Lymphangitis-Associated Rickettsiosis • CID 2005:40 (15 May) • 1435

M A J O R A R T I C L E

Lymphangitis-Associated Rickettsiosis,a New Rickettsiosis Caused by Rickettsia sibiricamongolotimonae: Seven New Cases and Reviewof the Literature

Pierre-Edouard Fournier,1 Frédérique Gouriet,1 Philippe Brouqui,1,2 Frédéric Lucht,3 and Didier Raoult1,21Unité des Rickettsies, Université de la Méditerranée, Faculté de Médecine, and 2Service de Maladies Infectieuses et Tropicales, Hôpital Nord,Marseille, and 3Service de Maladies Infectieuses, Hôpital Bellevue, Saint-Etienne, France

Background. Rickettsia sibirica mongolotimonae has been found in Hyalomma ticks in Inner Mongolia (inChina) and Niger and in humans in France and South Africa. To date, only 3 cases of human infection have beenreported.

Methods. Patients received a diagnosis of R. sibirica mongolotimonae infection on the basis of culture and/orPCR results plus serological test results.

Results. From January 2000 to June 2004, R. sibirica mongolotimonae infection was diagnosed in 7 patients.In 3 patients, the bacterium was cultivated from the inoculation eschar. The other 4 patients had cases that werediagnosed with use of PCR of samples obtained from the eschar (2 patients) or blood (2 patients), plus specificWestern blot before (2 patients) and after (2 patients) cross-adsorption. The clinical presentation included fever(temperature, 138.5�C), a maculopapular rash, and �1 inoculation eschar in 6 patients, enlarged regional lymphnodes in 4 patients, and lymphangitis in 3 patients. On the basis of the study of 9 cases, R. sibirica mongolotimonaeinfection differed from other tick-borne rickettsioses in the Mediterranean area in the following ways: it involveda specific incidence in the spring, the presence of 2 eschars in 2 (22%) of the patients, the presence of a draininglymph node in 5 (55%) of the patients, and lymphangitis expanding from the inoculation eschar to the drainingnode in 4 (44%) of the patients. The most recent patient in our series received a clinical diagnosis on the basisof such findings. All patients recovered without any sequelae.

Conclusions. We propose that this new rickettsiosis be named “lymphangitis-associated rickettsiosis.” Lym-phangitis-associated rickettsiosis should be considered in the differential diagnosis of tick-borne rickettsioses inEurope, Africa, and Asia.

Species definition in Rickettsia species is subject to con-

troversy. On the basis of serotyping [1], Rickettsia si-

birica mongolotimonae, initially named strain HA-91

[2], was reported as the new species Rickettsia mon-

golotimonae [2, 3]. According to our recently published

gene sequence–based criteria to define Rickettsia spe-

cies, R. sibirica mongolotimonae belonged to the Rick-

ettsia sibirica species [4] (figure 1). However, this rick-

Received 26 October 2004; accepted 12 January 2005; electronically published7 April 2005.

Reprints or correspondence: Dr. Didier Raoult, Unité des Rickettsies, CNRSUPRESA 6020, IFR 48, Université de la Méditerranée, Faculté de Médecine, 27,Bd Jean Moulin, 13385 Marseille Cedex 05, France (didier.raoult@medecine.univ-mrs.fr).

Clinical Infectious Diseases 2005; 40:1435–44� 2005 by the Infectious Diseases Society of America. All rights reserved.1058-4838/2005/4010-0009$15.00

ettsia exhibits serotypic and ecological specificities

(table 1). In particular, it has been found in Asia, Eu-

rope, and Africa [2, 5, 6], whereas R. sibirica, the agent

of North Asian tick typhus, is confined to Siberia and

western China [7–9]. In this study, we will refer to it

as “R. sibirica mongolotimonae” and to the agent of

North Asian tick typhus as “R. sibirica sensu stricto.”

R. sibirica mongolotimonae was first isolated in Bei-

jing from Hyalomma asiaticum ticks collected in the

Alashian region of Inner Mongolia in 1991 [2]. In 1996,

we diagnosed the first case of human infection in south-

ern France in a woman who presented with a febrile

rash and a single inoculation eschar in the groin [6].

In 2000, we diagnosed a second human case in a man

who developed an inoculation eschar on the leg, fever,

and lymphangitis expanding from the eschar to an en-

larged and painful lymph node in the groin [3]. Re-

1436 • CID 2005:40 (15 May) • Fournier et al.

Figure 1. Unrooted tree showing the phylogenetic relationships among tick-associated rickettsiae, as inferred from sequence analysis of the ompBgene with use of the maximum-parsimony method. Rickettsiae of recognized pathogenicity are indicated by boldface type. Bootstrap values areindicated at the nodes.

Table 1. Comparison of the epidemiological and clinical characteristics associated with Rickettsia sibirica mongolotimonae andRickettsia sibirica sensu stricto.

