenic Malawian children with cerebral malaria induced
weak platelet clumping, whereas adjustment of the plasma platelet concentration to within a
normal range resulted in massive clumping. Conversely, Mayor et al. found that the
frequency of platelet clumps was inversely proportional to platelet count (146). Further
studies will be required to clarify the relationship between platelet count and frequency of
platelet mediated clumping in falciparum malaria, and to determine the role of platelet
mediated clumping in the pathogenesis of disease.
The mechanisms of platelet-mediated clumping have not been fully determined. Expression
of platelet surface molecule CD36 appears to be required (125), while P-selectin (145) and
gc1qR (147) have also been identified as receptors mediating adherence of infected
erythrocytes to platelets.
A possible protective role of platelets in the early stages of malaria infection has also been
recently identified, with mice and human platelets binding to RBCs parasitized with P.
chabaudi and P. falciparum respectively, and killing the intra-erythrocytic parasites (148).
Inhibition of platelet function by aspirin was found to prevent the lethal effect of human
platelets on P. falciparum, and mortality from P. chabaudi was increased in aspirin-treated or
platelet-deficient mice (148). An anti-parasitic role of platelets has been proposed to
account for the characteristic thrombocytopenia of knowlesi malaria (149), and studies
investigating the binding of platelets to infected and uninfected RBCs in knowlesi malaria are
underway.
51
4.5. Erythrocyte invasion by P. knowlesi merozoites
4.5.1. Duffy antigen and receptor ligands
In early studies P. knowlesi served as an important model for the study of erythrocyte
invasion by Plasmodium merozoites, with electron and video microscopy used to describe in
detail the processes of merozoite attachment, apical reorientation, junction formation, and
internalization of the parasite by membrane invagination (150-152). These steps are
mediated by specific receptor-ligand interactions, and the best described of these is the
interaction between the Duffy antigen (Fy), a RBC surface chemokine receptor, and P. vivax
and P. knowlesi Duffy-binding proteins (PvDBP and PkDBP) (153, 154). Three major Duffy
alleles exist: Fya and Fyb, and Fy-. In the latter, a point mutation ablates Duffy expression on
RBCs, and has been shown to confer resistance to knowlesi and vivax malaria (153-155). In
vivax malaria patients, the Fyb phenotype is associated with increased binding of the P.
vivax Duffy-binding protein (PvDBP) at the erythrocyte surface, compared to Fya, with a
consequent increase in susceptibility to P. vivax malaria (156). In P. knowlesi malaria
PkDBP binds more strongly to Fyb cells (157) however this has not been shown to be
associated with enhanced merozoite invasion, suggesting important differences between P.
vivax and P. knowlesi (158). Further analysis of the Duffy phenotype of individuals infected
with P. knowlesi is required to determine if this factor plays a role in susceptibility to and
severity of disease.
More recently, Semenya et al. identified two P. knowlesi reticulocyte binding-like (RBL)
proteins, normocyte binding proteins PkNBPXa and PkNBPXb, that may play a role in the
attachment of P. knowlesi merozoites to erythrocytes (159). While PkNBXa was shown to
adhere to erythrocytes of rhesus macaques, PkNBPXb bound only human erythrocytes,
suggesting that these ligand may play a role in host-specificity of P. knowlesi.
52
4.5.2. Influence of the age of red blood cells on invasion of erythrocytes by
P. knowlesi merozoites
Among the human malaria species, erythrocyte invasion by merozoites may be dependent
on the age of the erythrocyte. While P. vivax (160) and P. ovale (161) preferentially invade
and replicate within reticulocytes, P. malariae invades only older erythrocytes (162), with this
restriction to a subpopulation of erythrocytes limiting the parasite densities reached. In
contrast, P. falciparum is able to invade erythrocytes of all ages (163), and consequently
very high parasitemia infections may occur. In P. knowlesi infections, although erythrocytes
of all ages are invaded in long-tailed (164) and rhesus macaques (165), Lim et al. recently
demonstrated that in human blood a primate strain of P. knowlesi invaded only young
erythrocytes (166). However, after prolonged culture in human blood this parasite strain was
found to invade erythrocytes of all ages, demonstrating an ability of P. knowlesi to adapt to
proliferation within the human host. Mathematical modeling in this study demonstrated that
this adaption of P. knowlesi to invade erythrocytes of all ages would lead to parasitemias
comparable to those seen in infections with P. falciparum.
4.6. Inflammatory response to P. knowlesi malaria
Only one study to date has described the host inflammatory response to P. knowlesi. In this
study, Cox-Singh et al. described the chemokine and cytokine profiles of 94 patients with P.
knowlesi, and compared these to those of 20 patients with P. vivax and 22 patients with P.
falciparum malaria (166). Among these, 8 (8.5%), 1 (5%) and 5 (23%) patients with
knowlesi, vivax and falciparum malaria respectively had severe disease. Among patients
with knowlesi malaria, neutrophil count, and inflammatory cytokines including TNF, IL-6, and
IL-8 were all strongly associated with parasitemia, and were higher among patients with
53
severe disease compared to those with non-severe disease. This finding is consistent with
studies involving African children and Asian adults with falciparum malaria, which also found
TNF to be associated with increased disease severity (167-169). Among patients with
knowlesi malaria the anti-inflammatory cytokines IL-10 and IL-1ra were also associated with
parasitemia and elevated in severe disease, and were associated with the percentage of
immature trophozoites, which the authors suggested was a possible response to schizont
rupture (149). Markers of macrophage activation, MIP-1β and MCP-1 were also associated
with P. knowlesi parasitemia and were higher in patients with severe disease, however were
lower than among patients with falciparum malaria.
4.7. Knowledge gaps identified from the literature search
Very little is known regarding the pathogenesis of knowlesi malaria, with only one autopsy
report performed (27), one study reporting on the human cytokine response to P. knowlesi
(32), and one small study evaluating cytoadherence of P. knowlesi to endothelial receptors
(105).
Understanding the pathogenesis of human P. knowlesi malaria will require investigation of
the features that underlie severe falciparum malaria, as discussed above. Additional
pathological studies involving electron microscopy will be required to further determine the
extent that P. knowlesi accumulates in microvasculature, and whether endothelial
cytoadherence occurs. Mechanisms by which P. knowlesi accumulates in the
microvasculature will require studies of the rheological features of knowlesi malaria,
including investigation of RBC deformability, and investigation for the presence of increased
aggregation of RBCs as a result of platelet-mediated clumping, rosetting and auto-
agglutination as occur in falciparum malaria (121-125).
54
The role of platelets in the pathogenesis of knowlesi malaria requires further investigation,
including the ability of platelets to bind to infected and uninfected RBCs, and the possibility
that platelets may play an anti-parasitic role in knowlesi malaria.
Finally, additional studies are required to further describe the host inflammatory response to
knowlesi malaria, including the role of endothelial activation, the role of Weibel-Palade Body
contents (particularly angiopoietin-2 or von Willenbrand factor), and the role of platelet
inflammatory mediators in knowlesi malaria.
While Chapter 12 of this thesis describes a study evaluating RBC deformability in patients
with knowlesi and falciparum malaria, further study of the pathogenesis of knowlesi malaria
is beyond the scope of this thesis.
55
Prior to the commencement of this thesis, the bulk of evidence regarding epidemiological
and clinical features of knowlesi malaria in humans had come from Kapit, Sarawak (3, 24).
This included the first seminal report of a large focus of human knowlesi malaria cases,
which included demographic data of 106 patients (3). Of these, 9 (8%) were children <15
years, and the youngest child reported was 10 years. Clinical features of the children were
not described. A subsequent prospective study conducted at the same site, described the
clinical and laboratory features of 107 adults, while eight children with knowlesi malaria were
excluded (24). Data regarding the clinical features of knowlesi malaria in children were
therefore lacking.
At Queen Elizabeth Hospital, an adult tertiary-referral hospital in Sabah, increased
surveillance for P. knowlesi had commenced following the discovery of this parasite as a
cause of human malaria in Malaysian Borneo, and a retrospective study at this hospital
described 56 adult patients with PCR-confirmed P. knowlesi monoinfection during 2007 –
2009 (31). During this time it was noted that a large proportion of patients with knowlesi
malaria had been referred from Kudat District Hospital, and this location was therefore
chosen as a site for further epidemiological and clinical studies.
Kudat District is located on the north-east tip of Borneo and covers an area of 1,287 km2,
with a population of 83,000 people of predominantly Rungus ethnicity. The district has been
substantially deforested, with significant areas of rubber, palm oil and coconut plantations.
Located 220 km north of the Equator, it has a tropical climate with no dry season and the
maximum rainfall generally corresponds to Sabah’s north-east monsoon season from
November to March. Temperatures are fairly constant throughout the year with an average
56
monthly mean of 27°C. Macaques are numerous throughout Kudat District, including Kudat
Town, and are frequently found close to houses.
In 2009 I conducted a retrospective review of the laboratory microscopy records at Kudat
District Hospital from January-November 2009. I recorded basic demographic details for all
patients with a blood film positive for malaria parasites, and reviewed medical records for
children <15 years of age.
Figure 6. Kudat District Hospital
Plasmodium knowlesi can cause severe malaria in adults; however, descriptions of clinical disease in children are lacking. We reviewed case records of children (age <15 years) with a malaria diagnosis at Kudat District Hospital, serving a largely deforested area of Sabah, Malaysia, during January–November 2009. Sixteen children with PCR-con� rmed P. knowlesi monoinfection were compared with 14 children with P. falciparum monoinfection diagnosed by microscopy or PCR. Four children with knowlesi malaria had a hemoglobin level at admission of <10.0 g/dL (lowest minimum level 6.4 g/dL). Minimum level platelet counts were lower in knowlesi than in falciparum malaria (median 76,500/�L vs. 156,000/�L; p = 0.01). Most (81%) children with P. knowlesi malaria received chloroquine and primaquine; median parasite clearance time was 2 days (range 1–5 days). P. knowlesi is the most common cause of childhood malaria in Kudat. Although infection is generally uncomplicated, anemia is common and thrombocytopenia universal. Transmission dynamics in this region require additional investigation.
The simian malaria parasite Plasmodium knowlesi is increasingly recognized as a frequent cause of
potentially fatal human malaria in adults in Malaysian Borneo (1–4). The infection has also been reported in peninsular Malaysia (5) and in other Southeast Asian
countries, including Thailand (6,7), Myanmar (8,9), Vietnam (10), the Philippines (11), Indonesian Borneo (12–14), and Singapore (15,16). Until recently, P. knowlesi had been almost uniformly misdiagnosed by microscopy as P. malariae because of its morphologic similarities, leading to underestimations of prevalence (1,17). Accurate diagnosis therefore requires molecular methods.
The clinical and laboratory features of P. knowlesi infections in adults have been described in Kapit, Sarawak, where 107 (70%) of 152 adults with malaria were infected with P. knowlesi (3). Although P. knowlesi malaria was diagnosed in 8 children, the clinical and laboratory features were not described. All previously reported P.knowlesi infections that caused clinical disease have been in adults (1,2,6,8,11–13,15,18–20). In malaria caused by P.falciparum (21) and P. vivax (22), the 2 species that cause the greatest number of human malaria cases, well-described differences exist between adults and children in terms of the clinical epidemiology, disease spectrum, and laboratory manifestations of disease. We report the demographic, clinical, and laboratory features of P. knowlesi infection in children in Kudat, Sabah, a rural coastal farming area with little remaining primary rainforest, an epidemiologic setting that contrasts with the previously described forested areas of Sarawak.
Methods
Study SettingThe study was conducted at Kudat District Hospital
(KDH) on the northeast tip of Sabah, Malaysia, a coastal rural area which has been largely deforested. KDH services 5 subdistricts (Tigapapan, Dualog, Matunggung, Tambuluran, and the island of Banggi), with a total population of 85,000 persons. The Rungus are the most
Plasmodium knowlesi Malaria in Children
Bridget E. Barber, Timothy William, Mohammad Jikal, Jenarun Jilip, Prabakaran Dhararaj, Jayaram Menon, Tsin W. Yeo, and Nicholas M. Anstey
RESEARCH
814 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 17, No. 5, May 2011
Author af� liations: Menzies School of Health Research, Darwin, Northern Territory, Australia (B.E. Barber, T.W. Yeo, N.M. Anstey); Charles Darwin University, Darwin, Northern Territory (B.E. Barber, T.W. Yeo, N.M. Anstey); Queen Elizabeth Hospital, Kota Kinabalu, Sabah, Malaysia (B.E. Barber, T. William, J. Menon); Sabah Department of Health, Kota Kinabalu (T. William, M. Jikal, J. Jilip, J. Menon); Kudat District Hospital, Kudat, Sabah, Malaysia (P. Dhararaj); and Royal Darwin Hospital, Darwin (T.W. Yeo, N.M. Anstey)
DOI: 10.3201/eid1705.101489
Plasmodium knowlesi Malaria in Children
common ethnic group on the mainland, and minority groups include ethnic Chinese, Bajaus, Dusuns, and Balabaks.
Ministry of Health policy in Sabah requires that all patients with a blood � lm result that indicates malaria be admitted to the hospital and discharged only after smear results are negative for 2 consecutive days. Since January 2009, in response to increasing reports of P. knowlesi infections and the dif� culties of diagnosing this species by microscopy, it became policy at KDH to send slides that had been determined by microscopy to show presence of P. malariae to the Sabah State Reference Laboratory for PCR con� rmation. In addition, KDH sends �15% of all blood � lms positive for other Plasmodium spp. for PCR con� rmation.
Retrospective Case ReviewWe retrospectively searched laboratory microscopy
records for all blood smear results positive for Plasmodium spp. during January 1–November 30, 2009. The patient’s age, sex, ethnicity, and address were recorded for all positive samples. Microscopy results were matched with PCR results.
Medical records were retrieved for all children <15 years of age who had received a diagnosis of malaria on the basis of microscopy results. Demographic, clinical, and laboratory details were extracted by using standardized data forms, which also included disease response to antimalarial treatment. Parasite clearance time was de� ned as the number of days until negative smear. Anemia and severe anemia were de� ned as hemoglobin levels <11 g/dL and <7.1 g/dL (3), respectively.
Laboratory ProceduresBlood � lms were examined by experienced laboratory
microscopists at KDH with the parasite count being classi� ed in most on a scale of 1 to 4 (1 = 4–40 parasites/�L, 2 = 41–400 parasites/�L, 3 = 401–4,000 parasites/�L, 4 = >4,000 parasites/�L), with accurate quantitation per microliter being recorded for most blood � lms that showed P. knowlesi, but for only a limited number that showed P. falciparum. Hemoglobin level and leukocyte and thrombocyte counts were measured on site by using automated systems (Sysmex XT1800 [Sysmex Corp., Mundelein, IL, USA] and CELL-DYN Sapphire [Abbott Diagnostics, Abbott Park, IL, USA]) At the Sabah State Reference Laboratory, parasite DNA was extracted, and nested PCR was performed for P. falciparum, P. vivax, P. malariae, P. ovale, and P. knowlesi by methods described (1,23).
Statistical AnalysisData were analyzed by using Stata statistical software,
version 10 (StataCorp LP, College Station, TX, USA).
Proportions were compared by using the �2 or Fisher exact test. Normally distributed and non–normally distributed variables were compared by using Student t test and Wilcoxon rank-sum test, respectively. For comparison between P. falciparum and P. knowlesi cases, the analysis included children with PCR-con� rmed P. knowlesi infection, and children with P. falciparum infection diagnosed by either microscopy or PCR. Children with mixed Plasmodium infections were excluded from analysis, as were children whose medical records could not be located.
Results
Malaria in All Age GroupsFrom January 1 through November 30, 2009, 220
patients at KDH were given a diagnosis of malaria on the basis of microscopy results (Figure 1). Of these, 196 (89%) had P. malariae monoinfection or mixed infection. PCR was performed on samples from 157 (80%) of these patients and results were positive for P. knowlesi in 137 (87%); 125 (91%) of these were P. knowlesi monoinfections. For the remaining 20 patients who had been given a diagnosis of P.malariae monoinfection or mixed infection by microscopy, PCR found undifferentiated Plasmodium spp., negative results for Plasmodium spp., and 7 cases of P. falciparum infection.
To estimate the � nal numbers of malaria species, we used positive PCR results when possible and microscopy results when PCR had not been performed (replacing P.malariae with P. knowlesi). Microscopy results were also used if PCR result was negative (5 P. knowlesi infections and 1 P. knowlesi/P. falciparum infection) or if PCR result was positive for Plasmodium spp. (7 P. knowlesi infections). Using this method, we found 172 (78%) P. knowlesi monoinfections, 29 (13%) P. falciparum monoinfections, and 19 (9%) mixed infections (Figure 1). A greater proportion of patients with P. knowlesi malaria were male (123/172 [72%]) than were those with P.falciparum malaria (15/29 [52%]; p = 0.03). Median age was higher for those with P. knowlesi infection (median 32 years, interquartile range [IQR] 19–49 years) than for those with P. falciparum infection (median 11 years, IQR 6–30 years).
Malaria in ChildrenOf 220 patients with positive results by microscopy, 41
(19%) were <15 years of age. Microscopy showed that 24 (59%) children had P. malariae monoinfection, 10 (24%) had P. falciparum monoinfection, and 7 (17%) had mixed infections (Figure 2). Samples from 17 children with P.malariae monoinfection underwent PCR; 16 (94%) showed P. knowlesi, and 1 showed mixed P. knowlesi/P. vivax infection. Again, � nal numbers of malaria species were
Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 17, No. 5, May 2011 815
RESEARCH
estimated by using PCR results (if available) or microscopy (if PCR had not been performed). Accordingly, 24 cases (59%) were P. knowlesi, 15 (37%) were P. falciparum, and 2 were mixed P. knowlesi/P. vivax infections. Children thus represented 14% (24/172) of all cases of P. knowlesi monoinfection.
Further analysis was con� ned to those children with PCR-con� rmed P. knowlesi monoinfection and children with P. falciparum monoinfection diagnosed by either microscopy or PCR. Medical records were unavailable for 2 children (1 with P. knowlesi and 1 with P. falciparum infection), leaving 16 children with PCR-con� rmed P.knowlesi infection and 14 with P. falciparum infection for comparison (Table 1).
Demographic Characteristics of Children with P. knowlesi Malaria
Half of the children with PCR-con� rmed P. knowlesi malaria were male (Table 1). The mean age was 8.9 (95% con� dence interval 7.6–10.3) years. The youngest child with P. knowlesi infection was 4 years; all others (15/16; 94%) were 7–14 years of age. In contrast, children with P. falciparum malaria were signi� cantly younger, with a mean age of 5.4 years (95% con� dence interval 3.5–7.3; range 9 months–11 years; p = 0.002).
Clinical Features of P. knowlesiand P. falciparum Malaria
All children had fever or a history of fever. The duration of fever before hospital admission appeared shorter with knowlesi malaria than with falciparum malaria (median 4 days vs. 6.5 days), although this difference was not statistically signi� cant. Although documentation of clinical features was limited, no child was in a state of shock, and no child was reported to be dyspneic, to have bleeding complications, or to have any other clinical feature or laboratory results that indicated severe malaria (24). None of the children who had either knowlesi or falciparum malaria died.
AnemiaIn 11 (69%) children with P. knowlesi malaria, the
parasite density was accurately determined at hospital admission, with a median of 2,240 (IQR 480–7,200) parasites/�L. Only 2 children with P. falciparum malaria had parasite densities accurately determined at admission (200 and 1,600 parasites/�L). Anemia was common with P. knowlesi and P. falciparum infections. At admission, 9 (56%) children with P. knowlesi infection were anemic (hemoglobin level <11.0 g/dL), and 4 (25%) had a hemoglobin level <10 g/dL. All children with P. knowlesi infection had a hemoglobin minimum level <11.0 g/dL, and 10 (63%) had a hemoglobin minimum level <10 g/dL. The lowest hemoglobin level was found in an 8-year-old boy: 8.3 g/dL at admission, 6.4 g/dL on day 3, and 7.2 g/dL on day 6, the day before discharge. His parasite count at
816 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 17, No. 5, May 2011
Figure 1. Distribution of malaria cases diagnosed by microscopy and PCR among all age groups, Kudat, Malaysia, January 1–November 30, 2009. P.m., Plasmodium malariae; P.f., P. falciparum; P.v., P. vivax; P.k., P. knowlesi; neg, negative.
Figure 2. Distribution of malaria cases diagnosed by microscopy and PCR among children <15 years of age, Kudat, Malaysia, January 1–November 30, 2009. P.m., Plasmodium malariae; P.f., P. falciparum; P.v., P. vivax; P.k., P. knowlesi.
Plasmodium knowlesi Malaria in Children
admission (14,400 parasites/�L) was the highest recorded in this study.
In addition, severe anemia (hemoglobin 6.0 g/dL at admission; minimum level 4.9 g/dL on day 1 requiring transfusion) occurred in a second child with positive results for P. knowlesi by microscopy (this child was not included in the � nal analysis because of lack of PCR con� rmation). Among children with P. knowlesi infection, a lower hemoglobin minimum level was associated with higher parasite density at admission (p = 0.001). All children with P. falciparum infection had hemoglobin levels at admission <11.0 g/dL; 10 (71%) had hemoglobin levels <10 g/dL, and 5 (36%) had severe anemia (hemoglobin <7.1 g/dL). Severe anemia developed in a sixth child after admission. Four children received transfusions, with hemoglobin levels of 4.8–6.1 g/dL at admission. Children with P. knowlesi malaria took longer to reach their hemoglobin minimum level than did children with P. falciparum malaria (2.6 vs 1.5 days; p = 0.02).
ThrombocytopeniaAll children with P. knowlesi malaria had
thrombocytopenia. Fifteen children (94%) had a platelet count <150,000/�L at admission, and the remaining child exhibited thrombocytopenia within 1 day. The lowest platelet count recorded was 28,000/�L. Platelet count was not correlated with hemoglobin level. Thrombocytopenia was less common in children with P. falciparum; 7 (50%) had platelet counts <150,000/�L at admission or during hospitalization and the lowest platelet count was 47,000/�L.
Children with P. knowlesi malaria had a lower lymphocyte count minimum than those with P. falciparum malaria (1.6 vs. 2.4 � 103/�L; p = 0.01).
Response to TreatmentMost (81%) children with P. knowlesi malaria were
given oral chloroquine and primaquine for 3 days, and these children had a median parasite clearance time of 2 days (Table 2). The longest parasite clearance time with chloroquine and primaquine was 5 days in the aforementioned 8-year-old boy with the highest admission parasitemia level (14,400/�L) and the lowest minimum hemoglobin level (6.4 g/dL). Three children with P.knowlesi infection were given oral quinine, and parasites cleared within 2 days. Among the 11 children with P.knowlesi and parasite densities accurately assessed at admission, the correlation between parasite density and parasite clearance time was signi� cant (p = 0.002).
Children with P. falciparum malaria were given a variety of treatment regimens. Most (71%) received a regimen that contained quinine, 7 (50%) children received sulfadoxine/pyrimethamine, 6 (43%) received primaquine, and 5 (36%) were given chloroquine. Only 5 (36%) received artemisinin-based combination therapy. Parasite clearance times were signi� cantly slower among children with P. falciparum malaria, with a median of 5 days until the � rst negative smear; in 4 children (29%), it took >10 days for parasites be cleared. Children with P. knowlesi malaria had signi� cantly shorter hospital admissions (median 4 days, IQR 4–5 days) than did
Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 17, No. 5, May 2011 817
Table 1. Demographic data, clinical features of infection, and laboratory values in 14 cases of Plasmodium falciparum malaria and 16 cases of PCR-confirmed P. knowlesi malaria in children, Kudat, Malaysia, January 1–November 30, 2009* Characteristic P. knowlesi, n =16 P. falciparum, n =14 p valueAge, y Mean (SD) 8.9 (2.5) 5.4 (3.2) 0.002 Range 4–14 0.75–11 Male sex, no. (%) 8 (50) 7 (50) 1.00Living in Banggi, no. (%) 0 (0) 10 (71.4) <0.001Duration of fever, d, median (IQR) 4 (3–6) 6.5 (4–7) 0.12 Examination findings at admission Temperature, oC, median (IQR) 37.9 (37.0–39.1) 37.0 (37.0–37.4) 0.07 Heart rate, beats/min, mean (SD) 110 (24) 119 (15) 0.24 Respiratory rate, beats/min, mean (SD) 27.8 (3.9) 32 (3.8) 0.01 Systolic blood pressure, mm Hg, mean (SD) 105 (8.3) 98 (11.2) 0.06Laboratory results Hemoglobin d 0, g/dL, median (IQR) 10.7 (10.0–11.4) 8.5 (6.1–10.5) 0.001 Hemoglobin minimum level, g/dL, median (IQR) 9.7 (8.4–10.2) 7.15 (6.1–9.3) 0.035 Day of hemoglobin minimum level, mean (SD) 2.6 (1.03) 1.5 (1.3) 0.022 Lymphocyte count d 0, 103/�L, median (IQR) 2.0 (1.7–2.3) 2.5 (2.0–4.6) 0.09 Lymphocyte minimum level, 103/�L, median (IQR) 1.6 (1.3–2.2) 2.4 (1.9–3.9) 0.01 Day of lymphocyte minimum level, median (IQR) 1 (0–1) 1.5 (0–3) 0.19 Thrombocyte count d 0, 103/�L, median (IQR) 89.5 (72.0–118.5) 171.5 (98–271) 0.038 Thrombocyte count minimum level, 103/�L, median (IQR) 76.5 (68.5–110.0) 156 (98–227) 0.014 Day of thrombocyte count minimum level, median (IQR) 1 (1–1) 0.5 (0–2) 0.67*IQR, interquartile range. Boldface indicates statistical significance.
RESEARCH
children with P. falciparum malaria (median 7 days, IQR 5–10 days).
DiscussionP. knowlesi was the most common cause of malaria in
adults and children in the Kudat region in Sabah, Malaysian Borneo. Although those with P. knowlesi infections had an older age distribution than did those with P. falciparum infections, the species still caused 63% of all malaria cases among children <15 years. Although nearly all previous reported cases of knowlesi malaria have been in adults, 14% of all knowlesi cases in Kudat occurred in children. In children, P. knowlesi most often caused uncomplicated malaria, which responded well to chloroquine. Nevertheless, knowlesi malaria was associated with substantial illness in children, with all PCR-con� rmed P. knowlesi–infected children being anemic at admission or during hospital stay.
In the only other report of knowlesi malaria in children, a prospective study of adult knowlesi malaria in Sarawak reported the exclusion of 8 children <15 years with P.knowlesi infection, comprising 7% of all knowlesi cases (3). The clinical and laboratory features of P. knowlesi malaria in children have not been described. Consistent with the reported features of P. knowlesi malaria in adults (3), the disease in most children was uncomplicated. In adults with knowlesi malaria, increasing age is a risk factor for severe disease. Although the numbers are small, none of the children had severe manifestations of knowlesi malaria that have been reported in adults (3), such as acute lung injury or acute renal failure. In falciparum malaria, these
conditions are also largely con� ned to adults (21); anemia, coma, and acidosis-related respiratory distress are the major manifestations of severe falciparum malaria in children. No child with knowlesi malaria exhibited coma or respiratory distress; however, anemia developed in all children with knowlesi malaria, with the hemoglobin concentration in 1 patient (6%) falling to <7.0 g/dL. Anemia was more common in children than has been previously described in knowlesi malaria in adults (3).
Thrombocytopenia was found at admission in nearly all (94%) children with P. knowlesi malaria, in contrast to only half of the children with P. falciparum malaria. Although the cause is unclear, thrombocytopenia is also nearly universal in infected adults (3), which makes it a characteristic feature of P. knowlesi infection across all age groups. The role of thrombocytopenia and platelet activation in the pathogenesis of knowlesi malaria requires further investigation.
Most children with P. knowlesi malaria had an adequate response to a 3-day regimen of treatment with chloroquine and primaquine, although the mean parasite clearance time of 2 days was longer than the 90% parasite clearance time of 10.3 hours that was recently reported in adults (25). One child who received chloroquine and primaquine, and had a high parasite count at admission, required 5 days to clear parasites. Standard Ministry of Health pediatric dosing regimens of chloroquine are used in Kudat; however, posttreatment vomiting or inadequate blood concentrations could not be excluded. Prospective studies that evaluate the response to chloroquine and artemisinin-based combination
818 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 17, No. 5, May 2011
Table 2. Malaria treatment according to Plasmodium species and response to treatment in children, Kudat, Malaysia, January 1–November 30, 2009* Species and treatment regimen No. (%) patients Days until negative smear† P. knowlesi, n = 16 3 days CQ/PQ 13 (81) 2 (2–3, 1–5) Oral Q 2 (13) 2, 2 IV Q 1 (6) 2 All 2 (2–3, 1–5) P. falciparum, n = 14 3 days CQ/PQ oral Q 2 (14) 5, 5 Oral Q AL 1 (7) 3 SP/PQ AM 1 (7) 8 SP/PQ 2 (14) 1, 2 SP/PQ CQ/Q AL 1 (7) 3 SP/PQ CQ/IV Q AL 1 (7) 5 SP/PQ CQ/IV Q oral Q 1 (7) 3 SP/PQ oral Q IV Q oral Q 1 (7) 16 IV Q oral Q 2 (14) 1, 10 IV Q oral Q/PQ 1 (7) 11 Oral Q/PQ AM 1 (7) 10 All 5 (3–10, 1–16)‡ *IQR, interquartile range; CQ, chloroquine; PQ, primaquine; Q, quinine; IV, intravenous; AL, artemether/lumefantrine; SP, sulfamethoxazole/pyrimethamine; AM, artesunate/mefloquine. †Values are mean (IQR, range) by category or days by patient. ‡p value = 0.009 for days until negative smear for P. falciparum vs. P. knowlesi.
Plasmodium knowlesi Malaria in Children
therapy in pediatric knowlesi malaria are required. Children with P. falciparum malaria received many different treatment regimens and took signi� cantly longer to clear their parasites. Less than half received the recommended artemisinin-based combination therapy, and only after alternative treatments failed. Children with P. falciparum malaria had signi� cantly longer hospital stays, likely related at least in part to suboptimal treatment regimens. This � nding highlights the importance of increasing the usage of artemisinin-based combination therapy for falciparum malaria in district hospitals in Sabah and elsewhere.
Our study had several limitations. First, PCR was only performed for 73% of cases across all age groups, and this limited our ability to accurately determine the true proportion of disease caused by P. knowlesi. Furthermore, although PCR was performed on samples from most children with P.knowlesi malaria, only 1 child with P. falciparum malaria had PCR performed. Our analysis, therefore, compared children with PCR-con� rmed knowlesi malaria to children with falciparum malaria diagnosed by either microscopy or PCR. Some of those children with a diagnosis of falciparum malaria may have actually had knowlesi malaria, and the differences found between these 2 species may therefore be minimum estimates. The retrospective design of our study also limited our ability to collect standardized clinical information.
This study demonstrates that P. knowlesi has become the predominant malaria species in the Kudat region, estimated by results of microscopy, PCR, or both, as contributing to 87% of all malaria cases, a higher proportion than that reported elsewhere in Malaysian Borneo. The emerging dominance of P. knowlesi in Malaysian Borneo has been hypothesized to result from the following factors: changing patterns of human exposure to monkeys and vectors (26) because of deforestation, and potentially reduced competition or cross-species protection from P. vivax and P. falciparum as a result of a 40-fold reduction in the prevalence of these species in Sabah and Sarawak during 1960–2006 following intensive malaria control efforts (27). The paucity of P. vivax in Kudat was particularly notable.
Previous reports of adult disease have been from communities adjacent to rainforests (3). Kudat is a coastal rural farming area with varied land use and vegetation patterns and with minimal remaining regrowth forest. Although our retrospective study did not gather detailed travel or exposure histories, it is likely that most pediatric infections were locally acquired and that infections with P. knowlesi did not occur solely in those spending time in forested areas. Macaque monkeys, the natural hosts of P. knowlesi, are widely distributed in different habitats throughout the Kudat area and are frequently domesticated. The major vectors of P. knowlesi in forested areas of
Sarawak, Anopheles latens mosquitoes, disappear with deforestation, but vectors capable of transmitting P.knowlesi, An. balabacencis mosquitoes (26), do persist at lower densities in largely deforested areas of Sabah (28,29). In notable contrast to P. falciparum malaria, pediatric knowlesi malaria was restricted to children of school age. Further studies will be required to characterize the transmission patterns, vectors, and risk factors for P.knowlesi in deforested areas of Malaysia.
ConclusionsP. knowlesi is the most common cause of malaria
in adults and children in the Kudat region of Sabah, a rural coastal deforested region. Consistent with previous studies in adults (3), we found that P. knowlesi in children most often caused uncomplicated malaria that responded adequately to chloroquine and primaquine. Anemia was common in children and knowlesi infection was associated with moderately severe anemia. Thrombocytopenia was universal and is characteristic of knowlesi malaria across all age groups. Larger prospective clinical studies are needed to describe more fully the epidemiology of P.knowlesi malaria in children, the full spectrum of clinical disease and the transmission patterns in nonforested areas such as Kudat.
AcknowledgmentsWe thank Fread Anderios, Ella Curry, and the laboratory
staff at Kudat District Hospital for their help with this study, and Jeffrey Hii for helpful discussions. We also thank the Director General of Health (Malaysia) for permission to publish this article.
Research supported by the National Health and Medical Research Council of Australia (fellowships to T.W.Y. and N.M.A.; scholarship to B.E.B.; and Program Grant 496600).
Dr Barber is an infectious diseases physician and PhD scholar at Menzies School of Health Research, Darwin, and Clinical Research Centre, Kota Kinabalu, Malaysia. Her primary research interest is P. knowlesi malaria.
References
1. Cox-Singh J, Davis TM, Lee KS, Shamsul SS, Matusop A, Ratnam S, et al. Plasmodium knowlesi malaria in humans is widely distribut-ed and potentially life threatening. Clin Infect Dis. 2008;46:165–71. doi:10.1086/524888
2. Cox-Singh J, Hiu J, Lucas S, Divis P, Zulkarnaen M, Chandran P, et al. Severe malaria-a case of fatal Plasmodium knowlesi infection with post-mortem � ndings. Malar J. 2010;9:10. doi:10.1186/1475-2875-9-10
3. Daneshvar C, Davis TM, Cox-Singh J, Rafa’ee M, Zakaria S, Divis P, et al. Clinical and laboratory features of human Plasmodium knowle-si infection. Clin Infect Dis. 2009;49:852–60. doi:10.1086/605439
4. William T, Menon J, Rajahram G, Chan L, Ma G, Donaldson S, et al. Severe Plasmodium knowlesi malaria in a tertiary hospital, Sabah, Malaysia. Emerg Infect Dis. In press.
Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 17, No. 5, May 2011 819
RESEARCH
5. Vythilingam I, Noorazian YM, Huat TC, Jiram AI, Yusri YM, Aza-hari AH, et al. Plasmodium knowlesi in humans, macaques and mosquitoes in peninsular Malaysia. Parasit Vectors. 2008;1:26. doi:10.1186/1756-3305-1-26
6. Jongwutiwes S, Putaporntip C, Iwasaki T, Sata T, Kanbara H. Natu-rally acquired Plasmodium knowlesi malaria in human, Thailand. Emerg Infect Dis. 2004;10:2211–3.
7. Putaporntip C, Hongsrimuang T, Seethamchai S, Kobasa T, Limkit-tikul K, Cui L, et al. Differential prevalence of Plasmodium infec-tions and cryptic Plasmodium knowlesi malaria in humans in Thai-land. J Infect Dis. 2009;199:1143–50. doi:10.1086/597414
8. Zhu HM, Li J, Zheng H. Human natural infection of Plasmodiumknowlesi [in Chinese]. Zhongguo Ji Sheng Chong Xue Yu Ji Sheng Chong Bing Za Zhi. 2006;24:70–1.
9. Jiang N, Chang Q, Sun X, Lu H, Yin J, Zhang Z, et al. Co-infections with Plasmodium knowlesi and other malaria parasites, Myanmar. Emerg Infect Dis. 2010;16:1476. doi:10.3201/eid1609.100339
10. Van den Eede P, Van HN, Van Overmeir C, Vythilingam I, Duc TN, Hung LX, et al. Human Plasmodium knowlesi infections in young children in central Vietnam. Malar J. 2009;8:249. doi:10.1186/1475-2875-8-249
11. Luchavez J, Espino F, Curameng P, Espina R, Bell D, Chiodini P, et al. Human Infections with Plasmodium knowlesi, the Philippines. Emerg Infect Dis. 2008;14:811–3. doi:10.3201/eid1405.071407
12. Figtree M, Lee R, Bain L, Kennedy T, Mackertich S, Urban M, et al. Plasmodium knowlesi in human, Indonesian Borneo. Emerg Infect Dis. 2010;16:672.
13. Bronner U, Divis P, Färnert A, Singh B. Swedish traveller with Plas-modium knowlesi malaria after visiting Malaysian Borneo. Malar J. 2009;8:15. doi:10.1186/1475-2875-8-15
14. Sulistyaningsih E, Fitri L, Löscher T, Berens-Riha N. Diagnostic dif� culties with Plasmodium knowlesi infection in humans. Emerg Infect Dis. 2010;16:1033.
15. Ng OT, Ooi E, Lee C, Lee P, Ng L, Pei S, et al. Naturally acquired human Plasmodium knowlesi infection, Singapore. Emerg Infect Dis. 2008;14:814–6. doi:10.3201/eid1405.070863
16. Lee YC, Tang CS, Ang LW, Han HK, James L, Goh KT. Epidemio-logical characteristics of imported and locally-acquired malaria in Singapore. Ann Acad Med Singapore. 2009;38:840–9.
17. Lee K-S, Cox-Singh J, Brooke G, Matusop A, Singh B. Plasmo-dium knowlesi from archival blood � lms: further evidence that hu-man infections are widely distributed and not newly emergent in Malaysian Borneo. Int J Parasitol. 2009;39:1125–8. doi:10.1016/j.ijpara.2009.03.003
18. Galinski MR, Barnwell JW. Monkey malaria kills four humans. Trends Parasitol. 2009;25:200–4. doi:10.1016/j.pt.2009.02.002
19. Kantele A, Marti H, Felger I, Muller D, Jokiranta T. Monkey malaria in a European traveler returning from Malaysia. Emerg Infect Dis. 2008;14:1434–6. doi:10.3201/eid1409.080170
20. Thuy-Huong T, Ana S, Marwa A, María M, Marta L, Eduardo A, et al. First case of detection of Plasmodium knowlesi in Spain by real time PCR in a traveller from Southeast Asia. Malar J. 2010;9:219. doi:10.1186/1475-2875-9-219
21. Dondorp AM, Lee SJ, Faiz MA, Mishra S, Price R, Tjitra E, et al. The relationship between age and the manifestations of and mortal-ity associated with severe malaria. Clin Infect Dis. 2008;47:151–7. doi:10.1086/589287
22. Tjitra E, Anstey N, Sugiarto P, Warikar N, Kenangalem E, Kary-ana M, et al. Multidrug-resistant Plasmodium vivax associated with severe and fatal malaria: a prospective study in Papua, Indonesia. PLoS Med. 2008;5:e128. doi:10.1371/journal.pmed.0050128
23. Singh B, Sung L, Matusop A, Radhakrishnan A, Shamsul S, Cox-Singh J, et al. A large focus of naturally acquired Plasmodiumknowlesi infections in human beings. Lancet. 2004;363:1017–24. doi:10.1016/S0140-6736(04)15836-4
24. World Health Organization. Severe falciparum malaria. World Health Organization, Communicable Diseases Cluster. Trans R Soc Trop Med Hyg. 2000;94:S1–90.
25. Daneshvar C, Davis T, Cox-Singh J, Rafa’ee M, Zakaria S, Divis P, et al. Clinical and parasitological response to oral chloroquine and primaquine in uncomplicated human Plasmodium knowlesi infec-tions. Malar J. 2010;9:238. doi:10.1186/1475-2875-9-238
26. Indra V. Plasmodium knowlesi in humans: a review on the role of its vectors in Malaysia. Trop Biomed. 2010;27:1.
27. Cox-Singh J, Singh B. Knowlesi malaria: newly emergent and of public health importance? Trends Parasitol. 2008;24:406–10. doi:10.1016/j.pt.2008.06.001
28. Hii J, Macdonald M, Sang V. The use of CDC light traps for ma-lariometric entomology surveys in Sabah, Malaysia. Trop Biomed. 1986;3:39–48.
29. Vythilingam I, Tan C, Asmad M, Chan S, Lee K, Singh B. Natu-ral transmission of Plasmodium knowlesi to humans by Anoph-eles latens in Sarawak, Malaysia. Trans R Soc Trop Med Hyg. 2006;100:1087–8. doi:10.1016/j.trstmh.2006.02.006
Address for correspondence: Bridget E. Barber, Global Health Division, Menzies School of Health Research, PO Box 41096, Casuarina, Darwin, Northern Territory, Australia; email: [email protected]
820 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 17, No. 5, May 2011
http://www.facebook.com/
57
Implications of and questions arising from this paper
In this study we found that P. knowlesi was the most common cause of human malaria at
Kudat District Hospital. However, our study only included data from 2009, and so we had no
information on whether this was a trend that was changing over time. Furthermore, we had
no information on whether this predominance of P. knowlesi infections was confined to
Kudat, or existed in other districts of Sabah. These unanswered questions led us to perform
a state-wide review of all available P. knowlesi notification data, presented in Chapter 6.
In this paper we also report that in Kudat in 2009 P. knowlesi was the most common cause
of malaria in children, and that 24/172 (14%) cases of P. knowlesi occurred in children. This
contrasts with the two studies of knowlesi malaria from Kapit, Sarawak, that reported that
8/121 (6.6%) and 9/106 (8.5%) cases of knowlesi malaria occurred in children (3, 24). Chi-
squared analysis demonstrates that when the results of the two Kapit studies are pooled, the
proportion of children among knowlesi malaria patients in Kapit (17/127; 7.5%) is smaller
than that seen among those in the Kudat study (p=0.035). This suggested that the
epidemiology of knowlesi malaria in Kudat may differ from that described in Kapit, and led us
to perform the subsequent and more detailed review of the epidemiology of knowlesi malaria
in Kudat presented in the next chapter.
Finally, in this paper we provide some information on the clinical features of knowlesi malaria
in children. We found that most infections in children appeared uncomplicated, although all
children were thrombocytopenic and anaemia was common. Other features of severe
disease described in adults, including jaundice, shock, respiratory distress, and renal failure,
were not documented to occur in this study. Our clinical descriptions however were limited
in this retrospective study by the completeness of patients’ medical records. In particular,
we were unable to exclude the possibility that there may have been cases of unrecognized
58
severe malaria. In addition, while we report that anaemia was common among children with
knowlesi malaria, we were unable to exclude other causes of anaemia such as helminth
infections. Finally, while we report the occurrence of a delayed parasite clearance in one
patient with knowlesi malaria treated with chloroquine, this retrospective study may have
failed to detect other factors known to influence parasite clearance, such as a history of
splenectomy or concomitant illness.
Subsequent to the publication of this study, additional clinical features of knowlesi malaria in
8 children admitted to Kapit District Hospital were described by Daneshvar et al (170).
These children were aged 9 – 12 years (median 11 years), had a median parasite count of
940/μL (range 440 – 26,270/μL), and were all thrombocytopenic. Two were mildly anaemic,
two had elevated liver transaminases, and one had jaundice (bilirubin 38 μmol/L). One
patient with a parasite count of 27,270/μL had a petechial rash over the shins and retinal
haemorrhages on examination. All children responded to treatment.
Larger prospective studies will be required to describe in more detail the clinical spectrum of
knowlesi malaria in children, including the risk of severe disease.
59
In our first study in Kudat, Sabah, we estimated that 87% of all patients admitted with
malaria to Kudat District Hospital in 2009 were infected with P. knowlesi, and that 14% of
patients with knowlesi malaria were aged <15 years old (171). In order to further investigate
the epidemiology of knowlesi malaria in Kudat, we extended our original study to include a
retrospective review of the laboratory microscopy records for an additional two years, with
demographic data collected for each patient with a blood film positive for malaria parasites.
In addition, we reviewed the medical records of all patients suspected of representing family
clusters, and, where possible, contacted these families for additional information.
RESEARCH Open Access
Epidemiology of Plasmodium knowlesi malaria innorth-east Sabah, Malaysia: family clusters andwide age distributionBridget E Barber1,2, Timothy William2,3, Prabakaran Dhararaj4, Fread Anderios5, Matthew J Grigg1,2,Tsin W Yeo1,6 and Nicholas M Anstey1,6*
Abstract
Background: The simian parasite Plasmodium knowlesi is a common cause of human malaria in Malaysian Borneo,with a particularly high incidence in Kudat, Sabah. Little is known however about the epidemiology in thissubstantially deforested region.
Methods: Malaria microscopy records at Kudat District Hospital were retrospectively reviewed from January2009-November 2011. Demographics, and PCR results if available, were recorded for each positive result. Medicalrecords were reviewed for patients suspected of representing family clusters, and families contacted for furtherinformation. Rainfall data were obtained from the Malaysian Meteorological Department.
Results: “Plasmodium malariae” mixed or mono-infection was diagnosed by microscopy in 517/653 (79%) patients.Of these, PCR was performed in 445 (86%) and was positive for P. knowlesi mono-infection in 339 (76%). Patientswith knowlesi malaria demonstrated a wide age distribution (median 33, IQR 20–50, range 0.7-89 years) with P.knowlesi predominating in all age groups except those <5 years old, where numbers approximated those ofPlasmodium falciparum and Plasmodium vivax. Two contemporaneous family clusters were identified: a father withtwo children (aged 10–11 years); and three brothers (aged one-11 years), all with PCR-confirmed knowlesi malaria.Cases of P. knowlesi demonstrated significant seasonal variation, and correlated with rainfall in the preceding threeto five months.
Conclusions: Plasmodium knowlesi is the most common cause of malaria admissions to Kudat District Hospital. Thewide age distribution and presence of family clusters suggest that transmission may be occurring close to or insidepeople’s homes, in contrast to previous reports from densely forested areas of Sarawak. These findings havesignificant implications for malaria control. Prospective studies of risk factors, vectors and transmission dynamics ofP. knowlesi in Sabah, including potential for human-to-human transmission, are needed.
Keywords: Plasmodium knowlesi, Malaria, Epidemiology
BackgroundIn recent decades Malaysia has achieved a dramatic re-duction in malaria prevalence as a result of intensivecontrol efforts, particularly indoor residual spraying anddistribution of insecticide-treated nets, with cases fallingfrom 60,000 in 1995 to 6,650 in 2010 [1]. Consequently,
Malaysia is in the pre-elimination phase of malaria con-trol and aims to be malaria-free by 2020 [2]. Recentlyhowever, the zoonotic infection Plasmodium knowlesihas emerged as a common and potentially fatal cause ofhuman malaria in Malaysian Borneo [3-7], and presentsan increasing threat to malaria control. Predominantly aparasite of the long-tailed and pig-tailed macaques andtransmitted by the forest-dwelling Anopheles leuco-sphyrus group of mosquitoes, P. knowlesi is now themost common cause of human malaria in several dis-tricts throughout Sabah and Sarawak [3,4,8,9]. The
* Correspondence: [email protected] Health Division, Menzies School of Health Research, PO Box 41096,Casuarina 0810,, Northern Territory, Australia6Royal Darwin Hospital, Darwin, NT, AustraliaFull list of author information is available at the end of the article
© 2012 Barber et al.; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the CreativeCommons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, andreproduction in any medium, provided the original work is properly cited.
Barber et al. Malaria Journal 2012, 11:401http://www.malariajournal.com/content/11/1/401
highest proportion has been reported at Kudat DistrictHospital (KDH), on the north-east tip of Sabah, where87% of patients admitted with malaria in 2009 wereinfected with P. knowlesi [8].Little is known however about the epidemiology of P.
knowlesi in this region. Recent studies conducted in thedensely forested Kapit District in Sarawak found that P.knowlesi primarily infected adults with a recent historyof forest exposure, and that clustering of cases did notoccur [4,10]. These findings suggest that in this regiontransmission occurs in forested areas away from people’shomes, and that human-to-human transmission, al-though demonstrated experimentally [11], does not ap-pear to be occurring. In contrast to Kapit however,Kudat District has undergone significant deforestationand hence represents a different environmental setting.In addition, mosquito vectors differ between the tworegions. In Kapit, Anopheles latens has been identified asthe P. knowlesi vector [12], while in Sabah, although theP. knowlesi vector is not yet known, recent studies sug-gest that Anopheles balabacensis and Anopheles donaldiare the primary vectors of human malaria [13,14].Transmission dynamics in Kudat may therefore differfrom those previously described in Sarawak. The epi-demiology of malaria in Kudat from 2009–2011 wastherefore investigated.
MethodsStudy settingKudat District (Figure 1) is located on the north-east tipof Borneo and covers an area of 1,287 km2, with a
population of 83,000 people of predominantly Rungusethnicity. The district has been substantially deforested,with significant areas of rubber, palm oil and coconutplantations. Located 220 km north of the Equator, it hasa tropical climate with no dry season and the maximumrainfall generally corresponds to Sabah’s north-east mon-soon season from November to March. Temperaturesare fairly constant throughout the year with an averagemonthly mean of 27°C. Kudat District incorporates thedensely forested Pulau Banggi group of islands, the lar-gest of which has an area of 440 km2, a population ofapproximately 20,000 and a main town located 24 km,or one hour by boat, from Kudat. Macaques are numer-ous throughout Kudat District, including Kudat Town,and are frequently found close to houses.KDH is the only hospital in the district, and serves as
the referral base for five subsector government clinics.All clinics have microscopy facilities and are obliged torefer patients with slides positive for malaria parasites toKDH for admission.
Laboratory proceduresIn this retrospective study blood slides were taken fromall febrile patients seen at health clinics, or KDH emer-gency or outpatient departments, and examined byexperienced laboratory microscopists at the clinics orKDH respectively. For clinic patients referred to KDH,slides were repeated and malaria treatment commencedon arrival at KDH. In Sabah, blood slides with para-sites resembling Plasmodium malariae/P. knowlesi arereported most commonly as “P. malariae”, but sometimes
Figure 1 Map of Sarawak and Sabah.
Barber et al. Malaria Journal 2012, 11:401 Page 2 of 8http://www.malariajournal.com/content/11/1/401
as “P. malariae/P. knowlesi” or “P. malariae (?P. know-lesi)”. For simplicity these microscopy reports are all re-ferred to in this paper as “P. malariae”. During the studyperiod policy at KDH was to send all these slides to SabahState Reference Laboratory for PCR confirmation. How-ever, PCR was not routinely performed on the majority ofpatients diagnosed with other malaria species. At theSabah State Reference Laboratory Plasmodium knowlesiwas identified by nested PCR as described elsewhere [10]until June 2011 when a real-time PCR assay was insti-tuted [15]. Plasmodium falciparum, P. vivax, P. malariaeand Plasmodium ovale were identified using a species-specific real-time PCR assay [16]. Measures taken tominimize contamination in the laboratory included theuse of separate workstations for DNA extraction, master-mix preparation, and addition of DNA template; and theuse of filter micropipette tips. One negative and one posi-tive control (P. vivax or P. falciparum) were used duringDNA extraction, and a reagent control was used duringpreparation of the PCR mastermix. Plasmodium falcip-arum, P. vivax and P. malariae positive controls weresourced from CDC Atlanta (Stephanie Johnston), whilethe P. knowlesi positive control was sourced from the In-stitute of Medical Research, Kuala Lumpur.
Retrospective reviewLaboratory microscopy records were reviewed for totalnumber of blood slides taken and all blood slides posi-tive for malaria parasites from January 2009 to Novem-ber 2011 (with results from 2009 briefly reported in part[8]). Date, age, and sex were recorded for all positiveresults. In addition, PCR results from Sabah State Refer-ence Laboratory were recorded if available. Medicalrecords were reviewed for those patients with P. know-lesi suspected of representing family clusters, based onpatients with the same family name presenting within afour-week period, and these families were contacted forfurther information.Rainfall was recorded using a rain gauge at Kudat
Meteorological Station, located one km from the hospital.Monthly rainfall recorded during the study period wasobtained from the Malaysian Meteorological Department.
Statistical analysisData were analysed using Stata statistical software, ver-sion 10.0 (StataCorp LP, College Station, TX, USA). Allanalyses were performed on patients with PCR-confirmed P. knowlesi mono-infection, and microscopy-diagnosed P. falciparum or P. vivax mono-infection, dueto the small number of PCR assays performed on theselatter species. Five patients with microscopy-diagnosedP. falciparum but found to have P. knowlesi by PCRwere excluded from analyses of P. falciparum cases. Me-dian ages of patients infected with each of the malaria
species were compared using Wilcoxon rank-sum test,and proportions were assessed using the Chi-square test.The Chi-square test was also used to determine inter-annual variation in malaria cases (as a proportion of thetotal number of slides taken), while seasonality of P.knowlesi cases was assessed using Edward’s test. Spear-man’s correlation was used to assess the association be-tween rainfall and monthly P. knowlesi cases, with cross-correlations analysed to determine the time lag (up tosix months) at which the strongest association occurred.
Ethics statementThe study was approved by the Medical Research Sub-Committee of the Malaysian Ministry of Health and theHealth Research Ethics Committee of Menzies School ofHealth Research, Australia. Ethical approval to contactpatients by phone was obtained.
ResultsMalaria species distributionFrom January 2009 to November 2011, 653 (3.4%)patients had slides positive for malaria parasites (Table 1),out of a total of 18,993 slides. “Plasmodium malariae”mixed or mono-infection was reported in 517 patients(79%). Of these, PCR was performed in 445 (86%), andwas positive for P. knowlesi mono-infection in 339(76%). The remaining 24% included 40 (9%) P. knowlesimixed infections, 23 (5%) P. vivax mono-infections, 13(3%) P. falciparum mono-infections, 20 (4%) positiveonly for Plasmodium genus, nine (2%) negative results,and one mixed P. falciparum/P. malariae. Of the total445 patients diagnosed by microscopy with P. malariaemixed or mono-infection and who had PCR performed,only four (1%) were P. malariae by PCR (one P. falcip-arum/P.malariae, one P. malariae/P. knowlesi and twoP. vivax/P. malariae/P. knowlesi). Only 35 patients diag-nosed by microscopy with P. falciparum or P. vivaxmono-infection had PCR performed. Of these, P. know-lesi mono-infection was identified in 5/16 (31%) patientswith microscopy-diagnosed P. falciparum mono-infec-tion, and mixed P. knowlesi/P. vivax infection was iden-tified in 6/19 (32%) patients with microscopy-diagnosedP. vivax mono-infection. There were no P. knowlesi/P.vivax mixed infections diagnosed by the real-time PCRassay subsequent to June 2011 despite an increase in P.vivax mixed and mono-infections diagnosed by micros-copy over this time. One additional P. knowlesi mono-infection was identified by PCR in a patient withmicroscopy-diagnosed P. falciparum/P. vivax mixed in-fection, giving a total of 345 patients with PCR-confirmed P. knowlesi mono-infection.Based on these results, and using the population of the
Kudat Division, the estimated incidence of malaria from2009 to 2011 was 2.6/1,000 people/year, with a
Barber et al. Malaria Journal 2012, 11:401 Page 3 of 8http://www.malariajournal.com/content/11/1/401
minimum incidence of PCR-confirmed P. knowlesimono-infections of 1.4 infections/1,000 people/year.
Age and sex distributionPatients with knowlesi malaria demonstrated a wide agedistribution with the species being most common in allage groups except those <5 years old, where numbersapproximated those of P. falciparum and P. vivax(Figure 2). Although the median age of patients withPCR-confirmed P. knowlesi (33 years, IQR 20–50 years)was significantly older than patients diagnosed by mi-croscopy with P. falciparum (19 years, IQR 9–31 years,p < 0.001) or P. vivax (19 years, IQR 7–32 years, p <0.001), 17% of all PCR-confirmed P. knowlesi cases oc-curred in children <15 years old. Eight children <5 yearsold had PCR-confirmed knowlesi malaria (with one ofthese already reported [8]), in addition to two withmicroscopy-diagnosed “P. malariae”. The youngest childwith PCR-confirmed knowlesi malaria was eight monthsold.Among patients with PCR-confirmed P. knowlesi mal-
aria, 252 (73%) were male. This proportion was loweramong children (<15 years) than it was among adults(33/57 [58%] vs 219/288 [76%], p = 0.005), and highestamong adults aged 15–40 years, of whom 133/157 (85%)were male compared to 119/188 (63%) of patients out-side this age range (p < 0.001).Overall the median age of females with PCR-
confirmed knowlesi malaria (39 years, IQR 14–53 years)was not significantly different than that of males (32years, IQR 21.5-49 years, p = 0.770), however amongadults with knowlesi malaria, females were significantlyolder than males (median age 45 years [IQR 30–57years] vs 37 years [IQR 25–52 years], p = 0.004).
Temporal variationThe number of patients diagnosed with PCR-confirmedP. knowlesi mono-infection as a proportion of the totalnumber of slides taken demonstrated significant inter-
annual variation (Figure 3), with 139 (2.1%), 21 (0.4%)and 185 (2.8%) cases diagnosed in 2009, 2010 and 2011respectively (χ2 = 107, p < 0.001). Significant seasonalitywas also demonstrated (p < 0.001), with increased trans-mission from March to July in 2009 and April to Augustin 2011. In both years the number of cases peaked inMay, with the peak being particularly marked in 2011.Remarkably few cases of knowlesi malaria occurred in2010.Rainfall recorded at Kudat meteorological station was
strongly correlated with the number of cases of P. know-lesi in the proceeding three to five months (Table 2).The number of P. knowlesi cases peaked during the fifthmonth following the rainfall, with a correlation coeffi-cient of 0.50. Total rainfall in 2010 (1,830 mm) was sig-nificantly less than in 2009 (2,848 mm) and 2011 (3,643mm).The number of cases of microscopy-diagnosed P. fal-
ciparum mono-infections (excluding the five patientsfound to be P. knowlesi by PCR) decreased slightlyin the last year of the study period, with 26 (0.4%),29 (0.5%) and 16 (0.2%) cases diagnosed in 2009, 2010and 2011 respectively (χ2 = 6, p = 0.056). In contrast,microscopy-diagnosed P. vivax mono-infections increased,with 45 (0.7%) cases in 2011 compared to zero and eight(0.1%) in 2009 and 2010 respectively (χ2 = 60, p < 0.001).
Family clustersTwo family clusters of patients with PCR-confirmed P.knowlesi monoinfection were identified. The first con-sisted of a 37 year-old man with his 10 year-old daughterand 11 year-old son, who all presented on the same daywith the same duration of illness. The family lived intown, the children attended the local school, and thefather worked as a clerk in a town resort. Although theyhad not travelled into forested areas, the family reportedseeing macaques close to their home. The second familyincluded three brothers aged one, five and 11 years, allpresenting on the same day. They lived on the island of
Table 1 Microscopy and PCR results, January 2009 to November 2011
PCR
Microscopy Pf Pv Pk Pv/Pk Pf/Pk Pf/Pv/Pk Pm/Pk Pf/Pm Pv/Pm/Pk P.genus negative Not done Total
Pf 10 0 5 0 0 0 0 0 0 0 1 60 76
Pv 0 9 0 6 0 0 0 0 0 4 0 34 53
Pm 3 11 306 18 6 2 1 0 0 18 7 57 429
Pm/Pv 3 12 10 11 0 0 0 0 1 2 1 7 47
Pf/Pm 7 0 22 0 0 0 0 1 1 0 1 8 40
Pf/Pv 1 2 1 1 0 0 0 0 0 0 0 2 7
Pf/Pv/Pm 0 0 1 0 0 0 0 0 0 0 0 0 1
Total 24 34 345 36 6 2 1 1 2 24 10 168 653
*Blood slides with parasites resembling P. malariae were sometimes reported as “Pm/Pk” or “Pm(?Pk)” but for simplicity are all referred to here as “Pm”.Abbreviations: Pf=P. falciparum, Pv=P. vivax, Pm=P. malariae, Pk=P. knowlesi.
Barber et al. Malaria Journal 2012, 11:401 Page 4 of 8http://www.malariajournal.com/content/11/1/401
Banggi, a forested area with numerous macaques. The11 year-old boy had chronic myeloid leukaemia forwhich he took imatinib. His uncomplicated knowlesimalaria was unremarkable except for a notable absenceof thrombocytopenia.
DiscussionPlasmodium knowlesi is the most common cause of mal-aria admissions to Kudat District Hospital and affects allages from young children to the elderly. Although thegreater proportion of males, particularly among adults,is likely to be consistent with occupational forest orplantation exposure as a risk factor for knowlesi malaria,the wide age distribution suggests that this is not theonly determinant of transmission. Furthermore, this
report of two family clusters, one of which had not trav-elled outside Kudat town, suggests that transmissionmay be occurring close to or inside people’s homes.These findings differ from those reported from two stud-ies in Sarawak, where a smaller proportion of knowlesimalaria cases occurred in children (8/121 [6.6%] in onestudy [4] and 9/106 [8.5%] in the other [10]), and noclustering of cases was reported despite the presence ofcommunal longhouses in the study areas [4,10]. Epi-demiological risk factors of knowlesi malaria have notbeen previously investigated in either of these areas andrequire further evaluation.It was notable that in contrast to the age-distribution
of other malaria species, 25% of all P. knowlesi cases oc-curred in adults over 50 years of age. This is consistent
Figure 2 Age distribution of patients with malaria by microscopy and PCR, January 2009 to November 2011. A. Age distribution ofpatients with malaria diagnosed by microscopy; B. Age distribution of patients with malaria diagnosed by PCR. Abbreviations: Pf=P. falciparum,Pv=P. vivax, Pm=P. malariae, Pk=P. knowlesi.
Barber et al. Malaria Journal 2012, 11:401 Page 5 of 8http://www.malariajournal.com/content/11/1/401
with a lack of past immunity to P. knowlesi in this agegroup and/or a relatively recent increase in the risk ofexposure to this species in the Kudat region. It may alsorelate to greater forest exposure among older indivi-duals, with farmers and plantation workers over-represented among this age group [7]. The finding thatadult females with knowlesi malaria are older than adultmales is consistent with a previous report [7] andrequires further investigation, but may relate to less for-est and/or vector exposure among younger females. Theolder age of female adults with P. knowlesi may also ex-plain the finding from two previous studies that femalepatients are at greater risk of severe disease [4,6].
In Kapit, Sarawak, An. latens was recently identified asthe primary P. knowlesi vector, with the mosquitoes bit-ing both monkeys and humans, and four (0.4%) infectedmosquitoes found in forest and farm locations [12]. Al-though 126 (11.7%) An. latens were found in longhouses,none were infected, suggesting transmission occurs awayfrom homes [12]. A subsequent study found evidencethat in Kapit P. knowlesi remains a zoonosis, based onan extremely high prevalence (87%) of P. knowlesi inlong-tailed macaques, and sequencing of the csp geneand mtDNA showing a greater number of P. knowlesigenotypes per monkey infection than human infection,with genotypes common to both hosts and no genotypeexclusive to either [17].In Kudat, the wide age distribution and clustering of
cases suggest that the P. knowlesi vector may be bitinghumans close to or inside people’s houses. PreviouslyAn. balabacensis was known to be the primary vector ofhuman malaria in Sabah [18], with Anopheles flavirostrisalso identified as a potent P. falciparum vector onBanggi Island [19]. Anopheles balabacensis have beenshown to readily bite monkeys at the canopy level [20],and a P. knowlesi-infected An balabacensis was recentlyfound in Ranau [13]. Of significance for potentialhuman-to-human malaria transmission, An. balabacen-sis have been shown to exhibit “learning behaviour” inrelation to host preference, and may also demonstrate
Table 2 Correlation between monthly rainfall (mm) andnumber of P. knowlesi cases
Lag time between rainfall andP. knowlesi cases (months)
Correlationcoefficient
P value
0 −0.246 0.156
1 0.122 0.484
2 0.280 0.103
3 0.469 0.005
4 0.491 0.003
5 0.504 0.002
6 0.331 0.060
Figure 3 Monthly rainfall and malaria cases. *excludes five patients diagnosed by microscopy with P. falciparum but found to have P. knowlesiby PCR. Abbreviations: Pf=P. falciparum, Pv=P. vivax, Pm=P. malariae, Pk=P. knowlesi.
Barber et al. Malaria Journal 2012, 11:401 Page 6 of 8http://www.malariajournal.com/content/11/1/401
“habitat loyalty” by returning to previous feeding sites[21]. Furthermore, An. balabacensis has been shown tobe a highly efficient vector [13], and is closely related toAnopheles dirus, the primary malaria vector in the Me-kong region, which maintains high levels of malaria en-demicity at low population densities [22], is capable ofadapting to deforestation [23], and in Southern Vietnamwas found to be positive for P. knowlesi DNA [24].More recently however, An. donaldi was shown to
have replaced An. balabacensis as the primary species inthe Kinabatangan region of Sabah [14]. The peak out-door biting time for this mosquito occurs from 18.00-19.00 hours, when humans are often outside theirhomes, and indoor biting occurs throughout the night.Although sporozoites were detected in An. donaldi inthe Kinabatangan region [14], species identification wasnot performed and further studies are required to deter-mine the role of this species in the transmission ofknowlesi malaria.A strong association was demonstrated between know-
lesi malaria cases and rainfall, with a lag time of three tofive months. Anopheles mosquitoes depend on rainfallto provide aquatic breeding sites, and An. balabacensisand An. donaldi have both been shown to haveincreased parity rates during the months of greatestrainfall [14]. An association between rainfall and falcip-arum and vivax malaria prevalence has been well docu-mented [25-29], and can be used to assist with malariaearly warning systems and prediction of epidemics[30,31]. This association should also be considered whenplanning for control of knowlesi malaria. However, alter-native explanations for seasonal variation in knowlesimalaria such as other climatic variables and seasonalplantation work were not explored in this study and re-quire further investigation.Although the numbers of P. vivax and P. falciparum
in this study were small, and could only be assessed overa three-year period, the number of P. vivax casesincreased during the study period along with P. knowlesi,while a modest decrease was seen in P. falciparum cases.Larger prospective studies will be required to furtherevaluate the interaction between malaria species, whichmay have implications for malaria control.This study had several limitations. First, the retro-
spective design did not allow the collection of informa-tion on where patients acquired their malaria infection.This prevented detailed assessment of forest exposure asa risk factor for knowlesi malaria, and may also haveinfluenced the association between malaria cases andrainfall recorded at Kudat Meteorological Station. Inparticular, information was lacking on which patientscame from Banggi Island, which, although only a shortdistance from Kudat town, may represent a differentgeographical setting with different transmission
dynamics. Second, identification of case clusters wasbased only on patients with the same family namerecorded in microscopy books, and this may have led toan underestimation of case clustering. Third, only asmall number of patients with microscopy-diagnosed P.falciparum or P. vivax had PCR performed, and micros-copy reports were therefore relied on for analyses involv-ing these species. It is possible that some of the patientsdiagnosed by microscopy as P. falciparum or P. vivaxhad P. knowlesi. However, despite the potential for thisto dilute age-related differences among species, a signifi-cant difference in age distribution was noted between P.knowlesi and microscopy-diagnosed P .falciparum or P.vivax, making the estimates of age-related differencesconservative. Finally, this study involved the use, prior toJune 2011, of a nested PCR assay (using Pmk8-Pmk9primers) that has been associated with cross-reactivitywith P. vivax DNA, and false-positive PCR findings ofmixed P. knowlesi/P. vivax in true P. vivax mono-infections [32]. In this study this may account for theone third of patients diagnosed by microscopy with P.vivax mono-infection prior to June 2011, and subse-quently found to have mixed P. vivax/P. knowlesi infec-tions by PCR. Cross-reactivity with P. vivax isolates hasnot been reported with the real-time PCR assay [15],and it was notable that after the introduction of thismethod there were no P. knowlesi/P. vivax mixed infec-tions diagnosed by PCR despite an increase inmicroscopy-diagnosed P. vivax mixed and mono-infections over this time.Malaysia has made impressive progress in reducing
malaria prevalence in recent years; however, P. knowlesicases appear to be increasing, with the species nowaccounting for the majority of malaria admissions toKudat Hospital as well as other district hospitalsthroughout Sabah and Sarawak [3,4,9]. The simian hostreservoir of this zoonotic species presents particularchallenges for malaria control, and will hinder the pro-gress of malaria elimination in Malaysia. In addition, theincreasing dominance of P. knowlesi and the possibilityof transmission occurring close to or inside people’shomes may increase the possibility of a switch to thehuman host. Prospective studies on epidemiological riskfactors, vectors and transmission dynamics of P. know-lesi in Sabah, including the possibility of human-to-human transmission, are required in order to developstrategies for knowlesi malaria control.
Competing interestsThe authors declare that they have no competing interests.
Authors’ contributionsBEB, NMA, TWY, TW, PD and MJG conceived and designed the study. FAperformed the PCR assays at Sabah State Reference Laboratory. BEBcollected and analysed the data. BEB and NMA wrote the manuscript. Allauthors read and approved the final manuscript.
Barber et al. Malaria Journal 2012, 11:401 Page 7 of 8http://www.malariajournal.com/content/11/1/401
AcknowledgementsWe thank the laboratory staff at Kudat District Hospital for their help withdata collection, Mark Chatfield for his help with the statistical methods, andJeffrey Hii for helpful comments. We also thank the Director General ofHealth (Malaysia) for permission to publish this study.
Author details1Global Health Division, Menzies School of Health Research, PO Box 41096,Casuarina 0810,, Northern Territory, Australia. 2Infectious DiseasesDepartment, Queen Elizabeth Hospital, Karung Berkunci No. 2029, JalanPenampang, Kota Kinabalu 88560, Sabah, Malaysia. 3Sabah Department ofHealth, Kota Kinabalu, Sabah, Malaysia. 4Kudat District Hospital, Peti Surat No.22, 89057, Kudat, Sabah, Malaysia. 5Sabah Public Health Reference Laboratory,Bukit Padang, Jalan Kolam, 88850, Kota Kinabalu, Sabah, Malaysia. 6RoyalDarwin Hospital, Darwin, NT, Australia.
Received: 11 October 2012 Accepted: 28 November 2012Published: 5 December 2012
References1. World Health Organization: World Malaria Report 2011. Available: http://
www.who.int/malaria/world_malaria_report_2011/9789241564403_eng.pdf.Accessed 3rd May 2012.
2. Rundi C: Malaria elimination in Malaysia. Third annual meeting of the AsiaPacific Malaria Elimination Network (APMEN). Sabah, Malaysia: Kota Kinabalu;2011. Available: http://apmen.org/apmen-iii-meeting-proceedings/.Accessed 3rd May 2012.
3. Cox-Singh J, Davis TM, Lee KS, Shamsul SS, Matusop A, Ratnam S, RahmanHA, Conway DJ, Singh B: Plasmodium knowlesi malaria in humans iswidely distributed and potentially life threatening. Clin Infect Dis 2008,46:165–171.
4. Daneshvar C, Davis TM, Cox-Singh J, Rafa'ee M, Zakaria S, Divis P, Singh B:Clinical and laboratory features of human Plasmodium knowlesi infection.Clin Infect Dis 2009, 49:852–860.
5. Rajahram G, Barber BE, William T, Menon J, Anstey NM, Yeo TW: Deaths dueto Plasmodium knowlesi malaria in Sabah, Malaysia: association withreporting as P. malariae and delayed parenteral artesunate. Malar J 2012,11:284.
6. William T, Menon J, Rajahram G, Chan L, Ma G, Donaldson S, Khoo S,Fredrick C, Jilip J, Anstey NM, Yeo TW: Severe Plasmodium knowlesimalaria in a tertiary hospital, Sabah. Malaysia Emerg Infect Dis 2011,17:1248–1255.
7. Barber BE, William T, Grigg MJ, Menon J, Auburn S, Marfurt J, Anstey NM,Yeo TW: A prospective comparative study of knowlesi, falciparum andvivax malaria in Sabah, Malaysia: high proportion with severe diseasefrom Plasmodium knowlesi and P. vivax but no mortality with earlyreferral and artesunate therapy. Clin Infect Dis 2012, doi:10.1093/cid/cis902.Advance Access published October 19.
8. Barber BE, William T, Jikal M, Jilip J, Dhararaj P, Menon J, Yeo TW, AnsteyNM: Plasmodium knowlesi malaria in children. Emerg Infect Dis 2011,17:814–820.
9. Joveen-Neoh WF, Chong KL, Wong CM, Lau TY: Incidence of malaria inthe Interior Division of Sabah, Malaysian Borneo, based on nested PCR.J Parasitol Res 2011, 2011. doi:10.1155/2011/104284.
10. Singh B, Sung LK, Matusop A, Radhakrishnan A, Shamsul SSG, Cox-Singh J,Thomas A, Conway DJ: A large focus of naturally acquired Plasmodiumknowlesi infections in human beings. Lancet 2004, 363:1017–1024.
11. Chin W, Contacos P, Collins W, Jeter M, Alpert E: Experimentalmosquito-transmission of Plasmodium knowlesi to man and monkey.Am J Trop Med Hyg 1968, 17:355.
12. Tan C, Vythilingam I, Matusop A, Chan S, Singh B: Bionomics of Anopheleslatens in Kapit, Sarawak, Malaysian Borneo in relation to the transmissionof zoonotic simian malaria parasite. Plasmodium knowlesi Malar J 2008,7:52.
13. Vythilingam I: Plasmodium knowlesi. in humans: a review on the role ofits vectors in Malaysia. Trop Biomed 2010, 27:1–12.
14. Vythilingam I, Chan S, Shanmugratnam C, Tanrang H, Chooi KH: The impactof development and malaria control activities on its vectors in theKinabatangan area of Sabah, East Malaysia. Acta Trop 2005, 96:24–30.
15. Divis P, Shokoples SE, Singh B, Yanow SK: A TaqMan real-time PCR assayfor the detection and quantitation of Plasmodium knowlesi. Malar J 2010,9:344.
16. Rougemont M, Van Saanen M, Sahli R, Hinrikson HP, Bille J, Jaton K:Detection of four Plasmodium species in blood from humans by 18SrRNA gene subunit-based and species-specific real-time PCR assays.J Clin Microbiol 2004, 42:5636.
17. Lee K-S, Divis PCS, Zakaria SK, Matusop A, Julin RA, Conway DJ, Cox-Singh J,Singh B: Plasmodium knowlesi: reservoir hosts and tracking theemergence in humans and macaques. PLoS Pathog 2011, 7:e1002105.
18. Hii J: Anopheles malaria vectors in Malaysia with reference to Sabah.In Anopheles malaria vectors in Malaysia with reference to Sabah. Edited byHarinasutu C, Reynolds D. Tokyo: Southeast Asian Medical InformationCentre; 1985:71–81.
19. Hii J, Kan S, Vun Y, Chin K, Tambakau S, Chan M, Lye M, Mak J:Transmission dynamics and estimates of malaria vectorial capacity forAnopheles balabacensis and An. flavirostris (Diptera: Culicidae) on Banggiisland, Sabah, Malaysia. Ann Trop Med Parasitol 1988, 82:91–101.
20. Eyles DE, Wharton R, Cheong W, Warren MW: Studies on malaria andAnopheles balabacensis in Cambodia. Bull World Health Organ 1964, 30:7.
21. Hii J, Chew M, Sang V, Munstermann L, Tan S, Panyim S, Yasothornsrikul S:Population genetic analysis of host seeking and resting behaviors in themalaria vector, Anopheles balabacensis (Diptera: Culiciae). J Med Entomol1991, 28:675–684.
22. Rosenberg R, Andre R, Somchit L: Highly efficient dry season transmissionof malaria in Thailand. Trans R Soc Trop Med Hyg 1990, 84:22–28.
23. Singhasivanon P, Thimasarn K, Yimsamran S, Linthicum K, Nualchawee K,Dawreang D, Kongrod S, Premmanisakul N, Maneeboonyang W, Salazar N:Malaria in tree crop plantations in south-eastern and western provincesof Thailand. Southeast Asian J Trop Med Public Health 1999, 30:399–404.
24. Nakazawa S, Marchand R, Quong N, Culleton R, Manh N, Maeno Y:Anopheles dirus co-infection with humans and monkey malariaparasites in Vietnam. Int J Parasitol 2009, 39:1533–1537.
25. Alemu A, Abebe G, Tsegaye W, Golassa L: Climatic variables and malariatransmission dynamics in Jimma town. South West Ethiopia. Parasit Vectors2011, 4:30.
26. Huang F, Zhou S, Zhang S, Wang H, Tang L: Temporal correlation analysisbetween malaria and meteorological factors in Motuo County. Tibet.Malar J 2011, 10:54.
27. Ngomane L, de Jager C: Changes in malaria morbidity and mortality inMpumalanga Province, South Africa (2001–2009): a retrospective study.Malar J 2012, 11:19.
28. Teklehaimanot HD, Lipsitch M, Teklehaimanot A, Schwartz J: Weather-basedprediction of Plasmodium falciparum malaria in epidemic-prone regionsof Ethiopia I, Patterns of lagged weather effects reflect biologicalmechanisms. Malar J 2004, 3:41.
29. Zhou G, Minakawa N, Githeko AK, Yan G: Association between climatevariability and malaria epidemics in the East African highlands. Proc NatlAcad Sci U S A 2004, 101:2375.
30. Hay SI, Were EC, Renshaw M, Noor AM, Ochola SA, Olusanmi I, Alipui N,Snow RW: Forecasting, warning, and detection of malaria epidemics: acase study. Lancet 2003, 361:1705–1706.
31. Gomez-Elipe A, Otero A, Van Herp M, Aguirre-Jaime A: Forecasting malariaincidence based on monthly case reports and environmental factors inKaruzi, Burundi, 1997–2003. Malar J 2007, 6:129.
32. Imwong M, Tanomsing N, Pukrittayakamee S, Day NPJ, White NJ, SnounouG: Spurious amplification of a Plasmodium vivax small-subunit RNA geneby use of primers currently used to detect P. knowlesi. J Clin Microbiol2009, 47:4173.
doi:10.1186/1475-2875-11-401Cite this article as: Barber et al.: Epidemiology of Plasmodium knowlesimalaria in north-east Sabah, Malaysia: family clusters and wide agedistribution. Malaria Journal 2012 11:401.
Barber et al. Malaria Journal 2012, 11:401 Page 8 of 8http://www.malariajournal.com/content/11/1/401
60
Implications of and questions arising from this paper
This study involved 345 patients with PCR-confirmed P. knowlesi malaria and hence
represented the largest cohort of such patients described, providing the opportunity to
describe in detail the notable age distribution affected by this parasite. In this paper we
report that in contrast to P. vivax and P. falciparum, P. knowlesi affects an older age group,
with 25% of cases occurring in adults >50 years. The reason for this older age distribution
however has not been determined, and requires further investigation. A case-control study
is currently underway in Kudat, and may help to answer this question.
This older age distribution may also impact on the clinical presentation of knowlesi malaria.
In falciparum malaria, increasing age has been associated with increasing risk of severe
disease (172). In addition, manifestations of severe disease in falciparum malaria depend
on age, with hyperparasitemia, jaundice and acute kidney injury all occurring more
commonly among older patients, while anaemia and convulsions occur less commonly
(104). Increasing age has also been associated with increased mortality among patients
with severe falciparum malaria (104).
In this paper we also report the existence of two family clusters. Both of these clusters
involved children, and one family lived in Kudat town and reported minimal forest exposure.
This is in contrast to the findings of two previous studies in Kapit, Sarawak, where no case-
clustering was reported despite the presence of long-houses in the study area (3, 24). This
finding suggests that the epidemiology of knowlesi malaria in these two areas may differ. In
Sarawak molecular studies involving sequencing of the circumsporozoite protein (csp) gene
and mitochondrial DNA of P. knowlesi isolates from humans and macaques found that P.
knowlesi remains primarily a zoonosis (53). Similar studies will be required in the Kudat
region to determine if the situation there differs.
61
Finally, this paper provides a description of the relationship between rainfall and knowlesi
malaria, with cases increasing 2 – 5 months following periods of heavy rainfall. More
detailed prospective studies are required to confirm this finding, which may have implications
for malaria control strategies in the region.
62
-
Our two retrospective studies conducted in Kudat, Sabah (58, 171), demonstrated that P.
knowlesi is now overwhelmingly the most common cause of malaria admissions to Kudat
District Hospital. Increasing evidence is also emerging from other regions of Sabah,
suggesting that the predominance of knowlesi malaria may be widespread across the state.
In one such study, P. knowlesi accounted for 59% of malaria blood films in the Interior
Division in 2009 (25), and in another, also in 2009, P. knowlesi DNA was detected in
353/445 (77%) samples referred to the Sabah Public Health Reference Laboratory for
confirmation of Plasmodium infection (26). However, the timing of the emergence of P.
knowlesi as a common cause of human malaria in Sabah was not known at the time of these
studies. In 2001, only 96/6050 (1.6%) malaria slides referred to the Sabah State Public
Health Laboratory were diagnosed as “P. malariae” monoinfection by microscopy, with this
proportion increasing to 59/2741 (2.2%) in 2004 (26), suggesting that the increase in
reported “P. malariae” has occurred only recently.
To investigate the trends of “P. malariae/P. knowlesi” notifications in Sabah I reviewed all
available Sabah Department of Health malaria notification records from 1992 – 2002.
Increasing Incidence of Plasmodium knowlesi Malariafollowing Control of P. falciparum and P. vivax Malaria inSabah, MalaysiaTimothy William1,2, Hasan A. Rahman3, Jenarun Jelip4, Mohammad Y. Ibrahim4, Jayaram Menon2,4,
Matthew J. Grigg1,5, Tsin W. Yeo5,6, Nicholas M. Anstey5,6*, Bridget E. Barber1,5
1 Infectious Diseases Unit, Department of Medicine, Queen Elizabeth Hospital, Kota Kinabalu, Sabah, Malaysia, 2Department of Medicine, Queen Elizabeth Hospital, Kota
Kinabalu, Sabah, Malaysia, 3Ministry of Health, Putrajaya, Malaysia, 4 Sabah Department of Health, Kota Kinabalu, Sabah, Malaysia, 5Global Health Division, Menzies
School of Health Research and Charles Darwin University, Darwin, Northern Territory, Australia, 6 Royal Darwin Hospital, Darwin, Northern Territory, Australia
Abstract
Background: The simian parasite Plasmodium knowlesi is a common cause of human malaria in Malaysian Borneo andthreatens the prospect of malaria elimination. However, little is known about the emergence of P. knowlesi, particularly inSabah. We reviewed Sabah Department of Health records to investigate the trend of each malaria species over time.
Methods: Reporting of microscopy-diagnosed malaria cases in Sabah is mandatory. We reviewed all available Departmentof Health malaria notification records from 1992–2011. Notifications of P. malariae and P. knowlesi were considered as asingle group due to microscopic near-identity.
Results: From 1992–2011 total malaria notifications decreased dramatically, with P. falciparum peaking at 33,153 in 1994and decreasing 55-fold to 605 in 2011, and P. vivax peaking at 15,857 in 1995 and decreasing 25-fold to 628 in 2011.Notifications of P. malariae/P. knowlesi also demonstrated a peak in the mid-1990s (614 in 1994) before decreasing to<100/year in the late 1990s/early 2000s. However, P. malariae/P. knowlesi notifications increased .10-fold between 2004 (n = 59)and 2011 (n = 703). In 1992 P. falciparum, P. vivax and P. malariae/P. knowlesi monoinfections accounted for 70%, 24% and1% respectively of malaria notifications, compared to 30%, 31% and 35% in 2011. The increase in P. malariae/P. knowlesinotifications occurred state-wide, appearing to have begun in the southwest and progressed north-easterly.
Conclusions: A significant recent increase has occurred in P. knowlesi notifications following reduced transmission of thehuman Plasmodium species, and this trend threatens malaria elimination. Determination of transmission dynamics and riskfactors for knowlesi malaria is required to guide measures to control this rising incidence.
Citation: William T, Rahman HA, Jelip J, Ibrahim MY, Menon J, et al. (2013) Increasing Incidence of Plasmodium knowlesi Malaria following Control of P. falciparumand P. vivax Malaria in Sabah, Malaysia. PLoS Negl Trop Dis 7(1): e2026. doi:10.1371/journal.pntd.0002026
Editor: J. Kevin Baird, Eijkman-Oxford Clinical Research Unit, Indonesia
Received October 9, 2012; Accepted December 6, 2012; Published January 24, 2013
Copyright: � 2013 William et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permitsunrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This work was supported by the Australian National Health and Medical Research Council (Program Grant 496600, fellowships to NMA and TWY, andscholarship to BEB). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing Interests: The authors have declared that no competing interests exist.
* E-mail: [email protected]
Introduction
Malaria elimination is now a goal of many countries in
Southeast Asia and the Western Pacific, and large reductions in
malaria prevalence have been achieved [1]. However, significant
challenges remain, and while the threat of artemisinin resistance
has been the focus of much international concern, zoonotic
malaria species have received less consideration. Malaysia has had
one of the most successful malaria control programs in the region,
and aims to be malaria-free by 2020 [1,2]. However, the simian
parasite Plasmodium knowlesi, transmitted by the forest-dwelling
Anopheles leucosphyrus group of mosquitoes, is now a common cause
of human malaria in the eastern states of Sabah and Sarawak, and
presents an increasing threat to malaria elimination [3,4,5,6,7].
Documentation of the emergence of this species over time is
limited by the inability to distinguish P. knowlesi from P. malariae by
microscopy. Although the first naturally acquired case of human
knowlesi malaria was reported from Peninsular Malaysia in 1965
[8], with a second probable case several years later [9], it was not
until the early 2000s that a large focus of human infections was
described in Kapit, Sarawak [10]. Since this time an increasing
number of cases have been reported, and P. knowlesi is now the
most common cause of human malaria in several districts
throughout Sabah and Sarawak [3,4,5,6]. The highest proportion
has been reported at Kudat District Hospital (KDH), on the
northeast tip of Sabah, where 87% of patients admitted with
malaria in 2009 were infected with P. knowlesi [3].
Whether this apparent increase in cases however is due to a true
emergence of the species or increasing recognition remains
uncertain. Evolutionary analyses of sequence data from samples
obtained from Sarawak indicate that P. knowlesi existed in
macaques in Southeast Asia more than 100,000 years ago, with
PLOS Neglected Tropical Diseases | www.plosntds.org 1 January 2013 | Volume 7 | Issue 1 | e2026
infection in humans likely occurring from the time of human
arrival in the region [11]. In the earliest documented malaria
survey conducted in Sarawak, in 1952, one third of all malaria
cases were reported as P. malariae by microscopy [12]. Given the
evidence of very few cases of P. malariae in Sarawak when PCR
methods are used [4,5,10], it seems likely that at least some of
these cases were P. knowlesi. When PCR was performed on the
earliest P. malariae slides available, taken in 1996, 35/36 (97%)
were positive for P. knowlesi, with only one being positive for P.malariae [13]. In 1999, ‘‘P. malariae’’ accounted for 9% of all
malaria notifications in Sarawak, and 20% of cases in the Kapit
district [14].
In Sabah, limited available evidence suggests that the situation
may differ from that of Sarawak, and that P. knowlesi infection in
humans may have increased only recently. In 2001, only 96/6050
(1.6%) malaria slides referred to the Sabah State Public Health
Laboratory were diagnosed as P. malariae monoinfection by
microscopy, with the proportion increasing to 59/2741 (2.2%) in
2004 [15]. In contrast, microscopy-diagnosed ‘‘P. malariae’’accounted for 621/1872 (33%) of malaria cases reported to the
Sabah Department of Health in 2011 (unpublished data from
Sabah Department of Health records).
In this study, we reviewed the Sabah Department of Health records
of malaria notifications from 1992–2011, in order to investigate the
trend of each malaria species over time, and in particular to determine
if P. knowlesi represents an emerging infection in humans.
Methods
Ethics statementThe study was approved by the Medical Research Sub-Committee
of the MalaysianMinistry of Health and theMenzies School of Health
Research, Australia. All data analysed were anonymised.
Study siteThe north-eastern Malaysian state of Sabah has an area of
73,600 km2 and a population of 3.2 million [16]. Situated between
4u and 7u north of the equator, Sabah has a mostly tropical
climate, with high humidity and rainfall throughout the year and
temperatures of 25–35uC. The southwest interior of Sabah is
mountainous, with the Crocker Range separating west coast
lowlands from the rest of the state and extending north to Mount
Kinabalu at 4095 meters above sea level. Sabah was previously
covered almost entirely in dense primary rainforest, however
extensive deforestation occurred throughout the 1970s and 1980s,
reducing forest cover to 44–63% of the state [17,18,19]. Cleared
areas have been partly replaced by plantations, with palm oil
estates comprising 16% of Sabah’s land area [19].