Characteristic, by class R. sibirica mongolotimonae (PR) [3, 6] R. sibirica sensu stricto [7–9]

Epidemiological characteristicRecognized tick vectors Hyalomma asiaticum [2], Hyalomma

truncatum [38]Dermacentor marginatus, Dermacentor nuttali,

Dermacentor silvarum [39], Dermacentorpictus [7], Dermacentor sinicus, Dermacen-tor auratus, Haemaphysalis concinna,Hyalomma wellingtoni, Hyalomma yeni [40]

Geographic area(s) of endemicity Algeria (PR), China [31], France (PR),Niger [38], and South Africa [41]

Siberia [8] and Western China [42, 43]

Outbreak season(s) for human infection Springa Spring and summerClinical characteristic

Headache 55 100Fever 100 YesRash 78 100Enlarged lymph nodes 55 YesLymphangitis 44 4Eschar 89 77Multiple eschars 22 0Fatal outcome 0 Rare

NOTE. Data are percentage of reported patients with the specified characteristic, unless otherwise indicated. PR, present report.a In France.

cently, a third case was diagnosed in South Africa in a man

who developed an inoculation eschar on a toe, fever, headache,

and lymphangitis expanding from the eschar to an enlarged

inguinal lymph node [10]. Since 2000, we have diagnosed in

our laboratory an additional 7 human cases by culture and/or

PCR plus serological testing. The aim of this study was to

describe the epidemiological and clinical characteristics of our

7 new patients, together with those from the literature, and to

Table 2. Epidemiological, clinical, and microbiological characteristics associated with patients infected with Rickettsia sibirica mongolotimonae.

Characteristic Patient 1 Patient 2 Patient 3 Patient 4 Patient 5 Patient 6 Patient 7

Sex M F M F M M FAge, years 53 40 59 21 70 55 62Month of onset May Early July April May May June AprilAt-risk activity Gardening Gardening Gardening Gardening Gardening, contact

with birdsWalk in Camargue

National ParkTravel to southern

AlgeriaIncubation period, days 3 5 6 NA NA NA NAReport of tick bite Yes Yes Yes No No No NoBody temperature, �C 39.5 38.8 39.0 39.2 39.0 39.5 39.5Symptom(s)

Headache Yes No Yes No Yes No YesMyalgias Yes Yes No Yes Yes Yes YesNo. of eschar (location) 1 (left knee) 0 1 (back) 1 (right heel) 2 (right forearm and

abdomen)1 (right arm) 2 (left foot and

hypochondrium)Enlarged regional lymph nodes No Yes No Yes Yes Yes NoLymphangitis No Yes No Yes No Yes NoMaculopapular rash Yes Yes Yes No Yes Yes Yes

Microimmunofluorescence serological test results,IgG titer/IgM titer

First sample !64/!32 256/64 !64/!32 !64/!32 !64/!32 !64/!32 !64/!32Second sample NA 512/128a !64/!32 !64/!32 1024/!32 64/!32 64/32

Western blot showing antibodies to R. sibiricamongolotimonae only

Yes No No Yes Yes No No

Cross-adsorption followed by Western blot show-ing antibodies to R. sibirica mongolotimonaeonly

ND Yes ND ND Yes Yes Yes

Results of nested PCR of serum samples ND Positive ND Positive ND ND PositiveEschar culture Negative ND Positive ND Positive ND PositiveResults of PCR of eschar samples Positive ND Positive ND Positive Positive Positive

NOTE. IgG titers !1:64 and IgM titers !1:32 were reported as negative. IgG and IgM titers are expressed as reciprocals. Culture and PCR results positive for R. sibirica mongolotimonae arereported as positive. NA, specimen not available; ND, not done.

a Specific against R. sibirica mongolotimonae.

1438 • CID 2005:40 (15 May) • Fournier et al.

Table 3. Epidemiological and clinical characteristics of the main tick-borne spotted fever rickettsioses worldwide for which theavailability of detailed features allows a statistical comparison with lymphangitis-associated rickettsiosis.

Characteristic, by classRickettsia sibirica

mongolotimonae [3, 6]Rickettsia

conorii (PR)Rickettsia

slovaca (PR)

Epidemiological characteristicMain tick vector Unknown (Hyalomma

species suspected)Rhipicephalus species Dermacentor species

Geographic area(s) of endemicity Southern France, Africa,and Asia

The Mediterranean, the BlackSea, and India

Europe

Outbreak seasona Spring Summer Oct–MayClinical characteristic

No. of patients 9 85 58Incidence, by seasona,b

Autumn 0 14.1% (P p .8) 25.9% (P p .3)Winter 0 0 34.5% (P p .1)Spring 87.5% 4.7% (P ! .01c) 36.2% (P p .02c)Summer 12.5% 81.2% (P ! .01c) 3.4% (P p .8)

Sex ratio, M:F 1.25 1.57 (P p .5) 0.48 (P p .2)Children !10 years old 0% 3% (P p .7) 21% (P p .1)Symptom

Headache 55% 48% (P p .5) 33% (P p .2)Fever 100% 100% (P p 1.0) 38% (P ! .01c)Rash 78% 93% (P p .2) 10% (P ! .01c)Enlarged lymph nodes 55% 1% (P ! .01c) 100% (P ! .01c)Lymphangitis 44% 0% (P ! .01c) 0% (P ! .01c)Eschar 89% 94% (P p .5) 100% (P p .1)Multiple eschars 22% !1% (P ! .01c) 0% (P p .01c)

Fatality rate 0% 2.3% (P p .8) 0% (P p 1.0)

NOTE. Data are percentage of patients in the indicated report(s) with the specified characteristic, unless otherwise indicated. NA, not available; PR, presentreport.

a The seasons were defined as follows: autumn, 21 September–20 December; winter, 21 December–20 March; spring, 21 March–20 June; and summer, 21June–20 September.

b The seasonality of the disease in France was estimated by comparing only patients who were infected in France.c Indicates statistical significance ( ).P ! .05

compare these characteristics with those associated with the

major tick-borne rickettsioses worldwide to further characterize

the pathogenic role of R. sibirica mongolotimonae.

PATIENTS AND METHODS

Study design. Appropriate informed consent was obtained

from all patients. From January 1996 to June 2004, we pro-

spectively studied serum specimens and, when available, skin

biopsy and/or whole-blood samples obtained from patients

with a suspected arthropod-borne infection referred to our

laboratory. Then, we compared the epidemiological and clinical

features of patients who fulfilled the diagnostic criteria de-

scribed below for R. sibirica mongolotimonae infection with

those of patients with the main rickettsioses diagnosed in

France at our center (i.e., Mediterranean spotted fever [MSF]

and tick-borne lymphadenopathy [TIBOLA]). We also com-

pared these features with those associated with African tick-

bite fever (ATBF), which is the most frequent traveler-associated

rickettsiosis diagnosed in our laboratory. Subsequently, we also

compared R. sibirica mongolotimonae infection with large pub-

lished series of the main tick-borne rickettsioses worldwide to

define its clinical characteristics.

Case definition. A definite diagnosis of infection with R.

sibirica mongolotimonae was made on the basis of the isolation

of this rickettsia from clinical specimens or the association of

a PCR result and serological test result positive for R. sibirica

mongolotimonae. Patients with MSF, TIBOLA, and ATBF were

classified as having definite cases when they fulfilled the pre-

viously described diagnostic criteria, including epidemiological

and clinical criteria, and had positive culture and serological

test results [11–14]. PCR results were also considered among

diagnostic criteria [15]. For each patient, epidemiological and

clinical data were collected by the consulting physician with

use of a standardized questionnaire at the time of clinical

examination.

Serological tests. For each patient, an acute-phase serum

sample was obtained within 2 weeks after the onset of symp-

toms and, when possible, a convalescent-phase serum sample

Lymphangitis-Associated Rickettsiosis • CID 2005:40 (15 May) • 1439

Rickettsiaafricae (PR)

Rickettsiarickettsii [44]

Rickettsiajaponica [38]

Rickettsiaaustralis [45]

Rickettsiaheilongjiangensis [41]

Amblyomma species Dermacentor species Haemaphysalis longicornisDermacentor taiwanensis

Ixodes holocyclus,I. tasmani, I. cornuatus

Dermacentor silvarum

Sub-Saharan Africa andthe West Indies

The Americas Japan Eastern Australia China and the RussianFar East

All year Apr–Aug Apr–Oct Jun–Nov Summer

139 262 31 37 13

… … … … …… … … … …… … … … …… … … … …

1.48 (P p .5) 1.22 (P p .6) 0.40 (P p .1) 2.7 (P p .5) 1.6 (P p .6)0.7% (P p .9) NA Rare NA 0% (1.0)

15% (P ! .01c) 91% (P ! .01c) 80% (P p .1) 90% (P p .06) 100% (P p .01c)89% (P p .4) 99% (P p .9) 100% (P p 1.0) 100% (P p 1.0) 100% (P p 1.0)51% (P p .1) 88% (P p .3) 100% (P p .04c) 94% (P p .3) 92% (P p .3)49% (P p .5) 27% (P p .07) No 84% (P p .1) 77% (P p .3)0.7% (P ! .01c) 0% (P ! .01c) 0% (P ! .01c) 0% (P ! .01c) 15% (P p .1)

99% (P p .1) Rare 90% (P p .6) 65% (P p .3) 92% (P p .7)46% (P p .1) 0% (P ! .01c) 0% (P p .04c) 0% (P p .04c) 0% (P p .1)0% (P p 1.0) 4% (P p .7) Rare 2% (P p .4) 0% (P p 1.0)

(i.e., one collected 12 weeks after onset of symptoms) was also

obtained. IgG and IgM antibody titers were estimated with use

of the microimmunofluorescence (MIF) assay, as reported else-

where [16]. We used the following antigens: for patients infected

in France, we used R. sibirica mongolotimonae strain HA-91,

ATCC VR-1526 [2], Rickettsia conorii strain Malish, ATCC VR-

613 [17], Rickettsia slovaca strain 13B [18], Rickettsia helvetica

strain C9P9 [19], Rickettsia massiliae strain Mtu1 [20], and

Rickettsia felis strain Marseille, ATCC VR-1525 [21]; for patients

returning from sub-Saharan Africa, we used Rickettsia africae

strain ESF-5, R. conorii strain Malish, Rickettsia aeschlimannii

strain MC16, R. sibirica mongolotimonae strain HA-91, Rick-

ettsia akari strain MK, and R. felis strain Marseille; in addition,

to estimate the degree of cross-reactivity within the R. sibirica

species, we also used R. sibirica sensu stricto strain 246, ATCC

VR151T [22]. Titers of 1:64 for IgG and 1:32 for IgM were

used as cutoff values. Western blotting procedures were per-

formed as described elsewhere [23] with use of R. sibirica mon-

golotimonae and R. sibirica sensu stricto antigens.

Serological evidence of infection with R. sibirica mongoloti-

monae was considered to be present when IgG and IgM titers

were at least 2 serial dilutions higher for R. sibirica mongolo-

timonae than for other tested rickettsiae, including R. sibirica

sensu stricto, or when the western blot profile showed only

antibodies to R. sibirica mongolotimonae.