Malaysia has a long history of malaria control programs dating
back to the early 1900s, with an initial focus on environmental
management techniques. The launch of the Malaria Eradication
Program in 1967, followed by state-wide malaria control programs
during the 1970s and 1980s, led to large reductions in malaria
prevalence, with cases falling from 240,000 in 1961 to around
50,000/year during the 1980s [20,21]. Further scale-up of malaria
control activities began in 1992, consisting of increased surveil-
lance, vector control, training of community volunteers, and early
diagnosis and treatment [21]. Use of insecticide-treated nets and
indoor residual spraying was implemented in 1995, with nation-
wide coverage of the high-risk population reported to be .50%
and 25–50% respectively in 2010 [1]. In addition, Malaysia
reports 100% confirmatory testing of suspected malaria cases and
mandatory notification of detected cases [1].
Mosquito vectors in Sabah include An. balabacensis and An.donaldi [22], and the P. knowlesi hosts, the long-tailed and pig-tailed
macaques, are found throughout the state.
Review of malaria notification recordsIn Sabah mandatory reporting of all malaria cases to the Sabah
State Health Department is generally done by nursing staff, with
species normally reported according to microscopy results. Blood
slides with parasites resembling P. malariae/P. knowlesi are mostly
reported, and hence notified, as P. malariae.
We reviewed all available malaria notification records held by the
Sabah State Health Department. Hard copy summaries of annual
malaria notifications by species and by district were available from
1992. From 2007 yearly Excel databases were also available that
included limited demographic/epidemiological information for
each malaria notification. We therefore recorded the number of
notifications of each Plasmodium species annually for each district in
Sabah from 1992–2011, in addition to the age and sex distribution
and seasonal variation of each species from 2007–2011.
Data analysisData were analysed using Stata statistical software, version 10.0
(StataCorp LP, College Station, TX, USA). Spearman’s correla-
tion coefficient was used to analyse the association between annual
notification rates of the Plasmodium species. Median ages were
compared using Wilcoxon rank-sum test, and proportions were
assessed using the Chi-square test. Edwards’ test was used to assess
seasonality of the Plasmodium species.
Notifications of P. malariae and P. knowlesi were considered as a single
group (‘‘P. malariae/P. knowlesi’’), due to the inability to distinguish these
species bymicroscopy.Mixed-species infections were recorded as a single
group, with analysis of these cases limited to annual notification rates.
Results
Malaria trends in Sabah state, 1992–2011Between 1992 and 2011 the total number of malaria
notifications to the Sabah State Health Department decreased
Author Summary
The simian parasite Plasmodium knowlesi is a commoncause of malaria in Malaysian Borneo; however, little isknown about its emergence over time, particularly inSabah. We reviewed all available Sabah Department ofhealth malaria notification records from 1992–2011, andconsidered notifications of P. malariae and P. knowlesi as asingle group due to their microscopic similarity. We foundthat malaria notifications in Sabah have decreaseddramatically, with P. falciparum and P. vivax notificationspeaking at 33,153 and 15,877 respectively during 1994–1995, and falling to 605 and 628 respectively in 2011.Notifications of P. malariae/P. knowlesi fell from a peak of614 in 1994 to <100/year in the late 1990s/early 2000s,however increased .10-fold between 2004 (n= 59) and2011 (n= 703). In 1992 P. falciparum, P. vivax and P.malariae/P. knowlesi monoinfections accounted for 70%,24% and 1% respectively of malaria notifications, com-pared to 30%, 31% and 35% in 2011. The increase in P.malariae/P. knowlesi notifications occurred state-wide,appearing to have begun in the southwest and progressednorth-easterly. This significant recent increase in P.knowlesi notifications following reduced transmission ofthe human Plasmodium species threatens malaria elimina-tion; further research is required to determine transmissiondynamics and risk factors for knowlesi malaria.
Increasing Incidence of P. knowlesi in Sabah
PLOS Neglected Tropical Diseases | www.plosntds.org 2 January 2013 | Volume 7 | Issue 1 | e2026
dramatically, with P. falciparum notifications peaking at 33,153 in
1994 and decreasing 55-fold to 605 in 2011, while P. vivax
notifications peaked at 15,857 in 1995 and decreased 25-fold to
628 in 2011 (Figure 1). Notifications of P. malariae/P. knowlesi
also demonstrated a peak in the mid-1990s (increasing from 200
in 1992 to 614 in 1994), before decreasing to around 100/year
in the late 1990s and early 2000s. Until 2003, annual
notifications of P. malariae/P. knowlesi strongly correlated with
those of P. falciparum (Spearman’s correlation coefficient 0.94,
p,0.0001; Figure 2). However, the relationship between the
species began to change in the early 2000s, with P. falciparum
notifications steadily decreasing (from 3264 in 2002 to 605 in
2011) while P. malariae/P. knowlesi notifications remained stable
from the late 1990s to 2006, and then increased markedly from
2007 (Figure 1.B). An inverse correlation was demonstrated
between P. falciparum notifications and P. malariae/P. knowlesi
notifications between 2004 and 2011 (Spearman’s correlation
coefficient 20.76, p = 0.028; Figure 2).
Notifications of P. vivax generally correlated with those of P.
falciparum (Spearman’s correlation coefficient from 1992–
2011= 0.90, p,0.0001), and with P. malariae/P. knowlesi notifica-
tions until around 2008 (Spearman’s correlation coefficient 0.91,
p,0.0001). Since 2008 P. vivax notifications decreased while P.
malariae/P. knowlesi notifications increased, although this relation-
ship was not statistically significant.
Using Sabah population estimates based on the 1991, 2000 and
2010 Population and Housing Censuses of Malaysia [23,24], the
incidences of P. falciparum and P. vivax peaked at 16.0 and 7.36/
1000 people/year respectively during 1994–1995, and decreased
to 0.18 and 0.19/1000 people respectively in 2011 (Figure 1.C). In
contrast the incidence of P. malariae/P. knowlesi peaked at 0.28/
1000 people in 1995, decreased to <0.02–0.04/1000 people from
2000–2006, and increased to 0.21/1000 people in 2011.
The relative proportions of the Plasmodium species changed
significantly over the past two decades, with P. falciparum, P. vivax
and P. malariae/P. knowlesi monoinfections accounting for 70%,
Figure 1. Malaria trends in Sabah, 1992–2011. A. Annual malaria notifications by species 1992–2011; B. Annual malaria notifications by species2001–2011; C. Malaria incidence by species 2001–2011; D. Annual notifications of species as a percentage of total malaria notifications, 1992–2011.*Population projections based on adjusted 2000 and 2010 population [24].doi:10.1371/journal.pntd.0002026.g001
Increasing Incidence of P. knowlesi in Sabah
PLOS Neglected Tropical Diseases | www.plosntds.org 3 January 2013 | Volume 7 | Issue 1 | e2026
24% and 1% respectively of total malaria notifications in 1992,
compared to 30%, 31% and 35% in 2011 (Figure 1.D). A total of
4.4% of all malaria notifications were mixed-species infections,
with this percentage increasing slightly over the years from a
median of 3.98% from 1992–2001 to 5.20% from 2002–2011
(p = 0.049).
Malaria trends by district, 1992–2011The 23 districts of Sabah (Figure 3) in general have experienced
similar malaria trends over the past two decades, with P. falciparum
and P. vivax notifications falling dramatically in all districts
(Figure 4). P. malariae/P. knowlesi notifications mostly remained at
low stable levels throughout the 1990s, accounting for ,5% of
total notifications in 87% of district-years from 1992–1999.
Exceptions included Tambunan from 1993–1994 and Beluran
from 1995–1998, where P. malariae/P. knowlesi accounted for 38/
255 (15%) and 701/6980 (10%) of malaria notifications respec-
tively, and Tenom from 1998–1999 and Tuaran in 1994 and
1999, where approximately 6% of malaria notifications were P.
malariae/P. knowlesi.
Since the early 2000s most districts have experienced an
increase in notifications of P. malariae/P. knowlesi (Figure 3 and
Figure 4.B). This increase appears to have begun initially in the
Interior Division, in the southwest of the state adjacent to
Sarawak, where notifications nearly doubled between 2003
(n = 28) and 2005 (n = 55), and more than doubled between
2005 and 2007 (n = 136), before increasing at a slower rate
through to 2011. In the West Coast Division to the northeast
notifications appear to have increased later, remaining below 20
per year from 2001–2006 and then increasing to 45 in 2007 and
102 in 2009. Continuing northeast to the tip of Borneo, Kudat
Division has experienced the most remarkable and recent
increase in P. malariae/P. knowlesi notifications, with cases
increasing from 2–11 per year from 2001–2007, to 106 in
2008, 245 in 2009, and 276 in 2011. In the eastern districts of
Sabah (Sandakan and Tawau Division) notifications of P.
malariae/P. knowlesi have been fewer, although have been
increasing since 2008.
In 2011 Kudat district accounted for the highest number of P.
malariae/P. knowlesi notifications (184, 26%), followed by Ranau
(121, 17%), Keningau (65, 9%), Tenom (62, 9%) and Kota
Marudu (52, 7.4%).
Age and sex distribution of malaria notificationsEpidemiological characteristics of notifications according to
species were assessed from 2007–2011, when relevant data were
recorded for each notification. This time period included 16,011
malaria notifications, although species was not recorded for 373
(2.3%). The overall median age of patients with P. malariae/P.
knowlesi (31 years) was significantly higher than that of patients with
P. vivax or P. falciparum (median ages 23 years for both, p = 0.001).
Males with P. malariae/P. knowlesi demonstrated an approximately
normal age distribution, with a mean, median and interquartile
range of 33, 30 and 20–45 years respectively (Figure 5). In contrast
females with P. malariae/P. knowlesi appeared to demonstrate a
bimodal age distribution, with local maxima at 9–12 and 50 years
(Figure 5). While most males (71%) with P. malariae/P. knowlesi
were between the ages of 15 and 50 years, with 13% of cases
occurring in children ,15 years and 17% occurring in adults .50
years, only half (50%) of female cases were aged 15–50 years, with
28% occurring in children ,15 years old and 24% occurring in
adults .50 years. Among adults ($15 years) with P. malariae/P.
knowlesi, females were significantly older than males (median age
43 years vs. 33 years, p,0.0001).
Among patients with P. vivax and P. falciparum the overall
median age was lower among females than it was among males
(median age 20 and 24 years for females and males respectively
with P. vivax, p,0.0001; and 17.5 and 24 years for females and
males respectively with P. falciparum, p,0.0001). As with P.
malariae/P. knowlesi however, adult females with P. vivax were older
than adult males (median ages 30 and 27 years respectively,
p = 0.002).
Figure 2. Ratio of P. malariae/P. knowlesi to P. falciparum notifications, 1992–2011. Pm/Pk= P. malariae/P. knowlesi, Pf = P. falciparum.doi:10.1371/journal.pntd.0002026.g002
Increasing Incidence of P. knowlesi in Sabah
PLOS Neglected Tropical Diseases | www.plosntds.org 4 January 2013 | Volume 7 | Issue 1 | e2026
The median age of all malaria patients increased progressively
from a median of 24 years in 2007 to 27 years in 2011 (Spearman’s
correlation coefficient 0.04, p,0.0001). The proportion of patients
.50 years old also increased, from 244/3191 (7.7%) in 2007, to
364/4135 (8.8%), 345/4009 (8.6%), 244/2644 (9.2%) and 263/
2032 (12.9%) in the years 2008, 2009, 2010 and 2011 respectively
(p,0.0001). Among patients .50 years old, P. malariae/P. knowlesi
cases as a proportion of all malaria notifications increased from
43/244 (17.6%) in 2007 to 131/263 (49.8%) in 2011 (p,0.0001).
A greater proportion of patients with P. malariae/P. knowlesi were
male (77% compared to 73% of patients with P. vivax and P.
falciparum, p = 0.0007), and this proportion increased among those
aged 15–60 years, of whom 82% were male, compared to 63%
outside this age range (p,0.0001).
Seasonal variationFrom 2007–2011 significant seasonality was demonstrated for
all Plasmodium species, with maximum notifications occurring in
July, April and June for P. falciparum (p = 0.0001), P. vivax
(p = 0.002) and P. malariae/P. knowlesi (p = 0.0001) respectively
(Figure 6).
Discussion
Although P. knowlesi is now well documented in Sabah, the
emergence of this species over time has not been previously
described. In this study, we found that while cases of P. knowlesi
(reported as ‘‘P. malariae’’) may have been prevalent at low levels
for decades, a significant increase in notifications has occurred
over the past decade. This increase follows a dramatic reduction in
notification rates of P. vivax and P. falciparum. In fact over the past
decade, a strong inverse correlation has occurred between
notification rates of ‘‘P. malariae/P. knowlesi’’ and P. falciparum.
Available evidence does not allow us to determine what
proportion of ‘‘P. malariae/P. knowlesi’’ notifications during the last
two decades is actually P. knowlesi, with PCR testing only instituted
at the Sabah State Reference Laboratory in 2005 [15], and no
PCR results available from Sabah blood samples prior to 2003 [4].
In the 1990s when prevalence of P. falciparum and P. vivax was high,
it is possible that a significant number of P. malariae cases also
occurred. However recent studies demonstrate that, at least since
2007, PCR-confirmed P. malariae in Sabah is rare. Although eight
cases of P. malariae were detected by PCR from 49 ‘‘P. malariae’’
blood films taken from Sabah during 2003–2005 (with six of these
from Kudat) [4], four subsequent studies identified only eight
(0.6%) PCR-confirmed P. malariae infections among 1286 patients
with PCR-confirmed Plasmodium infection in Sabah from 2007 to
2011 [6,7,15,25]. In one of these studies only four (0.8%) P.
malariae infections were identified from 475 patients with PCR-
confirmed Plasmodium infections in Kudat from 2009–2011,
including 365 with microscopy-diagnosed ‘‘P. malariae’’ [25]. In
another, P. malariae was detected by nested PCR in only two of 318
(0.6%) microscopy-diagnosed P. malariae cases referred to the
Sabah State Public Health Laboratory in 2009 [15]. Furthermore,
Figure 3. P. malariae/P. knowlesi notifications by division. Map shows districts and divisions of Sabah. Bar graphs show annual P. malariae/P.knowlesi notifications, by division, from 2001–2011. Population of Sabah Divisions in 2010 [37]: Interior 424,524; West Coast 1,011,725; Kudat 192,457;Sandakan 702,207; Tawau 819,955.doi:10.1371/journal.pntd.0002026.g003
Increasing Incidence of P. knowlesi in Sabah
PLOS Neglected Tropical Diseases | www.plosntds.org 5 January 2013 | Volume 7 | Issue 1 | e2026
Figure 4. Malaria notifications by district. A. Malaria notifications by district 1992–2011; B. Malaria notifications by district 2001–2011. Int.:Interior Division; W.C.: West Coast Division; Kud.: Kudat Division; Sand.: Sandakan Division; Taw.: Tawau Division.doi:10.1371/journal.pntd.0002026.g004
Figure 5. Age distribution of P. falciparum, P. vivax and P. malariae/P. knowlesi, 2007–2011.doi:10.1371/journal.pntd.0002026.g005
Increasing Incidence of P. knowlesi in Sabah
PLOS Neglected Tropical Diseases | www.plosntds.org 6 January 2013 | Volume 7 | Issue 1 | e2026
the age and sex distributions of ‘‘P. malariae/P. knowlesi’’
notifications since 2007 in the current study are very similar to
those described in a previous study in Kudat, in which 345 patients
with PCR-confirmed P. knowlesi were analysed [25]. Given the
unique age distribution of P. knowlesi, this strongly suggests that a
large majority of ‘‘P. malariae/P. knowlesi’’ notifications, at least
since 2007, are indeed P. knowlesi cases.
The reason for the older age group affected by P. knowlesi in this
and previous studies [5,7,25] remains unclear, however may relate
to greater forest exposure among older individuals, with farmers
and plantation workers over-represented in this age group [7]. The
bimodal age distribution of females affected by P. knowlesi requires
further investigation, but may possibly relate to lower forest
exposure among young adult females; this may also account for
the finding in this and other studies [7,25] that, among adults with
knowlesi malaria, females are older than males. Concurrent
zoonotic and human-human transmission may also explain a
bimodal age distribution.
There are several possible explanations for the emergence of
P. knowlesi. Firstly, increased recognition of the species may
account for increased reporting by microscopists. Although this
possibility cannot be excluded, the previous high prevalence
rates of malaria in Sabah ensured that microscopy skill levels
were maintained at high levels. It seems unlikely therefore that
large numbers of ‘‘P. malariae’’ slides would have been
misdiagnosed as P. vivax or P. falciparum. In fact, in a study
involving blood films obtained from 243 patients with PCR-
confirmed P. knowlesi in Sarawak between 2001–2006, only
4.5% and 6.6% were misdiagnosed by microscopy as P.falciparum and P. vivax respectively [4], and it is likely that a
majority of these blood films would have been reported prior to
the increased awareness of P. knowlesi. The consistency of
notification trends across districts further supports the overall
reliability of the microscopy reports and the State Department
records. Furthermore, the number and proportion of all malaria
patients aged .50 years increased significantly between 2007
and 2011. Given that this age group is over-represented among
patients with knowlesi malaria [7,25], this finding is consistent
with a true increase in the proportion of P. knowlesi cases and
cannot be attributed to increased recognition.
We believe, therefore, that the prevalence of P. knowlesi in Sabah
has increased, and that this has occurred as a result of
environmental change together with reducing rates of the other
human malaria species. The extensive deforestation that has
occurred in Sabah has led to encroachment of humans into
previously forested areas, resulting in increased interaction with
mosquito vectors and simian hosts. Furthermore, the removal of
habitat together with malaria control activities may have led to a
change in vector behaviour, or a vector shift, as has been seen in
the Kinabatangan region where the previously dominant malaria
vector An. balabacensis appears to have been displaced by An. donaldi[22]. Both these factors may increase the chance of human
acquisition of P. knowlesi, although further research regarding P.knowlesi vectors in Sabah is needed.
Finally, the finding in this study that the prevalence of P.knowlesiappears to have increased very recently, long after Sabah’s most
extensive period of deforestation during the 1970s and early 1980s
[17], suggests that decreasing rates of P. vivax and P. falciparum are
likely to have contributed directly to this trend. Possible
explanations for this may be derived from examining the
relationship between P. falciparum and P. vivax, as in other regions
prevalence of P. vivax has increased as rates of P. falciparum decrease
[26,27]. In addition, studies of P. vivax and P. falciparum have
demonstrated lower than expected rates of mixed infections [28]
and the occurrence of reciprocal seasonality between the two
species [29]. These observations suggest an inhibitory interaction
between P. falciparum and P. vivax, a phenomenon also demon-
strated in early syphilis studies in which P. falciparum was found to
suppress P. vivax parasitemia when both species were inoculated
simultaneously [30,31]. More recently, Bruce et al. reported that
asymptomatic children living in a highly endemic area demon-
strated relatively stable total parasite density counts despite
changes in the density of individual species, suggesting density-
dependent regulation that transcends species [28]. Similar
interactions between P. knowlesi and either P. falciparum or P. vivax
may explain the malaria trends in Sabah, with density-dependent
Figure 6. Monthly malaria notifications, 2007–2011.doi:10.1371/journal.pntd.0002026.g006
Increasing Incidence of P. knowlesi in Sabah
PLOS Neglected Tropical Diseases | www.plosntds.org 7 January 2013 | Volume 7 | Issue 1 | e2026
regulation possibly accounting for previously low rates of
symptomatic P. knowlesi. The occurrence of density-dependent
regulation may also explain the lack of earlier reports of severe ‘‘P.malariae’’, similar to reports from other regions that cases of severe
vivax malaria increased as the prevalence of P. falciparum reduced
[27].
In addition, it is possible that cross-species immunity may play a
role in the malaria prevalence patterns observed in Sabah.
Although heterologous immunity does not generally occur
between human malaria species, it has been argued that a degree
of cross-resistance may be more likely to occur between species
infecting different hosts [32]. In a study involving sera from
Gambian adults highly immune to P. falciparum, antibodies werefound to bind to the surface of P.knowlesi merozoites, although
erythrocyte invasion was not prevented [33]. In addition, data
from neurosyphilis malariotherapy series demonstrated that
patients who had been previously infected with P. vivax were less
susceptible to infection with P. knowlesi [34]. Loss of cross-
protection provided by immunity to P. falciparum or P. vivax may be
particularly relevant given that P. knowlesi tends to effect older
individuals; frequent exposure to P. falciparum and P. vivax may
previously have protected this age group from infection with P.knowlesi.The finding that notification rates of P. knowlesi have increased
following decreasing prevalence of the other malaria species has
implications for malaria control in any country where P. knowlesi isknown to occur, which includes nearly every country in Southeast
Asia [35]. In Sabah, P. knowlesi is now the most common cause of
malaria, and based on current trends, is likely to become
increasingly dominant and may extend to previously unaffected
districts. Furthermore, human-to-human transmission, if not
already occurring, may become more likely as prevalence
continues to increase. Close monitoring of P. knowlesi in Sabah
and elsewhere is therefore essential, including accurate reporting
of microscopy-diagnosed ‘‘P. malariae’’ as P. knowlesi, as has been
previously recommended [4,36], in addition to PCR-confirmation
of suspected cases. Moreover, further research is required to
determine the risk factors for knowlesi malaria, in order that
malaria control programs can include strategies to address the
increasing prevalence of this species. Although Malaysia has been
highly successful in reducing rates of P. falciparum and P. vivax,
malaria elimination will not be achieved unless control of knowlesi
malaria is addressed.
Acknowledgments
We thank the Ministry of Health, Malaysia, for permission to publish this
study.
Author Contributions
Conceived and designed the experiments: TW HAR JJ MYI JM MJG
TWY NMA BEB. Performed the experiments: BEB. Analyzed the data:
BEB. Wrote the paper: BEB NMA.
References
1. World Health Organization (2011) World Malaria Report 2011. Geneva: World
Health Organization.
2. Rundi C (2011) Malaria Elimination in Malaysia. Third annual APMEN
technical and business meeting, 9–12 May 2011; Kota Kinabalu, Malaysia.Available: http://apmen.org/apmen-iii-meeting-proceedings/. Accessed 3 May
2012.
3. Barber BE, William T, Jikal M, Jilip J, Dhararaj P, et al. (2011) Plasmodiumknowlesi malaria in children. Emerg Infect Dis 17: 814–820.
4. Cox-Singh J, Davis TM, Lee KS, Shamsul SS, Matusop A, et al. (2008)Plasmodium knowlesi malaria in humans is widely distributed and potentially life
threatening. Clin Infect Dis 46: 165–171.
5. Daneshvar C, Davis TM, Cox-Singh J, Rafa’ee M, Zakaria S, et al. (2009)Clinical and laboratory features of human Plasmodium knowlesi infection. ClinInfect Dis 49: 852–860.
6. Joveen-Neoh WF, Chong KL, Wong CM, Lau TY (2011) Incidence of malaria
in the Interior Division of Sabah, Malaysian Borneo, based on nested PCR.
J Parasitol Res 2011: DOI: 10.1155/2011/104284.
7. Barber BE, William T, Grigg MJ, Menon J, Auburn S, et al. (2012) A
prospective comparative study of knowlesi, falciparum and vivax malaria inSabah, Malaysia: high proportion with severe disease from Plasmodium knowlesiand P. vivax but no mortality with early referral and artesunate therapy. Clin
Infect Dis, first published online October 19: doi:10.1093/cid/cis902.
8. Chin W, Contacos P, Coatney R, Kimball H (1965) A naturally acquired
quotidian-type malaria in man transferable to monkeys. Science 149: 865.
9. Fong Y, Cadigan F, Coatney G (1971) A presumptive case of naturally occurring
Plasmodium knowlesi malaria in man in Malaysia. Trans R Soc Trop Med Hyg 65:
839–840.
10. Singh B, Sung LK, Matusop A, Radhakrishnan A, Shamsul SSG, et al. (2004) A
large focus of naturally acquired Plasmodium knowlesi infections in human beings.Lancet 363: 1017–1024.
11. Lee K-S, Divis PCS, Zakaria SK, Matusop A, Julin RA, et al. (2011) Plasmodiumknowlesi: reservoir hosts and tracking the emergence in humans and macaques.PLoS Pathog 7: e1002105.
12. de Zulueta J (1956) Malaria in Sarawak and Brunei. Bull World Health Organ15: 651.
13. Lee K-S, Cox-Singh J, Brooke G,Matusop A, Singh B (2009) Plasmodium knowlesi fromarchival blood films: Further evidence that human infections are widely distributedand not newly emergent in Malaysian Borneo. Int J Parasitol 39: 1125–1128.
14. Singh B (2011) P. knowlesi and the elimination agenda. Third annual APMENtechnical and business meeting, 9–12 May 2011; Kota Kinabalu, Malaysia.
Available: http://apmen.org/apmen-iii-meeting-proceedings/. Accessed 3 May
2012.
15. Naing DKS, Anderios F, Lin Z (2011) Geographic and ethnic distribution of Pknowlesi infection in Sabah, Malaysia. International Journal of CollaborativeResearch on Internal Medicine & Public Health 3: 391–400.
16. Department of Statistics, Malaysia (2011) Population Distribution and Basic
Demographic Characteristics, Population and Housing Census of Malaysia2010. Kuala Lumpur: Department of Statistics, Malaysia.
17. McMorrow J, Talip MA (2001) Decline of forest area in Sabah, Malaysia:Relationship to state policies, land code and land capability. Global
Environmental Change 11: 217–230.
18. Sabah Forestry Department. Geography of Sabah. Available: http://www.forest.sabah.gov.my/resources/geography.asp. Accessed 25th August 2012.
19. World Wildlife Foundation (2005) Borneo: treasure island at risk: World WildlifeFoundation Germany.
20. Vector Borne Disease Control Program, Ministry of Health Malaysia (2008)
Vector control in Malaysia. Consultative meeting/workshop on management ofvector control programs, 11–15 August 2008; Siem Reap, Cambodia. Available:
http://www.actmalaria.net/files/vector_control/vc_policy/vc_malaysia.pdf.Accessed 10 December 2012.
21. Asian Collaborative Training Network for Malaria. Malaysia-Country Profile.
Available: http://www.actmalaria.net/downloads/pdf/info/2008/Malaysia%20Malaria%20Country%20Profile%202008.pdf. Accessed 25 August
2012.
22. Vythilingam I, Chan S, Shanmugratnam C, Tanrang H, Chooi KH (2005) The
impact of development and malaria control activities on its vectors in the
Kinabatangan area of Sabah, East Malaysia. Acta Trop 96: 24–30.
23. Department of Statistics, Malaysia. Available: http://www.statistics.gov.my/
portal/index.php?option = com_content&view = article&id = 363&Itemid =169&lang = en. Accessed June 18th 2012.
24. Sabah Department of Statistics (2008) Revised intercensual mid-year population
estimates by states, Malaysia 1992–1999. Kuala Lumpur: Sabah Department ofStatistics.
25. Barber BE, William T, Dhararaj P, Anderios F, Grigg MJ, et al. (2012)Epidemiology of Plasmodium knowlesi malaria in northeast Sabah, Malaysia:
family clusters and wide age distribution. Malar J 11: 401.
26. Nosten F, Van Vugt M, Price R, Luxemburger C, Thway K, et al. (2000) Effectsof artesunate-mefloquine combination on incidence of Plasmodium falciparummalaria and mefloquine resistance in western Thailand: a prospective study.Lancet 356: 297–302.
27. Oliveira-Ferreira J, Lacerda MVG, Brasil P, Ladislau JLB, Tauil PL, et al.
(2010) Malaria in Brazil: an overview. Malar J 9: 115.
28. Bruce MC, Donnelly CA, Alpers MP, Galinski MR, Barnwell JW, et al. (2000)
Cross-species interactions between malaria parasites in humans. Science 287:845–848.
29. Maitland K, Williams T, Bennett S, Newbold C, Peto T, et al. (1996) The
interaction between Plasmodium falciparum and P. vivax in children on EspirituSanto island, Vanuatu. Trans R Soc Trop Med Hyg 90: 614–620.
30. Boyd MF, Kitchen S (1937) Simultaneous inoculation with Plasmodium vivax andPlasmodium falciparum. Am J Trop Med Hyg 1: 855–861.
Increasing Incidence of P. knowlesi in Sabah
PLOS Neglected Tropical Diseases | www.plosntds.org 8 January 2013 | Volume 7 | Issue 1 | e2026
31. Mayne B, Young M (1938) Antagonism between species of malaria parasites in
induced mixed infections. Public Health Rep 53: 1289–1291.32. Richie T (1988) Interactions between malaria parasites infecting the same
vertebrate host. Parasitology 96: 607–639.
33. Miller LH, Johnson JG, Schmidt-Ullrich R, Haynes J, Wallach D, et al. (1980)Determinants on surface proteins of Plasmodium knowlesi merozoites common to
Plasmodium falciparum schizonts. The Journal of Experimental Medicine 151: 790–798.34. van Rooyen CE, Pile G (1935) Observations on infection by Plasmodium knowlesi (ape
malaria) in the treatment of general paralysis of the insane. Br Med J 2: 662–666.
35. Cox-Singh J (2012) Zoonotic malaria: Plasmodium knowlesi, an emerging pathogen.
Curr Opin Infect Dis 25: DOI: 10.1097/OCO.1090b1013e3283558780.36. Rajahram G, Barber BE, William T, Menon J, Anstey NM, et al. (2012) Deaths
due to Plasmodium knowlesi malaria in Sabah, Malaysia: association with reporting
as P. malariae and delayed parenteral artesunate. Malar J 11: 284.37. Department of Statistics, Malaysia. Population by Administrative District,
Sabah . Ava i lab le : h t tp ://www.s ta t i s t i c s .gov .my/porta l/ index .php?option= com_wrapper&view=wrapper&Itemid= 129. Accessed 18 June
2012.
Increasing Incidence of P. knowlesi in Sabah
PLOS Neglected Tropical Diseases | www.plosntds.org 9 January 2013 | Volume 7 | Issue 1 | e2026
63
Implications of and questions arising from this paper
In this study we investigated the apparent emergence of human cases of P. knowlesi in
Sabah over time. We found that while cases of P. falciparum and P. vivax malaria have
fallen markedly over the last two decades, notifications of “P. malariae/P. knowlesi” have
increased >10-fold over the last decade, and in 2011 accounted for a greater number of
notifications than P. falciparum or P. vivax.
This study however was associated with significant limitations. Most importantly, Sabah
Health Department notification records were based on microscopic diagnoses, and the exact
proportion of P. knowlesi cases among “P. malariae/P. knowlesi” notifications is therefore
uncertain. Furthermore, the possibility that the increasing notification rate of P. knowlesi is
due to increased recognition of this species cannot be discounted. However, recent studies
have reported that a large proportion of ‘P. malariae/P. knowlesi” cases are confirmed to be
P. knowlesi when PCR methods are used, and that PCR-confirmed P. malariae in Sabah is
rare (25, 26, 171). In addition, in this paper we argue that, given the older age group
affected by P. knowlesi, the recent increase in the age of all malaria notifications is
consistent with a true increase in cases of P. knowlesi malaria.
The explanation for this marked and recent increase in incidence of P. knowlesi in Sabah
remains uncertain. In this paper we postulate that it has occurred in part as a result of
environmental change, which has led to increased interaction between humans, simian
hosts and mosquito vectors. Alternatively, it is possible that there has been a change in the
P. knowlesi vector, as a result of habitat removal together with malaria control activities, or
that there has been a change in behavior or migration of the simian hosts. Studies involving
the P. knowlesi vector or the macaque hosts were not included in this thesis, and are
required to determine whether these factors have influenced the changing epidemiology of
knowlesi malaria.
64
In addition, we suggest that the very recent increase in incidence of P.knowlesi may indicate
that decreasing rates of P. vivax and P. falciparum have contributed directly to this trend,
possibly as a result of loss of cross-protective immunity. This could also explain the
relatively high rates of knowlesi malaria in Malaysian Borneo, where prevalence of P. vivax
and P. falciparum is now low, compared to other southeast Asian countries where
prevalence of these latter species is higher and P. knowlesi is less common (7, 8, 29, 61,
68).
Finally, the recent increase in human cases of P. knowlesi may represent an adaption of the
parasite to human hosts. In a recent paper Lim et al. demonstrated that a simian strain of P.
knowlesi had the ability to adapt to human blood, invading blood cells of all ages after
several months of continuous culture in human blood, whereas initially invasion had been
restricted to young red cells (166). Mathematical modeling demonstrated that this adaption
would result in substantially higher parasite counts, which presumably would be associated
with a higher chance of human to human transmission.
The uncertainty around these issues highlights the need for further research, including,
ideally, confirming the increasing rate of P. knowlesi in Sabah by performing molecular
methods of diagnosis on all malaria cases over several years. In addition, further research
is required to investigate the interaction between P. knowlesi and the other human malaria
species, and the possibility of human-human transmission.
65
8.1 Background
Following the seminal paper by Singh et al. documenting a large focus of human P. knowlesi
cases in Kapit, Sarawak (3), increased surveillance for cases of knowlesi malaria began at
Queen Elizabeth Hospital (QEH) in Sabah, with PCR increasingly conducted on malaria
slides diagnosed as “P. malariae”. Consistent with the report from Sarawak, the majority of
“P. malariae” cases admitted to QEH were found to be P. knowlesi when tested by PCR. In
addition, a significant proportion of these patients were noted to have severe disease, and
several patients had died. At this time published reports of severe knowlesi malaria were
limited. Only one prospective study of knowlesi malaria had been conducted, and in this
district hospital study, 10 (10%) of 107 patients with knowlesi malaria had severe disease,
and 2 deaths occurred (24). An additional five fatal cases had been described in two reports
(23, 27).
QEH, being a tertiary referral hospital, provided an opportunity to describe in more detail the
clinical features of patients with severe knowlesi malaria, and in 2010 William et al.
conducted a retrospective review of 56 patients with PCR-confirmed knowlesi malaria
admitted to QEH during 2007 – 2009 (31). In keeping with the referral-hospital setting,
William et al. found higher rates of severe disease than had been reported in the prospective
district hospital study, with severe disease occurring in 22 (39%) of 56 patients hospitalized
with knowlesi malaria, and with 6 (27%) of these patients dying. This retrospective study
also reported for the first time the use of intravenous artesunate for the treatment of severe
knowlesi malaria, and the use of oral artemisinin therapy for the treatment of uncomplicated
knowlesi malaria. These regimens were introduced to QEH following hospital policy change
66
in December 2008. Although an observational study, with no randomization, patients with
severe knowlesi malaria treated with intravenous artesunate were found to have faster
parasite clearance than those treated with intravenous quinine, and fewer patients treated
with intravenous artesunate died. Among patients with uncomplicated knowlesi malaria,
parasite clearance times were faster with artemether-lumefantrine than they were with
chloroquine or quinine. Following this study, intravenous artesunate became the
recommended treatment for all patients with severe knowlesi malaria at QEH and at district
hospitals within the QEH catchment area. In addition, clinicians at these district hospitals
were encouraged to refer to QEH all patients with symptoms or signs suggestive of severe
malaria.
While this study provided important additional information regarding the clinical features of
patients with severe knowlesi malaria, the clinical descriptions were limited by the
retrospective nature of the study. In particular, parasite counts had not been quantitated,
and so risk of severe disease according to parasitemia was unable to be assessed. In
addition, the clinical features of patients with knowlesi malaria were not compared to those of
patients with falciparum or vivax malaria. While Daneshvar et al. included patients with
falciparum and vivax malaria in their prospective study, these cases were mostly imported
and hence could not be directly compared to patients with locally-acquired knowlesi malaria
(24). Hence while the clinical and laboratory findings of naturally acquired P. knowlesi were
described prospectively for the first time in this study, comparative descriptions of knowlesi,
falciparum and vivax malaria were not possible.
From September 2010 to October 2011 we therefore conducted a prospective comparative
study of all patients admitted to QEH with a microscopy diagnosis of malaria.
67
8.2 Study Site
QEH is located in Kota Kinabalu, which is the capital of the north-eastern Malaysian state of
Sabah. Sabah has an area of 73,600 km2 and a population of 3.2 million (173). Situated
between 4o and 7o north of the equator, Sabah has a mostly tropical climate, with high
humidity and rainfall throughout the year and temperatures of 25-35oC. The southwest
interior of Sabah is mountainous, with the Crocker Range separating west coast lowlands
from the rest of the state and extending north to Mount Kinabalu at 4095 meters above sea
level. Sabah was previously covered almost entirely in dense primary rainforest, however
extensive deforestation occurred throughout the 1970s and 1980s, reducing forest cover to
44 - 63% of the state (174-176). Cleared areas have been partly replaced by plantations,
with palm oil estates comprising 16% of Sabah’s land area (176).
Sabah is comprised of 5 divisions: the Interior Division in the southwest of the state adjacent
to Sarawak; the West Coast Division; the Kudat Division in the northeast; and the Sandakan
and Tawau Divisions in the east of the state (Figure 2). The capital Kota Kinabalu is located
in the most populous West Coast Division. QEH, a 600 bed adult tertiary-referral hospital
with modern intensive care facilities, services all of this Division as well as the Kudat
Division. These two Divisions have a population of 1.14 million, and contain 6 district
hospitals which are located up to 3 hour’s drive from QEH.
68
Figure 7. Map of Sabah showing Districts and Divisions
Kudat Division
West Coast Division Sandakan Division
Tawau Division Interior Division
69
8.3 Patient enrolment and study procedures
Malaria patients in this study were referred to our research team from the QEH emergency
department, from hospital wards if already admitted, or from primary care clinics within the
Kota Kinabalu region. In addition, patients suspected of having severe malaria were referred
from the 6 district hospitals within the West Coast and Kudat Divisions. During the study
period, our research team was in frequent contact with clinicians at district hospitals to
ensure that this referral procedure was adhered to. In addition, we regularly visited the QEH
emergency department and medical wards to ensure that we were being notified of all
malaria patients. We also regularly surveyed the QEH laboratory microscopy record book to
ensure that there no positive malaria slides that we had not been notified of.
Consistent with Sabah Ministry of Health policy, all malaria patients were admitted to
hospital and discharged only after having 2 negative blood smears on two consecutive days.
All malaria patients were assessed by myself for eligibility, and enrolled into the study if they
were within 18 hours of having commenced antimalarial treatment, did not have significant
comorbidities or concurrent illness, and were not pregnant. On enrolment all patients
underwent a detailed history and examination (by myself), and had blood taken for baseline
haematological and biochemical parameters. Lactate and acid-base parameters were
assessed at the bedside using iSTAT. Thick and thin blood films were prepared for all
patients on enrolment, and for those who had already commenced treatment, a pre-
treatment blood film or EDTA blood sample was retrieved from the QEH or referring hospital
laboratory. All patients were assessed daily until discharge (by me). Blood films and full
blood count were performed daily, with additional investigations as clinically indicated.
70
Figure 8. Queen Elizabeth Hospital Emergency Department
Figure 9. Lingzhi Infectious Diseases Ward
Figure 10. Beatrice Wong processing bloods at Lingzhi
Figure 11. Bridget Barber reviewing patients on the medical ward, Queen Elizabeth Hospital
71
8.4 Significance of the paper
This paper represents the largest prospective study of the clinical and laboratory features of
knowlesi malaria of any severity, and the largest series of severe knowlesi malaria reported
to date. It is also the only study to date to compare in detail and in the same population, the
clinical and biochemical features of P. knowlesi, P. vivax and P. falciparum. This study
therefore provided the opportunity to assess independent risk factors for severity among
patients with knowlesi malaria, and to compare these to those of the other human malaria
species. This study is also the first prospective study to report the early efficacy of oral
artemisinin therapy for the treatment of uncomplicated knowlesi malaria, and intravenous
artesunate for the treatment of severe knowlesi malaria.
Important findings from this study include:
1. The use of intravenous artesunate for the treatment of severe knowlesi malaria was
highly effective. Our study was conducted at a tertiary referral hospital and a high rate of
severity was seen among patients with knowlesi malaria, with 38 (29%) of 130 patients
having severe disease. Despite this, no deaths occurred in our study. Although our study
was not a controlled trial, intravenous artesunate nonetheless appeared effective for the
treatment of severe knowlesi malaria. This, together with earlier findings from the QEH
retrospective study (31), have been incorporated into the 2012 WHO guidelines for the
management of severe malaria, which now recommends intravenous artesunate for the
treatment of all severe malaria regardless of species (110).
2. Among patients with knowlesi malaria, parasitemia is the major independent risk
factor for severity. In our study a parasite count of >20,000 parasites/μL was associated
with an 11-fold increased risk of severity, while a parasite count of >100,000 parasites/μL
was associated with a 29-fold increased risk of severe disease. Only 3 of 18 patients with
72
parasite counts >100,000/μL did not have other criteria of severe malaria. Our findings are
consistent with those reported by Wilman et al., who in a case-control study published
around the same time as our study, found a parasite count of >35,000/μL to be associated
with a 10-fold increased risk of severity (102). This finding has important implications for the
management of patients with knowlesi malaria. As a minimum, it suggests that patients with
a parasite count >100,000/μL are at high risk of complications, and should be closely
monitored and treated with immediate intravenous artesunate, as per WHO guidelines for
the treatment of severe malaria. However, the treatment of patients with P. knowlesi
parasite counts <100,000/μL is less certain. Oral artemisinin therapy is associated with
rapid parasite clearance of other Plasmodium species, and in our study appeared highly
effective for the treatment of non-severe knowlesi malaria. However, a large proportion of
patients with non-severe knowlesi malaria in our study received one or more doses of
intravenous artesunate, particularly those referred from district hospitals with “4+” but
otherwise uncomplicated malaria. Hence the efficacy of oral artemisinin therapy alone could
not be adequately assessed in our study, and requires further research. This uncertainty
regarding the optimal management of patients with P. knowlesi parasitemia of <100,000/μL
is reflected in the latest WHO guidelines for the management of severe malaria (110).
These guidelines state that patients with knowlesi malaria should be treated with parenteral
therapy if they have any features of severe disease, or if they have a parasitemia
>100,000/μL. In addition, the guidelines recommend that patients with parasitemia
>20,000/μL should be treated with parenteral therapy “if testing for laboratory criteria for
severe malaria is not readily available”. In our study, five patients with knowlesi malaria and
parasitemia >20,000/μL (range 25,873/μL – 60.840/μL) were treated with oral artemisinin
therapy alone with good outcomes, including one patient with jaundice (bilirubin 48 μmol/L)
as a sole severity criterion.
73
3. Among patients with knowlesi malaria, increasing age is strongly correlated with
increasing parasitemia. Although previous studies have demonstrated an association
between age and risk of severity (24, 31), our study demonstrates that this association is
accounted for by the strong association between age and P. knowlesi parasitemia. Patients
≥50 years old had a median parasitemia of 17,834/μL, compared to a median of 5430/μL
among patients <50 years old. Accordingly, 53% of patients ≥50 years old had severe
disease, compared to only 15% of those <50 years old. While an association between age
and increasing incidence of hyperparasitemia has also been reported among Asian patients
with severe falciparum malaria (104), the explanation for this association has not been
determined.
In our study, it is possible that epidemiological factors contribute to the higher parasitemias
seen in the older age groups. Farmers and plantation workers were overrepresented among
the older age groups, and these occupations were also independently associated with a
higher parasite count. The explanation for this finding however remains uncertain. No
difference occurred in fever duration according to age group, or farming/plantation work,
suggesting that treatment-seeking behavior did not account for the higher parasitemias seen
in these patients. Alternatively, physiological factors may account for the higher parasitemias
seen among older patients, such as an impaired inflammatory response to P. knowlesi
parasitemia among older individuals, the influence of greater prior exposure in this age
group in the distant past to non-knowlesi Plasmodium spp, or an age-related effect on the
ability of P. knowlesi to replicate within human blood. Further studies are required to
investigate these factors.
4. P. knowlesi malaria was associated with an increased risk of severe malaria
compared to P. falciparum. Our finding that a greater proportion of patients with knowlesi
malaria had severe disease compared to those with falciparum malaria was influenced by
74
the fact that, due to differences in the geographical distribution of P. knowlesi and P.
falciparum malaria, the majority of patients with knowlesi malaria were referred from district
hospitals whereas most patients with falciparum malaria presented directly to the QEH
emergency department or were referred from primary care clinics nearby the hospital.
However, in multivariate analysis controlling for district-hospital referral, P. knowlesi was still
associated with a 3-fold increased risk of severe malaria compared to P. falciparum.
Pathogenic mechanisms of disease in knowlesi malaria have not been investigated in detail,
but are likely to differ from those of falciparum malaria. Further studies are required to
investigate the comparative pathophysiology of these two species, in order to identify
conserved and species-specific mechanisms of disease which may account for differences
in severity and different clinical syndromes. In addition, studies investigating the occurrence
of human-human transmission of knowlesi malaria are required, as this may also influence
pathogenicity. In early syphilis studies increased passage of P. knowlesi through humans
was associated with increasing parasitemias and virulence of infection (45, 46), while Lim et
al. recently reported that after continuous culture in human blood P. knowlesi demonstrated
an ability to adapt and invade red blood cells of all ages, leading to increased parasite
densities (166).