Culture. Attempted cultivation of rickettsiae from skin bi-

opsy specimens and heparinized blood samples was performed

using the shell-vial cell culture technique, as previously reported

[24].

PCR amplification and sequencing. DNA was extracted

from EDTA blood specimens and ground skin biopsy samples

by using the QIAamp Tissue Kit (Qiagen) according to the

manufacturer’s recommendations. These extracts were used as

templates in previously described PCR assays incorporating the

primers 190–70 and 190–701, which amplify a 630-bp fragment

of the ompA gene [25], and 877F and 1258R, which amplify a

381-bp fragment of the gltA gene [26]. As negative controls,

we used sterile water processed as described above and DNA

extracted from a heart valve from a patient with degenerative

valvulopathy that was incorporated into every 6 specimens. As

positive control, we used DNA from Rickettsia montanensis

strain M/5–6 [27]. Testing was done blindly. When regular PCR

performed using the skin biopsy sample had negative results

or when an acute-phase serum sample (but no skin biopsy

sample) was available, we performed a nested PCR incorpo-

rating the ompA-amplifying primer sets AF1F–AF1R and AF2F–

1440 • CID 2005:40 (15 May) • Fournier et al.

Figure 2. Two inoculation eschars (arrows), one on the forearm andthe other on the abdomen, observed in a patient infected with Rickettsiasibirica mongolotimonae.

AF2R for the nested amplification, as described elsewhere [12,

15]. We incorporated the above-described negative controls in

every 6 specimens. To avoid contamination, we did not include

any positive control in this assay. All positive PCR products

were sequenced in both directions, as described elsewhere [28].

The epidemiological, clinical, and microbiologic features of the

7 patients are detailed in table 2.

Histopathological and immunohistochemical testing.

The eschar biopsy sample from patient 5 was formalin-fixed,

paraffin-embedded, and then cut to 3-mm thickness and stained

with hematoxylin-eosin-saffron with use of routine staining

methods. Serial sections were also obtained to perform im-

munohistochemical investigations, as described elsewhere [29],

using mouse antiserum produced against R. sibirica mongolo-

timonae [2]. A skin biopsy sample obtained from a patient with

psoriasis was used as a negative control and was processed as

described above.

Statistical analysis. To estimate the specificity of clinical

characteristics of R. sibirica mongolotimonae infection, we con-

ducted a statistical comparison of patients with MSF, TIBOLA,

and ATBF whose diagnosis was made in our laboratory (Unite

des Rickettsies; Marseille, France) with a large published series

of patients with RMSF, Siberian tick typhus, Japanese spotted

fever, Queensland tick typhus, and Rickettsia heilongjiangensis

infection (table 3) using Fisher’s exact test. The seasonality of

rickettsioses in France was further studied by comparing R.

sibirica mongolotimonae infection with MSF and TIBOLA after

stratification of patients according to the season during which

they developed the disease. Observed differences were consid-

ered significant when P was !.05 for 2-tailed tests.

RESULTS

R. sibirica mongolotimonae infection. From January 1996

through June 2004, 9 patients had cases that we diagnosed as

fulfilling our case definition for R. sibirica mongolotimonae in-

fection, including 2 patients whose cases have previously been

reported [3, 6]. In the 7 patients whose cases had not previously

been reported, the diagnosis of infection with R. sibirica mon-

golotimonae was made on the basis of a positive culture result

in 3 cases and on an association of positive PCR results and a

specific antibody response in 4 cases (table 2). Six of the cases

occurred during the spring, and only 1 infection occurred dur-

ing the summer (in early July). Six patients lived in southern

France, and 1 patient had recently returned from a trip to

southern Algeria. Four patients were male, and 3 patients were

female. The median age of the patients was 55 years (range,

21–70 years). A Tick bite or tick-handling was reported by 3

patients, but no tick was collected for further examination. In

France, contact with ticks occurred for 5 patients in their garden

and for 1 patient during a walk in the Camargue National Park.

In Algeria, the patient had contacts with camel ticks but did

not remember receiving any tick bites. The median incubation

time was 5 days (range, 3–6 days). Symptoms at onset included

fever in all patients (temperature range, 38.8�C–39.5�C), my-

algias in 6 patients, and headache in 4 patients. Four patients

developed a single inoculation eschar on the lower or upper

limbs or on the trunk. Two patients presented with 2 eschars

(figure 2). One patient developed no eschar but did develop

an enlarged lymph node and lymphangitis in the territory

draining the tick-bite site on the leg. Another 3 patients each

presented with an enlarged lymph node in the territory draining

the eschar, including 2 patients who also had lymphangitis

expanding from the eschar to the draining node. Our initial

diagnosis for the first 5 patients from France was spotted fever

rickettsiosis. The presence of lymphangitis expanding from the

eschar to the draining lymph node in patient 6 (table 2) led 2

of the investigators (D.R. and P.B.) to clinically suspect infection

with R. sibirica mongolotimonae. Five patients developed a gen-

eralized maculopapular rash involving the palms and soles but

not the face, and 1 patient had a dozen maculopapular spots

Lymphangitis-Associated Rickettsiosis • CID 2005:40 (15 May) • 1441

Figure 3. Western immunoblot results for patient 6 before and after cross-adsorption with Rickettsia conorii, Rickettsia sibirica mongolotimonae,or Rickettsia sibirica sensu stricto, showing an antibody response directed against an outer membrane protein of R. sibirica mongolotimonae only.Lanes 1, 4, 7, and 10, R. conorii antigen; lanes 2, 5, 8,, and 11, R. sibirica mongolotimonae antigen; lanes 3, 6, 9, and 12, R. sibirica sensu strictoantigen; lanes 1–3, untreated serum; lanes 4–6, serum adsorbed with R. conorii (antibodies to both antigens remain); lanes 7–9, serum adsorbed withR. sibirica mongolotimonae (no antibodies remain); lanes 10–12, serum adsorbed with R. sibirica sensu stricto (antibodies to R. sibirica mongolotimonaeremain). MM, molecular mass.