A limitation of our study was that blood cultures were not performed in all patients, and
hence we cannot exclude the fact that co-infection may have been present in some patients
and may confounded our results. However, blood cultures were performed in 7 of 13
patients with severe falciparum malaria, and 24 of 38 patients with severe knowlesi malaria,
and all were negative. It is therefore unlikely that a significant proportion of patients with
falciparum or knowlesi malaria were bacteremic. Further studies evaluating blood culture
results in a larger number of patients with knowlesi and falciparum malaria are ongoing, and
will assist with evaluating the prevalence of bacteremia among these patients.
75
5. P. vivax was associated with a high risk of severity. In this study, disease meeting
modified 2010 WHO criteria for severity occurred among 7 (16%) of 43 patients with vivax
malaria. Hypotension was the most common severity criterion, occurring in 5 patients, while
jaundice, respiratory distress, abnormal bleeding and multiple convulsions also occurred. In
contrast to patients with severe knowlesi and falciparum malaria who had a median of 2 and
3 severity criteria, respectively, patients with vivax malaria had a median of 1 severity
criterion (p=0.04 by Kruskal-Wallis test). A major limitation of the assessment of patients
with severe vivax malaria in our study was that only 3 of 7 patients had pre-antibiotic blood
cultures performed. Blood cultures were negative in two of these patients, and positive for
Streptococcus pneumoniae in the third patient. Four patients did not have pre-antibiotic
blood cultures performed, and concurrent bacterial infection could therefore not be excluded
and may have contributed to the complications occurring in these patients. Nevertheless,
even if concurrent bacterial infection was common among these cases, this would not
necessarily indicate that the P. vivax parasitemia was incidental and unrelated to disease
severity. In falciparum malaria bacterial infections have been shown to be biologically
associated with severe disease (177), and similar processes may occur in severe disease in
association with P. vivax.
A relatively high risk of severe disease among patients with vivax malaria is supported by
other recent case series, particularly from Papua New Guinea (178, 179), Indonesia (180),
India (181), and Brazil (182, 183), that also document high rates of complications among
adults and children with vivax malaria. While severe anaemia (178, 180-182, 184, 185) and
respiratory distress (166, 168-170, 173-176) are the most consistently reported
complications in these case series, shock (181, 182, 185), jaundice (181, 182, 185), acute
kidney injury (181, 182, 185), and coma (178, 180, 181, 185, 186) have also been reported.
As with our study, many of these case series are limited by incomplete investigation of
76
concurrent infections and other comorbidities, which may contribute to the complications
reported (187).
77
In the previous chapter, the clinical features of 295 patients admitted to QEH with PCR-
confirmed Plasmodium monoinfection were described, including 130 with P. knowlesi, 122
with P. falciparum, and 43 with P. vivax. Although the proportion of patients with severe
disease was high, particularly among patients with knowlesi malaria, no deaths occurred in
our study.
However, during 2010-2011, 14 malaria deaths were reported in Sabah. The following
paper presents a case-note review of these 14 fatal cases, comprising seven patients with
falciparum malaria, six with knowlesi malaria and one patient with vivax malaria.
RESEARCH Open Access
Deaths due to Plasmodium knowlesi malaria inSabah, Malaysia: association with reporting asPlasmodium malariae and delayed parenteralartesunateGiri S Rajahram1, Bridget E Barber1,2, Timothy William1,3,4*, Jayaram Menon3,4, Nicholas M Anstey2,5
and Tsin W Yeo2,5
Abstract
Background: The simian parasite Plasmodium knowlesi is recognized as a common cause of severe and fatalhuman malaria in Sabah, Malaysia, but is morphologically indistinguishable from and still commonly reported asPlasmodium malariae, despite the paucity of this species in Sabah. Since December 2008 Sabah Department ofHealth has recommended intravenous artesunate and referral to a general hospital for all severe malaria cases ofany species. This paper reviews all malaria deaths in Sabah subsequent to the introduction of these measures.Reporting of malaria deaths in Malaysia is mandatory.
Methods: Details of reported malaria deaths during 2010-2011 were reviewed to determine the proportion of eachPlasmodium species. Demographics, clinical presentations and management of severe malaria caused by eachspecies were compared.
Results: Fourteen malaria deaths were reported, comprising seven Plasmodium falciparum, six P. knowlesi and onePlasmodium vivax (all PCR-confirmed). Of the six P. knowlesi deaths, five were attributable to knowlesi malaria andone was attributable to P. knowlesi-associated enterobacter sepsis. Patients with directly attributable P. knowlesideaths (N = 5) were older than those with P. falciparum (median age 51 [IQR 50-65] vs 22 [IQR 9-55] years, p = 0.06).Complications in fatal P. knowlesi included respiratory distress (N = 5, 100%), hypotension (N = 4, 80%), and renalfailure (N = 4, 80%). All patients with P. knowlesi were reported as P. malariae by microscopy. Only two of fivepatients with severe knowlesi malaria on presentation received immediate parenteral anti-malarial treatment. Thepatient with P. vivax-associated severe illness did not receive parenteral treatment. In contrast six of seven patientswith severe falciparum malaria received immediate parenteral treatment.
Conclusion: Plasmodium knowlesi was responsible, either directly or through gram-negative bacteraemia, for almosthalf of malaria deaths in Sabah. Patients with severe non-falciparum malaria were less likely to receive immediateparenteral therapy. This highlights the need in Sabah for microscopically diagnosed P. malariae to be reported asP. knowlesi to improve recognition and management of this potentially fatal species. Clinicians need to be betterinformed of the potential for severe and fatal malaria from non-falciparum species, and the need to treat all severemalaria with immediate intravenous artesunate.
Keywords: Malaria, Plasmodium knowlesi
* Correspondence: [email protected] Diseases Department, Queen Elizabeth Hospital, Karung BerkunciNo. 2029, Jalan Penampang, Kota Kinabalu, 88560, Sabah, Malaysia3Department of Medicine, Queen Elizabeth Hospital, Kota Kinabalu, Sabah,MalaysiaFull list of author information is available at the end of the article
© 2012 Rajahram et al.; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the CreativeCommons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, andreproduction in any medium, provided the original work is properly cited.
Rajahram et al. Malaria Journal 2012, 11:284http://www.malariajournal.com/content/11/1/284
BackgroundThe simian parasite Plasmodium knowlesi is commonlymisdiagnosed as Plasmodium malariae by microscopydue to its near-identical appearance. However, in contrastto the relatively benign clinical course of P. malariae,P. knowlesi is now recognized as a common cause of se-vere and fatal human malaria in Malaysian Borneo [1].Cases have also been reported in West Malaysia [2] andnearly all countries in Southeast Asia [3-10]. In Borneo,P. knowlesi mono-infection accounted for 64% of malariaadmissions in Kapit, Sarawak [11], and 78% of malariaadmissions in Kudat, Sabah [12]. Severe disease has beenreported from Sarawak [1,11], Sabah [13,14], and WestMalaysia [2], including 13 fatal cases [1,11,13,14]. In aretrospective study conducted from December 2007 toNovember 2009 at Queen Elizabeth Hospital (QEH), a ter-tiary referral hospital in Sabah, 22/56 (39%) patients ad-mitted with PCR-confirmed knowlesi malaria had severedisease by WHO criteria, and six (27%) died [14].The 24-hour replication cycle of P. knowlesi may be
associated with rapid increases in parasitaemia and con-sequent complications, and hence prompt diagnosis andinitiation of effective treatment is essential. The optimaltreatment however has not been determined. Whilechloroquine was shown to be effective for uncompli-cated knowlesi malaria in Kapit [15], the retrospectivestudy at QEH found faster parasite clearance timeswith an oral artemisinin combination therapy (ACT),artemether-lumefantrine [14]. Among patients with severeknowlesi malaria, parasite clearance times were faster withintravenous artesunate than with intravenous quinine, andfewer patients who received artesunate died [14].In Sabah, intravenous artesunate has been the recom-
mended treatment for severe malaria from any speciessince December 2008. In addition, referral to a generalhospital is advised for patients with symptoms or signssuggestive of severe disease. Despite these measures, 14deaths from malaria were reported in Sabah during2010-2011. In this study the case notes of all thesepatients were reviewed to determine the Plasmodiumspecies causing fatal malaria in Sabah, and to identifyany notable differences in demographics, clinical featuresand management of fatal malaria caused by the differentspecies.
MethodsSettingThe north-eastern Malaysian state of Sabah has an areaof 73,600 km2 and a population of 3.2 million [16]. Sit-uated between 4° and 7° above the equator, Sabah has amostly tropical climate, with high humidity and rainfallthroughout the year and temperatures of 25-35°C. Ma-laria incidence is estimated at 0.78/1000 persons/year[17]. Sabah’s government health system comprises one
tertiary-referral hospital offering specialist and sub-specialist care, three general hospitals offering specialistcare, 18 district hospitals, 77 health clinics and 189 ruralclinics.
Case seriesAll deaths due to malaria in Sabah must be reported tothe Sabah Ministry of Health where they are reviewed.Details of reported malaria deaths during 2010-2011were obtained from the Sabah Department of Health.Approval to review the case notes of fatal malaria caseswas obtained from the Medical Research sub-Committeeof the Malaysian Ministry of Health and the HealthResearch Ethics Committee of Menzies School of HealthResearch. Case notes were retrieved from district hospitalsand reviewed for clinical details, laboratory results andcause of death.Blood slides for malaria parasites were reported accor-
ding to a scale of 1+ to 4+ (1 + = 1-10 parasites/100 highpower microscopy fields [HPMFs] or� 4-40 parasites/μL,2 + = 11-100 parasites/100 HPMFs or� 41-400 para-sites/μL, 3 + = 1-10 parasites/HPMF or� 401-4,000parasites/μL, and 4+ = >10 parasites/HPMF or >4,000parasites/μL). PCR was performed by the Sabah StateReference Laboratory by methods previously published[1,18]. The diagnosis of all fatal malaria cases was con-firmed by PCR. Laboratory investigations and clinicaldetails are listed in Table 1.
ResultsFourteen deaths were reported during 2010-2011, in-cluding five due to P. knowlesi mono-infection, oneP. knowlesi-associated fatality in which gram-negativeseptic shock was thought to be the primary cause ofdeath, one death associated with P. vivax and sevenfrom P. falciparum.
Cases 1-5: fatal cases primarily attributed to PCR-confirmed plasmodium knowlesi monoinfectionCase oneA 71-year-old man with a history of hypertension andchronic obstructive pulmonary disease presented toa district hospital with a seven-day history of feverand dyspnoea, and reduced conscious state just priorto presentation. His Glasgow Coma Score (GCS) wasrecorded on arrival by a paramedic as 3/15 and shortlyafterwards as 10/15; his blood pressure was 75/49 mmHgand his oxygen saturation was 75% on room air. Chestauscultation noted prolonged expiratory phase butno crackles or wheeze. Blood film was reported asPlasmodium falciparum/P. malariae “3+”, and renalfailure was present. Chest radiograph showed no infil-trates, and arterial blood gas was not performed. He wascommenced on intravenous fluids and quinine in addition
Rajahram et al. Malaria Journal 2012, 11:284 Page 2 of 7http://www.malariajournal.com/content/11/1/284
to oral primaquine, chloroquine and doxycycline. His oxy-gen saturation and blood pressure continued to deterio-rate and he suffered a cardiac arrest 40 minutes afterpresentation. PCR confirmed P. knowlesimono-infection.
Case twoA 65-year-old man with a history of type 2 diabetes andhypertension presented to a district hospital with a two-day history of fever, lethargy, myalgia and retro-orbitalpain. On examination he was alert and orientated but
hypoxic with an oxygen saturation of 92% on room air.Initial blood film was reported as negative for malariaparasites, and the patient was given a provisional diagno-sis of dengue fever and commenced on intravenousfluids. On day 2 a repeat blood film was reported asP. malariae “3+” and oral chloroquine was started. Thefollowing day the patient was noted to be tachypnoeicand hypoxic (oxygen saturation 85% on room air), andchest radiograph showed diffuse infiltrates. Emergency in-tubation was performed however cardiac arrest occurred
Table 1 Demographic, clinical and laboratory features of reported Plasmodium knowlesi deaths, on admission
Details Case 1 Case 2 Case 3 Case 4 Case 5 P. knowlesi-associatedgram-negative septicshock
Age, years 71 65 51 50 49 36
Sex Male Male Male Male Female Female
Time to death, hours 1 75 16 56 63 84
Blood pressure, mmHg 75/49 117/73 70/40 83/51 112/51 131/87
Heart rate per minute 127 58 110 89 113 110
Oxygen saturation onroom air, %
75 92 90 60 88 85
PaO2:FiO2 ratio NA NA 66 115 91 NA
Axillary temperature, °C 39.4 36.8 37.6 37.5 36.7 37.8
Haemoglobin, g/dL (females12.0-16.0, males 13.5-17.5)
14.9 14.6 9.4 10.9 12.6 11.1
WBC count, x 103 cells/μL(4.5-11)
10.8 4.8 6.8 7.81 12.1 7.8
Platelet count, x 103 cells/μL(150-450)
88 58 8 3 32 53
Serum creatinine, μmol/L(63-133)
1451 NA 578 330 283 NA
Serum urea, mmol/L (1.0–8.3) 81.5 6.1 44 38.5 25 10.8
Total serum bilirubin, μmol/L(<17)
NA NA 146 74 25 NA
Serum aspartateaminotransferase, U/L (<37)
42 NA 53 NA 39 NA
Serum alanine aminotransferaseconcentration, U/L (<40)
NA NA 28 49 20 NA
Serum albumin, g/L (35-60) NA NA 23 19 28 NA
Serum bicarbonate, mmol/L(18-23)
NA NA 14 16.2 14 6.5
Serum lactate mmol/L (0.5-2.2) NA NA 6.4 6.8 NA NA
Blood cultures negative Not done negative negative Not done Enterobacter cloacae
Initial microscopic diagnosis P. falciparum”3+”P. malariae “4+”
P. malariae “3+” P. malariae “4+” P. vivax “4+” P. malariae “4+” P. malariae “1+”
PCR result* P. knowlesi P. knowlesi P. knowlesi P. knowlesi P. knowlesi P. knowlesi
Initial therapy received IV quinine/oralchloroquine,primaquine anddoxycycline
Oral chloroquine IV artesunate Oral chloroquineand primaquine
Oral chloroquineand primaquine
Oral sulfadoxine/pyrimethamineand primaquine
NOTE. Laboratory reference ranges are given in parentheses.NA, not available; IV, intravenous.*Blood samples used for PCR-confirmation were taken on the day of hospital admission for cases 1, 3 and 5; day 2 for case 2; day 3 for case 4, and day 1 for case 6.
Rajahram et al. Malaria Journal 2012, 11:284 Page 3 of 7http://www.malariajournal.com/content/11/1/284
during the procedure. Cause of death was reported as se-vere malaria with acute respiratory distress syndrome.
Case threeA 51-year-old man with no known medical history pre-sented to a clinic with a four-day history of fever, rigours,myalgia and arthralgia. On examination he was drowsy,jaundiced, hypotensive (blood pressure 70/40 mmHg),tachycardic (heart rate 110 beats/minute) and hypoxic(oxygen saturation 90% on room air). Hepatomegaly wasnoted on abdominal examination. Blood film was reportedas P. malariae “4+”, and he had renal failure with a crea-tinine of 578 μmol/L. Chest radiograph revealed patchyconsolidation in both lung fields. He was commenced onintravenous fluids, oxygen supplementation, inotropicsupport, intravenous antibiotics and intravenous artesu-nate, and transferred to a general hospital, where he wasintubated and ventilated and commenced on haemodialy-sis. He died the following day from acute respiratory dis-tress syndrome. Blood cultures were negative.
Case fourA 50-year-old man with no known medical history pre-sented to a district hospital with a seven-day history offever and rigours and a three-day history of cough. Onexamination he was alert and orientated but hypotensive(blood pressure 83/51 mmHg) and hypoxic (oxygen sat-uration 70% on 10 L oxygen via high flow mask), andwheeze was heard on chest auscultation. Blood film wasreported as Plasmodium vivax “4+”. The patient wascommenced on intravenous fluids, antibiotics and oralchloroquine, and transferred to a general hospital, wherehe was intubated and ventilated, and commenced on ino-tropic support. Chest radiograph showed diffuse infiltrates.The patient had renal failure (creatinine 330 μmol/L) andsevere thrombocytopenia (3 x 103 platelets/μL) althoughno bleeding complications were noted. Haemodialysis wascommenced and platelet transfusion given. On day 3 a re-peat blood film was reported as P. malariae “4+”. Intra-venous artesunate was commenced, however the patientremained on maximum inotropic and ventilator support,and further haemodialysis was temporarily unavailable. Hedied nine hours later with multiple organ failure. Causeof death was reported as severe malaria. Blood culturesand dengue serology were negative, and PCR performedon a blood sample taken on day 3 confirmed P. knowlesimonoinfection.
Case fiveA 49-year-old woman with no known medical illnesspresented to a district hospital with a four-day history offever, cough and dyspnoea. On examination she was ta-chypnoeic and hypoxic (oxygen saturation on room air88%) and wheeze was heard on chest auscultation. Blood
film was reported as P. malariae “4+”, and renal failurewas present (creatinine 283 μmol/L). Bronchodilators,intravenous antibiotics, oral chloroquine and primaquinewere commenced in addition to inotropic support andtransfer to a tertiary hospital. Chest radiograph on ar-rival revealed bilateral lower zone infiltrates, and the pa-tient was intubated and ventilated however developedrefractory hypotension and died the following morning.Cause of death was stated as septic shock, althoughblood cultures were not performed.
Case 6: fatality attributed to Plasmodium knowlesi-associated gram-negative sepsisA 36-year-old woman with no known medical illness pre-sented to a district hospital with a seven-day history offever, cough and myalgia. Examination was unremarkableand no features of severe malaria were evident. Blood filmwas reported as P. malariae “1+” and she was treated withoral sulphadoxine/pyrimethamine (SP) and primaquine.On day 3, she was aparasitaemic, but had become ta-chypnoeic (respiratory rate 44 breaths/minute), hypoxic(oxygen saturation 85% on room air) and hypotensive, andwidespread wheeze was heard on chest auscultation. Chestradiograph showed no infiltrates and arterial blood gasrevealed metabolic acidosis (PaO2 110 mmHg, pH=7.21,bicarbonate 6.5 mmol/L). She was commenced onbronchodilators, intravenous ceftriaxone and inotropicsupport and transferred to a tertiary hospital where shewas intubated and ventilated and given intravenous arte-sunate. Further investigation results at the tertiary hospitalincluded haemoglobin 8.0 g/dL, platelets 65 x 103/μL,white cell count 6.6 x 103/mL, creatinine 149 μmol/L andaspartate aminotransferase 795 U/L. Following admissionthe patient’s blood pressure deteriorated further and shedeveloped a tachyarrthymia. Cardioversion was unsuccess-ful and the patient died from cardiac arrest four hoursafter arrival. Cause of death was reported as severe ma-laria, however on review, blood cultures taken on admis-sion to the district hospital were noted to be positive forEnterobacter cloacae (reported as sensitive to ceftriaxone),untreated until progression to septic shock. Repeat cul-tures taken after antibiotics at the referral hospital werenegative.
Case 7: fatality associated with vivax malariaAn 85 year-old-woman presented to a district hospitalwith fever, rigours and abdominal pain. On examinationshe was pale, tachypnoeic and hypoxic with oxygen sa-turation of 88% on room air. Her blood pressure was137/88 mmHg, and chest auscultation was clear. No ar-terial blood was available. On abdominal examination atender pulsating mass was noted, and bedside ultra-sound confirmed a 7.6 cm x 5.6 cm abdominal masswith minimal free fluid. No computed tomography scan
Rajahram et al. Malaria Journal 2012, 11:284 Page 4 of 7http://www.malariajournal.com/content/11/1/284
was done. A provisional diagnosis of aortic dissectionwas made, and after discussion with family, conservativemanagement was planned. Blood investigations revea-led anaemia (haemoglobin 8.8 g/dL), thrombocytopenia(platelets 15 x 103/μL) and acute kidney injury (creatinine149 μmol/L, urea 28 mmol/L). Blood film was reported asP. malariae “4+”, and oral chloroquine was commenced.Her condition deteriorated with increasing oxygenrequirements and on day 3 her blood pressure became un-recordable. The cause of death was reported as dissectingaortic aneurysm. PCR identified P. vivax mono-infection.
Cases 8-14: fatal falciparum malariaThe falciparum deaths comprised three children (agedeight to 11 years) and four adults (aged 22–60 years). Allthe children and two adults were Filipino. All met theWHO criteria for severe malaria [19] on presentation,with severity criteria including jaundice (N= 5) renalfailure (N = 4), respiratory distress (N = 4), anaemia(N= 4), hypotension (N= 3) and cerebral malaria (N = 1).One patient died without receiving anti-malarial treat-ment due to a delayed diagnosis, but all others weregiven intravenous quinine (N= 3) or artesunate (N = 3)within two hours of malaria diagnosis. Three patientswere intubated and ventilated, three received inotropesand two were dialyzed. Four patients died within oneday of presentation while three died at days 2-5. Bloodcultures in one child were positive for Enterobacteraerogenes.
DiscussionThis case series highlights the misdiagnosis of severeknowlesi malaria due to continued reporting of P. knowlesias P. malariae. It also highlights the lack of recognition ofnon-falciparum Plasmodium species as potential causes ofsevere and fatal malaria, and the fatal consequences of ini-tial oral therapy for severe malaria of any species, particu-larly P. knowlesi.In this series patients with deaths directly attributable to
P. knowlesi were older than those with fatal P. falciparum(median age 51 [IQR 50-65] vs 22 [IQR 9-55] years,p = 0.06). The older age group affected, and the complica-tions experienced by the fatal knowlesi malaria patients,are consistent with those already reported [1,11,14].In particular, respiratory distress (oxygen saturation <94%and respiratory rate >30 breaths/minute) occurs common-ly in knowlesi malaria [11,14], and was present in allpatients. Diffuse infiltrates were seen on chest radiographin four patients, and three of these patients met the cri-teria for acute respiratory distress syndrome (ratio of thepartial pressure of oxygen to the fraction of inspired oxy-gen [PaO2:FiO2] <200). All patients required ventilatorysupport, and respiratory failure contributed directly tocause of death in at least four patients. Renal failure has
been reported to occur in 30–55% of patients with severeknowlesi malaria [11,14], and occurred in four of five(80%) patients in this series.Hypotension is also a common complication of severe
knowlesi malaria, and occurred in four of five (80%)patients with deaths directly attributable to P. knowlesi.Pre-antibiotic blood cultures were taken in three of thesepatients and were all negative. Relatively low rates ofclinically significant bacteraemia have been reported pre-viously in severe knowlesi malaria [14], suggesting thatbacteraemia is unlikely to account for hypotension inmost patients with P. knowlesi. Nevertheless, one patientin this series who presented with uncomplicated know-lesi malaria had Enterobacter cloacae bacteraemia on ad-mission, with bacterial septic shock the likely cause ofdeath. This species was also identified in the only otherreport of clinically significant bacteraemia in a patientwith severe knowlesi malaria [14]. Gram-negative bacte-raemia is a well-recognized complication of severe fal-ciparum malaria in children, and is associated withincreased mortality [20-25]. The proposed mechanismsof the association between falciparum malaria andbacteraemia include an increased risk of non-typhoidalSalmonella due to malaria-induced haemolysis and neu-trophil dysfunction [26], impaired macrophage functiondue to haemozoin deposition in monocytes [27-29], andnitric oxide quenching [30]. Bacterial translocation intothe blood stream has also been hypothesized as a resultof microvascular sequestration of P. falciparum in gutmucosa, and parasite accumulation in multiple organswas demonstrated in the single P. knowlesi autopsy report.Further studies are required to investigate mechanisms ofbacteraemia in P. knowlesi infection.Coma has not been reported to occur in knowlesi ma-
laria, despite the only autopsy report demonstrating cere-bral accumulation of parasitized red blood cells [13]. Onepatient in this study was reported to have had reducedconsciousness and hypotension just prior to death. Hisaltered conscious state however likely reflected his agonalstate and is not consistent with cerebral malaria.In contrast to other human malarias, P. knowlesi has
a 24-hour replication cycle, which can lead to rapidincreases in parasitaemia. Diagnosis of this malaria spe-cies is therefore critical in order that its potential tocause severe disease is recognized, appropriate treatmentinstituted without delay, and further complications andfatalities avoided. In this series, no patient was correctlydiagnosed with P. knowlesi. Rather, all received a diagno-sis of P. malariae, despite a very low incidence of thisspecies in Sabah, with P. malariae detected by nestedPCR in only 2 of 318 (0.6%) microscopy-diagnosedP. malariae cases referred to the Sabah State PublicHealth Laboratory in 2009 [31]. On thick film micros-copy, P. knowlesi is indistinguishable from P. malariae.
Rajahram et al. Malaria Journal 2012, 11:284 Page 5 of 7http://www.malariajournal.com/content/11/1/284
While there may be very subtle morphological differencesbetween the late stages of P. knowlesi and P. malariae onthin film microscopy, these are not consistent and cannotbe relied upon in clinical practice [32]. P. malariae causesa relatively benign acute illness, with low parasitaemia andonly very rare reports of acutely severe malaria [33,34].Current textbook descriptions and treatment guidelines re-flect this. As has been previously recommended [1], thiscase series highlights the need for blood films positive formalaria parasites resembling P. malariae to be reported asP. knowlesi in Sabah, in order to direct clinicians in the re-cognition and management of this potentially fatal species.Moreover, this case series illustrates the importance of
a unified treatment policy for all patients with severemalaria, regardless of species [35]. Although the optimaltreatment of knowlesi malaria has not been determined,severe disease should be treated as for severe falciparummalaria, including immediate institution of parenteralartesunate. Plasmodium knowlesi is also sometimes mis-diagnosed as P. vivax (as occurred initially in one patientin this series) and P. vivax can also cause severe andfatal disease [36,37]. For both these reasons, P. vivaxshould also be treated with intravenous artesunate ifsigns of severity are present, particularly as oral chloro-quine may be poorly absorbed in acutely unwell patientsand chloroquine-resistant P. vivax is increasingly pre-valent in Southeast Asia [38]. This unified treatmentstrategy for severe malaria from all species, includingP. knowlesi, has been Sabah state policy since 2008 andhas been adopted in the latest WHO severe malariamanagement guidelines [35]. Despite this policy, in thiscase series only two of five patients with knowlesi ma-laria who met severity criteria on admission received im-mediate parenteral treatment, with the others receivingoral chloroquine. Similarly, the one patient with vivaxmalaria, who also met severity criteria on admission, wastreated only with oral chloroquine. In contrast, six ofseven patients with severe falciparum malaria receivedparenteral treatment within two hours of diagnosis(p = 0.07 for fatal P. falciparum vs non-falciparum; Yatescorrected Chi2). These findings suggest a lack of recog-nition of the potential of the non-falciparum malarias tocause severe disease and the need for treatment withintravenous artesunate for all severe malaria patients.This study had several limitations, the major one being
its retrospective design resulting in unavoidably incom-plete laboratory and clinical data. Blood culture resultswere not available for two patients, and for these pa-tients bacterial sepsis cannot be excluded as contributingto poor outcome. Lack of accurate parasite density countsalso made assessment of parasite burden difficult. Post-mortem examination was not performed for any patient,preventing detailed description of the pathogenic mecha-nisms of severe malaria and death.
ConclusionPlasmodium knowlesi is a common cause of malaria inSabah, and in this series was responsible, either directlyor through bacteraemia, for almost half of malariadeaths throughout the state. These results highlight theimportance of accurate reporting of P. knowlesi to im-prove recognition and management of this species. Cli-nicians need to be better informed of the potential forsevere and fatal malaria from non-falciparum species, es-pecially P. knowlesi, and the need to treat all severe ma-laria with immediate intravenous artesunate regardlessof species.
Competing interestsThe authors declare that they have no competing interests.
Authors’ contributionsGR performed the case note review with assistance from TW and BB. TW andJM facilitated the case review. GR, BB, NA and TY wrote the paper. Allauthors read and approved the final manuscript.
AcknowledgementsWe would like to thank Fread Anderios for performing the PCR assays at theSabah Public Health Laboratory, the medical record departments at allinvolved hospitals for assistance with retrieving medical records, and theDirector General, Malaysian Ministry of Health, for kind permission to publishthis article. BEB, TWY and NMA are supported by the Australian NationalHealth and Medical Research Council.
Author details1Infectious Diseases Department, Queen Elizabeth Hospital, Karung BerkunciNo. 2029, Jalan Penampang, Kota Kinabalu, 88560, Sabah, Malaysia. 2MenziesSchool of Health Research and Charles Darwin University, Darwin, NT,Australia. 3Department of Medicine, Queen Elizabeth Hospital, Kota Kinabalu,Sabah, Malaysia. 4Sabah Department of Health, Kota Kinabalu, Sabah,Malaysia. 5Royal Darwin Hospital, Darwin, NT, Australia.
Received: 19 April 2012 Accepted: 13 July 2012Published: 20 August 2012
References1. Cox-Singh J, Davis TM, Lee KS, Shamsul SS, Matusop A, Ratnam S, Rahman HA,
Conway DJ, Singh B: Plasmodium knowlesi malaria in humans is widelydistributed and potentially life threatening. Clin Infect Dis 2008, 46:165–171.
2. Lee CE, Adeeba K, Freigang G: Human Plasmodium knowlesi infections inKlang Valley, Peninsular Malaysia: a case series. Med J Malaysia 2010,65:63–65.
3. Figtree M, Lee R, Bain L, Kennedy T, Mackertich S, Urban M, Cheng Q,Hudson BJ: Plasmodium knowlesi in human. Indonesian Borneo. EmergInfect Dis 2010, 16:672.
4. Jiang N, Chang Q, Sun X, Lu H, Yin J, Zhang Z, Wahlgren M, Chen Q:Co-infections with Plasmodium knowlesi and other malaria parasites.Myanmar. Emerg Infect Dis 2010, 16:1476.
5. Jongwutiwes S, Putaporntip C, Iwasaki T, Sata T, Kanbara H: Naturallyacquired Plasmodium knowlesi malaria in human, Thailand. Emerg InfectDis 2004, 10:2211–2213.
6. Khim N, Siv S, Kim S, Mueller T, Fleischmann E, Singh B, Cliff P, Divis S,Steenkeste N, Duval L, Bouchier C, Duong S, Ariey F, Ménard D: Plasmodiumknowlesi infection in humans, Cambodia, 2007-2010. Emerg Infect Dis2011, 17:1900–1902.
7. Luchavez J, Espino F, Curameng P, Espina R, Bell D, Chiodini P, Nolder D,Sutherland C, Lee K, Singh B: Human infections with Plasmodium knowlesi,the Philippines. Emerg Infect Dis 2008, 14:811–813.
8. Ng OT, Ooi EE, Lee CC, Lee PJ, Ng LC, Pei SW, Tu TM, Loh JP, Leo YS:Naturally acquired human Plasmodium knowlesi infection, Singapore.Emerg Infect Dis 2008, 14:814–816.
9. Putaporntip C, Hongsrimuang T, Seethamchai S, Kobasa T, Limkittikul K, Cui L,Jongwutiwes S: Differential prevalence of Plasmodium infections and cryptic
Rajahram et al. Malaria Journal 2012, 11:284 Page 6 of 7http://www.malariajournal.com/content/11/1/284
Plasmodium knowlesimalaria in humans in Thailand. J Infect Dis 2009,199:1143–1150.
10. Eede P, Van H, Van Overmeir C, Vythilingam I, Duc T, Hung L, Manh H, Anne J,D'Alessandro U, Erhart A: Human Plasmodium knowlesi infections in youngchildren in central Vietnam. Malar J 2009, 8:249.
11. Daneshvar C, Davis TM, Cox-Singh J, Rafa'ee M, Zakaria S, Divis P, Singh B:Clinical and laboratory features of human Plasmodium knowlesi infection.Clin Infect Dis 2009, 49:852–860.
12. Barber BE, William T, Jikal M, Jilip J, Dhararaj P, Menon J, Yeo TW, Anstey NM:Plasmodium knowlesimalaria in children. Emerg Infect Dis 2011, 17:814–820.
13. Cox-Singh J, Hiu J, Lucas SB, Divis PC, Zulkarnaen M, Chandran P, Wong KT,Adem P, Zaki SR, Singh B: Severe malaria-a case of fatal Plasmodiumknowlesi infection with post-mortem findings. Malar J 2010, 9:10.
14. William T, Menon J, Rajahram G, Chan L, Ma G, Donaldson S, Khoo S,Fredrick C, Jilip J, Anstey NM, Yeo TW: Severe Plasmodium knowlesi malariain a tertiary hospital, Sabah, Malaysia. Emerg Infect Dis 2011, 17:1248–1255.
15. Daneshvar C, Davis TM, Cox-Singh J, Rafa'ee M, Zakaria S, Divis P, Singh B:Clinical and parasitological response to oral chloroquine and primaquinein uncomplicated human Plasmodium knowlesi infections. Malar J 2010,9:238.
16. Department of Statistics, Malaysia: Population Distribution and BasicDemographic Characteristics. Kuala Lumpur: Population and Housing Censusof Malaysia 2010; 2011. Available: http://www.statistics.gov.my/portal/download_Population/files/census2010/Taburan_Penduduk_dan_Ciri-ciri_Asas_Demografi.pdf. Accessed 18th June 2012.
17. Rundi C: Malaria Elimination in Malaysia. Kota Kinabalu, Sabah, Malaysia:Third annual meeting of the Asia Pacific Malaria Elimination Network(APMEN); 2011. Available: http://apmen.org/apmen-iii-meeting-proceedings/.Accessed 3rd May 2012.
18. Singh B, Sung LK, Matusop A, Radhakrishnan A, Shamsul SSG, Cox-Singh J,Thomas A, Conway DJ: A large focus of naturally acquired Plasmodiumknowlesi infections in human beings. Lancet 2004, 363:1017–1024.
19. World Health Organization: Guidelines for the treatment of malaria. 2ndedition. Geneva: World Health Organization; 2010.
20. Were T, Davenport GC, Hittner JB, Ouma C, Vulule JM, Ong'echa JM, Perkins DJ:Bacteremia in Kenyan children presenting with malaria. J Clin Microbiol2011, 49:671.
21. Berkley J, Mwarumba S, Bramham K, Lowe B, Marsh K: Bacteraemiacomplicating severe malaria in children. Trans R Soc Trop Med Hyg 1999,93:283–286.
22. Bronzan RN, Taylor TE, Mwenechanya J, Tembo M, Kayira K, Bwanaisa L,Njobvu A, Kondowe W, Chalira C, Walsh AL: Bacteremia in Malawianchildren with severe malaria: prevalence, etiology, HIV coinfection, andoutcome. J Infect Dis 2007, 195:895.
23. Graham SM, Walsh AL, Molyneux EM, Phiri AJ, Molyneux ME: Clinicalpresentation of non-typhoidal Salmonella bacteraemia in Malawianchildren. Trans Roy Soc Trop Med Hyg 2000, 94:310–314.
24. Walsh AL, Phiri AJ, Graham SM, Molyneux EM, Molyneux ME: Bacteremia infebrile Malawian children: clinical and microbiologic features. PediatrInfect Dis J 2000, 19:312.
25. Scott JAG, Berkley JA, Mwangi I, Ochola L, Uyoga S, Macharia A, Ndila C,Lowe BS, Mwarumba S, Bauni E, Marsh K, Williams TN: Relation betweenfalciparum malaria and bacteraemia in Kenyan children: a population-based, case–control study and a longitudinal study. Lancet 2011,378:1316–1323.
26. Cunnington AJ, de Souza JB, Walther M, Riley EM: Malaria impairsresistance to Salmonella through heme- and heme oxygenase-dependent dysfunctional granulocyte mobilization. Nat Med 2012,18:120–127.
27. Schwarzer E, Skorokhod OA, Barrera V, Arese P: Hemozoin and the humanmonocyte-a brief review of their interactions. Parassitologia 2008, 50:143.
28. Awandare GA, Ouma Y, Ouma C, Were T, Otieno R, Keller CC, Davenport GC,Hittner JB, Vulule J, Ferrell R: Role of monocyte-acquired hemozoin insuppression of macrophage migration inhibitory factor in children withsevere malarial anemia. Infect Immun 2007, 75:201–210.
29. Carney CK, Schrimpe AC, Halfpenny K, Harry RS, Miller CM, Broncel M, Sewell SL,Schaff JE, Deol R, Carter MD: The basis of the immunomodulatory activity ofmalaria pigment (hemozoin). J Biol Inorganic Chem 2006, 11:917–929.
30. Yeo TW, Lampah DA, Tjitra E, Gitawati R, Kenangalem E, Piera K, Granger DL,Lopansri BK, Weinberg JB, Price RN, Dufful SB, Celermajer DS, Anstey NM:Relationship of cell-free haemoglobin to impaired nitric oxide
bioavailability and perfusion in severe falciparum malaria. J Infect Dis2009, 200:1522–1529.
31. Naing DKS, Anderios F, Lin Z: Geographic and ethnic distribution of Pknowlesi infection in Sabah, Malaysia. Int J Collab Res Intern Med PublicHealth 2011, 3:391–400.
32. World Health Organization: Meeting report. Kota Samarahan, Sawarak,Malaysia: Informal consultation on the public health importance ofPlasmodium knowlesi; 2011.
33. Descheemaeker PN, Mira JP, Bruneel F, Houzé S, Tanguy M, Gangneux J-P,Flecher E, Rousseau C, Bras JL, Mallédant Y: Near-fatal multiple organdysfunction syndrome induced by Plasmodium malariae. Emerg Infect Dis2009, 15:832–834.
34. Imbert P, Rapp C, Buffet PA: Pathological rupture of the spleen in malaria:analysis of 55 cases (1958-2008). Travel Med Infect Dis 2009, 7:147–159.
35. World Health Organization: Management of severe malaria. 3rd edition.Geneva: A practical handbook; In Press.
36. Tjitra E, Anstey NM, Sugiarto P, Warikar N, Kenangalem E, Karyana M,Lampah DA, Price RN: Multidrug-resistant Plasmodium vivax associatedwith severe and fatal malaria: a prospective study in Papua. Indonesia.PLoS Med 2008, 5:e128.
37. Genton B, D'Acremont V, Rare L, Baea K, Reeder JC, Alpers MP, Müller I:Plasmodium vivax and mixed infections are associated with severemalaria in children: a prospective cohort study from Papua New Guinea.PLoS Med 2008, 5:e127.
38. Douglas NM, Anstey NM, Angus BJ, Nosten F, Price RN: Artemisinincombination therapy for vivax malaria. Lancet Infect Dis 2010, 10:405–416.
doi:10.1186/1475-2875-11-284Cite this article as: Rajahram et al.: Deaths due to Plasmodium knowlesimalaria in Sabah, Malaysia: association with reporting as Plasmodiummalariae and delayed parenteral artesunate. Malaria Journal 2012 11:284.
Submit your next manuscript to BioMed Centraland take full advantage of:
• Convenient online submission
• Thorough peer review
• No space constraints or color figure charges
• Immediate publication on acceptance
• Inclusion in PubMed, CAS, Scopus and Google Scholar
• Research which is freely available for redistribution
Submit your manuscript at www.biomedcentral.com/submit
Rajahram et al. Malaria Journal 2012, 11:284 Page 7 of 7http://www.malariajournal.com/content/11/1/284
78
Implications of this paper
This paper reviews 14 malaria deaths that occurred in Sabah during 2010-2011, including
six cases with P. knowlesi, eight with P. falciparum and one with P. vivax. It describes in
detail the P. knowlesi cases, and reports differences between the clinical features and
management of patients according to malaria species.
Importantly, we found that all patients with P. knowlesi were reported, and hence diagnosed,
as P. malariae, despite the current rarity of P. malariae in Sabah (25, 26, 58, 171). P.
malariae generally causes benign malaria, with parasitemia rarely exceeding 5000
parasites/μL. In contrast, P. knowlesi can be associated with high parasitemia, multi-organ
failure and death (23, 24, 27, 31, 103). Misdiagnosis of P. knowlesi as P. malariae may
therefore result in clinicians failing to recognize risk of severe disease, with consequent
delay in appropriate management, as was evident in this review. This paper therefore
recommends that in Sabah, blood slides with parasites resembling P. malariae be reported
as P. knowlesi, in order to improve recognition and management of this species.
Moreover, this paper highlights the tendency of clinicians to treat patients with severe
malaria from non-falciparum species with less urgency than they treat those with severe
falciparum malaria. While P. falciparum continues to account for the majority of malaria
deaths world-wide (70), severe disease from non-falciparum malaria species is increasingly
recognized (178-182, 185, 186). Patients with malaria from any species may therefore be at
risk of severe disease, and symptoms or signs suggestive of complications should prompt
urgent initiation of intravenous artesunate and other supportive care (110).
79
Microscopy has been previously reported to have low sensitivity and variable specificity in
areas where P. falciparum and P. vivax are co-endemic (188, 189). The studies described in
the earlier chapters of this thesis demonstrate that Sabah represents a unique
epidemiological situation where P. vivax, P. falciparum and P. knowlesi all commonly occur,
and in approximately equal proportions. We hypothesized that this situation may present
particular difficulties for microscopists in Sabah. Sabah has experienced a marked reduction
of falciparum and vivax malaria in recent years and hence microscopists are exposed to far
fewer malaria slides than previously. Despite this, they must now distinguish between three
commonly occurring malaria species, one of which may not have been included as part of
their training.
The microscopic features of P. knowlesi were described in Chapter 3. The difficulties of
distinguishing P. knowlesi and P. falciparum have been well-described, with the young ring
forms of P. knowlesi known to resemble those of P. falciparum (57). In addition, a single
study in Sarawak reported that 10% of cases of PCR-confirmed P. vivax were misdiagnosed
as “P. malariae” (23). Misdiagnosis of P.vivax as “P. malariae/P. knowlesi” would have
important clinical implications, as patients may not be treated with primaquine and hence
would be at risk of relapse. In addition, frequent microscopic misdiagnoses may influence
the reliability of Sabah Department of Health malaria notification data.
An understanding of the inaccuracies of microscopy is therefore important when planning
strategies for treatment, diagnosis, and surveillance of the different malaria species in
Sabah. We therefore evaluated the accuracy of routine hospital-based microscopy, and
microscopy performed by our experienced research microscopist, for the diagnosis of P.
RESEARCH Open Access
Limitations of microscopy to differentiatePlasmodium species in a region co-endemic forPlasmodium falciparum, Plasmodium vivax andPlasmodium knowlesiBridget E Barber1,2*, Timothy William2,3, Matthew J Grigg1,2, Tsin W Yeo1,4 and Nicholas M Anstey1,4
Abstract
Background: In areas co-endemic for multiple Plasmodium species, correct diagnosis is crucial for appropriatetreatment and surveillance. Species misidentification by microscopy has been reported in areas co-endemic forvivax and falciparum malaria, and may be more frequent in regions where Plasmodium knowlesi also commonlyoccurs.
Methods: This prospective study in Sabah, Malaysia, evaluated the accuracy of routine district and referralhospital-based microscopy, and microscopy performed by an experienced research microscopist, for the diagnosisof PCR-confirmed Plasmodium falciparum, P. knowlesi, and Plasmodium vivax malaria.
Results: A total of 304 patients with PCR-confirmed Plasmodium infection were enrolled, including 130 with P.knowlesi, 122 with P. falciparum, 43 with P. vivax, one with Plasmodium malariae and eight with mixed speciesinfections. Among patients with P. knowlesi mono-infection, routine and cross-check microscopy both identified 94(72%) patients as “P. malariae/P. knowlesi”; 17 (13%) and 28 (22%) respectively were identified as P. falciparum, and13 (10%) and two (1.5%) as P. vivax. Among patients with PCR-confirmed P. falciparum, routine and cross-checkmicroscopy identified 110/122 (90%) and 112/118 (95%) patients respectively as P. falciparum, and 8/122 (6.6%) and5/118 (4.2%) as “P. malariae/P. knowlesi”. Among those with P. vivax, 23/43 (53%) and 34/40 (85%) were correctlydiagnosed by routine and cross-check microscopy respectively, while 13/43 (30%) and 3/40 (7.5%) patients werediagnosed as “P. malariae/P. knowlesi”. Four of 13 patients with PCR-confirmed P. vivax and misdiagnosed by routinemicroscopy as “P. malariae/P. knowlesi” were subsequently re-admitted with P. vivax malaria.