on the trunk. Six patients were administered oral doxycycline,

and 1 patient received amoxicillin. The 7 patients recovered

without any sequelae.

The results of serological testing with MIF were positive for

4 of the 7 patients (table 2). One of the 4 patients with positive

titers (patient 2) had a 4-fold higher titer to R. sibirica mon-

golotimonae (IgG titer, 1:512; IgM titer, 1:128) than to other

tested rickettsiae, including R. sibirica sensu stricto (IgG titer,

1:128; IgM titer, 1:128). In the other 3 cases, MIF titers to all

tested antigens were identical. In 3 patients, the Western blot

demonstrated the presence of antibodies to high–molecular

weight proteins (rOmpA and rOmpB) of R. sibirica mongolo-

timonae only. Another 3 patients had antibodies specifically

directed against R. sibirica mongolotimonae after serum cross-

adsorption with R. sibirica sensu stricto (figure 3). Altogether,

6 of 7 patients exhibited a specific serological reaction to R.

sibirica mongolotimonae. Skin biopsy samples were the most

useful specimens for both PCR (with 5 of 5 patients having

positive results) and culture (with 3 of 5 patients positive for

R. sibirica mongolotimonae). The histopathological features of

the inoculation eschar from patient 5 were dominated by the

presence of endothelial swelling, mural and occlusive fibrin

thrombi observed in a few blood vessels, dermal edema, and

cutaneous necrosis. Intramural and perivascular infiltration by

polymorphonuclear leukocytes, small lymphocytes, and mac-

rophages were also observed. Rickettsiae were detected by im-

munohistochemical analysis in the endothelium and in inflam-

matory cells organized in and around the blood vessels (figure

4). The rickettsia was also cultivated from blood samples ob-

tained from patient 7. In addition, nested PCR of serum sam-

ples from 3 patients (including 2 patients for whom no skin

biopsy sample was available) yielded positive results. R. sibirica

mongolotimonae was identified on the basis of a 100% similarity

in ompA nucleotide sequence for French isolates and amplicons

and a 99.9% similarity for the Algerian isolate, with sequences

available in GenBank for R. sibirica mongolotimonae.

Other tick-borne rickettsioses. From January 1996 through

June 2004, we also diagnosed 121 cases of definite MSF (85 of

which occurred in patients who were infected in southern

France), 68 cases of definite TIBOLA (58 of which occurred in

patients who were infected in southern France), and 139 cases

of definite ATBF. The epidemiological and clinical character-

istics of the patients are detailed in table 3.

Statistical comparison. In comparison with MSF and TI-

BOLA, R. sibirica mongolotimonae infection exhibited specific

features, including occurrence in the spring ( andP ! .01 P p

, respectively), occurrence of lymphangitis ( for both.02 P ! .01

diseases), and the presence of multiple eschars ( andP ! .01

, respectively) (table 3). The presence of enlarged lymphP p .01

nodes was significantly more common in patients with R. si-

birica mongolotimonae infection than in patients with MSF

( ) and was significantly less common in patients with R.P ! .01

sibirica mongolotimonae infection than in patients with TIBOLA

( ). TIBOLA was also associated with significantly fewerP ! .01

1442 • CID 2005:40 (15 May) • Fournier et al.

Figure 4. Immunohistochemical detection of Rickettsia sibirica mongolotimonae in the inoculation eschar of patient 5. Note the location of thebacteria (arrows) in the inflammatory cells present in the dermis (mouse anti–R. sibirica mongolotimonae antiserum and hematoxylin counterstain;original magnification, �400).

instances of fever ( ) and rash ( ) than was R. sibiricaP ! .01 P ! .01

mongolotimonae infection. When compared with other major

rickettsioses worldwide, 2 features of R. sibirica mongolotimonae

infection appeared unusual: the presence of lymphangitis

(which occurs in no other major rickettsioses except for R.

heilongjiangensis infection) and multiple eschars (which occurs

in no other major rickettsioses except for ATBF). In addition,

the development of a headache was significantly more common

in patients with R. sibirica mongolotimonae infection than in

patients with ATBF ( ), but it was significantly less com-P ! .01

mon in patients with R. sibirica mongolotimonae infection than

in patients with Rocky Mountain spotted fever ( ) or R.P p .03

heilongjiangensis infection ( ); significantly more patientsP p .04

with R. sibirica mongolotimonae infection than patients with

TIBOLA were febrile ( ); and patients with R. sibiricaP ! .01

mongolotimonae infection developed a cutaneous rash signifi-

cantly more frequently than did patients with TIBOLA (P !