Conclusions: Microscopy does not reliably distinguish between P. falciparum, P. vivax and P. knowlesi in a regionwhere all three species frequently occur. Misdiagnosis of P. knowlesi as both P. vivax and P. falciparum, and viceversa, is common, potentially leading to inappropriate treatment, including chloroquine therapy for P. falciparumand a lack of anti-relapse therapy for P. vivax. The limitations of microscopy in P. knowlesi-endemic areas supportsthe use of unified blood-stage treatment strategies for all Plasmodium species, the development of accurate rapiddiagnostic tests suitable for all species, and the use of PCR-confirmation for accurate surveillance.
Keywords: Plasmodium knowlesi, Malaria, Microscopy, Diagnosis
* Correspondence: [email protected] School of Health Research and Charles Darwin University, Darwin,Northern Territory, Australia2Infectious Diseases Unit, Queen Elizabeth Hospital, Kota Kinabalu, Sabah,MalaysiaFull list of author information is available at the end of the article
© 2013 Barber et al.; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the CreativeCommons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, andreproduction in any medium, provided the original work is properly cited.
Barber et al. Malaria Journal 2013, 12:8http://www.malariajournal.com/content/12/1/8
BackgroundDespite recent progress towards elimination, malaria con-tinues to affect over 200 million people per year with anestimated 655,000 deaths [1]. Although most deaths arecaused by Plasmodium falciparum, the relative contribu-tion of the non-falciparum Plasmodium species to theglobal malaria burden is increasing as incidence of P.falciparum falls [2-4]. The most widely distributed of thesespecies is Plasmodium vivax, which accounts for half ofthe world’s malaria and is increasingly recognized as acause of severe and potentially fatal disease [5-8]. More-over, reducing transmission of P. vivax has proved moredifficult than P. falciparum, with successful malaria con-trol programmes in some countries leading to an increasein incidence of P. vivax as overall malaria rates drop [2,3].More recently, the simian parasite Plasmodium knowlesihas been identified as the most common cause of humanmalaria in parts of Malaysia [9-14], with its emergencealso associated with reduction in incidence of the humanPlasmodium species [12]. Plasmodium knowlesi is capableof causing severe disease and death [9,13,15-18], and isincreasingly reported in other Southeast Asian countries[19].In areas co-endemic for P. falciparum, P. vivax and P.
knowlesi, species-differentiation at the time of diagnosisis crucial for directing appropriate treatment, particu-larly in settings which have separate treatment policiesfor different species, most commonly artemisinin-combination treatment (ACT) for P. falciparum andchloroquine for non-falciparum species. Even in regionssuch as Papua, Indonesia, which have adopted a unifiedtreatment strategy of ACT for all malaria [1], diagnosisof P. vivax is still required to allow administration ofanti-hypnozoite treatment to prevent relapses, with mis-diagnosis of this species potentially leading to increasedmorbidity and transmission. In areas also endemic for P.knowlesi, accurate diagnosis is important for epidemio-logical surveillance of this potentially fatal emerging zoo-notic infection.Microscopy of stained blood smears remains the stan-
dard method of malaria diagnosis in most parts of themalaria-endemic world, and ideally allows differentiationof species. However, with the difficulty in distinguishingyoung ring-stage parasites, frequent misdiagnosis has beenreported in areas co-endemic for P. falciparum and P.vivax [20,21]. It is well established that microscopy cannotreliably distinguish P. knowlesi from Plasmodium mala-riae [22,23], but misdiagnosis of P. knowlesi with otherspecies may also be frequent [13]. This study thereforeevaluated the accuracy of both routine hospital micros-copy and microscopy performed by an experienced re-search microscopist, for the diagnosis of P. falciparum, P.knowlesi, and P. vivax, in an area where all three speciescommonly occur.
MethodsStudy site and referral systemThe study was conducted at Queen Elizabeth Hospital(QEH), an adult tertiary referral hospital in Kota Kinabalu,Sabah, Malaysia. The hospital services the West Coast andKudat Divisions of Sabah, with six district hospitals and apopulation of 1.14 million. From September 2010, in re-sponse to ongoing malaria deaths in Sabah [16], new treat-ment and referral guidelines were instituted and includedtertiary hospital referral for all malaria patients with athick blood film reported as “4+” (indicating >10 para-sites/high-power microscopy field [24]) or who had anyevidence of severe malaria. Treatment was commencedprior to transfer and a pre-treatment blood film was sentwith the patient. Local health clinics within the KotaKinabalu area were required to refer all malaria patients toQEH for admission, with treatment commenced onarrival.
SubjectsAll patients referred to or admitted directly to QEH with amicroscopic diagnosis of malaria were assessed for eligibil-ity from September 2010 to October 2011 as part of a pro-spective study of the epidemiology, clinical spectrum andpathophysiology of knowlesi malaria, reported elsewhere[15]. Non-pregnant patients ≥12 years old were enrolled ifthey were within 18 hours of commencing malaria treat-ment, had no major co-morbidities, and had not alreadybeen enrolled in the study. Patients who were PCR nega-tive were retrospectively excluded. Patients were classifiedas having severe malaria using modified 2010 WHOSevere Malaria Criteria [15,25,26]. Written informed con-sent was provided by patients or their relatives. Ethicsapproval was obtained from the Medical Research Sub-Committee of the Malaysian Ministry of Health and theHealth Research Ethics Committee of the Menzies Schoolof Health Research, Australia.
Study proceduresStandardized data forms were used to record demo-graphic and clinical information. Venous blood wascollected in a CTAD tube labelled with the patient’sstudy number, and thick and thin blood smears preparedusing Giemsa staining. Species identification using thickand thin blood films was performed initially by micro-scopists at referring district hospitals, or at QEH if pre-senting directly to this hospital (routine microscopy).Thick and thin films were later cross-checked by a re-search microscopist (cross-check microscopy) with morethan 15 years’ experience, who was blinded to the resultsof routine microscopy. Because reliable differentiation ofP. knowlesi from P. malariae is not possible [27], slidesreported as P. malariae, P. knowlesi, or P. malariae/P.knowlesi were all considered a single group, further
Barber et al. Malaria Journal 2013, 12:8 Page 2 of 6http://www.malariajournal.com/content/12/1/8
referred to as “P. malariae/P. knowlesi”. Parasite densitywas quantified by the research microscopist using pre-treatment slides, and reported as the number of para-sites per 200 white blood cells or per 1,000 red bloodcells and converted to parasites/μL using the patient’swhite blood cell count or haematocrit, respectively. If apre-treatment slide could not be reliably obtained (6% oftotal slides) the referring hospital’s microscopy reportwas used and the “1+ − 4+” grade converted into para-sites/μL using the relevant median parasite density. Para-site DNA was extracted and PCR performed usingpreviously described methods for P. falciparum, P. vivax,Plasmodium ovale, and P. malariae [28] and P. knowlesi[29]. PCR diagnosis was used as the gold standard.Patients were followed-up on days 14 and 28 if possible,and/or if readmitted to QEH.
Statistical analysisData were analysed using STATA version 10.1 (Stata-Corp LP, College Station, TX, USA). For continuousvariables intergroup differences were compared usingthe Kruskal-Wallis test, or the Mann–Whitney test forpairwise comparisons, while the χ2 test was used forintergroup differences between categorical variables. Lo-gistic regression was used to assess the associationbetween parasite count and correct identification ofspecies.
ResultsA total of 304 patients with PCR-confirmed Plasmodiuminfection were enrolled, including 130 with P. knowlesi,122 with P. falciparum, 43 with P. vivax, one with P.malariae, and eight with mixed-species infection. Demo-graphic and clinical features are reported separately [15].Half (51%) of all patients were referred from district hos-pitals, including 86 (66%) with knowlesi malaria, 47(39%) with P. falciparum, and 21 (49%) with P. vivaxmalaria. Severe malaria occurred in 38 (29%) patientswith P. knowlesi, 13 (11%) with P. falciparum, and seven(16%) with P. vivax [15]. Patients with non-severe P.knowlesi had lower parasite counts than those with non-severe P. falciparum (4,837 [IQR 1576–14,641] vs 10,500[IQR 4,014–32,267] parasites/μL, p < 0.01), but not P.vivax (median 4,753 [IQR 2369–10,316] parasites/μL,p = 0.95). Among patients with severe malaria, there wasno significant difference in median parasitaemia ofthose with P. knowlesi (80,359 [IQR 25,857–168,279]parasites/μL) and P. falciparum (72,270 [IQR 27,905–273,909] parasites/μL, p = 0.78).Microscopy and PCR results are shown in Tables 1 and
2. Slides were unavailable for cross-check microscopy foreight (2.6%) patients. Routine and cross-check microscopycorrectly identified the species in 229/304 (75%) and242/296 (82%) patients respectively (p = 0.055), with 188/
296 (64%) patients correctly diagnosed by both micro-scopic methods. Among patients with PCR-confirmed P.knowlesi mono-infection, routine and cross-check micros-copy each correctly identified 94 (72%) patients as “P.malariae/P. knowlesi”, with 66 (51%) patients correctlydiagnosed by both microscopy readings. Routine micros-copy diagnosed 17 (13%) and 13 (10%) patients with PCR-confirmed P. knowlesi mono-infection as P. falciparumand P. vivax respectively, while cross-check microscopydiagnosed 28 (22%) and two (1.5%) as P. falciparum andP. vivax respectively. Four (3%) patients with PCR-confirmed P. knowlesi were diagnosed by both routine andcross-check microscopy as P. falciparum, and no patientwas diagnosed by both readings as P. vivax. Among the 38patients with severe knowlesi malaria, routine microscopycorrectly diagnosed 33 (87%) as “P. malariae/P. knowlesi”.Three were diagnosed as P. falciparum, one as P. vivaxand one as P. falciparum/“P. malariae/P. knowlesi”. Themedian parasite count among the five patients with severeknowlesi malaria and misdiagnosed by routine microscopy
Table 1 PCR results compared with routine microscopy
Microscopy PCR
Pf Pv Pk Pm Pf/Pv Pf/Pk Pf/Pm Total
Pf 110 3 17 0 2 2 0 134
Pv 1 23 13 0 1 0 0 38
"Pm/Pk" 8 13 94 1 0 0 1 117
"Pm/Pk"/Pv 0 3 2 0 0 1 0 6
"Pm/Pk"/Pf 2 0 3 0 0 0 0 5
Pf/Pv 1 1 0 0 1 0 0 2
“P species” 0 0 1 0 0 0 0 1
Total 122 43 130 1 4 3 1 304
Abbreviations: Pf = Plasmodium falciparum, Pv = Plasmodium vivax, Pk =Plasmodium knowlesi, Pm = Plasmodium malariae, P = Plasmodium.
Table 2 PCR results compared with cross-checkmicroscopy
Microscopy PCR
Pf Pv Pk Pm Pf/Pv Pf/Pk Pf/Pm Total
Pf 112 0 28 0 1 3 0 144
Pv 1 34 2 0 1 0 0 38
"Pm/Pk" 5 3 94 1 0 0 1 104
Pf/Pv 0 1 2 0 1 0 0 4
Po 0 2 0 0 0 0 0 2
Negative 0 0 4* 0 0 0 0 4
Total 118 40 130 1 3 3 1 296
Abbreviations: Pf = Plasmodium falciparum, Pv = Plasmodium vivax, Pk =Plasmodium knowlesi, Pm = Plasmodium malariae, Po = Plasmodium ovale.Note: Slides were unavailable for cross-check microscopy for eight patients.Cross-check microscopy was performed on pre-treatment slides for 280 (95%)of 296 patients.*Includes one post-treatment slide. Parasite counts (according to routinemicroscopy results) of remaining three patients were 32, 55 and 1574parasites/μL.
Barber et al. Malaria Journal 2013, 12:8 Page 3 of 6http://www.malariajournal.com/content/12/1/8
was 168,279 (range 26,368 – 506,218) parasites/μL, andwas not significantly different to those correctly diagnosed.Among patients with PCR-confirmed P. falciparum,
routine and cross-check microscopy correctly identified110/122 (90%) and 112/118 (95%) as P. falciparum,while 8/122 (6.6%) and 5/118 (4.2%) patients respectivelywere diagnosed as “P. malariae/P. knowlesi”. All 13patients with severe falciparum malaria were correctlydiagnosed by routine microscopy.Among those with PCR-confirmed P. vivax, 23/43
(53%) and 34/40 (85%) were correctly diagnosed byroutine and cross-check microscopy respectively, while13/43 (30%) and 3/40 (7.5%) patients were diagnosed as“P. malariae/P. knowlesi”. Among the 13 patients withPCR-confirmed P. vivax and misdiagnosed by routinemicroscopy as “P. malariae/P. knowlesi”, four (31%) werere-admitted to QEH within four months with PCR-confirmed P. vivax infection. Among the seven patientswith severe vivax malaria, routine microscopy diagnosedthree as P. vivax, two as “P. malariae/P. knowlesi”, oneas P. falciparum/P. vivax, and one as “P. malariae/P.knowlesi”/P. vivax.Among the eight patients with mixed-species infection
by PCR, one patient with P. falciparum/P. vivax wascorrectly identified by routine microscopy, and anotherwith P. falciparum/P. vivax was correctly identified bycross-check microscopy. All others were misdiagnosedby both microscopic readings (with one patient’s slideunavailable for cross-check microscopy).An association was found between parasite count and
correct identification of P. vivax by routine microscopy(OR [log increase] 1.64 [95% CI 1.01 – 2.68], p = 0.046),with a similar trend also seen with identification of P.vivax by cross-check microscopy (OR [log increase] 1.73[95% CI 0.96 – 3.10], p = 0.067). No association howeveroccurred between parasite count and correct identifica-tion of P. knowlesi or P. falciparum, by either routine orcross-check microscopy.
DiscussionThis study highlights the difficulties of microscopic diag-nosis of Plasmodium species in an area where P. falcip-arum, P. vivax and P. knowlesi all commonly occur.Misdiagnosis of P. knowlesi as both P. vivax and P.falciparum, and vice versa, is common. Only 72% ofpatients with PCR-confirmed P. knowlesi received anaccurate diagnosis by routine or by cross-check micros-copy, and correlation between the microscopic methodswas poor, with even fewer patients receiving an accuratediagnosis of P. knowlesi by both methods. These findingsoccurred despite considering P. knowlesi and P. malariaeas a single group, and so were not a consequence of thewell described near impossibility of distinguishing thesetwo species. Rather, patients with PCR-confirmed P.
knowlesi were commonly misdiagnosed as having eitherP. falciparum or P. vivax malaria, with misdiagnosis ofP. vivax as “P. malariae/P. knowlesi” also common.The difficulty with distinguishing P. knowlesi from P.
falciparum by microscopy has been previously described,due to similarities between the young rings of P. knowlesiand ring forms of P. falciparum, including double chroma-tin dots, multiple-infected erythrocytes and appliqueforms [22,23]. In a previous study in Sarawak, 11/216 (5%)patients diagnosed by microscopy as “P. malariae” wereactually P. falciparum by PCR, and 33/312 (11%)microscopy-diagnosed P. falciparum cases were P. know-lesi by PCR [13]. In this previous study and in the currentstudy, however, the difficulty differentiating P. knowlesiand P. vivax was also notable. In the current study 30% ofpatients with PCR-confirmed P. vivax were misdiagnosedby routine microscopy as “P. malariae/P. knowlesi”, withfour patients subsequently re-admitted with presumedvivax relapses due to lack of primaquine treatment. In theSarawak study, 43 of 440 (10%) patients with PCR-confirmed P. vivax malaria were misdiagnosed as “P.malariae” [13]. In a series of malaria deaths in Sabah, oneof six fatal cases of P. knowlesi was misdiagnosed as P.vivax by microscopy and a fatal case of P. vivax was mis-diagnosed as “P. malariae” [18].This frequent misdiagnosis of P. vivax as “P. malariae/
P. knowlesi” has significant implications for malaria con-trol, as failure to administer anti-hypnozoite treatmentmay lead to increased transmission and may hamperefforts to eliminate vivax malaria in regions where P.knowlesi is common. These findings support the currentSabah Ministry of Health policy for the performance ofreference centre PCR on all patients with a microscopicdiagnosis of “P. malariae/P. knowlesi” to enable adminis-tration of primaquine to those found to have misdiag-nosed P. vivax. In knowlesi-endemic areas, wherelogistically possible, PCR should also be performed on atleast a proportion of slides diagnosed as P. falciparumor P. vivax, to allow monitoring of the accuracy ofmicroscopic diagnoses at different clinical sites, and toidentify areas where additional training of microscopistsmay be required. Given the inaccuracies of microscopicdiagnosis, performance of PCR is also essential to main-tain accurate surveillance, particularly monitoring theemergence of P. knowlesi.The inaccuracy of microscopy for differentiating
Plasmodium species creates difficulties with basingtreatment decisions on microscopic results. The 2008Malaysian Ministry of Health malaria treatment guide-lines recommend chloroquine for uncomplicated P.malariae/P. knowlesi malaria; chloroquine plus prima-quine for uncomplicated P. vivax malaria; artemisinincombination treatment (ACT; artesunate/mefloquine[ArtequineW] or artemether/lumefantrine [RiametW]) for
Barber et al. Malaria Journal 2013, 12:8 Page 4 of 6http://www.malariajournal.com/content/12/1/8
uncomplicated P. falciparum malaria; and intravenousartesunate, or intravenous quinine plus oral doxycycline,for severe P. falciparum or severe P. malariae/P. knowlesimalaria [30]. No recommendations are given for treat-ment of severe vivax malaria. At Queen Elizabeth Hos-pital and referral hospitals in its catchment area, updatedtreatment guidelines recommend ACT for uncomplicatedP. falciparum or P. malariae/P. knowlesi malaria; chloro-quine plus primaquine, or ACT, for uncomplicated P. vivaxmalaria; and intravenous artesunate (followed by oral the-rapy as above) for severe malaria from any species [15].Given the different treatment recommendations for
each species in both of these guidelines, inappropriatetreatment decisions may be made as a result of incorrectmicroscopic diagnoses. Misdiagnosis of P. falciparum asP. knowlesi would, under current national guidelines,result in administration of chloroquine for P. falciparum.Given widespread resistance of P. falciparum to chloro-quine, this would lead to increased risk of complicationsand/or fatal outcome, as well as increased transmission.Patients with severe P. knowlesi malaria misdiagnosed asP. vivax may fail to receive immediate parenteral treat-ment if national guidelines are followed, and this sce-nario has been previously associated with fatal outcomes[18]. Even if the updated hospital guidelines are fol-lowed, misdiagnosis of severe P. knowlesi as P. vivaxmay still lead to treatment with oral chloroquine if signsof severity are missed, potentially leading to adverse out-comes given the slower parasite clearance associatedwith chloroquine as compared to oral ACT [16].These results therefore support the argument for a uni-
fied treatment strategy of ACT for uncomplicated malariafrom all Plasmodium species in knowlesi-endemic areas,an approach increasingly recommended for regions co-endemic for P. falciparum and P. vivax [31]. These dataalso support the 2012 WHO recommendation for intra-venous artesunate to be given to any patient meeting se-vere malaria criteria [25]. Even if signs of severity areoverlooked among patients with knowlesi malaria, treat-ment with oral ACT may ensure more rapid parasiteclearance and may lead to improved outcomes comparedto treatment with oral chloroquine [16], although the opti-mal oral agent for uncomplicated P.knowlesi remainsundetermined.An additional advantage of this unified treatment
approach would be the avoidance of inadvertent use ofchloroquine for P. falciparum misdiagnosed as anotherspecies. Furthermore, chloroquine-resistant P. vivax isan increasing problem throughout Southeast Asia [31],and has been previously documented in Sabah [32]and Peninsular Malaysia [33-35]. Use of chloroquinefor P. vivax malaria may therefore be associated withtreatment failures, and may potentiate spread of chloro-quine resistance.
Finally, this study highlights the need to develop rapiddiagnostic tests (RDTs) that have the ability to distin-guish between Plasmodium species, in order that reliableresults can be obtained more quickly and cheaply than ispossible with PCR. Although histidine-rich protein 2(HRP2)-based RDTs are able to diagnose P. falciparum,RDTs that distinguish P. knowlesi from P. vivax are notyet available. Limited data suggest that P. knowlesi cross-reacts with both P. falciparum and P. vivax-specificpLDH [36-39], and RDTs that combine these antigenswith HRP2 may therefore allow differentiation betweenP. vivax, P. falciparum and P. knowlesi mono-infections.However, while a pLDH RDT has shown high sensitivityfor the diagnosis of severe malaria from all three of thesespecies, neither pLDH- or aldolase-based RDTs havedemonstrated sufficiently high sensitivity for uncompli-cated P. knowlesi [40]. Prospective evaluation of moresensitive RDTs in knowlesi-endemic areas is needed.This study has found that microscopy does not reliably
distinguish between P. falciparum, P. vivax and P. knowlesiin areas co-endemic for all three species, with misdiag-nosis of P. vivax and P. knowlesi particularly common.In P. knowlesi-endemic areas these limitations of micro-scopic diagnosis must be considered when developingstrategies to monitor the prevalence of the different mal-aria species, and when developing treatment guidelines.
Competing interestsThe authors declare that they have no competing interests.
Authors’ contributionsNMA, TWY, TW and BEB conceived and designed the study; BEB collectedand analysed the data. BEB and NMA wrote the manuscript. MJG assistedwith data collection. All authors read and approved the final manuscript.
AcknowledgementsWe thank the clinical staff involved in patient care and microscopy staff atthe district hospitals and QEH; Rita Wong, Beatrice Wong and Ann Wee fortheir help with clinical and laboratory study procedures; Ferryanto Chalfeinfor performing cross-check microscopy; Jutta Marfurt, Sarah Auburn andNadine Kurz for performing the PCR assays; and the Director General ofHealth, Malaysia, for permission to publish this paper.
Author details1Menzies School of Health Research and Charles Darwin University, Darwin,Northern Territory, Australia. 2Infectious Diseases Unit, Queen ElizabethHospital, Kota Kinabalu, Sabah, Malaysia. 3Sabah Department of Health, KotaKinabalu, Sabah, Malaysia. 4Division of Medicine, Royal Darwin Hospital,Darwin, Northern Territory, Australia.
Received: 17 October 2012 Accepted: 20 December 2012Published: 8 January 2013
References1. World Health Organization: World Malaria Report 2011. Available: http://
www.who.int/malaria/world_malaria_report_2011/9789241564403_eng.pdf.Accessed 3rd May.
2. Nosten F, Van Vugt M, Price R, Luxemburger C, Thway K, Brockman A,McGready R, Ter Kuile F: Effects of artesunate-mefloquine combination onincidence of Plasmodium falciparum malaria and mefloquine resistancein western Thailand: a prospective study. Lancet 2000, 356:297–302.
3. Oliveira-Ferreira J, Lacerda MVG, Brasil P, Ladislau JLB, Tauil PL, Daniel-RibeiroCT: Malaria in Brazil: an overview. Malar J 2010, 9:115.
Barber et al. Malaria Journal 2013, 12:8 Page 5 of 6http://www.malariajournal.com/content/12/1/8
4. Gething PW, Elyazar IRF, Moyes CL, Smith DL, Battle KE, Guerra CA, Patil AP,Tatem AJ, Howes RE, Myers MF: A long neglected world malaria map:Plasmodium vivax endemicity in 2010. PLoS Negl Trop Dis 2012, 6:e1814.
5. Tjitra E, Anstey NM, Sugiarto P, Warikar N, Kenangalem E, Karyana M,Lampah DA, Price RN: Multidrug-resistant Plasmodium vivax associatedwith severe and fatal malaria: a prospective study in Papua Indonesia.PLoS Med 2008, 5:e128.
6. Genton B, D'Acremont V, Rare L, Baea K, Reeder JC, Alpers MP, Müller I:Plasmodium vivax and mixed infections are associated with severemalaria in children: a prospective cohort study from Papua New Guinea.PLoS Med 2008, 5:e127.
7. Price RN, Douglas NM, Anstey NM: New developments in Plasmodiumvivax malaria: severe disease and the rise of chloroquine resistance.Curr Opin Infect Dis 2009, 22:430.
8. Anstey NM, Douglas NM, Poespoprodjo JR, Price R: Plasmodium vivax:clinical spectrum, risk factors and pathogenesis. Adv Parasitol 2012,80:151–201.
9. Daneshvar C, Davis TM, Cox-Singh J, Rafa'ee M, Zakaria S, Divis P, Singh B:Clinical and laboratory features of human Plasmodium knowlesi infection.Clin Infect Dis 2009, 49:852–860.
10. Barber BE, William T, Dhararaj P, Anderios F, Grigg MJ, Yeo TW, Anstey NM:Epidemiology of Plasmodium knowlesi malaria in northeast SabahMalaysia: family clusters and wide age distribution. Malar J 2012, 11:401.
11. Barber BE, William T, Jikal M, Jilip J, Dhararaj P, Menon J, Yeo TW, AnsteyNM: Plasmodium knowlesi malaria in children. Emerg Infect Dis 2011,17:814–820.
12. William T, Rahman HA, Jelip J, Ibrahim MY, Menon J, Grigg MJ, Yeo TW,Anstey NM, Barber BE: Increasing incidence of Plasmodium knowlesimalaria following control of P. falciparum and P. vivax malaria in Sabah,Malaysia. PLoS Negl Trop Dis 2013. In press.
13. Cox-Singh J, Davis TM, Lee KS, Shamsul SS, Matusop A, Ratnam S, RahmanHA, Conway DJ, Singh B: Plasmodium knowlesi malaria in humans iswidely distributed and potentially life threatening. Clin Infect Dis 2008,46:165–171.
14. Joveen-Neoh WF, Chong KL, Wong CM, Lau TY: Incidence of malaria in theInterior Division of Sabah, Malaysian Borneo, based on nested PCR.J Parasitol Res 2011, doi:10.1155/2011/104284.
15. Barber BE, William T, Grigg MJ, Menon J, Auburn S, Marfurt J, Anstey NM,Yeo TW: A prospective comparative study of knowlesi, falciparum andvivax malaria in Sabah, Malaysia: high proportion with severe diseasefrom Plasmodium knowlesi and P. vivax but no mortality with earlyreferral and artesunate therapy. Clin Infect Dis Advanced Access 2012.doi:10.1093/cid/cis902.
16. William T, Menon J, Rajahram G, Chan L, Ma G, Donaldson S, Khoo S,Fredrick C, Jilip J, Anstey NM, Yeo TW: Severe Plasmodium knowlesi malariain a tertiary hospital, Sabah Malaysia. Emerg Infect Dis 2011, 17:1248–1255.
17. Cox-Singh J, Hiu J, Lucas SB, Divis PC, Zulkarnaen M, Chandran P, Wong KT,Adem P, Zaki SR, Singh B: Severe malaria-a case of fatal Plasmodiumknowlesi infection with post-mortem findings. Malar J 2010, 9:10.
18. Rajahram G, Barber BE, William T, Menon J, Anstey NM, Yeo TW: Deaths dueto Plasmodium knowlesi malaria in Sabah. Malaysia: association withreporting as P. malariae and delayed parenteral artesunate. Malar J 2012,11:284.
19. Cox-Singh J: Zoonotic malaria: Plasmodium knowlesi, an emergingpathogen. Curr Opin Infect Dis 2012, 25. doi:10.1097/OCO.1090b1013e3283558780.
20. Coleman RE, Maneechai N, Rachaphaew N, Kumpitak C, Miller RS, SoysengV, Thimasarn K, Sattabongkot J: Comparison of field and expert laboratorymicroscopy for active surveillance for asymptomatic Plasmodiumfalciparum and Plasmodium vivax in western Thailand. Am J Trop Med Hyg2002, 67:141–144.
21. McKenzies FE, Sirichaisinthop J, Miller RS, Gasser RA Jr, Wongsrichanalai C:Dependence of malaria detection and species diagnosis by microscopyon parasite density. Am J Trop Med Hyg 2003, 69:372–376.
22. Coatney GR, Collins WE, Warren MW, Contacos PG: The Primate Malarias.Washington: US Government Printing Office; 1971.
23. Lee K-S, Cox-Singh J, Singh B: Morphological features and differentialcounts of Plasmodium knowlesi parasites in naturally acquired humaninfections. Malar J 2009, 8:73.
24. World Health Organization: Basic Malaria Microscopy. Geneva: World HealthOrganization; 1991.
25. World Health Organization: Management of severe malaria: a practicalhandbook. 3rd edition. Geneva: World Health Organization; 2012.
26. World Health Organization: Guidelines for the Treatment of Malaria. 2ndedition. Geneva: World Health Organization; 2010.
27. World Health Organization: Meeting Report: Informal Consultation on thePublic Health Importance of Plasmodium knowlesi. Kuching, Sarawak,Malaysia: 2011.
28. Padley D, Moody A, Chiodini P, Saldanha J: Use of a rapid, single-round,multiplex PCR to detect malarial parasites and identify the speciespresent. Ann Trop Med Parasitol 2003, 97:131–137.
29. Imwong M, Tanomsing N, Pukrittayakamee S, Day NPJ, White NJ, SnounouG: Spurious amplification of a Plasmodium vivax small-subunit RNA geneby use of primers currently used to detect P. knowlesi. J Clin Microbiol2009, 47:4173.
30. Ministry of Health Malaysia: National antibiotic guideline 2008. Available:http://www.pharmacy.gov.my/aeimages//File/National_Antibiotic_Guideline_2008_edit%282%29.pdf.Accessed 3rd August 2012.
31. Douglas NM, Anstey NM, Angus BJ, Nosten F, Price RN: Artemisinincombination therapy for vivax malaria. Lancet Infect Dis 2010, 10:405–416.
32. Ahlm C, Wiström J, Carlsson H: Chloroquine resistant Plasmodium vivaxmalaria in Borneo. J Travel Med 1996, 3:124–124.
33. Baird JK: Chloroquine resistance in Plasmodium vivax. Antimicrob AgentsChemother 2004, 48:4075–4083.
34. Jamaiah I, Rohela M, Nissapatorn V, Amriana A, Sumaiyah M, Lila M,Norzawati A: Malaria: A retrospective study in Hospital Tengku AmpuanRahimah (HTAR), Klang, Selangor, Malaysia (2004–2006). Southeast AsianJ Trop Med Public Health 2007, 38:1.
35. Jamaiah I, Rohela M, Nissapatorn V, Khoo B, Khoo P, Radhiyah M, Aisyah A:Malaria: a 10-year (1994–2003) retrospective study at University MalayaMedical Center (UMMC), Kuala Lumpur, Malaysia. Southeast Asian J TropMed Public Health 2005, 36:60–63.
36. Tang T, Salas A, Ali-Tammam M, Martinez M, Lanza M, Arroyo A, Rubio J:First case of detection of Plasmodium knowlesi in Spain by Real TimePCR in a traveller from Southeast Asia. Malar J 2010, 9:219.
37. van Hellemond J, Rutten M, Koelewijn R, Zeeman A, Verweij J, Wismans P,Kocken C, van Genderen P: Human Plasmodium knowlesi infectiondetected by rapid diagnostic tests for malaria. Emerg Infect Dis 2009,15:1478.
38. Kawai S, Hirai M, Haruki K, Tanabe K, Chigusa Y: Cross-reactivity in rapiddiagnostic tests between human malaria and zoonotic simian malariaparasite Plasmodium knowlesi infections. Parasitol Int 2009, 58:300–302.
39. McCutchan T, Piper R, Makler M: Use of malaria rapid diagnostic test toidentify Plasmodium knowlesi infection. Emerg Infect Dis 2008, 14:1750.
40. Barber BE, William T, Grigg M, Piera KA, Yeo TW, Anstey NM: Evaluation ofthe sensitivity of a pLDH-based and an aldolase-based Rapid DiagnosticTest for the diagnosis of uncomplicated and severe malaria caused byPCR-confirmed Plasmodium knowlesi, P. falciparum and P. vivax. underreview.
doi:10.1186/1475-2875-12-8Cite this article as: Barber et al.: Limitations of microscopy todifferentiate Plasmodium species in a region co-endemic forPlasmodium falciparum, Plasmodium vivax and Plasmodium knowlesi.Malaria Journal 2013 12:8.
Submit your next manuscript to BioMed Centraland take full advantage of:
• Convenient online submission
• Thorough peer review
• No space constraints or color figure charges
• Immediate publication on acceptance
• Inclusion in PubMed, CAS, Scopus and Google Scholar
• Research which is freely available for redistribution
Submit your manuscript at www.biomedcentral.com/submit
Barber et al. Malaria Journal 2013, 12:8 Page 6 of 6http://www.malariajournal.com/content/12/1/8
80
knowlesi, P. falciparum, and P. vivax malaria. This study was performed alongside the
prospective study discussed in Chapter 8.
81
Implications of this paper
In this study we found that, in Sabah, microscopy does not reliably distinguish between P.
knowlesi, P. falciparum, and P. vivax malaria. P. knowlesi was commonly misdiagnosed as
either P. vivax or P. falciparum, while P. vivax was commonly diagnosed as P. knowlesi.
These findings have the following implications:
1. Inappropriate treatment decisions may be made on the basis of microscopic
misdiagnosis. In particular, this could include the use of chloroquine to treat P.
falciparum misdiagnosed as P. knowlesi. Cases of severe P. knowlesi misdiagnosed
as P. vivax would also potentially receive chloroquine if signs of severity were
missed. These findings support the use of a unified treatment strategy of ACT for all
malaria species. To support such a policy change, a randomized control trial of ACT
(artesunate-mefloquine) versus chloroquine for the treatment of uncomplicated
knowlesi and vivax malaria is currently underway in Sabah (ClinicalTrials.gov
Identifier: NCT01708876).
2. Misdiagnosis of P. vivax as P. knowlesi has particular implications for malaria control,
as patients with P. vivax may not be given anti-hypnozoite treatment, potentially
leading to increased transmission of P. vivax. Clear morphological differences exist
between P. vivax and P. knowlesi however, including the normal size of P. knowlesi-
infected RBCs in comparison to the enlarged RBCs infected by P. vivax. In our
study, feed-back to Queen Elizabeth Hospital laboratory staff resulted in a noticeable
reduction in misdiagnosis of P. vivax as P. knowlesi (data not reported). This
highlights the importance of regular audits of the performance of microscopy at all
laboratories in Sabah, with additional training provided as required. In addition, all
patients diagnosed with P. knowlesi malaria should have PCR performed in order
82
that patients with misdiagnosed P. vivax malaria can receive anti-hypnozoite
treatment.
3. Inaccuracies of microscopy limit the ability to use microscopy data for the
surveillance of P. knowlesi and other malaria species. This supports the use of
molecular methods of diagnosis for all patients diagnosed as P. knowlesi, and at
least a proportion of patients diagnosed with P. falciparum and P. vivax malaria.
Our study presented in Chapter 7 discusses the emergence of P. knowlesi in Sabah
over the past 20 years, and our conclusion that the incidence of P. knowlesi has
recently increased relies solely on microscopy data. While this is a major limitation of
this study, we do not believe that the increasing rate of notifications of “P. malariae/P.
knowlesi” across all districts of Sabah is entirely explained by microscopic over-
diagnosis of “P. malariae/P. knowlesi”. Although our results suggest that a proportion
of microscopy-diagnosed “P. malariae/P. knowlesi” cases may in fact have been P.
vivax cases, PCR-confirmed P. knowlesi cases were also commonly misdiagnosed
by microscopy as P. falciparum or P. vivax. Hence, it appears that in Sabah over-
diagnosis and under-diagnosis of P. knowlesi both commonly occur. Moreover, the
finding from our study in Chapter 7 that the overall age of all malaria cases has
increased over the same time period as notifications of “P. malariae/P. knowlesi”
have increased, cannot be explained by misdiagnosis of P. vivax or P. falciparum
and, given the rarity of PCR-confirmed P. malariae in Sabah, suggests a true
increase in incidence of P. knowlesi. However, further longitudinal surveillance of
malaria cases in Sabah using molecular methods of diagnosis is required to confirm
the trend of increasing incidence of P. knowlesi. Such studies are planned in coming
years.
83
4. The limitations of microscopy outlined in this paper support the need to develop rapid
diagnostic tests that are sensitive for P. knowlesi, and have the ability to distinguish
between P. knowlesi, P. falciparum, and P. vivax malaria.
84
The previous chapter discussed the inaccuracies of microscopy for the diagnosis of P.
knowlesi, P. falciparum and P. vivax malaria. Additional limitations of microscopy include the
need for well-maintained microscopes, and the need for well-trained microscopists.
Maintaining skill levels of microscopists can be particularly difficult in countries such as
Malaysia, where there have been dramatic reductions in malaria prevalence.
Rapid Diagnostic Tests (RDTs) provide an alternative method of diagnosis, but have not
been systematically evaluated for the diagnosis of knowlesi malaria. We therefore evaluated
two RDTs for the diagnosis of non-severe and severe knowlesi, falciparum, and vivax
malaria. This study was conducted at Queen Elizabeth Hospital, alongside the prospective
study discussed in Chapter 8.
Evaluation of the Sensitivity of a pLDH-Based and an Aldolase-BasedRapid Diagnostic Test for Diagnosis of Uncomplicated and SevereMalaria Caused by PCR-Confirmed Plasmodium knowlesi, Plasmodiumfalciparum, and Plasmodium vivax
Bridget E. Barber,a,b Timothy William,b,c Matthew J. Grigg,a,b Kim Piera,a Tsin W. Yeo,a,d Nicholas M. Ansteya,d
Global Health Division, Menzies School of Health Research and Charles Darwin University, Darwin, Northern Territory, Australiaa; Infectious Diseases Unit, Queen Elizabeth
Hospital, Kota Kinabalu, Sabah, Malaysiab; Sabah Department of Health, Kota Kinabalu, Sabah, Malaysiac; Royal Darwin Hospital, Darwin, Northern Territory, Australiad
Plasmodium knowlesi can cause severe and fatal human malaria in Southeast Asia. Rapid diagnosis of all Plasmodium species isessential for initiation of effective treatment. Rapid diagnostic tests (RDTs) are sensitive for detection of uncomplicated and se-vere falciparum malaria but have not been systematically evaluated in knowlesi malaria. At a tertiary referral hospital in Sabah,Malaysia, we prospectively evaluated the sensitivity of two combination RDTs for the diagnosis of uncomplicated and severemalaria from all three potentially fatal Plasmodium species, using a pan-Plasmodium lactate dehydrogenase (pLDH)-P. falcipa-rum histidine-rich protein 2 (PfHRP2) RDT (First Response) and a pan-Plasmodium aldolase-PfHRP2 RDT (ParaHIT). Among293 hospitalized adults with PCR-confirmed Plasmodium monoinfection, the sensitivity of the pLDH component of the pLDH-PfHRP2 RDT was 74% (95/129; 95% confidence interval [CI], 65 to 80%), 91% (110/121; 95% CI, 84 to 95%), and 95% (41/43;95% CI, 85 to 99%) for PCR-confirmed P. knowlesi, P. falciparum, and P. vivax infections, respectively, and 88% (30/34; 95% CI,73 to 95%), 90% (38/42; 95% CI, 78 to 96%), and 100% (12/12; 95% CI, 76 to 100%) among patients tested before antimalarialtreatment was begun. Sensitivity in severe malaria was 95% (36/38; 95% CI, 83 to 99), 100% (13/13; 95% CI, 77 to 100), and 100%(7/7; 95% CI, 65 to 100%), respectively. The aldolase component of the aldolase-PfHRP2 RDT performed poorly in all Plasmo-dium species. The pLDH-based RDT was highly sensitive for the diagnosis of severe malaria from all species; however, neitherthe pLDH- nor aldolase-based RDT demonstrated sufficiently high overall sensitivity for P. knowlesi. More sensitive RDTs areneeded in regions of P. knowlesi endemicity.
The simian parasite Plasmodium knowlesi is a common cause ofhuman malaria in all age groups in Malaysian Borneo (1–5). In
Sabah, Malaysia, the incidence of knowlesi malaria has increasedwith the control of Plasmodium falciparum and Plasmodium vivax(6) and accounts for up to 80% of malaria admissions to districthospitals (1, 3–5). The geographic range of P. knowlesi corre-sponds to the overlapping distribution of the simian hosts and theforest-dwelling Anopheles leucosphyrus group of mosquitoes andextends across all of Southeast Asia from southern China, Taiwan,Philippines, and Indonesia to Bangladesh and eastern India. Hu-man P. knowlesi infection has been reported in many of thesecountries and in travelers returning to regions where the parasiteis not endemic (7, 8). The 24-h replication cycle of P. knowlesi maybe associated with rapidly increasing parasite counts with conse-quent complications and fatality rates comparable to those of P.falciparum (9, 10). However, risk of death appears to be low withearly initiation of intravenous artesunate and oral artemisinincombination therapy for severe and nonsevere disease, respec-tively (9, 11). Prompt diagnosis of P. knowlesi infection is there-fore essential to allow early commencement of effective treatment.
In Malaysia and much of Southeast Asia, P. knowlesi is coen-demic with the two major human Plasmodium species, P. falcipa-rum and P. vivax. The diagnosis of malaria relies primarily onmicroscopic examination of stained blood smears. However, mi-croscopy is associated with several limitations, including the needfor well-trained staff and appropriately maintained microscopes.Maintaining high skill levels among microscopists can be partic-ularly difficult in areas of low malaria transmission. Rapid diag-
nostic tests (RDTs) provide an alternative method of detectionwhich can be performed by staff with minimal training and arenow widely used in many areas of malaria endemicity. The anti-gens detected by RDTs include P. falciparum-specific histidine-rich protein 2 (PfHRP2), genus-specific aldolase (pan-aldolase),and pan-Plasmodium lactate dehydrogenase (pan-pLDH) (12), aswell as P. vivax-specific (13, 14) and P. falciparum-specific (15)pLDH. Limited data from case reports demonstrate that RDTsdetecting pan-aldolase and pan-pLDH may be used to detect P.knowlesi (16–20); however, the sensitivity of any RDT for the di-agnosis of P. knowlesi has not been evaluated.
Data on the sensitivity of RDTs in the diagnosis of severe ma-laria from any Plasmodium species are also limited. Only fourstudies have evaluated the use of PfHRP2-based RDTs in severefalciparum malaria (21–24), with one of these also reporting onthe use of a pLDH-based RDT (21) and another reporting on theuse of an aldolase-based RDT (24). Importantly, there are no data
Received 12 December 2012 Returned for modification 7 January 2013
Accepted 16 January 2013
Published ahead of print 23 January 2013
Address correspondence to Bridget E. Barber, [email protected].
Supplemental material for this article may be found at http://dx.doi.org/10.1128
/JCM.03285-12.
Copyright © 2013, American Society for Microbiology. All Rights Reserved.
doi:10.1128/JCM.03285-12
1118 jcm.asm.org Journal of Clinical Microbiology p. 1118–1123 April 2013 Volume 51 Number 4
on March 18, 2013 by guest
http://jcm.asm
.org/D
ownloaded from
on the sensitivity of the best available RDTs in the diagnosis ofsevere malaria from the other two major causes of severe malariain Asia, P. knowlesi and P. vivax. The latter species, previouslythought to be benign, is now recognized as a cause of severe andfatal malaria in areas of endemicity, particularly in the presence ofcomorbidities (25, 26). Determining the sensitivity of RDTs insevere malaria from any Plasmodium species is important as afalse-negative result may lead to delayed or inappropriate treat-ment with consequent death.
We prospectively evaluated the sensitivity of a pan-pLDH-PfHRP2 RDT and a pan-aldolase-PfHRP2 RDT for the diagnosisof severe and nonsevere knowlesi, falciparum, and vivax malaria ata tertiary referral hospital. We chose the best-performing pan-pLDH and pan-aldolase RDTs, respectively, from the WHO RDTproduct testing rounds 1 and 2 that were commercially available atthe time of the study (27).
MATERIALS AND METHODSStudy site and referral system. This study was conducted at Queen Eliz-abeth Hospital (QEH), an adult tertiary referral hospital in Kota Kinabalu,Sabah, Malaysia. The hospital services the West Coast and Kudat Divi-sions of Sabah, with six district hospitals and a population of 1.14 million.From mid-2010, in response to ongoing malaria deaths in Sabah (9, 10),new guidelines were implemented, including tertiary hospital referral forall malaria patients with a thick blood film reported as 4� (indicating �10parasites/high-power microscopy field) or with any evidence of severemalaria. Treatment was commenced pretransfer, and a pretreatmentblood film was sent with the patient. Local health clinics within the KotaKinabalu area were required to refer all malaria patients to QEH for ad-mission, with treatment commencing on arrival.
Subjects. All patients admitted to QEH with a microscopic diagnosisof malaria were assessed for eligibility from September 2010 to October2011 as part of a prospective study of the epidemiology, clinical spectrum,and pathophysiology of knowlesi malaria, reported separately (11). Non-pregnant patients of �12 years old were enrolled if they were within 18 hof commencing malaria treatment (with 18 h chosen for both pathophys-iological and logistical purposes), had no major comorbidities, and hadnot previously been enrolled in the study. Patients who had mixed speciesinfection (n � 16) or were PCR negative were retrospectively excluded.Patients were classified as having severe malaria using modified 2010WHO severe malaria criteria (11, 28, 29). Written informed consent wasprovided by patients or their relatives. Approvals were obtained from theEthics Committees of the Malaysian Ministry of Health and MenziesSchool of Health Research.
RDT selection. WHO RDT product testing performance results fromrounds 1 and 2 were used to identify the RDTs with the highest reportedsensitivities for both P. falciparum and P. vivax (27). We then selected thepan-pLDH and pan-aldolase RDT with the highest reported sensitivities(27) that could be sourced commercially at the time of study: the FirstResponse Malaria Antigen pLDH/HRP2 Combo Card Test (First Re-sponse; Premier Medical Corporation Ltd., Mumbai, India), which de-tects pan-pLDH and PfHRP2, and the ParaHIT Total Dipstick (ParaHIT;SPAN Diagnostics, Ltd., Surat, India), which detects pan-aldolase andPfHRP2.
The panel detection scores (PDSs; a composite index of test positivityas well as of intertest and interlot consistency in performance [30]) for theFirst Response RDT were 84% and 100% for detection of P. falciparum atlow (200 parasites/�l) and high (2,000 to 5,000 parasites/�l) densities,respectively, and 75% and 100% for the detection of P. vivax at low andhigh densities, respectively. For the ParaHIT RDT, the PDSs at low andhigh densities were 64% and 99%, respectively, for detection of P. falcip-arum and 10% and 98%, respectively, for the detection of P. vivax.
Study procedures. Standardized data forms were used to record de-mographic and clinical information. Venous blood was collected in a
labeled tube containing citrate, theophylline, adenosine, and dipyridam-ole (CTAD), and thick and thin blood smears were prepared. RDTs wereperformed on enrolment according to the manufacturers’ instructions,and results were recorded by one of two research laboratory techniciansunaware of the microscopy result. Results were recorded as follows: neg-ative, no clearly visible band; 1, faint band; 2, a band darker than 1 butlighter than that of the control; 3, same as the control band; 4, a banddarker than the control. Results were cross-checked regularly by the studyclinician to ensure consistent reporting between laboratory technicians.The ParaHIT RDT was unavailable for a 4-week period during April toMay 2011, with its use then discontinued in September 2011.
Blood smear examination was performed by microscopists at referringdistrict hospitals or at QEH, with slides later cross-checked by an experi-enced research microscopist. Parasite density was quantified by the re-search microscopist using pretreatment slides and reported as parasitesper 200 leukocytes or per 1,000 erythrocytes and converted to the numberof parasites/�l using the patient’s leukocyte count or hematocrit, respec-tively. Where pretreatment slides were unavailable (6%), referring hospi-tal microscopy was used, and the grades 1� to 4� were converted intonumbers of parasites/�l using the relevant median parasite density. Par-asite DNA was extracted and PCR was performed as reported previously(11), using previously described methods for P. falciparum, P. vivax, Plas-modium ovale, and Plasmodium malariae (31), and P. knowlesi (32) detec-tion.
Statistical analysis. Data were analyzed using STATA, version 10.1(StataCorp LP, College Station, TX). Intergroup differences were com-pared using a Kruskal-Wallis test for nonnormally distributed continuousvariables, and a �2 test was used for categorical variables. The sensitivity ofRDTs for each species was defined as the total number of positive resultsdivided by the total number of samples positive for that species by PCR;95% confidence intervals (CIs) were estimated using Wilson’s method.Spearman’s correlation coefficient was used to assess the association be-tween parasite count and intensity of the RDT band.
RESULTSBaseline demographics and clinical features. From September2010 to October 2011, 293 patients had pan-pLDH-PfHRP2 withor without aldolase-PfHRP2 RDTs performed, including 129 pa-tients with P. knowlesi, 121 with P. falciparum, and 43 with P. vivaxinfections. Baseline demographics, reported previously (11), areshown in Table 1. Patients with knowlesi malaria were signifi-cantly older than those with falciparum or vivax malaria, with amedian age of 46 years versus 27 and 24 years, respectively (P �0.001). The majority (73%) of patients were male, and this did notdiffer between infecting species. Most patients (66%) withknowlesi malaria were referred from district hospitals, and 74%had commenced antimalarial treatment prior to enrolment in thestudy. The median time from commencing treatment to enrol-ment for patients with knowlesi malaria was 5.1 h (range, 0 to18 h). Severe malaria occurred in 38/129 (29%) patients with P.knowlesi, 13/121 (11%) with P. falciparum, and 7/43 (16%) with P.vivax infections (Table 1).
Pan-pLDH-PfHRP2 (First Response) RDT. Results of thepLDH component of the First Response RDT are shown in Fig. 1and 2 and in Table S1 in the supplemental material. Among allpatients with knowlesi malaria, 95/129 had positive results, givinga sensitivity of 74% (95% CI, 65 to 80%). The sensitivity of the testwas higher when only pretreatment samples were analyzed, with30/34 positive results (sensitivity, 88% [95% CI, 73 to 95%]) (seeTable S1). The four patients with pretreatment samples andnegative pLDH results had parasite counts of 75, 282, 320, and11,078 parasites/�l. The lowest parasite count among patients
RDTs for Diagnosis of Knowlesi Malaria
April 2013 Volume 51 Number 4 jcm.asm.org 1119
on March 18, 2013 by guest
http://jcm.asm
.org/D
ownloaded from
with pretreatment samples and a positive pLDH result was 907parasites/�l.
The sensitivity of the pLDH test for knowlesi malaria increasedwith increasing parasite count (Fig. 2). Among patients enrolledprior to treatment and with parasite counts of �1,000 parasites/
�l, sensitivity was 97% (95% CI, 83 to 99%), but it was only 25%among patients with parasite counts of �1,000 parasites/�l.Among the 38 patients with severe knowlesi malaria, 36 had apositive result (sensitivity of 95% [95% CI, 83 to 99%]); 31 (82%)
TABLE 1 Baseline demographic and clinical features
Parameter
Value by infecting organisma
P valueP. knowlesi (n � 129) P. falciparum (n � 121) P. vivax (n � 43)
Age (yr)Median (IQR) 46 (29–58) 27 (17–40) 24 (18–42) <0.001Range 14–83 13–78 13–79
No. of males (%) 96 (74) 87 (71) 33 (77) 0.769Source of referral (no. of patients [%]) <0.001
QEH or local health clinic 44 (34) 75 (61) 22 (51)District Hospital 86 (66) 47 (39) 21 (49)
No. of patients (%) enrolledpretreatment
34 (26) 42 (34) 12 (28) 0.231
Median time on malaria treatment(h [IQR])
5.13 (0–11.47) 4.04 (0–11.69) 4.81 (0–9.68) 0.726
No. of patients (%) with severemalaria
38 (29%) 13 (11%) 7 (46%) <0.001
Median parasite density count/�l(IQR [range])
All patients 8,537 (1,980–35,495 [32–584,015]) 11,610 (4,387–36,977 [16–959,383]) 4,938 (2,848–12,927 [72–84,403]) 0.009Nonsevere malaria 4,841 (1,581–14,994 [65–62,360]) 10,937 (4,002–32,326 [16–218,921]) 4,753 (2,369–10,316 [72–46,849]) <0.001Severe malaria 80,359 (25,874–168,279 [32–584,015]) 72,270 (27,905–273,909 [625–959,383]) 10,243 (4,387–40,895 [460–84,403]) 0.070
Severity criteria among patients withsevere malaria
Median no. of criteria (range) 2 (1–6) 3 (1–4) 1 (1–2) 0.039No. of patients (%) with:
Hyperparasitemia 18 (47) 2 (15) 0Respiratory distress 14 (37) 4 (31) 1 (14)Hypotension 13 (34) 6 (49) 3 (43)Jaundice 20 (53) 9 (69) 2 (29)Acute kidney injury 9 (24) 3 (23) 0Metabolic acidosis 4 (11) 4 (31) 0Severe anemia 2 (5) 2 (15) 0Abnormal bleeding 2 (5) 1 (8) 1 (14)Multiple convulsions 0 0 1 (14)Coma 0 0 0
a Table includes data previously reported (11).bBoldface indicates significant values.
FIG 1 Parasite count by pLDH-band intensity for all samples. Horizontallines indicate medians; boxes indicate interquartile ranges; vertical lines indi-cate ranges. Among patients with P. knowlesi infection, the pLDH band wasnegative in 34 (26%), 1� in 34 (26%), 2� in 17 (13%), 3� in 17 (13%), 4� in24 (18%), and recorded only as positive in 3 (2%). Among patients with P.falciparum infection, the pLDH band was negative in 11 (9%), 1� in 35 (29%),2� in 17 (14%), 3� in 24 (20%), 4� in 30 (25%), and recorded as positive in4 (3%). Among patients with P. vivax infection, the pLDH band was negativein 2 (5%), 1� in 4 (9%), 2� in 5 (12%), 3� in 7 (16%), 4� in 20 (47%), andrecorded as positive in 5 (12%).
FIG 2 Sensitivity of the pLDH-component of the First Response RDT byparasite count. Wide horizontal bars indicate sensitivity. Vertical lines repre-sent 95% confidence intervals. Among patients with P. knowlesi infection, 18(14%) had �103 parasites/�l, 52 (40%) had 103 to �104 parasites/�l, 41 (32%)had 104 to �105 parasites/�l, and 18 (14%) had �105 parasites/�l. Amongpatients with P. falciparum infection, 10 (8%) had �103 parasites/�l, 42 (35%)had 103 to �104 parasites/�l, 60 (50%) had 104 to �105 parasites/�l, and 9(7%) had �105 parasites/�l. Among patients with P. vivax infection, 8 (19%)had �103 parasites/�l, 21 (49%) had 103 to �104 parasites/�l, and 14 (33%)had 104 to �105 parasites/�l.
Barber et al.
1120 jcm.asm.org Journal of Clinical Microbiology
on March 18, 2013 by guest
http://jcm.asm
.org/D
ownloaded from
of these patients had commenced malaria treatment a median of5.7 h prior to enrolment. The two patients with severe knowlesimalaria and negative results had parasite counts of 3,486 and48,833 parasites/�l and were enrolled 13 and 9 h, respectively,after treatment.
The sensitivity of the pLDH test for P. falciparum was 91%(95% CI, 84 to 95%) and was not affected by the inclusion ofposttreatment samples. Sensitivity of the test decreased at parasitecounts of �1,000 parasites/�l, with only 6/10 (60%) positive re-sults, compared to 104/111 (94%) among patients with parasitecounts of �1,000 parasites/�l. The lowest parasite count detectedwas 16 parasites/�l; however, parasitemia ranged from 234 to25,308 (median, 1,698) parasites/�l among the 11 patients withnegative results. The sensitivity of the pLDH test among patientswith severe falciparum malaria was 100% (95% CI, 77 to 100%),with nine (69%) patients having already commenced malariatreatment a median of 3.8 h prior to enrolment.
The sensitivity of the pLDH test for vivax malaria was 95%(95% CI, 85 to 99%) when all samples were included and 100%(95% CI, 76 to 100) among the 12 pretreatment samples, whichincluded a parasitemia of 72 parasites/�l. The sensitivity of thepLDH test among the seven patients with severe vivax malaria was100% (95% CI, 65 to 100%); six of these patients commencedtreatment a median 9.2 h prior to enrolment.
The intensity of the pLDH band (grades 1� to 4�) was a goodindicator of parasitemia among patients with knowlesi malaria,with the median parasite count increasing with each successivegrade (Spearman’s correlation coefficient, 0.61; P � 0.0001) (seeTable S2 in the supplemental material). Patients with knowlesimalaria and a 4� pLDH band also had a 7.3-fold (95% CI, 2.3- to24-fold) higher risk of severe disease (defined using modified 2010WHO criteria [11]) than those with pLDH bands with intensitiesof 1� to 3� (P � 0.0001). Among these patients 17/24 (71%) hadsevere malaria, compared to 17/68 (25%) patients with knowlesimalaria and pLDH bands graded 1� to 3� (Table 2). The risk ofseverity increased further among patients of �50 years old withknowlesi malaria and a band intensity of 4�, of whom 11/13(85%) had severe disease. The median intensity score of the pLDHband was 4� (interquartile range [IQR], 2� to 4�) among allpatients with severe knowlesi malaria, compared to a grade of 1�(IQR, 0 to 3�) among patients with nonsevere knowlesi malaria(P � 0.0001). This difference was less marked among patientswith severe and nonsevere falciparum malaria (median intensityscores of 4 [IQR, 2 to 4] and 2 [IQR, 1 to 4], respectively; P �0.007) and among patients with severe and nonsevere vivax ma-laria (median intensity scores of 4 [IQR, 3 to 4] and 4 [IQR, 2.5 to4], respectively; P � 0.360).
The sensitivity of the First Response PfHRP2 RDT for falcipa-rum malaria was 98% overall and 100% among those with severedisease. The two patients with false-negative results were enrolled15 and 17 h after commencing antimalarial treatment and hadpretreatment parasite counts of 10,098 and 101,686 parasites/�l.No patient with knowlesi or vivax malaria had a positive HRP2result.
Pan-aldolase-PfHRP2 (ParaHIT) RDT. The sensitivity of thealdolase component of the ParaHIT RDT was low for all species(see Table S2 in the supplemental material). Only 22/96 (23%),37/84 (44%), and 20/36 (56%) patients with P. knowlesi, P. falcip-arum, and P. vivax infections, respectively, had positive aldolaseresults, with similar results obtained when posttreatment sampleswere excluded (sensitivities of 27%, 50%, and 73% for P. knowlesi,P. falciparum, and P. vivax, respectively). The sensitivity of thealdolase test among patients with severe malaria was 39% (95%CI, 24 to 56%), 50% (95% CI, 22 to 78%) and 83% (95% CI, 44 to97%) for patients with P. knowlesi (n � 31), P. falciparum (n � 8),and P. vivax (n � 6) infections, respectively. Sensitivity was loweven among patients with higher parasite counts. Among patientswith pretreatment samples and parasite counts of �10,000 parasites/�l, only 5/11 (45%) and 10/19 (56%) of those with P. knowlesi and P.falciparum infections, respectively, had positive results.
Among patients with P. falciparum infection, 79/84 (94%) hada positive PfHRP2 result, including all eight patients with severemalaria. Four of the five patients with nonsevere malaria and false-negative results were enrolled 4 to 16 h after treatment, with pre-treatment parasite counts of 444 to 15,503 parasites/�l, while onepatient was enrolled pretreatment with a parasitemia of 3,990 par-asites/�l. One patient with P. knowlesi and two with P. vivax in-fections had false-positive ParaHIT PfHRP2 results.
DISCUSSION
This is the first study to systematically evaluate the sensitivities ofRDTs for the diagnosis of knowlesi malaria and the first to evalu-ate the sensitivities of the best available RDTs for the diagnosis ofnonfalciparum severe malaria. The pLDH-based RDT performedbetter than the aldolase-based RDT for the diagnosis of knowlesimalaria of any severity and demonstrated good sensitivity at par-asite counts of �1,000 parasites/�l. This RDT performed wellamong patients with severe knowlesi malaria, among whom sensitiv-ity was 95%. Sensitivity was poor, however, at low parasite counts.Given that P. knowlesi is associated with lower parasite counts than P.falciparum among patients with nonsevere malaria (11), that the par-asite has a rapid replication cycle and is associated with relatively highrates of severe disease (11), and that late diagnosis has been associatedwith fatal outcomes (33), the poor sensitivity at low parasite counts ofboth pLDH- and aldolase-based RDTs limits their utility in areas ofknowlesi malaria endemicity.
Despite the suboptimal sensitivity of the pLDH-based test inknowlesi malaria, we found that the intensity of the pLDH bandcorrelated well with P. knowlesi parasitemia and risk of severedisease. In particular, a 4� band was associated with an increasedrisk of complications, particularly in older patients who are atgreater risk of severe disease (2, 9, 11). P. knowlesi parasite countsare strongly associated with complications, with the risk of severedisease increasing 11-fold and 28-fold with parasite counts of�20,000 parasites/�l and �100,000 parasites/�l, respectively(11). Despite this, in Sabah, malaria slides are often reported onlyon a scale of 1� to 4�, with slides with �10 parasites per high-
TABLE 2 Risk of severe malaria by intensity of First Response pLDHband
Infectingorganism
No. of severemalaria patients (%)by pLDH bandintensity
Odds ratio (95%CI) P value4� 1�–3�
P. knowlesi 17 (71) 17 (25) 7.29 (2.33–24.01) 0.0001P. falciparum 8 (27) 5 (6.7) 5.16 (1.31–22.85) 0.005P. vivax 4 (20) 2 (13) 1.75 (0.21–21.81) 0.549
RDTs for Diagnosis of Knowlesi Malaria
April 2013 Volume 51 Number 4 jcm.asm.org 1121
on March 18, 2013 by guest
http://jcm.asm
.org/D
ownloaded from
power microscopy field (or approximately 4,000 parasites/�l) re-ported only as 4�. Consequently, hyperparasitemia, with its asso-ciated risk of complications, may be underrecognized byclinicians, potentially leading to undertreatment and monitoring.Use of a pLDH-based RDT may assist clinicians to recognize pa-tients at high risk of severe disease in settings where parasitecounts are unavailable.
The pLDH-based RDT performed well for P. vivax, with anoverall sensitivity of 95% and sensitivity of 100% among thosewith severe disease. False-negative results occurred in only twopatients, both of whom had already commenced antimalarialtreatment. This finding confirms another recent report of the ex-cellent sensitivity of the First Response RDT for detecting P. vivax(34). In our study this RDT also performed well for P. falciparum;the sensitivity of the HRP2 component of the test was 98% overalland 100% among patients with severe disease. This provides con-firmation from Southeast Asia of recent reports of the very highsensitivity of HRP2-based RDTs in severe falciparum malaria inAfrica (21, 23) and India (22), despite the potential for geographicgenetic variability in PfHRP2 (34–37).
The aldolase component of the aldolase-based RDT performedpoorly for all species, with a sensitivity of �50% for knowlesi andfalciparum malaria. Sensitivity was poor even at higher parasitedensities, and less than half of patients with severe knowlesi ma-laria recorded a positive result. Using an aldolase-based RDT withpoorer performance characteristics on WHO product testing, arecent study also reported poor sensitivity for the diagnosis ofsevere vivax malaria (24). The similarly poor performance of al-dolase-based RDTs has also been reported in uncomplicated ma-laria (38, 39). Taking these results together, the use of aldolase-based RDTs for diagnosis of nonfalciparum species of any severitycannot therefore be recommended (40).
A limitation of both the RDTs evaluated in this study is thatthey do not differentiate between the nonfalciparum species. Thisis particularly problematic given the difficulties with microscopicdiagnosis of P. knowlesi, the ring forms of which resemble those ofP. falciparum while the more mature trophozoites are indistin-guishable from those of P. malariae. Consequently, patients withknowlesi malaria continue to be misdiagnosed with the more be-nign P. malariae infection, leading to inappropriate managementwith potentially fatal outcomes (10). Furthermore, misdiagnosisof P. vivax as P. knowlesi is also common (4, 10, 41), and this hasbeen associated with lack of antihypnozoite treatment and conse-quent relapses from P. vivax infection (41).
An RDT that differentiates between species is therefore ur-gently needed in areas of knowlesi malaria endemicity. P. knowlesihas been shown to cross-react with both P. falciparum-specific andP. vivax-specific pLDHs (17, 18, 20, 42), and RDTs that combinethese antigens with HRP2 may allow differentiation between P.vivax, P. falciparum, and P. knowlesi monoinfections. Prospectiveevaluation of the use of these RDTs in areas of knowlesi malariaendemicity is needed.
Our study had several limitations. Most importantly, the ma-jority of patients were referred from district hospitals and hadalready commenced malaria treatment prior to enrolment in thestudy. For these patients RDTs were done on posttreatment bloodsamples, possibly leading to an underestimation of sensitivity.However, the reported sensitivities of the RDTs under these cir-cumstances reflect the real-world utility of RDTs in referral hos-pitals, where prior treatment is common. Moreover, we also re-
port sensitivity calculations using only patients enrolled prior totreatment (i.e., on initial presentation to the hospital). A secondlimitation was that each RDT was read by a single observer; how-ever, regular cross-checks were performed to ensure consistentreporting. The high sensitivity of the PfHRP2 component of bothRDTs for diagnosing P. falciparum infection also supports thereliability of our results. Third, patients with knowlesi malariareferred to a tertiary hospital are likely to have had higher para-sitemias than those diagnosed in primary care settings (2), sug-gesting that the suboptimal sensitivities of the pLDH and aldolaseRDTs found for nonsevere P. knowlesi malaria may be even lowerthan reported here. Fourth, our study excluded pregnant womenand children of �12 years old; the performance of RDTs for thediagnosis of P. knowlesi infection in these groups will need to beevaluated in future studies. Finally, the design of our study did notallow evaluation of the specificity of the RDTs as only patientswith PCR-confirmed malaria were enrolled; further cross-sec-tional studies including patients with nonmalarial fevers are re-quired to provide this information.
In conclusion, this study found that the pLDH-based First Re-sponse RDT had high sensitivity for severe malaria from all threePlasmodium species evaluated and that among patients withknowlesi malaria, the intensity of the pLDH band predicted pa-tients at risk of complications. However, neither of the RDTs hadsufficient overall sensitivity for detecting P. knowlesi. The WorldHealth Organization (WHO) recommends a minimum sensitivityof 95% for P. falciparum densities of �100 parasites/�l (43); whilethere is no recommended threshold for nonfalciparum malaria, P.knowlesi has a 3-fold-greater likelihood of causing severe disease(11) and fatality rates comparable to those of P. falciparum, andhence the minimum sensitivity thresholds required for RDTsshould be at least as high for P. knowlesi as for P. falciparum.Further studies are needed to develop reliable RDTs for the diag-nosis of knowlesi malaria with high sensitivity and the ability todifferentiate between species.
ACKNOWLEDGMENTS
We thank all the patients enrolled in this study, all the clinical staff involved intheir care, Rita Wong, Beatrice Wong and Ann Wee for their help with clinicaland laboratory study procedures, Ferryanto Chalfein for performing micros-copy, Jutta Marfurt and Sarah Auburn for supervising the PCR assays, theClinical Research Centre, Sabah, for logistical support, and the Director Gen-eral of Health, Malaysia, for permission to publish this study.
The study was funded by the Australian National Health and MedicalResearch Council (program grant 496600; scholarship to B.E.B. and fel-lowships to N.M.A. and T.W.Y.)
REFERENCES1. Barber BE, William T, Dhararaj P, Anderios F, Grigg MJ, Yeo TW,
Anstey NM. 2012. Epidemiology of Plasmodium knowlesi malaria innortheast Sabah, Malaysia: family clusters and wide age distribution. Ma-lar. J. 11:401. doi:10.1186/1475-2875-11-401.
2. Daneshvar C, Davis TM, Cox-Singh J, Rafa’ee M, Zakaria S, Divis P,Singh B. 2009. Clinical and laboratory features of human Plasmodiumknowlesi infection. Clin. Infect. Dis. 49:852– 860.
3. Joveen-Neoh WF, Chong KL, Wong CM, Lau TY. 2011. Incidence ofmalaria in the Interior Division of Sabah, Malaysian Borneo, based onnested PCR. J. Parasitol. Res. 2011:104284. doi:10.1155/2011/104284.
4. Cox-Singh J, Davis TM, Lee KS, Shamsul SS, Matusop A, Ratnam S,Rahman HA, Conway DJ, Singh B. 2008. Plasmodium knowlesi malaria inhumans is widely distributed and potentially life threatening. Clin. Infect.Dis. 46:165–171.
5. Barber BE, William T, Jikal M, Jilip J, Dhararaj P, Menon J, Yeo TW,
Barber et al.
1122 jcm.asm.org Journal of Clinical Microbiology
on March 18, 2013 by guest
http://jcm.asm
.org/D
ownloaded from
Anstey NM. 2011. Plasmodium knowlesi malaria in children. Emerg. In-fect. Dis. 17:814 – 820.
6. William T, Rahman HA, Jelip J, Ibrahim MY, Menon J, Grigg MJ, YeoTW, Anstey NM, Barber BE. 24 January 2013. Increasing incidence ofPlasmodium knowlesi malaria following control of P. falciparum and P.vivax malaria in Sabah, Malaysia. PLoS Negl. Trop. Dis. [Epub ahead ofprint.] doi:10.1371/journal.pntd.0002026.
7. Kantele A, Sakari Jokiranta T. 2011. Review of cases with the emergingfifth human malaria parasite, Plasmodium knowlesi. Clin. Infect. Dis. 52:1356 –1362.
8. Cox-Singh J. 2012. Zoonotic malaria: Plasmodium knowlesi, an emergingpathogen. Curr. Opin. Infect. Dis. 25:530 –536.
9. William T, Menon J, Rajahram G, Chan L, Ma G, Donaldson S, KhooS, Fredrick C, Jilip J, Anstey NM, Yeo TW. 2011. Severe Plasmodiumknowlesi malaria in a tertiary hospital, Sabah, Malaysia. Emerg. Infect. Dis.17:1248 –1255.
10. Rajahram G, Barber BE, William T, Menon J, Anstey NM, Yeo TW.2012. Deaths due to Plasmodium knowlesi malaria in Sabah, Malaysia:association with reporting as P. malariae and delayed parenteral artesu-nate. Malar. J. 11:284. doi:10.1186/1475-2875-11-284.
11. Barber BE, William T, Grigg MJ, Menon J, Auburn S, Marfurt J, AnsteyNM, Yeo TW. 2013. A prospective comparative study of knowlesi, falcip-arum and vivax malaria in Sabah, Malaysia: high proportion with severedisease from Plasmodium knowlesi and P. vivax but no mortality with earlyreferral and artesunate therapy. Clin. Infect. Dis. 56:383–397. doi:10.1093/cid/cis902.
12. Murray CK, Gasser RA, Jr, Magill AJ, Miller RS. 2008. Update on rapiddiagnostic testing for malaria. Clin. Microbiol. Rev. 21:97–110.
13. Gillet P, Bosselaers K, Cnops L, Bottieau E, Van Esbroeck M, Jacobs J.2009. Evaluation of the SD FK70 malaria Ag Plasmodium vivax rapid di-agnostic test in a non-endemic setting. Malar. J. 8:129. doi:10.1186/1475-2875-8-129.
14. Gillet P, Van Dijk DPJ, Bottieau E, Cnops L, Van Esbroeck M, JacobsJ. 2009. Test characteristics of the SD FK80 Plasmodium falciparum/Plasmodium vivax malaria rapid diagnostic test in a non-endemic setting.Malar J. 8:262. doi:10.1186/1475-2875-8-262.
15. Palmer CJ, Lindo JF, Klaskala WI, Quesada JA, Kaminsky R, Baum MK,Ager AL. 1998. Evaluation of the OptiMAL test for rapid diagnosis ofPlasmodium vivax and Plasmodium falciparum malaria. J. Clin. Microbiol.36:203–206.
16. Berry A, Iriart X, Wilhelm N, Valentin A, Cassaing S, Witkowski B,Benoit-Vical F, Menard S, Olagnier D, Fillaux J, Sire S, Le CoustumierA, Magnaval JF. 2011. Case report: imported Plasmodium knowlesi ma-laria in a French tourist returning from Thailand. Am. J. Trop. Med. Hyg.84:535–538.
17. Ta TT, Salas A, Ali-Tammam M, Martinez Mdel C, Lanza M, Arroyo A,Rubio JM. 2010. First case of detection of Plasmodium knowlesi in Spainby real time PCR in a traveller from Southeast Asia. Malar. J. 9:219. doi:10.1186/1475-2875-9-219.
18. van Hellemond J, Rutten M, Koelewijn R, Zeeman A, Verweij J, Wis-mans P, Kocken C, van Genderen P. 2009. Human Plasmodium knowlesiinfection detected by rapid diagnostic tests for malaria. Emerg. Infect. Dis.15:1478 –1480.
19. Ong C, Lee S, Koh W, Ooi E, Tambyah P. 2009. Monkey malaria inhumans: a diagnostic dilemma with conflicting laboratory data. Am. J.Trop. Med. Hyg. 80:927–928.
20. Kawai S, Hirai M, Haruki K, Tanabe K, Chigusa Y. 2009. Cross-reactivity in rapid diagnostic tests between human malaria and zoonoticsimian malaria parasite Plasmodium knowlesi infections. Parasitol. Int.58:300 –302.
21. Hendriksen ICE, Mtove G, Pedro AJ, Gomes E, Silamut K, Lee SJ,Mwambuli A, Gesase S, Reyburn H, Day NPJ. 2011. Evaluation of aPfHRP2 and a pLDH-based rapid diagnostic test for the diagnosis of se-vere malaria in 2 populations of African children. Clin. Infect. Dis. 52:1100 –1107.
22. Shujatullah F, Malik A, Khan HM. 2006. Comparison of different diag-nostic techniques in Plasmodium falciparum cerebral malaria. J. VectorBorne Dis. 43:186 –190.
23. Birku Y, Welday D, Ayele D, Shepherd A. 1999. Rapid diagnosis ofsevere malaria based on the detection of Pf-Hrp-2 antigen. Ethiop. Med. J.37:173. doi:10.1097/QCO.0b013e32832f14c1.
24. Manning L, Laman M, Rosanas-Urgell A, Turlach B, Aipit S, Bona C,Warrell J, Siba P, Mueller I, Davis TME. 2012. Rapid antigen detection
tests for malaria diagnosis in severely ill Papua New Guinean children: acomparative study using Bayesian latent class models. PLoS One 7:e48701.doi:10.1371/journal.pone.0048701.
25. Price RN, Douglas NM, Anstey NM. 2009. New developments in Plas-modium vivax malaria: severe disease and the rise of chloroquine resis-tance. Curr. Opin. Infect. Dis. 22:430 – 435.
26. Anstey NM, Russell B, Yeo TW, Price RN. 2009. The pathophysiology ofvivax malaria. Trends Parasitol. 25:220 –227.
27. World Health Organization. 2010. Malaria rapid diagnostic test perfor-mance. Summary results of WHO product testing of malaria RDTs:rounds 1 and 2 (2008 –2009). WHO, Geneva, Switzerland. http://www2.wpro.who.int/NR/rdonlyres/E4823BBF-69FC-45DE-B924-1A9FCDD7406C/0/RoundIII_SummaryFINAL.pdf.
28. World Health Organization. 2012. Management of severe malaria: apractical handbook, 3rd ed. World Health Organization, Geneva, Switzer-land.
29. World Health Organization. 2010. Guidelines for the treatment of ma-laria, 2nd ed. World Health Organization, Geneva, Switzerland. http://www.who.int/malaria/publications/atoz/9789241547925/en/index.html.
30. WHO. 2012. Information note on recommended selection criteria forprocurement of malaria rapid diagnostic tests (RDTs). WHO Global Ma-laria Programme, Geneva, Switzerland. http://www.who.int/malaria/diagnosis_treatment/diagnosis/RDT_selection_criteria.pdf.
31. Padley D, Moody A, Chiodini P, Saldanha J. 2003. Use of a rapid,single-round, multiplex PCR to detect malarial parasites and identify thespecies present. Ann. Trop. Med. Parasitol. 97:131–137.
32. Imwong M, Tanomsing N, Pukrittayakamee S, Day NPJ, White NJ,Snounou G. 2009. Spurious amplification of a Plasmodium vivax small-subunit RNA gene by use of primers currently used to detect P. knowlesi. J.Clin. Microbiol. 47:4173– 4175.
33. Rajahram GS, Barber BE, Tan WM, Yeo TW, William T. 2013. Casereport: fatal Plasmodium knowlesi malaria following an atypical clinicalpresentation and delayed diagnosis. Med. J. Malaysia 68:48 – 49.
34. Maltha J, Gamboa D, Bendezu J, Sanchez L, Cnops L, Gillet P, JacobsJ. 2012. Rapid diagnostic tests for malaria diagnosis in the Peruvian Am-azon: impact of pfhrp2 gene deletions and cross-reactions. PLoS One7:e43094. doi:10.1371/journal.pone.0043094.
35. Gamboa D, Ho MF, Bendezu J, Torres K, Chiodini PL, Barnwell JW,Incardona S, Perkins M, Bell D, McCarthy J. 2010. A large proportion ofP. falciparum isolates in the Amazon region of Peru lack pfhrp2 andpfhrp3: implications for malaria rapid diagnostic tests. PLoS One 5:e8091.doi:10.1371/journal.pone.0008091.
36. Baker J, McCarthy J, Gatton M, Kyle DE, Belizario V, Luchavez J, BellD, Cheng Q. 2005. Genetic diversity of Plasmodium falciparum histidine-rich protein 2 (PfHRP2) and its effect on the performance of PfHRP2-based rapid diagnostic tests. J. Infect. Dis. 192:870 – 877.
37. Lee N, Baker J, Andrews KT, Gatton ML, Bell D, Cheng Q, McCarthyJ. 2006. Effect of sequence variation in Plasmodium falciparum histidine-rich protein 2 on binding of specific monoclonal antibodies: implicationsfor rapid diagnostic tests for malaria. J. Clin. Microbiol. 44:2773–2778.
38. Singh N, Shukla M, Shukla M, Mehra R, Sharma S, Bharti P, Singh M,Singh A, Gunasekar A. 2010. Field and laboratory comparative evalua-tion of rapid malaria diagnostic tests versus traditional and moleculartechniques in India. Malar. J. 9:191. doi:10.1186/1475-2875-9-191.
39. World Health Organization. 2011. Malaria rapid diagnostic test perfor-mance. Results of WHO product testing of malaria RDTs: round 3 (2010 –2011). World Health Organization, Geneva, Switzerland. http://www.who.int/tdr/publications/documents/rdt3.pdf.
40. Tjitra E, Suprianto S, Dyer M, Currie BJ, Anstey NM. 1999. Fieldevaluation of the ICT malaria Pf/Pv immunochromatographic test fordetection of Plasmodium falciparum and Plasmodium vivax in patientswith a presumptive clinical diagnosis of malaria in Eastern Indonesia. J.Clin. Microbiol. 37:2412–2417.
41. Barber BE, William T, Grigg MJ, Yeo TW, Anstey NM. 2013. Limita-tions of microscopy to differentiate Plasmodium species in a region co-endemic for Plasmodium falciparum, Plasmodium vivax and Plasmodiumknowlesi. Malar. J. 12:8. doi:10.1186/1475-2875-12-8.
42. McCutchan T, Piper R, Makler M. 2008. Use of malaria rapid diagnostictest to identify Plasmodium knowlesi infection. Emerg. Infect. Dis. 14:1750 –1752.
43. Bell D, Peeling RW. 2006. Evaluation of rapid diagnostic tests: malaria.Nat. Rev. Microbiol. 4:S34 –S38.
RDTs for Diagnosis of Knowlesi Malaria
April 2013 Volume 51 Number 4 jcm.asm.org 1123
on March 18, 2013 by guest
http://jcm.asm
.org/D
ownloaded from
85
Implications of and questions arising from this paper
This paper reports the first systematic evaluation of RDTs for the diagnosis of knowlesi
malaria. Findings from this paper include:
1. The aldolase-based RDT evaluated in this study performed poorly for P. knowlesi, P.
falciparum and P. vivax malaria, with a sensitivity of 23%, 44% and 56% respectively.
Sensitivity was poor among patients with severe disease, and among those with high
parasite densities. Similar results have been previously reported, and with better
performing non-aldolase based RDTs now widely available, these results suggest
that aldolase-based RDTs should not be used for the diagnosis of malaria.
2. The sensitivity of the pLDH-based RDT evaluated in this study was inadequate for
the diagnosis of low-parasitemia P. knowlesi infections. This is particularly
problematic given that among patients with uncomplicated malaria, P. knowlesi
parasite counts are lower than those of P. falciparum (Chapter 8). Furthermore, our
study was performed at a tertiary-referral hospital, where many patients were
referred specifically because of a malaria blood slide reported as “4+” parasitemia.
Parasite counts are therefore likely to be higher than would be seen at a district
hospital or primary care clinics, and hence the sensitivity of pLDH-based RDTs may
be even lower in these different settings. More sensitive RDTs are therefore required
for use in P. knowlesi-endemic areas.
3. Among patients with knowlesi malaria, the intensity of the pLDH band of the pLDH-
based RDT correlated well with parasitemia, and correlated with risk of severe
disease, with severe disease occurring in 71% of patients with a “4+” pLDH band. In
the absence of accurately quantitated parasite counts (as is often the case in Sabah),
pLDH-based RDTs may provide additional information to clinicians regarding which
86
patients are at risk of developing complications and therefore require more intensive
management.
4. A major limitation of the pLDH and aldolase-based RDTs evaluated in this study is
that they do not distinguish between P. knowlesi, P. falciparum, and P. vivax malaria.
Given the limitations of microscopy for differentiating these species, discussed in the
previous chapter, there is a need for development of RDTs that are specific for P.
knowlesi.
87
In our prospective comparative study of patients admitted with knowlesi, falciparum and
vivax malaria at Queen Elizabeth Hospital (QEH), described in Chapter 8, we found that P.
knowlesi appeared to be at least as pathogenic as P. falciparum. However, very little is
known about the pathogenic mechanisms of disease. As discussed in the introductory
Chapter 4, the key processes underlying severe disease in falciparum malaria include the
sequestration of parasitized RBCs within vital organs, decreased RBC deformability, and
increased aggregation of RBCs, with these processes leading to microvascular obstruction
and impaired organ perfusion. In the only autopsy report of human knowlesi malaria,
widespread accumulation of parasitized cells was also seen, including in cerebral vessels
(27). In addition, in early simian studies, impairment of microvascular flow appeared to be a
critical factor in the pathogenesis of knowlesi malaria (140, 141). However, there have been
no studies investigating microcirculatory flow in humans with knowlesi malaria, and, with the
exception of one report describing the ability of P. knowlesi to cytoadhere to ICAM-1 and
VCAM (105), there are no data regarding potential mechanisms by which P. knowlesi
accumulates within the microvasculature.
Decreased red blood cell deformability is one process which contributes to microcirculatory
dysfunction in falciparum malaria (134, 135, 190), and has also been demonstrated in
rhesus macaques infected with P. knowlesi (142). We therefore investigated the
deformability of RBCs among patients with knowlesi and falciparum malaria admitted to
QEH. This chapter discusses the first of several pathophysiological studies that we have, or
are in the process of, conducting at QEH, with the remainder being outside the scope of this
thesis. These are discussed in more detail in Chapter 13.
Manuscript in preparation:
Impaired red cell deformability in knowlesi malaria in proportion to
disease severity
1
Abstract
Plasmodium knowlesi commonly causes severe and fatal malaria in Malaysian Borneo, but
little is known about the pathogenesis of disease. In severe falciparum malaria,
sequestration of parasitized red cells results from cytoadherence to host endothelium and
decreased red blood cell deformability (RBC-D), leading to microvascular obstruction, tissue
hypoxia and organ dysfunction. In knowlesi malaria microvascular accumulation of
parasitized cells also occurs, however the mechanisms are unknown. Reduced deformability
with sludging of P. knowlesi-infected red cells has been demonstrated in rhesus macaques,
but has not been studied in humans. Using ektacytometry we measured RBC-D in adults
with severe (n=21) and non-severe (n=61) knowlesi malaria and severe (n=8) and non-
severe (n=82) falciparum malaria; and 15 healthy controls. At a shear stress of 30 Pascals,
RBC-D was reduced in patients with severe (elongation index [EI]=0.496, IQR 0.456-0.528)
and non-severe (EI=0.551, IQR 0.494-0.569) knowlesi malaria compared to controls
(EI=0.583, IQR 0.576-0.590; p<0.0001 for both comparisons), and reduced in severe
compared to non-severe knowlesi malaria (p=0.002). RBC-D was similar among patients
with severe (EI=0.510, IQR 0.496-0.539) and non-severe falciparum malaria (EI=0.516, IQR
0.475-0.558), but was reduced in both groups compared to controls (p=0.0045 and
p<0.0001 respectively). RBC-D did not differ significantly between patients with severe
knowlesi and severe falciparum malaria. Among patients with knowlesi, but not falciparum
malaria, RBC-D was inversely correlated with parasite count (spearman’s correlation
coefficient =-0.37, p=0.0006). Among patients with knowlesi malaria, reduced RBC-D may
contribute to microvascular sludging, microvascular accumulation of parasitized red cells and
impaired organ perfusion in severe disease.
2
Background
The simian parasite Plasmodium knowlesi causes severe and fatal human malaria in
Malaysian Borneo, with the risk of severity three times that of P. falciparum (1).
Manifestations of severe disease include acute respiratory distress syndrome, acute kidney
injury, shock, severe anaemia, jaundice and hyperparasitemia, with multiple organ failure
common. Little is known about the pathogenesis of severe disease in knowlesi malaria. In
falciparum malaria, the sequestration of parasitized red blood cells (RBCs) within vital
organs is a central process in severe disease, resulting from cytoadherence of parasitized
red cells to host endothelium, increased aggregation, and decreased RBC deformability, and
leading to microvascular obstruction and impaired organ perfusion (2, 3). In severe knowlesi
malaria, the only human autopsy report to date also noted widespread accumulation of
infected RBCs in capillaries and venules, including in the brain (4). Although electron
microscopy was not performed in this forensic autopsy, lack of cytoadherence was
suggested by the interspersion of infected RBCs with uninfected cells, and the lack of
marginalization of parasitized cells. Furthermore, ICAM-1, which mediates cytoadherence to
brain endothelial cells in falciparum malaria (5, 6), was not detected. Mechanisms by which
P. knowlesi accumulates in the microvasculature may differ from those of P. falciparum.
In an early study of P. knowlesi in rhesus macaques, parasitized and non-parasitized RBCs
were reported to be coated with a “precipitate”, binding RBCs together and changing blood
to a “thick, muck-like sludge”, leading to widespread tissue anoxia (7). In another study
involving rhesus macaques, P. knowlesi-infected RBCs were found to have reduced
deformability, with the reduction in deformability particularly marked at high parasitemia (8).
While parasite counts in human P. knowlesi infections are generally low, parasite count is
the major risk factor for severe disease. Among humans with severe knowlesi malaria
reduced deformability of RBCs may contribute to microvascular obstruction and organ
dysfunction, as occurs with severe falciparum malaria. We therefore investigated the
3
relationship between red cell deformability, parasitemia, and disease severity, among
patients with knowlesi malaria.
Methods
Study site and referral system
The study was conducted at Queen Elizabeth Hospital, an adult tertiary-referral hospital in
Kota Kinabalu, Sabah, Malaysia. This hospital has modern intensive care facilities for
invasive ventilation and renal replacement therapy, and serves as a referral hospital for the
West Coast and Kudat divisions, comprising 6 district hospitals and a population of 1.14
million. At the time of this study, referral criteria for malaria patients from district hospitals
had been standardized and included all malaria patients with a thick blood film reported as
“4+” (indicating >10 parasites/high-power microscopy field) or with any evidence of severe
malaria. Treatment was commenced pre- transfer and a pre-treatment blood film sent with
the patient. Local health clinics within the Kota Kinabalu area were required to refer all
malaria patients to QEH for admission, with treatment commenced on arrival.
Subjects
All patients admitted to QEH with a microscopic diagnosis of malaria were assessed for
eligibility from November 2010-May 2011, and September 2011–April 2012, using enrolment
criteria previously described (1). Patients enrolled through to October 2011 were also
included in a previously reported prospective study of the epidemiology and clinical features
of knowlesi malaria (1). The two enrolment periods coincided with the local availablility of a
Laser-assisted Optical Rotational Cell Analyzer (LORCA). Non-pregnant patients ≥12 years
old were enrolled if they had a blood film positive for P. knowlesi or P. falciparum, were
within 18 hours of commencing malaria treatment, had no major comorbidities, and had not
4
previously been enrolled in the study. Patients who had misdiagnosed P. vivax or mixed
species infection or were PCR-negative were retrospectively excluded. Controls were
healthy friends or relatives of enrolled patients, with no history of fever in the preceding two
weeks and a blood film negative for malaria parasites. Written informed consent was
provided by study participants or their relatives. Approvals were obtained from the Ethics
Committees of the Malaysian Ministry of Health and Menzies School of Health Research.
Study procedures
Standardized data forms were used to record demographic and clinical information.