) but significantly less frequently than did patients with Jap-.01

anese spotted fever ( ).P p .04

DISCUSSION

In this report, we describe the epidemiological and clinical

characteristics of R. sibirica mongolotimonae infection on the

basis of data from 9 patients who received that diagnosis in

our laboratory, including 7 new patients. Six of our 7 new

patients exhibited specific serological evidence of R. sibirica

mongolotimonae infection. Of these 6 patients, only 1 exhibited

an MIF antibody response specifically directed against R. sibirica

mongolotimonae. With use of Western blot, we observed in

another 3 patients an early antibody response specifically di-

rected against high–molecular weight proteins of R. sibirica

mongolotimonae. The higher specificity of early antibodies for

high–molecular weight, surface-exposed antigens has previ-

ously been observed for other rickettsioses [23, 30]. In 4 pa-

tients, the specificity of antibodies for these high–molecular

weight proteins of R. sibirica mongolotimonae was demonstrated

using cross-adsorption followed by Western blot (figure 3). We

confirm that high–molecular weight, surface-exposed proteins

of R. sibirica mongolotimonae exhibit an antigenic specificity

that elicits in humans, as in mice [31], a specific antibody

response. Culture and PCR of the eschar biopsy sample estab-

lished the diagnosis in 5 patients, including 1 patient in whom

the rickettsia was also isolated from blood samples. Nested PCR

of serum samples [15] led to the diagnosis in an additional 2

patients. With use of immunohistochemical tests, we could also

detect the rickettsia in 1 eschar biopsy sample.

In France, infection with R. sibirica mongolotimonae occurred

in the spring for 5 patients and in the beginning of summer

for 1 patient. The 2 French patients with previously reported

cases had also developed the disease in the spring [3, 6]. The

Lymphangitis-Associated Rickettsiosis • CID 2005:40 (15 May) • 1443

patient who was infected in Algeria was infected in April, and

the patient who was infected in South Africa was hospitalized

in September [10]. In southern France, the time of outbreak

of R. sibirica mongolotimonae infection (in the spring) was sig-

nificantly different from that of the other 2 rickettsioses en-

demic in the area, MSF due to R. conorii ( ) and TIBOLAP ! .01

due to R. slovaca ( ); these 2 diseases mostly occur inP p .02

the summer and the winter, respectively [13, 32]. In Algeria,

Mediterranean spotted fever, the main tick-borne spotted fever

rickettsiosis, also occurs in the summer.

When combining the clinical data from our 7 cases with

clinical data from the 2 previously published French cases [3,

6], the median incubation period was 6 days (range, 3–8 days).

The clinical presentation in most of the cases included fever

(temperature, 138.5�C), eschar(s), and a generalized maculo-

papular rash. When compared with definite cases of MSF and

TIBOLA diagnosed in our laboratory (table 3), R. sibirica mon-

golotimonae infection represented 6% of diagnosed rickettsioses

in the area and was characterized by the unusual features of

enlarged lymph nodes in the territory draining the eschar

( ) and lymphangitis expanding from the eschar to theP ! .01

draining node ( ). The South African patient also pre-P ! .01

sented with an enlarged lymph node and lymphangitis [10].

Therefore, lymphangitis, which was found in 5 (50%) of 10 of

the reported patients, may be considered relatively typical. In

fact, 2 of the investigators (D.R. and P.B.) clinically identified

the disease in a patient on the basis of the assumption that the

lymphangitis was specific. Therefore, we propose that this dis-

ease be named “lymphangitis-associated rickettsiosis.” In ad-

dition, 2 patients from our series presented with 2 inoculation

eschars (figure 2), a feature rarely encountered in patients with

MSF in our laboratory ( ) [32, 33], and in TIBOLAP ! .01

( ) [13]. No severe cases were noted, because all patientsP p .01

recovered without any sequelae. Infection with R. sibirica mon-

golotimonae also differed significantly from other rickettsioses

frequently encountered worldwide because it is the only one

(with the exception of R. heilongjiangensis infection) charac-

terized by the presence of lymphangitis (found in 15% of cases)

and the only one (with the exception of ATBF) characterized

by multiple eschars (table 3). Unfortunately, no statistical com-

parison was possible with North Asian tick typhus caused by

R. sibirica sensu stricto [7–9]. Nevertheless, the occurrence of

lymphangitis and multiple eschars appears to be differ between

patients with R. sibirica sensu stricto infection and those with

R. sibirica mongolotimonae infection (table 1). When compared

with nonrickettsial diseases, R. sibirica mongolotimonae should

be considered a differential diagnosis of tularemia in patients

who develop a nodular lymphangitis as a result of a tick bite

[34].

The presence of multiple eschars, as observed in 2 of our

patients, is common in patients with ATBF [12] because of the

hunter behavior of Amblyomma ticks (in response to stimuli,

they specifically converge on nearby hosts) [31], but it is un-

usual among patients with other rickettsioses. The presence of

2 eschars in our patients may be explained by the affinity of

the tick-vector for humans. However, the vector of R. sibirica

mongolotimonae has been identified in neither France nor South

Africa to date. This rickettsia was initially isolated from H.

asiaticum ticks in Inner Mongolia [2] and then isolated from

Hyalomma truncatum in Niger [5]. In South Africa, H. trun-

catum ticks, abundant in the region and known to feed on

humans, were suspected to be the vectors of R. sibirica mon-

golotimonae [35]. Hyalomma ticks are widely distributed in Asia

and Africa and are also prevalent in southern Europe, including

France [36]. Although the camel ticks that patient 7 had been

in contact with were not identified, they could have been Hy-

alomma dromedarii, which are common camel ticks and have

previously been shown to harbor rickettsiae [37].