Treatment was administered according to hospital guidelines, which included artemether-
lumefantrine for uncomplicated falciparum and knowlesi malaria, and intravenous artesunate
for severe malaria of any species, or if deemed warranted by the treating clinician. Patients
were classified as having severe malaria using modified 2010 WHO Severe Malaria Criteria
(1, 9, 10). Patients remained in hospital until they were afebrile and had negative blood films
on two consecutive days.
Laboratory procedures
Blood smear examination was performed by microscopists at referring district hospitals or at
QEH, with slides later cross-checked by an experienced research microscopist. Parasite
density was quantified by the research microscopist using pre-treatment slides, and reported
as parasites per 200 leukocytes or per 1000 erythrocytes and converted to parasites/μL
using the patient’s leukocyte count or haematocrit respectively. Full blood count, mean red
cell volume (MCV), haemoglobin electrophoresis, biochemistry, acid-base parameters, and
lactate (by bedside blood analysis; iSTAT System) were obtained on enrolment. Parasite
DNA was extracted and PCR performed, as reported (1), using previously described
methods for P. falciparum, P. vivax, P. ovale, and P.malariae (11); and P. knowlesi (12). To
quantitate total parasite biomass among patients with P. falciparum, plasma HRP2 was
5
measured by ELISA, as previously described (13). Because thalassaemia can effect RBC-D
(14), haemoglobin electrophoresis was performed on blood samples collected on enrolment.
RBC-D was measured on enrolment, and on day 3 for patients with knowlesi malaria, using
a Laser-assisted Optical Rotational Cell Analyzer (LORCA; Mechatronics, Hoorn,
Netherlands). With this method whole blood is added to a highly viscous medium (5%
polyvinylpyrrolidine [PVP] in phosphate-buffered saline) and the RBC suspension is sheared
between two concentric rotating cylinders at a constant temperature of 37oC. The increasing
rotation of the outer cylinder leads to an increasing shear stress that causes the RBCs to
elongate and align themselves in the fluid layer. A laser beam is directed through this fluid
layer and forms a diffraction pattern on the screen behind it. This diffraction pattern
undergoes computer analysis to produce an elongation index (EI), defined by the formula EI
= (L – W)/(L + W), where L and W are the length and width, respectively, of the diffraction
pattern. A lower EI at a given shear stress indicates reduced RBC-D. RBC-D was assessed
at shear stresses of 1.7 Pa and 30 Pa. Shear stresses of 1.7 Pa are encountered in the
capillaries (15); shear stresses of 30 Pa, although supraphysiological, provide information on
cell geometry, in particular surface area to volume ratios (16).
Statistical Analysis
Data were analyzed using Stata software, v 10.1. For continuous variables, the Kruskal-
Wallis test was used to compare intergroup differences, and the Mann-Whitney test was
used for post hoc pairwise comparisons. Categorical variables were compared using the χ2
or Fisher exact test. The association between RBC-D and other variables was compared
using Spearman’s correlation coefficient and multiple linear regression. Logistic regression
was used to assess predictors of malaria severity. Paired measurements of RBC-D were
assessed using the Wilcoxon signed-rank test.
6
Results
Demographic and clinical characteristics
Total enrolments included 61 patients with non-severe and 21 patients with severe knowlesi
malaria, 82 patients with non-severe and 8 patients with severe falciparum malaria. A total
of 120 had been enrolled into the prospective series described in Chapter 8 (1), and 52 were
additional cases (including 19 and 7 with non-severe and severe knowlesi malaria,
respectively, and 25 and one with non-severe and severe falciparum malaria, respectively).
A total of 15 healthy controls were enrolled. Baseline clinical and laboratory details are
listed in Table 1. Median parasite count among patients with non-severe knowlesi malaria
(2291 parasites/μL) was lower than among patients with non-severe falciparum malaria
(9210 parasites/μL; p=0.003), but did not differ significantly between patients with severe
knowlesi malaria (139,793 parasites/μL) and severe falciparum malaria (25,468
parasites/μL; p=0.157). Among the 21 patients with severe knowlesi malaria, most common
severity criteria included parasitemia >100,000 parasites/μL (n=12, 57%), jaundice (bilirubin
>43 μmol/L with either creatinine <132 μmol/L or parasitemia >20,000 parasites/μL; n=12,
57%), respiratory distress (n=10, 48%), hypotension (systolic blood pressure ≤ 80 mm Hg;
n=6, 29%) and acute kidney injury (n=5, 24%). Six patients one severity criterion, 6 had 2
criteria, 6 had 3, 2 had 4 and one had 6 severity criteria. Among patients with severe
falciparum malaria most common severity criteria included hypotension (n=4, 50%), jaundice
(bilirubin >43 μmol/L with either creatinine <132 μmol/L or parasitemia >100,000
parasites/μL; n=4, 50%), respiratory distress (n=3, 38%) and metabolic acidosis (n=3, 38%).
Three patients had 1 severity criterion, 4 had 3, and one had 4 severity criteria. No patient
had coma, and no deaths occurred from either parasite species in this study.
The MCV of RBCs did not differ significantly between patients with severe and non-severe
knowlesi malaria, or between patients with severe and non-severe falciparum malaria.
7
Among patients with severe knowlesi malaria 7 patients had microcytosis (MCV<80 fl). Of
these, hemoglobin electrophoresis was normal in 5 patients. One patient was heterozygous
for HbE, with an MCV of 63.7 fl and haemoglobin of 11.0 g/dL, and one, with MCV of 66.4 fl
and haemoglobin of 8.7 g/dL, did not have thalassemia screening performed. Five of 8
patients with severe falciparum malaria had an MCV<80 fl. Haemoglobin electrophoresis
was normal in 4 of these patients, but was not performed in one patient (MCV 79.7 fl,
haemoglobin 16.6 g/dL).
Red cell deformability
P. knowlesi malaria
RBC-D among severe and non-severe knowlesi malaria patients, and controls, is shown in
Figure 1. At a shear stress of 30 Pa, RBC-D was reduced in patients with severe (EI=0.496,
IQR 0.456-0.528) and non-severe (EI=0.551, IQR 0.494-0.569) knowlesi malaria compared
to controls (EI=0.583, IQR 0.576-0.590; p<0.0001 for both comparisons), and reduced in
severe compared to non-severe knowlesi malaria (p=0.002). The median EI among patients
with severe knowlesi malaria remained the same after exclusion of the patient heterozygous
for HbE and the patient with microcytic anaemia and unknown thalassemia status (median
EI=0.496, IQR 0.456 – 0.529 after exclusion of these patients). The difference in RBC-D
between patients with severe and non-severe knowlesi malaria remained significant after
exclusion of patients with microcytosis (p=0.003), as did the difference in RBC-D between
controls and patients with non-severe knowlesi malaria (p=0.020), and between controls and
patients with severe knowlesi malaria (p=0.003). Among patients with knowlesi malaria
RBC-D at 30 Pa was associated with parasite count (Spearman’s correlation coefficient = -
0.37, p=0.0006). In a logistic regression model including parasite count and RBC-D, only
parasite count predicted severe malaria. RBC-D was also associated with lactate (p=0.021),
platelet count (p=0.027), haemoglobin nadir (p=0.006), percentage of schizonts (p=0.002),
8
and number of severity criteria (p=0.065), however none of these associations remained
significant after adjusting for parasite count.
At a lower shear stress of 1.7 Pa, the difference in RBC-D between patients with severe
knowlesi malaria (EI=0.170, IQR 0.161 – 0.201), non-severe knowlesi malaria (EI=0.194,
IQR 0.168 – 0.223) and controls (EI=0.203, IQR 0.178 – 0.222; p=0.038), was not
statistically significant (p=0.131 by Kruskal-Wallis test).
No significant improvement was detected in RBC-D at 30 Pa among 27 patients with non-
severe knowlesi malaria who had RBC-D reassessed on day 3 (median baseline EI=0.529,
IQR 0.454 – 0.0.566; median day 3 EI=0.542, IQR 0.497 – 0.575; p=0.471). Among 12
patients with severe knowlesi malaria, RBC-D improved from a baseline median of 0.493
(IQR 0.449 – 0.515) to a day 3 median of 0.508 (IQR 0.497 – 0.536; p=0.072).
P. falciparum malaria
Among patients with falciparum malaria, at a shear stress of 30 Pa, RBC-D differed between
patients with severe malaria (EI=0.510, IQR 0.496 - 0.539), non-severe malaria (EI= 0.516,
IQR 0.475 - 0.558) and controls (EI=0.583, IQR 0.576-0.590; p=0.0002 by Kruskal-Wallis),
although the difference between patients with severe and non-severe falciparum malaria
was not significant on post-hoc analysis (Figure 1). At a lower shear stress of 1.7 Pa, RBC-
D was lower among patients with severe (EI=0.180, IQR 0.163 – 0.197) and non-severe
malaria (EI=0.195, IQR 0.177 – 0.214) compared to controls (p=0.033 by Kruskal Wallis),
however on post-hoc analysis only the difference between patients with non-severe
falciparum malaria and controls was significant (p=0.018).
Among patients with falciparum malaria there was no association between RBC-D and
parasite count at either 1.7 Pa or 30 Pa, however there was a weak negative association
9
between RBC-D and HRP2 at 30 Pa (Spearman’s correlation coefficient=-0.24, p=0.02). At
1.7 Pa, but not at 30 Pa, RBC-D was associated with haemoglobin nadir (Spearman’s
correlation coefficient=0.32, p=0.002).
RBC-D was lower among patients with non-severe falciparum malaria than it was among
patients with non-severe knowlesi malaria, at 1.7 Pa (p=0.02) and 30 Pa (p=0.05). Among
patients with severe malaria however there was no significant difference between patients
with falciparum malaria and those with knowlesi malaria.
Discussion
The ability of RBCs to deform as they pass through capillaries is a critical factor in
maintaining normal microvascular flow. In this study we demonstrate that deformability of
RBCs is reduced in knowlesi malaria, and is proportional to disease severity. Similar results
have been obtained from previous studies involving patients with severe falciparum malaria.
In these studies, also using LORCA, RBC-D was reduced among adults and children with
falciparum malaria, and was associated with severe anaemia and mortality (17-19). In our
study, deformability of RBCs was at least as low in severe knowlesi malaria as it was in
severe falciparum malaria.
In falciparum malaria, reduced deformability is known to be a characteristic of both
parasitized and unparasitized RBCs. Among parasitized cells, young ring forms cause a
reduction in deformability by increasing the sphericity of the RBC, hence reducing the
surface area to volume ratio (20, 21). With parasite maturation deformability reduces further,
with the decreased deformability of trophozoite- and schizont-infected RBCs resulting from
increased membrane rigidity, and increased internal viscosity from the parasites themselves
(20). This reduction in deformability of P. falciparum-infected RBCs is an important
contributor to the microvascular sequestration and obstruction that characterizes severe
10
falciparum malaria. However, measurements obtained using the LORCA primarily reflect the
deformability of unparasitized cells (19). These measurements are derived by computer
analysis of the diffraction pattern produced when a laser beam is directed through a RBC
suspension, and hence reflect the mean deformability of all red blood cells in the
suspension. Extremely high parasitemias would be required for the deformability of infected
red blood cells alone to have a substantial effect on the measurement provided by the
LORCA. In addition, the lack of association between P. falciparum parasitemia and RBC
deformability in previous studies (17) further supports the importance of reduced
deformability of unparasitized cells. In our study we also did not find an association between
RBC-D and parasitemia among patients with falciparum malaria, and found only a weak
association between RBC-D and HRP2, providing additional evidence that reduced RBC-D
among patients with falciparum malaria is primarily accounted for by reduced deformability of
unparasitized cells.
The mechanisms of increased rigidity of unparasitized cells in falciparum malaria have not
been fully determined. In previous studies, reduction in RBC-D was more marked at the
lower shear stresses (1.7 Pa) such as those found in capillaries, where intracellular viscosity
and membrane deformability are important determinants of RBC-D (17, 19). Dondorp et al.
reported that manipulation of internal viscosity did not result in restoration of deformability,
suggesting that membrane changes are likely responsible for the reduced RBC-D (22).
Membrane damage caused by heme products has been thought to be a possible contributor,
and hemin was recently shown to reduce RBC-D in a dose-dependent manner, with the
reduction in RBC-D prevented and reverted by the anti-oxidant N-acetylcysteine (23).
Reduced RBC-D has also been shown to result from binding of a P. falciparum exoantigen
to normal RBCs (24). In our study, reduction in RBC-D among patients with falciparum
malaria was less marked at a shear stress of 1.7 Pa compared to 30 Pa. However, our
numbers were small, and the range of measurements in our study was greater at the lower
shear stress; hence the significance of this finding is uncertain.
11
In P. knowlesi malaria, parasite counts are generally lower than those of patients with
falciparum malaria. It is therefore almost certain that the results obtained using the LORCA
in our study primarily reflect decreased deformability of unparasitized cells. However, in
contrast to P. falciparum, we found that RBC deformability was associated with P. knowlesi
parasitemia, suggesting that the mechanisms of reduced deformability of unparasitized cells
in knowlesi malaria may differ from those that occur in falciparum malaria.
Our findings of reduced RBC-D among patients with knowlesi malaria are consistent with an
early study that measured viscosity and filterability of P. knowlesi-infected red cell
suspensions among rhesus macaques (8). In that study the viscosity and resistance to flow
of infected blood was greater than that of controls, reflecting a decrease in RBC
deformability. Resistance to flow increased with increasing parasite count, and was greater
with more mature trophozoites compared to younger trophozoites. Furthermore, Miller et al.
demonstrated the exclusion of P.knowlesi schizonts from rouleaux, suggesting increased
rigidity of these cells. In other early simian studies, Knisely et al., although not measuring
RBC deformability directly, demonstrated that circulatory changes were a crucial contributor
to the pathogenesis of knowlesi malaria (7, 25). In these studies, parasitized and
unparasitized RBCs were noted to be coated with an “adhesive precipitate” that bound
parasitized and unparasitized RBCs together in clumps and changed blood to “thick, muck-
like sludge” that resisted flow through the microvasculature (7). This “sludge” resulted in
impaction of numerous small vessels with this process appearing to underlie coma and
death (25). Although coma has not been reported in the 21st century literature of human
knowlesi malaria to date, accumulation of parasitized RBCs within cerebral vessels was
noted in the single autopsy report of a fatal case of human knowlesi malaria (4). In
falciparum malaria, additional processes contributing to sequestration of parasitized cells
and reduced microvascular flow include adherence of parasitized cells to activated
endothelium (5, 26), to unparasitized cells to form rosettes (27-29), and to other parasitized
12
cells in platelet-mediated clumps (30). None of these processes have been investigated in
human knowlesi malaria, and require further investigation.
In conclusion, this study demonstrates that RBC-D is reduced among patients with knowlesi
malaria in proportion to disease severity, and in severe disease is at least as low as occurs
among patients with severe falciparum malaria. Reduced deformability may contribute to
microvascular sludging, microvascular accumulation of parasitized red cells and impaired
organ perfusion in severe knowlesi malaria in humans. Further studies will be required to
determine the relative reduction in deformability of unparasitized cells compared to
parasitized cells, to delineate the mechanisms of reduced deformability, and to fully
determine the pathophysiological consequences of reduced RBC deformability among
humans with knowlesi malaria.
13
Tabl
e 1.
Epi
dem
iolo
gica
l and
clin
ical
feat
ures
of m
alar
ia p
atie
nts
and
cont
rols
on
enro
lmen
t
Plas
mod
ium
kno
wle
si
Plas
mod
ium
falc
ipar
um
Co
ntro
ls (n
=15)
N
on-s
ever
e (n
=61)
Se
vere
(n=2
1)
P va
lue,
seve
re
vs. n
on-s
ever
e N
on-s
ever
e (n
=82)
Se
vere
Pf (
n=8)
P
valu
e, se
vere
vs.
no
n-se
vere
Ag
e, y
ears
Med
ian,
IQR
38 (2
2 - 4
5)
36 (2
6 - 5
2)
50 (3
8 - 6
1)
0.02
0 27
.5 (2
0 - 3
9)
22 (1
6.5
- 34)
0.
391
Rang
e 19
- 58
16
- 71
20
- 69
13 -
62
13 -
50
M
ale
sex,
n (%
) 11
(73)
46
(75)
18
(86)
0.
325
64 (7
8)
5 (6
3)
0.32
1 Pa
rasit
e co
unt,
para
sites
/μL
22
91 (9
30 -
13,6
80)
139,
793
(39,
564
- 23
7,64
8)
<0.0
001
9210
(215
4 - 3
2,26
7)
25,4
68 (4
307
- 103
,852
) 0.
234
Hb,
14.4
(13.
3 –
15.4
)*
12.9
(12.
2 –
13.7
) 12
.8 (1
1.8
– 14
.2)
0.71
8 13
.4 (1
1.3
– 14
.4)
13.1
(9.1
– 1
4.1)
0.
399
MCV
8
6.3
(79.
5 –
89.1
)*
81.
4 (7
5.1
– 85
) 8
2.8
(78.
4 –
86.7
) 0.
279
80.
4 (7
5.6
– 85
.7)
78.
8 (7
1.6
– 84
.1)
0.52
3 N
adir
Hb
11
.6 (1
0.7
- 12.
8)
10.4
(9.1
- 11
.4)
0.00
07
11.8
(10.
5 - 1
2.8)
10
.1 (8
.9 -
11.6
) 0.
059
Plat
elet
s
56 (4
1 - 8
3)
24 (1
6 - 3
3)
<0.0
001
74 (4
2 - 1
34)
29 (2
3 - 5
6)
0.00
7 Cr
eatin
ine
88
(76
- 109
) 12
2 (1
13 -
187)
<0
.000
1 86
.5 (7
2 - 1
05)
112.
5 (7
5.5
- 246
) 0.
158
Lact
ate
1.17
(0.8
5 - 1
.39)
1.
67 (1
.15
- 2.1
6)
0.00
14
1.18
(0.9
2 - 1
.55)
1.
53 (1
.2 -
2.04
) 0.
061
Bilir
ubin
15.7
(12.
1 - 2
4)
49.2
(26
- 94.
8)
<0.0
001
18.3
(11.
8 - 3
0.1)
53
.6 (2
3.6
- 75.
2)
0.00
9 N
umbe
r of s
ever
ity
crite
ria
2 (1
- 3)
3
(1 -
3)
Seve
rity
crite
ria, n
(%):
Hy
perp
aras
item
ia
12 (5
7)
1 (1
3)
Re
spira
tory
dist
ress
10 (4
8)
3 (3
8)
Hy
pote
nsio
n
6
(29)
4
(50)
Jaun
dice
12 (5
7)
4 (5
0)
Ac
ute
kidn
ey in
jury
5
(24)
2
(25)
Met
abol
ic a
cido
sis
2
(10)
3
(38)
seve
re a
naem
ia
1 (5
)
1
(12.
5)
Ab
norm
al b
leed
ing
3 (1
4)
1 (1
2.5)
RBC-
D at
SS
1.7
Pa^
0.20
3 (0
.178
- 0.
222)
0.
194
(0.1
68 -
0.22
3)
0.17
(0.1
61 -
0.20
1)
NA
0.17
95 (0
.163
- 0.
197)
a 0.1
95 (0
.177
- 0.
214)
0.
171
RBC-
D at
SS
30 P
a# 0.
583
(0.5
76 -
0.59
0)
0.55
1 (0
.494
- 0.
569)
b 0.
496
(0.4
56 -
0.52
8)b
0.00
18
0.51
55 (0
.475
- 0.
558)
c 0.
5095
(0.4
96 -
0.53
9)d
0.94
9
U
nles
s oth
erw
ise in
dica
ted,
dat
a ar
e m
edia
n, IQ
R. N
A =
Not
ass
esse
d *D
ata
wer
e m
issin
g fo
r 3 c
ontr
ols
^ Fo
r diff
eren
ce b
etw
een
cont
rols,
non
-sev
ere
mal
aria
and
seve
re m
alar
ia, u
sing
Krus
kal-W
allil
s tes
t, p=
0.13
1 fo
r P. k
now
lesi
and
p=0.
033
for P
. fal
cipa
rum
# F
or d
iffer
ence
bet
wee
n co
ntro
ls, n
on-s
ever
e m
alar
ia a
nd se
vere
mal
aria
, usin
g Kr
uska
l-Wal
lils t
est,
p=0.
0001
for P
. kno
wle
si an
d p=
0.00
02 fo
r P. f
alci
paru
m
a : p=0
.017
8 vs
con
trol
s; b : p
=0.0
001
vs c
ontr
ols;
c : p<0
.000
1 vs
con
trol
s; d
: p=0
.004
5 vs
con
trol
s
14
Figure 1. Red blood cell deformability among patients with P. knowlesi malaria, at a shear stress of 30 Pa. Wide horizontal lines indicate median, boxes indicate inter-quartile range, and short horizontal lines indicate range.
Elo
ng
ati
on
Ind
ex
C o n trols
N o n -se v e re
P k
S e v e reP k
C o n trols
N o n -se v e re
P f
S e v e reP f
0 .3
0 .4
0 .5
0 .6
0 .7
0 .0 0 0 1 0 .0 0 2
0 .0 0 0 1
< 0 .0 0 0 1 0 .9 5
0 .0 0 0 2
15
References
1. Barber BE, William T, Grigg MJ, Menon J, Auburn S, Marfurt J, et al. A prospective
comparative study of knowlesi, falciparum and vivax malaria in Sabah, Malaysia: high
proportion with severe disease from Plasmodium knowlesi and P. vivax but no
mortality with early referral and artesunate therapy. Clin Infect Dis. 2013;56(3):383-97.
2. Clark IA, Alleva LM, Mills AC, Cowden WB. Pathogenesis of malaria and clinically
similar syndromes. Clin Microbiol Rev. 2004;17:509-39.
3. Dondorp A, Pongponratn E, White N. Reduced microcirculatory flow in severe
falciparum malaria: pathophysiology and electron-microscopic pathology. Acta Trop.
2004;89(3):309-17.
4. Cox-Singh J, Hiu J, Lucas SB, Divis PC, Zulkarnaen M, Chandran P, et al. Severe
malaria-a case of fatal Plasmodium knowlesi infection with post-mortem findings.
Malar J. 2010;9:10.
5. Turner GDH, Morrison H, Jones M, Davis TME, Looareesuwan S, Buley ID, et al. An
immunohistochemical study of the pathology of fatal malaria: evidence for widespread
endothelial activation and a potential role for intercellular adhesion molecule-1 in
cerebral sequestration. Am J Pathol. 1994;145(5):1057.
6. Berendt A, Simmons D, Tansey J, Newbold C, Marsh K. Intercellular adhesion
molecule-1 is an endothelial cell adhesion receptor for Plasmodium falciparum. Nature.
1989;341(6237):57-9.
7. Knisely MH, Stratman-Thomas WK. Knowlesi malaria in monkeys; microscopic
pathological circulatory physiology of rhesus monkeys during acute Plasmodium
knowlesi malaria. J Natl Mal Soc. 1945;4:285.
8. Miller LH, Usami S, Chien S. Alteration in the rheological properties of Palsmodium
knowlesi-infected red cells. A possible mechanisms for capillary obstruction. J Clin
Invest. 1971;50:1451-5.
9. World Health Organization. Management of severe malaria: a practical handbook, 3rd
edition. Geneva: World Health Organization; 2012.
16
10. World Health Organization. Guidelines for the treatment of malaria - 2nd edition.
Geneva: World Health Organization; 2010.
11. Padley D, Moody A, Chiodini P, Saldanha J. Use of a rapid, single-round, multiplex
PCR to detect malarial parasites and identify the species present. Ann Trop Med
Parasitol. 2003;97(2):131-7.
12. Imwong M, Tanomsing N, Pukrittayakamee S, Day NPJ, White NJ, Snounou G.
Spurious amplification of a Plasmodium vivax small-subunit RNA gene by use of
primers currently used to detect P. knowlesi. J Clin Microbiol. 2009;47(12):4173.
13. Dondorp AM, Desakorn V, Pongtavornpinyo W, Sahassananda D, Silamut K,
Chotivanich K, et al. Estimation of the total parasite biomass in acute falciparum
malaria from plasma PfHRP2. PLoS Med. 2005;2(8):e204.
14. Dondorp A, Chotivanich K, Fucharoen S, Silamut K, Vreeken J, Kager P, et al. Red
cell deformability, splenic function and anaemia in thalassaemia. Br J Haematol.
1999;105(2):505-8.
15. Chien S. Physiological and pathophysiological significance of hemorheology. Clinical
Hemorheology: Springer; 1986. p. 125-64.
16. Bessis M, Mohandas N, Feo C. Automated ektacytometry: a new method of measuring
red cell deformability and red cell indices. Blood Cells. 1980;6(3):315.
17. Dondorp A, Angus B, Hardeman M, Chotivanich K, Silamut K, Ruangveerayuth R, et
al. Prognostic significance of reduced red blood cell deformability in severe falciparum
malaria. Am J Trop Med Hyg. 1997;57(5):507.
18. Dondorp A, Nyanoti M, Kager P, Mithwani S, Vreeken J, Marsh K. The role of reduced
red cell deformability in the pathogenesis of severe falciparum malaria and its
restoration by blood transfusion. Trans R Soc Trop Med Hyg. 2002;96(3):282-6.
19. Dondorp A, Angus B, Chotivanich K, Silamut K, Ruangveerayuth R, Hardeman M, et
al. Red blood cell deformability as a predictor of anemia in severe falciparum malaria.
Am J Trop Med Hyg. 1999;60(5):733.
17
20. Nash G, O'Brien E, Gordon-Smith E, Dormandy J. Abnormalities in the mechanical
properties of red blood cells caused by Plasmodium falciparum. Blood.
1989;74(2):855-61.
21. Paulitschke M, Nash G. Membrane rigidity of red blood cells parasitized by different
strains of Plasmodium falciparum. J Lab Clin Med. 1993;122(5):581.
22. Dondorp A, Kager P, Vreeken J, White N. Abnormal blood flow and red cell
deformability in severe falciparum malaria. Parasitology Today. 2000;16:228-32.
23. Nuchsongsin F, Chotivanich K, Charunwatthana P, Fausta O-S, Taramelli D, Day NP,
et al. Effects of malaria heme products on red blood cell deformability. Am J Trop Med
Hyg. 2007;77(4):617-22.
24. Naumann KM, Jones GL, Saul A, Smith R. Plasmodium falciparum exo-antigen alters
erythrocyte membrane deformability. FEBS Lett. 1991;292(1):95-7.
25. Knisely MH, Strathman-Thomas WK, Eliot TS, Bloch EH. Knowlesi malaria in monkeys
IIA: a first step in the separation of the mechanical pathologic circluatory factors of one
slude disease from possible specific toxic factors of that disease. Angiology.
1964;15(9):411-6.
26. Craig A, Scherf A. Molecules on the surface of the Plasmodium falciparum infected
erythrocyte and their role in malaria pathogenesis and immune evasion. Mol Biochem
Parasitol. 2001;115(2):129-43.
27. David PH, Handunnetti SM, Leech JH, Gamage P, Mendis KN. Rosetting: a new
cytoadherence property of malaria-infected erythrocytes. The American Journal of
Tropical Medicine and Hygiene. 1988;38(2):289.
28. Kaul D, Roth EJ, Nagel R, Howard R, Handunnetti S. Rosetting of Plasmodium
falciparum-infected red blood cells with uninfected red blood cells enhances
microvascular obstruction under flow conditions. Blood. 1991;78(3):812-9.
29. Udomsangpetch R, Wåhlin B, Carlson J, Berzins K, Torii M, Aikawa M, et al.
Plasmodium falciparum-infected erythrocytes form spontaneous erythrocyte rosettes. J
Exp Med. 1989;169(5):1835-40.
18
30. Pain A, Ferguson DJ, Kai O, Urban BC, Lowe B, Marsh K, et al. Platelet-mediated
clumping of Plasmodium falciparum-infected erythrocytes is a common adhesive
phenotype and is associated with severe malaria. Proceedings of the National
Academy of Sciences. 2001;98(4):1805-10.
88
Limitations of and questions arising from this paper
In this paper we found that deformability of RBCs was reduced in patients with knowlesi
malaria in proportion to disease severity. This finding is consistent with similar studies
involving patients with severe falciparum malaria, and is also consistent with early studies
involving rhesus macaques infected with knowlesi malaria. The results of our study indicate
that reduced deformability of RBCs among patients with knowlesi malaria may be an
important contributor to the pathogenesis of severe disease.
However, our study was associated with several important limitations. The number of
patients included in this study was small, particularly the number of patients with severe
knowlesi and severe falciparum malaria. This was partly due to the unavailability of the
LORCA during certain time periods, including May to September 2011 which coincided with
a period of peak enrollments into our QEH prospective study. Various technical problems
with the LORCA apparatus at certain times also prevented continuous enrolment of all
malaria patients.
A second limitation of our study is that there was considerable variation in RBC-D
measurements obtained, particularly at the lower shear stress of 1.7 Pa. Multiple factors
influence the measurements obtained using the LORCA, including technique of inserting the
RBC suspension into the rotating cylinders, adequate cleaning of the cup between
measurements and batch variability in viscosity of the PVP. These factors may have
contributed to the variability of our measurements at a low shear stress, and may have
limited our ability to detect significant differences between patient groups at this stress level.
The degree of reduced deformability according to shear stress is important as it provides an
indication of the mechanisms of reduced deformability. Larger studies using LORCA among
patients with P. knowlesi will be needed to determine the stress level at which reduced RBC-
D is most marked. These studies are currently underway.
89
Further studies will also be required to clarify whether the reduction in RBC deformability
among patients with knowlesi malaria is due to a reduction in deformability of parasitized
cells, or unparasitized cells, or both. In falciparum malaria micropipette aspiration
techniques have been used to evaluate the deformability of parasitized cells, including the
degree of increased rigidity according to parasite stage (113). These techniques have also
been used to evaluate the mechanisms of reduced deformability, including assessment of
cell surface area to volume ratio, and assessment of membrane rigidity (113). Studies
involving micropipette aspiration of P. knowlesi-infected and uninfected cells are planned,
and will provide important information regarding mechanisms of reduced RBC deformability
in human knowlesi malaria.
90
P. knowlesi is a common cause of human malaria in Malaysian Borneo, and is increasingly
reported in other Southeast Asian countries. Moreover, it causes severe and fatal disease;
with a severity rate 3-fold that of falciparum malaria. While this thesis has contributed new
information regarding the epidemiology, clinical features, diagnosis, treatment and
pathogenesis of knowlesi malaria, major questions remain unanswered and will require
additional research.
13.1 Epidemiology
This thesis has confirmed that P. knowlesi has become a major cause of human malaria in
Sabah. In Kudat District, northeast Sabah, it was overwhelmingly the most common cause
of all malaria hospital admissions during 2009-2011 (with the majority of cases confirmed by
PCR) (58, 171), while it was also the most commonly detected Plasmodium species, by
PCR, from blood slides collected in the Interior Division during 2010 (25). In our 20-year
review of Sabah Department of Health malaria notification data, we found that, while
notifications of P. falciparum and P. vivax have fallen markedly, notifications of “P.
malariae/P.knowlesi” have increased 10-fold between 2004 and 2011, with this increase
occurring state-wide. However, this latter study was associated with major limitations, in
particular, our reliance on microscopy data to support our conclusion that there has been a
significant and recent state-wide increase in incidence of knowlesi malaria. In order to
confirm this trend, larger prospective studies are required, ideally involving PCR-confirmation
of all microscopy-diagnosed malaria cases in Sabah. Plans are underway to conduct such a
study.
91
While this thesis has described an apparent recent emergence of P. knowlesi in Sabah, the
reasons for this trend have not been identified. It is unlikely that increased recognition of the
species is the sole explanation for the increase in reported incidence. Firstly, microscopy
skill levels would presumably have been higher in earlier decades in Sabah when malaria
prevalence was high, and it is unlikely that large numbers of P. knowlesi slides would have
been misdiagnosed as P. vivax or P. falciparum rather than “P. malariae”. Secondly, as we
report in Chapter 7, the recent increase in age of all patients with malaria, and the increase
in proportion of patients >50 years of age, is consistent with a true increase in the incidence
of P. knowlesi malaria.
Alternative possible explanations for the increasing incidence of P. knowlesi include: 1)
increased interaction between humans and the simian hosts, due to encroachment of
humans into macque habitat; 2) changes in the P. knowlesi vector, due to loss of habitat
and/or malaria control activities; 3) a decrease in cross-species immunity due to the
reduction in prevalence of P. falciparum and P. vivax, leading to increased susceptibility to
infection; or 4) an adaption of P. knowlesi to the human host, leading to higher parasitemia
infections with a consequent increase in the likelihood of human-human transmission (166).
These last two explanations could account for the apparent recent increase in cases of
severe knowlesi malaria, and the apparent restriction of severe cases of knowlesi malaria to
Malaysia. Further studies are required to investigate these factors.
Human to human transmission has not yet been demonstrated in the natural environment.
While molecular studies involving macaques and humans in Kapit, Sarawak, suggested that
P. knowlesi remained primarily a zoonosis (53), a more recent study in Thailand reported
that genetic diversity was greater among human P. knowlesi isolates than it was among
macaque P. knowlesi isolates, raising the possibility of human to human transmission (88).
Further studies involving sequence analysis of human and macaque P. knowlesi isolates in
92
Sabah and elsewhere may provide additional information as to whether human to human
transmission is occurring. The possible occurrence of transmission of P. knowlesi from
humans back to macaques may however limit the ability of such studies to provide firm
evidence of human-human transmission.
Finally, while we have demonstrated that P. knowlesi is a common cause of malaria among
patients presenting to clinics and hospitals in Sabah, little is known about the true burden of
infection. Cross-sectional studies utilizing highly sensitive molecular methods are required
to determine the prevalence of asymptomatic parasitemia among affected communities in
Sabah and elsewhere.
13.2 Diagnosis of Plasmodium knowlesi
In this thesis we evaluated the accuracy of microscopy for the diagnosis of knowlesi malaria,
and found that P. knowlesi was often confused with P. falciparum and P. vivax, and vice-
versa. However, while the difficulty of distinguishing the young trophozoites of P. falciparum
and P. knowlesi has been well reported, P. knowlesi is not thought to resemble P. vivax, and
hence the explanation for the frequent misdiagnosis between these two species requires
further investigation. Only one report has described in detail the morphological features of a
series of patients with naturally acquired knowlesi malaria. This report involved 10 patients,
all diagnosed at Kapit District Hospital, and most with only moderate parasitemia (median
5266 parasites/uL) (57). In the first description of P. knowlesi, Knowles and Das Gupta
reported that the morphological appearance of the parasite varied depending on the host,
resembling P. vivax when occurring in long-tailed macaques, with enlarged red cells and
Schuffner’s dots observed (37). Schuffner’s dots, which in P. vivax malaria represent
membrane-bound caveola-vesicle complexes (CVC) that contain malaria antigens (191),
93
have also been reported to occur in two case reports of human knowlesi malaria (16, 192).
The morphological appearance of P. knowlesi may differ between patient groups, particularly
if parasite adaption has occurred in certain settings. Further detailed descriptions of the
morphological appearance of P. knowlesi in large series of patients in different
epidemiological settings are therefore required. The findings from these studies will inform
the use of microscopy for the diagnosis of knowlesi malaria.
This thesis also evaluated the sensitivity of two rapid diagnostic tests for the diagnosis of
knowlesi malaria, and found suboptimal sensitivity for both tests evaluated, particularly at
low parasitemia. Given the difficulties associated with microscopy for the diagnosis of
knowlesi malaria, the development of sensitive RDTs that have the ability to distinguish
between species is urgently required in areas endemic for knowlesi malaria, particularly in
settings where management guidelines recommend different treatment strategies according
to species. In addition, diagnosis of knowlesi malaria would potentially be improved through
the development of alternative molecular methods of diagnosis, such as Loop Mediated
Isothermal Amplification (LAMP), that are low-cost and can be performed with simple
techniques that are applicable to field settings (96, 97).
13.3 Clinical features
Our prospective study conducted at QEH extends what is known about the clinical features
of knowlesi malaria. We found that, while the majority of patients with knowlesi malaria had
uncomplicated disease with low parasitemia, P. knowlesi was associated with a 3-fold
increased risk of severe disease compared to falciparum malaria, that parasite count was
the major independent risk factor for severe disease, and that parasite count was strongly
associated with increasing age. Features of severity were similar to those of falciparum
94
malaria, with jaundice, respiratory distress, shock and acute kidney injury all common.
However, at least in the series reported to date, coma does not appear to be a feature of
severe knowlesi malaria.
Many questions regarding the clinical features of knowlesi malaria remain unanswered. No
large prospective study has included children, and hence data regarding the clinical features
of paediatric knowlesi malaria are lacking. In our small retrospective study of knowlesi
malaria in children (58), severe anaemia (Hb 6.4 g/dL) occurred in one child with PCR-
confirmed knowlesi malaria, while another child with microscopy-diagnosed “P. malariae”
had a hemoglobin of 4.9 g/dL. Severe anaemia is relatively uncommon among adults with
severe knowlesi malaria, with 5% of patients in our prospective study having a haemoglobin
<7 g/dL and no patient having a haemoglobin <5 g/dL. Larger prospective studies involving
children are needed to determine if severe anaemia is a more common feature of paediatric
knowlesi malaria.
Our prospective study also excluded patients with significant comorbidities and concurrent
illnesses, including 2 patients with knowlesi malaria and end-stage renal failure, and one
patient with meningococcal meningitis. The impact of comorbidities on the clinical
presentation of knowlesi malaria may be particularly relevant for P. knowlesi given the older
age group affected, and will require further investigation. Our prospective study involved two
asplenic patients with knowlesi malaria; one with severe disease, and one with non-severe
disease. Further studies will be required to determine if asplenic patients are at greater risk
of acquiring knowlesi malaria, are at greater risk of severe disease, or have an altered
clinical presentation.
Concurrent infections such as dengue and leptospirosis may be common among patients
with knowlesi malaria given the overlapping geographic distribution, but have not been
95
adequately investigated. In addition, although now reported in two series, including the
prospective series described in this thesis (31, 103), the prevalence of bacteremia among
patients with knowlesi malaria has not been systematically evaluated. Finally, data are
lacking on the clinical features of knowlesi malaria in pregnant women.
13.4 Treatment
Although numerous antimalarial drugs appear to be effective for the treatment of
uncomplicated knowlesi malaria, the optimal treatment regimen is not known. While
Daneshvar et al. found chloroquine to be efficacious in a district-hospital study (107), in a
retrospective tertiary-referral hospital study William et al. found parasite clearance times to
be faster with artemether-lumefantrine than with chloroquine (31). A randomized control trial
of artemisinin combination therapy (artesunate-mefloquine) versus chloroquine is currently
underway in Sabah, Malaysia (ClinicalTrials.gov Identifier: NCT01708876), and will provide
important information regarding the most effective treatment for uncomplicated knowlesi
malaria.
For patients with severe malaria, WHO now recommends treatment with intravenous
artesunate regardless of species, and in our prospective study this treatment strategy was
associated with a zero mortality rate (103). However, uncertainty remains regarding the
optimal treatment strategy for patients with intermediate parasitemia in the absence of other
severity criteria. Whether these patients require intravenous artesunate, or can adequately
be treated with oral artemisinin combination therapies, has not been established.
Furthermore if high parasitemia alone is an indication for intravenous artesunate, the cut-off
parasite count mandating such treatment will need to be clarified. Finally, no data exist
regarding optimal strategies for managing complications of severe knowlesi malaria, such as
96
acute respiratory distress syndrome and acute kidney injury. The roles of blood transfusion,
platelet transfusion, inotropes and antibiotics are also unclear, and will only be determined
with larger series of patients with severe knowlesi malaria.
13.4 Pathogenesis
Our understanding of the clinical features of knowlesi malaria has improved substantially
since the report in 2004 of a cluster of human cases in Kapit, Sarawak (3). However, our
knowledge of the pathogenesis of disease in humans remains limited. Only one human
autopsy study has been performed (27), one study has reported on the cytokine response to
knowlesi malaria (32), and one study involving 5 patients has investigated the
cytoadherence properties of P. knowlesi infected erythrocytes ex vivo (105). In addition, in
this thesis we report the results of a study evaluating the deformability of red cells among
patients with knowlesi malaria.
Many pathogenic processes are yet to be investigated in P. knowlesi malaria. In falciparum
malaria a central process underlying the pathogenesis of disease is the sequestration of
parasitized erythrocytes within vital organs, leading to microvascular obstruction and
impaired organ perfusion (111-113). Reduced red blood cell deformability, as well as
adherence of parasitized cells to non-parasitized cells (rosetting) and to each other (platelet-
mediated clumping), further contributes to this process (111-113). In addition, endothelial
bioavailability of nitric-oxide (NO) is reduced in falciparum malaria in proportion to disease
severity, leading to endothelial dysfunction with increased expression of endothelial
receptors such as intercellular adhesion molecule-1 (ICAM-1) and E-selectin, impairment of
97
compensatory vascular responses, and further disturbance of microcirculatory flow (193,
194).
Early studies of knowlesi malaria in rhesus macaques indicate that, as with falciparum
malaria, microcirculatory dysfunction is fundamental to the development of severe disease
(140, 141). However, the rarity of coma in human knowlesi malaria suggests that the
mechanisms underlying microcirculatory dysfunction differ from those of falciparum malaria.
In the first of a series of studies of the pathogenesis of severe knowlesi malaria, we found
that red cell deformability was reduced in proportion to severe disease. This finding
suggests that reduced red cell deformability may play a role in microvascular obstruction in
knowlesi malaria. However, larger studies are required to confirm this finding, to determine if
the reduction in deformability affects parasitized or non-parasitized cells, and to determine
the mechanisms of reduced deformability. As with falciparum malaria, rosetting and platelet-
mediated clumping may contribute to impaired microcirculatory flow in knowlesi malaria, and
studies investigating these processes are planned. Opthalmoscopy has proven highly
informative in understanding pathophysiology in falciparum malaria (195-198). In retinal
photography studies at QEH outside the scope of this thesis, retinal changes characteristic
of severe falciparum malaria were not seen in 20 patients with severe knowlesi malaria
(Govindesamy et al, unpublished), again supporting potential differences in pathophysiology.
In additional pathophysiological studies being conducted at QEH, I have been investigating
endothelial NO bioavailability in patients with knowlesi malaria using peripheral arterial
tonometry (PAT). This measures reactive hyperemia (RH) after forearm ischemia induced
by inflation of a sphygmomanometer cuff to 200mmHg for 5 minutes. RH-PAT is at least
50% dependent on NO production, and in falciparum malaria is reduced in proportion to
disease severity, and is reversible with administration of L-arginine (193). Preliminary
results, outside the scope of this thesis, show that, as with falciparum malaria, NO-
98
dependent endothelial function is impaired among patients with severe knowlesi malaria. In
support of this, preliminary analysis suggests that markers of endothelial activation
(Angiopoietin-2 and von-Willibrand Factor) are increased in severe knowlesi malaria
compared to controls. In other studies of microvascular function I have been assessing
microvascular reactivity using Near Infrared Spectroscopy (NIRS). NIRS noninvasively
assesses tissue oxygenation by comparing absorption of near-infrared light by
oxyhaemoglobin and deoxyhaemoglobin in microcirculatory vessels, the vessels most
affected by malaria (199). By measuring tissue oxygenation before, during and after
ischaemic stress, NIRS has demonstrated impaired microvascular reactivity in patients with
severe falciparum malaria (194); initial analysis shows similar results with P. knowlesi.
These preliminary results suggest that in knowlesi malaria microvascular function is
impaired, and may contribute to microvascular obstruction and impaired organ perfusion
among patients with severe disease.
Orthogonal Polarising Spectroscopy (OPS) has been used to demonstrate microvascular
obstruction of rectal mucosal microvasculature in severe falciparum malaria (190), and
comparative studies of rectal OPS in severe knowlesi and falciparum malaria are in
progress.
While these additional studies will contribute to our understanding of the role of the
microcirculation in knowlesi malaria, information regarding cytoadherence of P. knowlesi
infected erythrocytes in the microvasculature, and the presence or absence of sequestration
in knowlesi malaria, will be difficult to obtain without additional autopsy reports of fatal cases.
Cultural factors limit the acceptability of autopsies in Malaysia however, and undertaking
further autopsy studies will therefore be challenging.
99
13.5 Summary
In this thesis I have demonstrated that P knowlesi is the most common cause of malaria in
several districts throughout Sabah and Sarawak, and that the incidence of P. knowlesi in
Sabah appears to be increasing as the other human malaria species are controlled. I have
described a wide age-distribution of patients affected by P. knowlesi, with patients older than
those with vivax or falciparum malaria, but with P. knowlesi still being a common cause of
malaria in children. I have described in detail the clinical features of knowlesi malaria in
adults, and demonstrated that P. knowlesi is associated with a severity risk 3-fold that of P.
falciparum. I have demonstrated that intravenous artesunate is highly efficacious for the
treatment of severe knowlesi malaria, and discussed the limitations of microscopy and rapid
diagnostic tests for the diagnosis of P. knowlesi. Finally, I have presented the results of a
study demonstrating reduced red cell deformability among patients with knowlesi malaria.
This thesis has confirmed the public health importance of P. knowlesi as an emerging
infectious disease in the region, and has highlighted the remaining knowledge gaps in the
epidemiology, clinical features, diagnosis, treatment and pathogenesis of knowlesi malaria.
100
1. Napier LE, Campbell HGM. Observations on a Plasmodium infection which causes
haemoglobinuria in certain species of monkey. Indian Medical Gazette. 1932;67:246-
9.
2. Chin W, Contacos P, Coatney R, Kimball H. A naturally acquired quotidian-type
malaria in man transferable to monkeys. Science. 1965;149:865.
3. Singh B, Sung LK, Matusop A, Radhakrishnan A, Shamsul SSG, Cox-Singh J, et al.
A large focus of naturally acquired Plasmodium knowlesi infections in human beings.
Lancet. 2004;363(9414):1017-24.
4. Lee K-S, Cox-Singh J, Brooke G, Matusop A, Singh B. Plasmodium knowlesi from
archival blood films: Further evidence that human infections are widely distributed
and not newly emergent in Malaysian Borneo. Int J Parasitol. 2009;39(10):1125-8.
5. Jongwutiwes S, Buppan P, Kosuvin R, Seethamchai S, Pattanawong U,
Sirichaisinthop J, et al. Plasmodium knowlesi malaria in humans and macaques,
Thailand. Emerg Infect Dis. 2011;17(10).
6. Jiang N, Chang Q, Sun X, Lu H, Yin J, Zhang Z, et al. Co-infections with Plasmodium
knowlesi and other malaria parasites, Myanmar. Emerg Infect Dis. 2010;16(9):1476.
7. Jongwutiwes S, Putaporntip C, Iwasaki T, Sata T, Kanbara H. Naturally acquired
Plasmodium knowlesi malaria in human, Thailand. Emerg Infect Dis.
2004;10(12):2211-3.
8. Khim N, Siv S, Kim S, Mueller T, Fleischmann E, Singh B, et al. Plasmodium
knowlesi infection in humans, Cambodia, 2007-2010. Emerg Infect Dis.
2011;17(10):1900-2.
9. Luchavez J, Espino F, Curameng P, Espina R, Bell D, Chiodini P, et al. Human
infections with Plasmodium knowlesi, the Philippines. Emerg Infect Dis.
2008;14(5):811 - 3.
101
10. Marchand R, Culleton R, Maeno Y, Quang N, Nakazawa S. Co-infections of
Plasmodium knowlesi, P. falciparum, and P. vivax among humans and Anopheles
dirus mosquitoes, southern Vietnam. Emerg Infect Dis. 2011;17(7):1232-9.
11. Ng OT, Ooi EE, Lee CC, Lee PJ, Ng LC, Pei SW, et al. Naturally acquired human
Plasmodium knowlesi infection, Singapore. Emerg Infect Dis. 2008;14:814 - 6.
12. Berry A, Iriart X, Wilhelm N, Valentin A, Cassaing S, witkowski B, et al. Case report:
imported Plasmodium knowlesi malaria in a French tourist returning from Thailand.
Am J Trop Med Hyg. 2011;84(4):535-8.
13. Bronner U, Divis P, Färnert A, Singh B. Swedish traveller with Plasmodium knowlesi
malaria after visiting Malaysian Borneo. Malar J. 2009;8(1):15.
14. Ennis J, Teal A, Habura A, Madison-Antenucci S, Keithly J, Arguin P, et al. Simian
malaria in a US traveler-New York, 2008. Morbidity and Mortality Weekly Report.
2009;58(9):229-32.
15. Fan L, Lee SY, Koay E, Harkensee C. Plasmodium knowlesi infection: a diagnostic
challenge. BMJ case reports. 2013;2013.
16. Figtree M, Lee R, Bain L, Kennedy T, Mackertich S, Urban M, et al. Plasmodium
knowlesi in human, Indonesian Borneo. Emerg Infect Dis. 2010;16(4):672.
17. Hoosen A, Shaw M. Plasmodium knowlesi in a traveller returning to New Zealand.
Travel Med Infect Dis. 2011.
18. Kantele A, Marti H, Felger I, Muller D, Jokiranta T. Monkey malaria in a European
traveler returning from Malaysia. Emerg Infect Dis. 2008;14:1434 - 6.
19. Kuo M-C, Chiang T-Y, Chan C-W, Tsai W-S, Ji D-D. A case report of simian malaria,
Plasmodium knowlesi, in a Taiwanese traveler from Palawan Island, the Philippines.
Taiwan Epidemiology Bulletin. 2009;25:178-91.
20. Tang T, Salas A, Ali-Tammam M, Martinez M, Lanza M, Arroyo A, et al. First case of
detection of Plasmodium knowlesi in Spain by Real Time PCR in a traveller from
Southeast Asia. Malar J. 2010;9:219.
102
21. Tanizaki R, Ujiie M, Kato Y, Iwagami M, Hashimoto A, Kutsuna S, et al. First case of
Plasmodium knowlesi infection in a Japanese traveller returning from Malaysia. Malar
J. 2013;12(1):1-4.
22. van Hellemond J, Rutten M, Koelewijn R, Zeeman A, Verweij J, Wismans P, et al.
Human Plasmodium knowlesi infection detected by rapid diagnostic tests for malaria.
Emerg Infect Dis. 2009;15(9):1478.
23. Cox-Singh J, Davis TM, Lee KS, Shamsul SS, Matusop A, Ratnam S, et al.
Plasmodium knowlesi malaria in humans is widely distributed and potentially life
threatening. Clin Infect Dis. 2008;46(2):165-71.
24. Daneshvar C, Davis TM, Cox-Singh J, Rafa'ee M, Zakaria S, Divis P, et al. Clinical
and laboratory features of human Plasmodium knowlesi infection. Clin Infect Dis.
2009;49(6):852-60.
25. Joveen-Neoh WF, Chong KL, Wong CM, Lau TY. Incidence of malaria in the Interior
Division of Sabah, Malaysian Borneo, based on nested PCR. J Parasitol Res.
2011;2011:DOI: 10.1155/2011/104284.
26. Naing DKS, Anderios F, Lin Z. Geographic and ethnic distribution of P knowlesi
infection in Sabah, Malaysia. International Journal of Collaborative Research on
Internal Medicine & Public Health. 2011;3(5):391-400.
27. Cox-Singh J, Hiu J, Lucas SB, Divis PC, Zulkarnaen M, Chandran P, et al. Severe
malaria-a case of fatal Plasmodium knowlesi infection with post-mortem findings.
Malar J. 2010;9:10.
28. Ong C, Lee S, Koh W, Ooi E, Tambyah P. Monkey malaria in humans: a diagnostic
dilemma with conflicting laboratory data. Am J Trop Med Hyg. 2009;80(6):927.
29. Van den Eede P, Van H, Van Overmeir C, Vythilingam I, Duc T, Hung L, et al.
Human Plasmodium knowlesi infections in young children in central Vietnam. Malar
J. 2009;8(1):249.
103
30. Lee CE, Adeeba K, Freigang G. Human Plasmodium knowlesi infections in Klang
Valley, Peninsular Malaysia: a case series. Med J Malaysia. 2010;65(1):63-5.
31. William T, Menon J, Rajahram G, Chan L, Ma G, Donaldson S, et al. Severe
Plasmodium knowlesi malaria in a tertiary hospital, Sabah, Malaysia. Emerg Infect
Dis. 2011;17(7):1248-55.
32. Cox Singh J, Singh B, Daneshvar C, Planche T, Parker-Williams J, Krishna S. Anti-
inflammatory cytokines predominate in acute human Plasmodium knowlesi infections.
PLoS ONE. 2011;6(6):e20541.
33. Antinori S, Milazzo L, Corbellino M. Plasmodium knowlesi: An Overlooked Italian
Discovery? Clin Infect Dis. 2011;53(8):849-.
34. Coatney GR, Collins WE, Warren MW, Contacos PG. The primate malarias.
Washington: US Government Printing Office; 1971.
35. Franchini G. Su di un plasmodio pigmentato di una scimmia. Arch Ital Sci Med Colon
Parassit. 1927;8:187-90.
36. Knowles R. Monkey malaria. Br Med J. 1935:1020.
37. Knowles R, Das Gupta, B.M. A study of monkey-malaria and its experimental
transmission to man. Indian Medical Gazette. 1932;67:246-9.
38. Sinton JA, Mulligan HW. A critical review of the literature relating to the identification
of the malarial parasites recorded from monkeys of the families Cercopithecidae and
Colobidae. Rec Malar Surv India. 1932;3:357-80.
39. Sinton JA, Mulligan HW. A critical review of the literature relating to the identification
of the malarial parasites recorded from monkeys of the families Cercopithecidae and
Colobidae. Rec Malar Surv India. 1933;3:381-443.
40. Garnham P, Lainson R, Cooper W. The tissue stages and sporogony of Plasmodium
knowlesi. Trans R Soc Trop Med Hyg. 1957;51(5):384-96.
41. van Rooyen, CE, Pile G. Observations on infection by Plasmodium knowlesi (ape
malaria) in the treatment of general paralysis of the insane. Br Med J. 1935;2:662-6.
104
42. Nicol W. Monkey malaria in GPI. Br Med J. 1935;2(3902):760.
43. Ciuca M, Ballif L, Chelarescu M, Lavrinenko M, Zotta E. Contributions a l'étude de
l'action pathogène de PI. knowlesi pour l'homme (considérations sur l'immunité
naturelle et l'immunité acquise contre cette espece de parasite). Bull Soc Path Exot.
1937;30:305-15.
44. Ciuca M, Tomescu P, Badenski G. Contribution a l'étude de la virulence du Pl.
knowlesi chez l'homme. Caractères de la maladie et biologie du parasite. Arch
Roumaines Path Experim Microbiol. 1937;10:5-28.
45. Ciuca M, Chelarescu M, Sofletea A, Constantenescu P, Teriteanu E, Cortez P, et al.
Contribution expérimentale a l'étude de l'immunité dans le paludisme. Bucarest:
L'Academia; 1955.
46. Milam DF, Kusch E. Observations on Plasmodium knowlesi malaria in general
paresis. South Med J. 1938;31(8):947-9.
47. Milam DF, Coggeshall LT. Duration of Plasmodium knowlesi infections in man. Am J
Trop Med Hyg. 1938;18:331-8.
48. Chin W, Contacos P, Collins W, Jeter M, Alpert E. Experimental mosquito-
transmission of Plasmodium knowlesi to man and monkey. Am J Trop Med Hyg.
1968;17(3):355.
49. Eyles D, Coatney R, Getz M. Vivax-type malaria parasite of macaques transmissible
to man. Science. 1960;132:1812-3.
50. Warren M, Cheong W, Fredericks H, Coatney R. Cycles of jungle malaria in West
Malaysia. Am J Trop Med Hyg. 1960;19(3):383-93.
51. Vythilingam I, NoorAzian Y, Huat T, Jiram A, Yusri Y, Azahari A, et al. Plasmodium
knowlesi in humans, macaques and mosquitoes in peninsular Malaysia. Parasit
Vectors. 2008;1:26.
105
52. Imwong M, Tanomsing N, Pukrittayakamee S, Day NPJ, White NJ, Snounou G.
Spurious amplification of a Plasmodium vivax small-subunit RNA gene by use of
primers currently used to detect P. knowlesi. J Clin Microbiol. 2009;47(12):4173.
53. Lee K-S, Divis PCS, Zakaria SK, Matusop A, Julin RA, Conway DJ, et al.
Plasmodium knowlesi: reservoir hosts and tracking the emergence in humans and
macaques. PLoS Pathog. 2011;7(4):e1002105.
54. McArthur J. An account of the work of the Malaria Research Department, North
Borneo, 1939-1942. London: Library of London School of Hygiene and Tropical
Medicine; 1942.
55. McArthur J. The transmission of malaria in Borneo. Trans R Soc Trop Med Hyg.
1947;40(5):537-58.
56. Gnémé A, Guelbéogo WM, Riehle MM, Tiono AB, Diarra A, Kabré GB, et al.
Plasmodium species occurrence, temporal distribution and interaction in a child-aged
population in rural Burkina Faso. Malar J. 2013;12(1):67.
57. Lee K-S, Cox-Singh J, Singh B. Morphological features and differential counts of
Plasmodium knowlesi parasites in naturally acquired human infections. Malar J.
2009;8:73.
58. Barber BE, William T, Jikal M, Jilip J, Dhararaj P, Menon J, et al. Plasmodium
knowlesi malaria in children. Emerg Infect Dis. 2011;17(5):814-20.
59. William T, Rahman HA, Jelip J, Ibrahim MY, Menon J, Grigg MJ, et al. Increasing
incidence of Plasmodium knowlesi malaria following control of P. falciparum and P.
vivax malaria in Sabah, Malaysia. PLoS Negl Trop Dis. 2013;7(1):e2026.
60. de Zulueta J. Malaria in Sarawak and Brunei. Bull World Health Organ. 1956;15(3-
5):651.
61. Putaporntip C, Hongsrimuang T, Seethamchai S, Kobasa T, Limkittikul K, Cui L, et al.
Differential prevalence of Plasmodium infections and cryptic Plasmodium knowlesi
malaria in humans in Thailand. J Infect Dis. 2009;199:1143 - 50.
106
62. Eede P, Van H, Van Overmeir C, Vythilingam I, Duc T, Hung L, et al. Human
Plasmodium knowlesi infections in young children in central Vietnam. Malar J.
2009;8(1):249.
63. Lee Y, Tang C, Ang L, Han H, James L, Goh K. Epidemiological characteristics of
imported and locally-acquired malaria in Singapore. Ann Acad Med Singapore.
2009;38:840-9.
64. Jeslyn WPS, Huat TC, Vernon L, Irene LMZ, Sung LK, Jarrod LP, et al. Molecular
epidemiological investigation of Plasmodium knowlesi in humans and macaques in
Singapore. Vector-Borne and Zoonotic Diseases. 2011;11(2):131-5.
65. Cox-Singh J, Singh B. Knowlesi malaria: newly emergent and of public health
importance? Trends Parasitol. 2008;24(9):406-10.
66. Ninan T, Nalees K, Newin M, Sultan Q, Than MM, Shinde S, et al. Plasmodium
knowlesi malaria infection in human. Brunei International Medical Journal.
2012;8(6):358-61.
67. Zhu H, Li J, Zheng H. Human natural infection of Plasmodium knowlesi. Chinese
Journal of Parasitology and Parasitic Diseases. 2006;24:70 - 1.
68. Sermwittayawong N, Singh B, Nishibuchi M, Sawangjaroen N, Vuddhakul V. Human
Plasmodium knowlesi infection in Ranong province, southwestern border of Thailand.
Malar J. 2012;11:36.
69. Tyagi RK, Das MK, Singh SS, Sharma YD. Discordance in drug resistance-
associated mutation patterns in marker genes of Plasmodium falciparum and
Plasmodium knowlesi during coinfections. J Antimicrob Chemother. 2013 May 1,
2013;68(5):1081-8.
70. World Health Organization. World Malaria Report 2011. Geneva: World Health
Organization; 2011. Available from:
http://www.who.int/malaria/world_malaria_report_2011/9789241564403_eng.pdf.
107
71. Looareesuwan S, White N, Bunnag D, Chittamas S, Harinasuta T. High rate of
Plasmodium vivax relapse following treatment of falciparum malaria in Thailand. The
Lancet. 1987;330(8567):1052-5.
72. Richie T. Interactions between malaria parasites infecting the same vertebrate host.
Parasitology. 1988;96:607-39.
73. Seethamchai S, Putaporntip C, Malaivijitnond S, Cui L, Jongwutiwes S. Malaria and
Hepatocystis species in wild macaques, southern Thailand. Am J Trop Med Hyg.
2008;78(4):646-53.
74. Putaporntip C, Jongwutiwes S, Thongaree S, Seethamchai S, Grynberg P, Hughes
AL. Ecology of malaria parasites infecting Southeast Asian macaques: evidence from
cytochrome b sequences. Mol Ecol. 2010;19(16):3466-76.
75. Lambrecht F, Dunn F, Eyles D. Isolation of Plasmodium knowlesi from Philippine
macaques. Nature. 1961;191:1117-8.
76. Tsukamoto M, Miyata A, Miyagi I. Surveys on simian malaria parasites and their
vector in Palawan Island, the Philippines. Tropical Biomedicine. 1978;20:39-50.
77. Eyles D, Laing A, Warren M, Sandoshan A. Malaria parasites of the Malayan leaf
monkeys of the genus Presbytis. Med J Malaya. 1962;17:85-6.
78. Wharton R, Eyles D. Anopheles hackeri, a vector of Plasmodium knowlesi in Malaya.
Science. 1961;134:279-80.
79. Rattanarithikul R, Harrison B, Harbach R, Panthusiri P, Coleman R. Illustrated keys
to the mosquitoes of Thailand. IV. Anopheles. Southeast Asian J Trop Med Public
Health. 2006;37 Suppl 2:1-128.
80. Sallum MAM, Peyton E, Harrison BA, Wilkerson RC. Revision of the Leucosphyrus
group of Anopheles (cellia)(diptera, culicidae). Rev Bras Entomol. 2005;49(Suppl
1):1-152.
108
81. Tan C, Vythilingam I, Matusop A, Chan S, Singh B. Bionomics of Anopheles latens in
Kapit, Sarawak, Malaysian Borneo in relation to the transmission of zoonotic simian
malaria parasite Plasmodium knowlesi. Malar J. 2008;7:52.
82. Vythilingam I. Plasmodium knowlesi in humans: a review on the role of its vectors in
Malaysia. Tropical Biomedicine. 2010;27(1):1-12.
83. Eyles DE, Wharton R, Cheong W, Warren MW. Studies on malaria and Anopheles
balabacensis in Cambodia. Bull World Health Organ. 1964;30(1):7.
84. Collins W, Contacos P, Guinn E. Studies on the transmission of simian malarias II.
Transmission of the H strain of Plasmodium knowlesi by Anopheles balabacensis
balabacensis. J Parasitol. 1967;53:841-4.
85. Vythilingam I, Chan S, Shanmugratnam C, Tanrang H, Chooi KH. The impact of
development and malaria control activities on its vectors in the Kinabatangan area of
Sabah, East Malaysia. Acta Trop. 2005;96:24-30.
86. Rhosenberg R, Andre R, Somchit L. Highly efficient dry season transmission of
malaria in Thailand. Transactions of the Royal Society of Tropical Medicine and
Hygiene. 1990;84(1):22-8.
87. Singhasivanon P, Thimasarn K, Yimsamran S, Linthicum K, Nualchawee K,
Dawreang D, et al. Malaria in tree crop plantations in south-eastern and western
provinces of Thailand. Southeast Asian J Trop Med Public Health. 1999;30(3):399-
404.
88. Putaporntip C, Thongaree S, Jongwutiwes S. Differential sequence diversity at
merozoite surface protein-1 locus of Plasmodium knowlesi from humans and
macaques in Thailand. Infect Genet Evol. 2013.
89. Kawai S, Hirai M, Haruki K, Tanabe K, Chigusa Y. Cross-reactivity in rapid diagnostic
tests between human malaria and zoonotic simian malaria parasite Plasmodium
knowlesi infections. Parasitol Int. 2009;58:300-2.
109
90. McCutchan T, Piper R, Makler M. Use of malaria rapid diagnostic test to identify
Plasmodium knowlesi infection. Emerg Infect Dis. 2008;14(11):1750.
91. Link L, Bart A, Verhaar N, van Gool T, Pronk M, Scharnhorst V. Molecular detection
of Plasmodium knowlesi in a Dutch traveler by real-time PCR. J Clin Microbiol.
2012;50(7):2523-4.
92. Divis P, Shokoples SE, Singh B, Yanow SK. A TaqMan real-time PCR assay for the
detection and quantitation of Plasmodium knowlesi. Malar J. 2010;9:344.
93. Babady NE, Sloan LM, Rosenblatt JE, Pritt BS. Detection of Plasmodium knowlesi by
real-time polymerase chain reaction. The American Journal of Tropical Medicine and
Hygiene. 2009;81(3):516.
94. Oddoux O, Debourgogne A, Kantele A, Kocken CH, Jokiranta TS, Vedy S, et al.
Identification of the five human Plasmodium species including P. knowlesi by real-
time polymerase chain reaction. Eur J Clin Microbiol Infect Dis. 2011;30(4):597-601.
95. Lucchi NW, Poorak M, Oberstaller J, DeBarry J, Srinivasamoorthy G, Goldman I, et
al. A new single-step PCR assay for the detection of the zoonotic malaria parasite
Plasmodium knowlesi. PLoS ONE. 2012;7(2):e31848.
96. Iseki H, Kawai S, Takahashi N, Hirai M, Tanabe K, Yokoyama N, et al. Evaluation of
a loop-mediated isothermal amplification method as a tool for diagnosis of infection
by the zoonotic simian malaria parasite Plasmodium knowlesi. J Clin Microbiol.
2010;48(7):2509.
97. Lau Y-L, Fong M-Y, Mahmud R, Chang P-Y, Palaeya V, Cheong F-W, et al. Specific,
sensitive and rapid detection of human Plasmodium knowlesi infection by loop-
mediated isothermal amplification (LAMP) in blood samples. Malar J. 2011;10(1):1-6.
98. Center for Disease Control and Prevention (CDC). Simian malaria in a U.S. traveler-
New York, 2008. Morb Mortal Wkly Rep. 2008;58:229-32.
110
99. Lee W-C, Chin P-W, Lau Y-L, Chin L-C, Fong M-Y, Yap C-J, et al. Hyperparasitaemic
human Plasmodium knowlesi infection with atypical morphology in peninsular
Malaysia. Malar J. 2013;12(1):88.
100. Rajahram GS, Barber BE, Tan WM, Yeo TW, William T. Case report: fatal
Plasmodium knowlesi malaria following an atypical clinical presentation and delayed
diagnosis. Med J Malaysia. 2013;68(1).
101. Azira N, Zairi N, Amry A, Zeehaida M. Case series of naturally acquired Plasmodium
knowlesi infection in a tertiary teaching hospital. Tropical Biomedicine.
2012;29(3):398-404.
102. Willmann M, Ahmed A, Siner A, Wong I, Woon L, Singh B, et al. Laboratory markers
of disease severity in Plasmodium knowlesi infection: a case control study. Malar J.
2012;11(1):363.
103. Barber BE, William T, Grigg MJ, Menon J, Auburn S, Marfurt J, et al. A prospective
comparative study of knowlesi, falciparum and vivax malaria in Sabah, Malaysia: high
proportion with severe disease from Plasmodium knowlesi and P. vivax but no
mortality with early referral and artesunate therapy. Clin Infect Dis. 2013;56(3):383-
97.
104. Dondorp AM, Lee SJ, Faiz M, Mishra S, Price R, Tjitra E, et al. The relationship
between age and the manifestations of and mortality associated with severe malaria.
Clin Infect Dis. 2008;47(2):151.
105. Fatih FA, Siner A, Ahmed A, Woon LC, Craig AG, Singh B, et al. Cytoadherence and
virulence–the case of Plasmodium knowlesi malaria. Malar J. 2012;11(1):33.
106. World Health Organization. Management of severe malaria: a practical handbook,
2nd edition. Geneva: World Health Organization; 2010.
107. Daneshvar C, Davis TM, Cox-Singh J, Rafa'ee M, Zakaria S, Divis P, et al. Clinical
and parasitological response to oral chloroquine and primaquine in uncomplicated
human Plasmodium knowlesi infections. Malar J. 2010;9(1):238.
111
108. Russell B, Chalfein F, Prasetyorini B, Kenangalem E, Piera K, Suwanarusk R, et al.
Determinants of in vitro drug susceptibility testing of Plasmodium vivax. Antimicrob
Agents Chemother. 2008;52(3):1040-5.
109. Pukrittayakamee S, Chantra A, Simpson JA, Vanijanonta S, Clemens R,
Looareesuwan S, et al. Therapeutic responses to different antimalarial drugs in vivax
malaria. Antimicrob Agents Chemother. 2000;44(6):1680-5.
110. World Health Organization. Management of severe malaria: a practical handbook,
3rd edition. Geneva: World Health Organization; 2012.
111. Clark IA, Alleva LM, Mills AC, Cowden WB. Pathogenesis of malaria and clinically
similar syndromes. Clin Microbiol Rev. 2004;17:509-39.
112. Dondorp A, Pongponratn E, White N. Reduced microcirculatory flow in severe
falciparum malaria: pathophysiology and electron-microscopic pathology. Acta Trop.
2004;89(3):309-17.
113. Nash G, O'Brien E, Gordon-Smith E, Dormandy J. Abnormalities in the mechanical
properties of red blood cells caused by Plasmodium falciparum. Blood.
1989;74(2):855-61.
114. Miller LH, Ackerman HC, Su X-z, Wellems TE. Malaria biology and disease
pathogenesis: insights for new treatments. Nat Med. 2013;19(2):156-67.
115. Schofield L, Grau G. Immunological processes in malaria pathogenesis. Nature
Reviews Immunology. 2005;5(9):722-35.
116. Lamikanra AA, Brown D, Potocnik A, Casals-Pascual C, Langhorne J, Roberts DJ.
Malarial anemia: of mice and men. Blood. 2007;110(1):18-28.
117. Menendez C, Fleming A, Alonso P. Malaria-related anaemia. Parasitology Today.
2000;16(11):469-76.
118. MacPherson G, Warrell M, White N, Looareesuwan S, Warrell D. Human cerebral
malaria. A quantitative ultrastructural analysis of parasitized erythrocyte
sequestration. Am J Pathol. 1985;119(3):385.
112
119. Turner GDH, Morrison H, Jones M, Davis TME, Looareesuwan S, Buley ID, et al. An
immunohistochemical study of the pathology of fatal malaria: evidence for
widespread endothelial activation and a potential role for intercellular adhesion
molecule-1 in cerebral sequestration. Am J Pathol. 1994;145(5):1057.
120. Craig A, Scherf A. Molecules on the surface of the Plasmodium falciparum infected
erythrocyte and their role in malaria pathogenesis and immune evasion. Mol Biochem
Parasitol. 2001;115(2):129-43.
121. Roberts DJ, Pain A, Kai O, Kortok M, Marsh K. Autoagglutination of malaria-infected
red blood cells and malaria severity. Lancet. 2000;355(9213):1427-8.
122. David PH, Handunnetti SM, Leech JH, Gamage P, Mendis KN. Rosetting: a new
cytoadherence property of malaria-infected erythrocytes. The American Journal of
Tropical Medicine and Hygiene. 1988;38(2):289.
123. Udomsangpetch R, Wåhlin B, Carlson J, Berzins K, Torii M, Aikawa M, et al.
Plasmodium falciparum-infected erythrocytes form spontaneous erythrocyte rosettes.
J Exp Med. 1989;169(5):1835-40.
124. Kaul D, Roth EJ, Nagel R, Howard R, Handunnetti S. Rosetting of Plasmodium
falciparum-infected red blood cells with uninfected red blood cells enhances
microvascular obstruction under flow conditions. Blood. 1991;78(3):812-9.
125. Pain A, Ferguson DJ, Kai O, Urban BC, Lowe B, Marsh K, et al. Platelet-mediated
clumping of Plasmodium falciparum-infected erythrocytes is a common adhesive
phenotype and is associated with severe malaria. Proceedings of the National
Academy of Sciences. 2001;98(4):1805-10.
126. Chen L, Li G, Lu Y, Luo Z. Histopathological changes of Macaca mulatta infected
with Plasmodium knowlesi. Chin Med J (Engl). 2001;114(10):1073-7.
127. Spangler WL, Gribble D, Abildgaard C, Harrison J. Plasmodium knowlesi malaria in
rhesus monkey. Vet Pathol. 1978;15:83-91.
113
128. Miller LH, Fremount HN, Luse SA. Deep vascular schizogony of Plasmodium
knowlesi in Macaca mulatta. Am J Trop Med Hyg. 1971;20(6):816-24.
129. Ozwara H, Langermans JA, Maamun J, Farah IO, Yole DS, Mwenda JM, et al.
Experimental infection of the olive baboon (Paplio anubis) with Plasmodium knowlesi:
severe disease accompanied by cerebral involvement. Am J Trop Med Hyg.
2003;69:188-94.
130. Berendt A, Simmons D, Tansey J, Newbold C, Marsh K. Intercellular adhesion
molecule-1 is an endothelial cell adhesion receptor for Plasmodium falciparum.
Nature. 1989;341(6237):57-9.
131. Dondorp A, Kager P, Vreeken J, White N. Abnormal blood flow and red cell
deformability in severe falciparum malaria. Parasitology Today. 2000;16:228-32.
132. Cranston HA, Boylan CW, Carroll GL, Sutera SP, Gluzman I, Krogstad D.
Plasmodium falciparum maturation abolishes physiologic red cell deformability.
Science. 1984;223(4634):400-3.
133. Paulitschke M, Nash G. Membrane rigidity of red blood cells parasitized by different
strains of Plasmodium falciparum. J Lab Clin Med. 1993;122(5):581.
134. Dondorp A, Angus B, Chotivanich K, Silamut K, Ruangveerayuth R, Hardeman M, et
al. Red blood cell deformability as a predictor of anemia in severe falciparum malaria.
Am J Trop Med Hyg. 1999;60(5):733.
135. Dondorp A, Angus B, Hardeman M, Chotivanich K, Silamut K, Ruangveerayuth R, et
al. Prognostic significance of reduced red blood cell deformability in severe
falciparum malaria. Am J Trop Med Hyg. 1997;57(5):507.
136. Dondorp A, Nyanoti M, Kager P, Mithwani S, Vreeken J, Marsh K. The role of
reduced red cell deformability in the pathogenesis of severe falciparum malaria and
its restoration by blood transfusion. Trans R Soc Trop Med Hyg. 2002;96(3):282-6.
137. Dondorp AM, Kager PA, Vreeken J, White NJ. Abnormal blood flow and red blood
cell deformability in severe malaria. Parasitology Today. 2000;16(6):228-32.
114
138. Naumann KM, Jones GL, Saul A, Smith R. Plasmodium falciparum exo-antigen alters
erythrocyte membrane deformability. FEBS Lett. 1991;292(1):95-7.
139. Nuchsongsin F, Chotivanich K, Charunwatthana P, Fausta O-S, Taramelli D, Day
NP, et al. Effects of malaria heme products on red blood cell deformability. Am J Trop
Med Hyg. 2007;77(4):617-22.
140. Knisely MH, Stratman-Thomas WK. Knowlesi malaria in monkeys; microscopic
pathological circulatory physiology of rhesus monkeys during acute Plasmodium
knowlesi malaria. J Natl Mal Soc. 1945;4:285.
141. Knisely MH, Strathman-Thomas WK, Eliot TS, Bloch EH. Knowlesi malaria in
monkeys IIA: a first step in the separation of the mechanical pathologic circluatory
factors of one slude disease from possible specific toxic factors of that disease.
Angiology. 1964;15(9):411-6.
142. Miller LH, Usami S, Chien S. Alteration in the rheological properties of Palsmodium
knowlesi-infected red cells. A possible mechanisms for capillary obstruction. J Clin
Invest. 1971;50:1451-5.
143. Grau GE, Mackenzie CD, Carr RA, Redard M, Pizzolato G, Allasia C, et al. Platelet
accumulation in brain microvessels in fatal pediatric cerebral malaria. J Infect Dis.
2003;187(3):461-6.
144. Chotivanich K, Sritabal J, Udomsangpetch R, Newton P, Stepniewska KA,
Ruangveerayuth R, et al. Platelet-induced autoagglutination of Plasmodium
falciparum-infected red blood cells and disease severity in Thailand. J Infect Dis.
2004;189(6):1052-5.
145. Wassmer SC, Taylor T, MacLennan CA, Kanjala M, Mukaka M, Molyneux ME, et al.
Platelet-induced clumping of Plasmodium falciparum-infected erythrocytes from
Malawian patients with cerebral malaria—possible modulation in vivo by
thrombocytopenia. J Infect Dis. 2008;197(1):72-8.
115
146. Mayor A, Hafiz A, Bassat Q, Rovira-Vallbona E, Sanz S, Machevo S, et al.
Association of severe malaria outcomes with platelet-mediated clumping and
adhesion to a novel host receptor. PLoS ONE. 2011;6(4):e19422.
147. Biswas AK, Hafiz A, Banerjee B, Kim KS, Datta K, Chitnis CE. Plasmodium
falciparum uses gC1qR/HABP1/p32 as a receptor to bind to vascular endothelium
and for platelet-mediated clumping. PLoS Pathog. 2007;3(9):e130.
148. McMorran BJ, Marshall VM, de Graaf C, Drysdale KE, Shabbar M, Smyth GK, et al.
Platelets kill intraerythrocytic malarial parasites and mediate survival to infection.
Science. 2009;323(5915):797-800.
149. Cox-Singh J, Singh B, Daneshvar C, Planche T, Parker-Williams J, Krishna S. Anti-
inflammatory cytokines predominate in acute human Plasmodium knowlesi infections.
PLoS ONE. 2011;6(6):e20541.
150. Dvorak JA, Miller LH, Whitehouse WC, Shiroishi T. Invasion of erythrocytes by
malaria merozoites. Science. 1975;187(4178):748-50.
151. Bannister L, Butcher G, Dennis E, Mitchell G. Structure and invasive behaviour of
Plasmodium knowlesi merozoites in vitro. Parasitology. 1975;71(03):483-91.
152. Dennis E, Mitchell G, Butcher G, Cohen S. In vitro isolation of Plasmodium knowlesi
merozoites using polycarbonate sieves. Parasitology. 1975;71(3):475-81.
153. Miller LH, Mason SJ, Clyde DF, McGinniss MH. The resistance factor to Plasmodium
vivax in blacks. N Engl J Med. 1976;295(6):302-4.
154. Miller LH, Mason SJ, Dvorak JA, McGinniss MH, Rothman IK. Erythrocyte receptors
for (Plasmodium knowlesi) malaria: Duffy blood group determinants. Science (New
York, NY). 1975;189(4202):561.
155. Singh AP, Ozwara H, Kocken CH, Puri SK, Thomas AW, Chitnis CE. Targeted
deletion of Plasmodium knowlesi Duffy binding protein confirms its role in junction
formation during invasion. Mol Microbiol. 2005;55(6):1925-34.
116
156. King CL, Adams JH, Xianli J, Grimberg BT, McHenry AM, Greenberg LJ, et al.
Fya/Fyb antigen polymorphism in human erythrocyte Duffy antigen affects
susceptibility to Plasmodium vivax malaria. Proceedings of the National Academy of
Sciences. 2011;108(50):20113-8.
157. Haynes JD, Dalton JP, Klotz FW, McGinniss M, Hadley T, Hudson D, et al. Receptor-
like specificity of a Plasmodium knowlesi malarial protein that binds to Duffy antigen
ligands on erythrocytes. J Exp Med. 1988;167(6):1873-81.
158. Moon RW, Hall J, Rangkuti F, Ho YS, Almond N, Mitchell GH, et al. Adaptation of the
genetically tractable malaria pathogen Plasmodium knowlesi to continuous culture in
human erythrocytes. Proceedings of the National Academy of Sciences.
2013;110(2):531-6.
159. Semenya AA, Tran TM, Meyer EV, Barnwell JW, Galinski MR. Two functional
reticulocyte binding-like (RBL) invasion ligands of zoonotic Plasmodium knowlesi
exhibit differential adhesion to monkey and human erythrocytes. Malar J.
2012;11(1):1-12.
160. Kitchen S. The infection of reticulocytes by Plasmodium vivax. The American Journal
of Tropical Medicine and Hygiene. 1938;1(4):347-59.
161. Collins WE, Jeffery GM. Plasmodium ovale: parasite and disease. Clin Microbiol Rev.
2005;18(3):570-81.
162. Kitchen S. The infection of mature and immature erythrocytes by Plasmodium
falciparum and Plasmodium malariae. The American Journal of Tropical Medicine
and Hygiene. 1939;1(1):47-62.
163. Pasvol G, Weatherall D, Wilson R. The increased susceptibility of young red cells to
invasion by the malarial parasite Plasmodium falciparum. Br J Haematol.
1980;45(2):285-95.
164. Hegner R. Relative frequency of ring-stage plasmodia in reticulocytes and mature
erythrocytes in man and monkey. Am J Epidemiol. 1938;27(3):690-718.
117
165. Miller LH, Chien S. Density distribution of red cells infected by Plasmodium knowlesi
and Plasmodium coatneyi Exp Parasitol. 1971;29(3):451-6.
166. Lim C, Hansen E, DeSimone TM, Moreno Y, Junker K, Bei A, et al. Expansion of host
cellular niche can drive adaptation of a zoonotic malaria parasite to humans. Nat
Commun. 2013;4:1638.
167. Grau GE, Taylor TE, Molyneux ME, Wirima JJ, Vassalli P, Hommel M, et al. Tumor
necrosis factor and disease severity in children with falciparum malaria. The New
England journal of medicine. 1989;320(24):1586.
168. Kwiatkowski D, Sambou I, Twumasi P, Greenwood B, Hill A, Manogue K, et al. TNF
concentration in fatal cerebral, non-fatal cerebral, and uncomplicated Plasmodium
falciparum malaria. The Lancet. 1990;336(8725):1201-4.
169. Day NP, Hien TT, Schollaardt T, Loc PP, Van Chuong L, Chau TTH, et al. The
prognostic and pathophysiologic role of pro-and antiinflammatory cytokines in severe
malaria. J Infect Dis. 1999;180(4):1288-97.
170. Singh B, Daneshvar C. Human infections and detection of Plasmodium knowlesi. Clin
Microbiol Rev. 2013;26(2):165-84.
171. Barber BE, William T, Dhararaj P, Anderios F, Grigg MJ, Yeo TW, et al. Epidemiology
of Plasmodium knowlesi malaria in northeast Sabah, Malaysia: family clusters and
wide age distribution. Malar J. 2012;11:401.
172. Phillips A, Bassett P, Szeki S, Newman S, Pasvol G. Risk factors for severe disease
in adults with falciparum malaria. Clin Infect Dis. 2009 April 1, 2009;48(7):871-8.
173. Department of Statistics, Malaysia. Population distribution and basic demographic
characteristics, population and housing census of Malaysia 2010. Kuala Lumpur:
Department of Statistics, Malaysia; 2011.
174. McMorrow J, Talip MA. Decline of forest area in Sabah, Malaysia: Relationship to
state policies, land code and land capability. Global Environmental Change.
2001;11(3):217-30.
118
175. Sabah Forestry Department. Geography of Sabah. [updated 23rd May 2012];
Available from: http://www.forest.sabah.gov.my/resources/geography.asp.
176. World Wildlife Foundation. Borneo: treasure island at risk: World Wildlife Foundation
Germany; 2005.
177. Berkley JA, Bejon P, Mwangi T, Gwer S, Maitland K, Williams TN, et al. HIV infection,
malnutrition, and invasive bacterial infection among children with severe malaria. Clin
Infect Dis. 2009;49(3):336-43.
178. Genton B, D'Acremont V, Rare L, Baea K, Reeder JC, Alpers MP, et al. Plasmodium
vivax and mixed infections are associated with severe malaria in children: a
prospective cohort study from Papua New Guinea. PLoS Med. 2008;5(6):e127.
179. Manning L, Laman M, Law I, Bona C, Aipit S, Teine D, et al. Features and prognosis
of severe malaria caused by Plasmodium falciparum, Plasmodium vivax and mixed
Plasmodium species in Papua New Guinean children. PLoS ONE.
2011;6(12):e29203.
180. Tjitra E, Anstey NM, Sugiarto P, Warikar N, Kenangalem E, Karyana M, et al.
Multidrug-resistant Plasmodium vivax associated with severe and fatal malaria: a
prospective study in Papua, Indonesia. PLoS Med. 2008;5(6):e128.
181. Kochar DK, Das A, Kochar SK, Saxena V, Sirohi P, Garg S, et al. Severe
Plasmodium vivax malaria: a report on serial cases from Bikaner in northwestern
India. Am J Trop Med Hyg. 2009;80(2):194-8.
182. Alexandre MA, Ferreira CO, Siqueira AM, Magalhães BL, Mourão MPG, Lacerda MV,
et al. Severe Plasmodium vivax Malaria, Brazilian Amazon. Emerg Infect Dis.
2010;16(10):1611-4.
183. Lacerda MVG, Fragoso SCP, Alecrim MGC, Alexandre MAA, Magalhães BML,
Siqueira AM, et al. Postmortem characterization of patients with clinical diagnosis of
Plasmodium vivax malaria: to what extent does this parasite kill? Clin Infect Dis.
2012.
119
184. Douglas N, Anstey N, Buffet P, Poespoprodjo J, Yeo T, White N, et al. The anaemia
of Plasmodium vivax malaria. Malar J. 2012;11(1):135.
185. Kaushik R, Kaushik RM, Kakkar R, Sharma A, Chandra H. Plasmodium vivax malaria
complicated by acute kidney injury: experience at a referral hospital in Uttarakhand,
India. Trans R Soc Trop Med Hyg. 2013;107(3):188-94.
186. Lampah D, Yeo TW, Hardianto SO, Tjitra E, Kenangalem E, Sugiarto P, et al. Coma
associated with microscopy-diagnosed Plasmodium vivax: a prospective study in
Papua, Indonesia. PLoS Negl Trop Dis. 2011;5(6):e1032.
187. Anstey NM, Douglas NM, Poespoprodjo JR, Price R. Plasmodium vivax: clinical
spectrum, risk factors and pathogenesis. Adv Parasitol. 2012;80:151-201.
188. Snounou G, Viriyakosol S, Jarra W, Thaithong S, Brown KN. Identification of the four
human malaria parasite species in field samples by the polymerase chain reaction
and detection of a high prevalence of mixed infections. Mol Biochem Parasitol.
1993;58(2):283-92.
189. Arai M, Kunisada K, Kim H-S, Miyake H, Mizukoshi C, Kakutani T, et al. A
colorimetric DNA diagnostic method for falciparum malaria and vivax malaria: a field
trial in the Solomon Islands. Nucleosides Nucleotides Nucleic Acids. 1996;15(1-
3):719-31.
190. Dondorp A, Ince C, Charunwatthana P, Hanson J, Van Kuijen A, Faiz M, et al. Direct
in vivo assessment of microcirculatory dysfunction in severe falciparum malaria. J
Infect Dis. 2008;197(1):79-84.
191. Barnwell JW, Ingravallo P, Galinski MR, Matsumoto Y, Aikawa M. Plasmodium vivax:
Malarial proteins associated with the membrane-bound caveola-vesicle complexes
and cytoplasmic cleft structures of infected erythrocytes. Exp Parasitol.
1990;70(1):85-99.
192. Abu Sayeed, A, Maude R, Hasan M, Mohammed N, Hoque M, Dondorp A, et al.
Malarial retinopathy in Bangladeshi adults. Am J Trop Med Hyg. 2011;84:141-7.
120
193. Yeo TW, Lampah DA, Gitawati R, Tjitra E, Kenangalem E, McNeil YR, et al. Impaired
nitric oxide bioavailability and L-arginine–reversible endothelial dysfunction in adults
with falciparum malaria. J Exp Med. 2007;204(11):2693-704.
194. Yeo TW, Lampah DA, Kenangalem E, Tjitra E, Price RN, Anstey NM. Impaired
skeletal muscle microvascular function and increased skeletal muscle oxygen
consumption in severe falciparum malaria. J Infect Dis. 2013;207(3):528-36.
195. Beare N, Harding S, Taylor T, Lewallen S, Molyneux M. Perfusion abnormalities in
children with cerebral malaria and malarial retinopathy. J Infect Dis. 2009;199:263-
71.
196. Beare N, Southern C, Chalira C, Taylor T. Prognostic significance and course of
retinopathy in children with severe malaria. Arch Ophtahlmol. 2004;122:1141-7.
197. Glover S, Harding S, ME M, Taylor T, Beare N. Retinal angiographic changes in
paediatric cerebral malaria. Am J Trop Med Hyg. 2009;81(Supp 1):54.
198. Lewallen S, White V, Whitten R, Gardine J, Hoar B, Lindley J, et al. Clinical-
histopathological correlation of the abnormal retinal vessels in cerebral malaria. Arch
Ophthalmol. 2000;118:924-8.
199. Mancini DM, Bolinger L, Li H, Kendrick K, Chance B, Wilson JR. Validation of near-
infrared spectroscopy in humans. J Appl Physiol. 1994;77(6):2740-7.