R. sibirica mongolotimonae causes a mild disease that we

propose to name lymphangitis-associated rickettsiosis. It may

be observed in Europe, Africa, and Asia. An active search for

the vector of R. sibirica mongolotimonae in France will be con-

ducted to complete our knowledge of the epidemiology of this

rickettsiosis.

Acknowledgments

We thank Annick Abeille and Betty Joseph, for their technical assistance;Philippe Parola, for providing clinical information; Hubert Lepidi, for per-forming histopathology; Oleg Mediannikov, for his expert advice; and Stan-ley J. Fenwick, for correcting our English.

Financial support. This work was supported by the “Programme derecherche fondamentale en microbiologie et maladies infectieuses et par-asitaires 2000” of the Ministère de l’éducation Nationale, de la Rechercheet de la Technologie, named “réseau pour l’identification des tiques et desagents pathogènes qu’elles transmettent à l’homme.”

Potential conflicts of interest. All authors: no conflicts.

References

1. Philip RN, Casper EA, Burgdorfer W, Gerloff RK, Hugues LE, Bell EJ.Serologic typing of rickettsiae of the spotted fever group by micro-immunofluorescence. J Immunol 1978; 121:1961–8.

2. Yu X, Jin Y, Fan M, Xu G, Liu Q, Raoult D. Genotypic and antigenicidentification of two new strains of spotted fever group rickettsiaeisolated from China. J Clin Microbiol 1993; 31:83–8.

3. Fournier PE, Tissot-Dupont H, Gallais H, Raoult D. Rickettsia mon-golotimonae: a rare pathogen in France. Emerg Infect Dis 2000; 6:290–2.

4. Fournier PE, Dumler JS, Greub G, Zhang J, Yimin W, Raoult D. Genesequence-based criteria for the identification of new Rickettsia isolatesand description of Rickettsia heilongjiangensis sp. nov. J Clin Microbiol2003; 41:5456–65.

5. Parola P, Inokuma H, Camicas JL, Brouqui P, Raoult D. Detection andidentification of spotted fever group Rickettsiae and Ehrlichiae in Af-rican ticks. Emerg Infect Dis 2001; 7:1014–7.

6. Raoult D, Brouqui P, Roux V. A new spotted-fever-group rickettsiosis.Lancet 1996; 348:412.

7. Lyskovtsev MM. Tickborne rickettsiosis. Miscellaneous Publications ofthe Entomological Society of America 1968; 6:42–140.

1444 • CID 2005:40 (15 May) • Fournier et al.

8. Rehacek J, Tarasevich IV. Acari-borne rickettsiae and rickettsioses inEurasia. Bratislava, Slovakia: Veda, Publishing House of the SlovakAcademy of Sciences, 1988:128–45.

9. Zdrodovskii PF, Golinevich HM. North Asian tick-borne rickettsiosisor tick-borne typhus fever. In: Zdrodovskii PF, Golinevich HM, eds.New York: Pergamon Press, 1960:311–32.

10. Pretorius AM, Birtles RJ. Rickettsia mongolotimonae: first human in-fection reported from South Africa. Emerg Infect Dis 2004; 10:125–6.

11. Jensenius M, Fournier PE, Vene S, et al. African tick-bite fever intravelers to rural sub-equatorial Africa. Clin Infect Dis 2003; 36:1411–7.

12. Raoult D, Fournier PE, Fenollar F, et al. Rickettsia africae, a tick-bornepathogen in travelers to sub-Saharan Africa. N Engl J Med 2001; 344:1504–10.

13. Raoult D, Lakos A, Fenollar F, Beytout J, Brouqui P, Fournier PE.Spotless rickettsiosis caused by Rickettsia slovaca and associated withDermatocentor ticks. Clin Infect Dis 2002; 34:1331–6.

14. Raoult D, Tissot-Dupont H, Caraco P, Brouqui P, Drancourt M, CharrelC. Mediterranean spotted fever in Marseille: descriptive epidemiologyand the influence of climatic factors. Eur J Epidemiol 1992; 8:192–7.

15. Fournier PE, Raoult D. Suicide PCR on skin biopsy specimens fordiagnosis of rickettsioses. J Clin Microbiol 2004; 42:3428–34.

16. La Scola B, Raoult D. Laboratory diagnosis of rickettsioses: currentapproaches to the diagnosis of old and new rickettsial diseases. J ClinMicrobiol 1997; 35:2715–27.

17. Brumpt E. Longévité du virus de la fièvre boutonneuse (Rickettsiaconorii n.sp.) chez la tique Rhipicephalus sanguineus. C R Séances SocBiol Fil 1932; 110:1199–209.

18. Rehacek J. Rickettsia slovaca, the organism and its ecology. Acta SCNat Brno 1984; 18:1–50.

19. Burgdorfer W, Aeschlimann A, Peter O, Hayes SF, Philip RN. Ixodesricinus: vector of a hitherto undescribed spotted fever group agent inSwitzerland. Acta Trop 1979; 36:357–67.

20. Beati L, Raoult L. Rickettsia massiliae sp.nov., a new spotted fever grouprickettsia. Int J Syst Bacteriol 1993; 43:839–40.

21. Raoult D, La Scola B, Enea M, et al. A flea-associated Rickettsia path-ogenic for humans. Emerg Infect Dis 2001; 7:73–81.

22. Bell EJ, Stoenner HG. Immunologic relationships among the spottedfever group of rickettsias determined by toxin neutralisation tests inmice with convalescent animal serums. J Immunol 1960; 84:171–82.

23. Teysseire N, Raoult D. Comparison of Western immunoblotting andmicroimmunofluoresence for diagnosis of Mediterranean spotted fever.J Clin Microbiol 1992; 30:455–60.

24. Marrero M, Raoult D. Centrifugation-shell vial technique for rapiddetection of Mediterranean spotted fever rickettsia in blood culture.Am J Trop Med Hyg 1989; 40:197–9.

25. Roux V, Fournier PE, Raoult D. Differentiation of spotted fever grouprickettsiae by sequencing and analysis of restriction fragment–lengthpolymorphism of PCR amplified DNA of the gene encoding the proteinrOmpA. J Clin Microbiol 1996; 34:2058–65.

26. Roux V, Rydkina E, Eremeeva M, Raoult D. Citrate synthase genecomparison, a new tool for phylogenetic analysis, and its applicationfor the rickettsiae. Int J Syst Bacteriol 1997; 47:252–61.

27. Bell EJ, Kohls GM, Stoenner HG, Lackman DB. Nonpathogenic rick-

ettsias related to the spotted fever group isolated from ticks, Derma-centor variabilis and Dermacentor Andersoni from Eastern Montana. JImmunol 1963; 90:770–81.

28. Fournier PE, Roux V, Raoult D. Phylogenetic analysis of spotted fevergroup rickettsiae by study of the outer surface protein rOmpA. Int JSyst Bacteriol 1998; 48:839–49.

29. Lepidi H, Fournier PE, Raoult D. Quantitative analysis of valvularlesions during Bartonella endocarditis: a case control study. Am J ClinPathol 2000; 114:880–9.

30. Fournier PE, Allombert C, Supputamongkol Y, Caruso G, Brouqui P,Raoult D. Aneruptive fever associated with antibodies to Rickettsiahelvetica in Europe and Thailand. J Clin Microbiol 2004; 42:816–8.

31. Sonenshine DE. Ecology of non-nidicolous ticks. In: Sonenshine DE,ed. Vol 2. Oxford, New York.: Oxford University Press, 1993:3–65.

32. Raoult D, Weiller PJ, Chagnon A, Chaudet H, Gallais H, Casanova P.Mediterranean spotted fever: clinical, laboratory and epidemiologicalfeatures of 199 cases. Am J Trop Med Hyg 1986; 35:845–50.

33. Tissot-Dupont H, Raoult D. Epidémiologie de la fièvre boutonneuseméditerraneenne en France. Med Mal Infect 1993; 23:485–90.

34. Kostman JR, DiNubile MJ. Nodular lymphangitis: a distinctive butoften unrecognized syndrome. Ann Intern Med 1993; 118:883–8.

35. Horak IG, Fourie LJ, Heyne H, Walker JB, Needham GR. Ixodid ticksfeeding on humans in South Africa: with notes on preferred hosts,geographic distribution, seasonal occurrence and transmission of path-ogens. Exp Appl Acarol 2002; 27:113–36.

36. Morel PC. Les Hyalomma (Acariens, Ixodidae) de France. Ann Par-asitol 1959; 34:552–5.

37. Lange JV, El Dessouky AG, Manor E, Merdan AI, Azad AF. Spottedfever rickettsiae in ticks from the Northern Sinai governate, Egypt. AmJ Trop Med Hyg 1992; 46:546–51.

38. Mahara F. Japanese spotted fever: report of 31 cases and review of theliterature. Emerg Infect Dis 1997; 3:105–11.

39. Eremeeva ME, Balayeva NM, Ignatovich VF, Raoult D. Proteinic andgenomic identification of spotted fever group rickettsiae isolated in theformer USSR. J Clin Microbiol 1993; 31:2625–33.

40. Chen M, Fan MY, Bi DZ, Zhang JZ, Huang YP. Detection of Rickettsiasibirica in ticks and small mammals collected in three different regionsof China. Acta Virol 1998; 42:61–4.

41. Mediannikov O, Sidelnikov Y, Ivanov L, et al. Acute tick-borne rick-ettsiosis caused by Rickettsia heilongjiangensis in the Russian Far East.Emerg Infect Dis 2004; 10:810–7.

42. Chen M, Fan MY, Bi DZ, Zhang JZ, Huang YP. Detection of Rickettsiasibirica in ticks and small mammals collected in three different regionsof China. Acta Virol 1998; 42:61–4.

43. Liu QH, Chen GY, Jin Y, et al. Evidence for a high prevalence of spottedfever group rickettsial infections in diverse ecologic zones of InnerMongolia. Epidemiol Infect 1995; 115:177–83.

44. Helmick CG, Bernard KW, D’Angelo LJ. Rocky mountain spotted fever:clinical, laboratory, and epidemiological features of 262 cases. J InfectDis 1984; 150:480–8.

45. Sexton DJ, Dwyer BW, Kemp R, Graves S. Spotted fever group rick-ettsial infections in Australia. Rev Infect Dis 1991; 13:876–86.