ARTlCLES

Development and testing of a rapid diagnostic test for bubonic and

pneumonic plague

Suzanne Chanteau, U/a Raha/ison, La/ao Ra/afiarisoa, Jeanine Fou/on, Mahery Ratsitorahina, La/a Ratsifasoamanana,

E/isabeth Carnie/, Farida Nato

Summary IntroductionRapid diagnostic tests (RDTs) for infectious diseases are

Background Plague is often fatal without prompt and of much value in facilitation of major improvements inappropriate treatment. It affects mainly poor and remate disease management, especially in developing countries.populations. Late diagnosis is one of the major causes of Some RDTs are based on immunochromatography,human death and spread of the disease, since it limits the with conjugated gold or latex particles used to detecteffectiveness of contrai measures. We aimed to develop and specific antigens from bacterial, viral, or parasiticassess a rapid diagnostic test (RDT) for plague. agents.'-3

Plague, a flea-borne rodent disease that is occasionallyMethods We developed a test that used monoclonal transmitted to man,' is still prevalent in more thanantibodies to the F1 antigen of Yersinia pestis. Sensitivity and 20 countries, mainly in Africa.' Madagascar is one ofspecificity were assessed with a range of bacterial cultures and the most active plague foci in the world.6-8 Promptclinical samples, and compared with findings from available diagnosis of this fulminant disease is of key importanceELlSA and bacteriological tests for plague. Samples from for control of the mortality rate and for effective controlopatients thought to have plague were tested with the RDT in the The diagnosis of bubonic or pneumonic plague is stilllaboratory and by health workers in 26 pilot sites in based on clinical symptoms in most endemic areas.Madagascar. However, such diagnosis is not reliable. Biological

samples are usually assessed by microscopy in regionalFindings The RDT detected concentrations of F1 antigen as low hospitaIs, but in remote areas such tests are oftenas 0.5 ngjmL in up to 15 min, and had a shelf life of 21 days not done at alI or are unreliable. Resource-poor regionsat 60°C. Its sensitivity and specificity were both 100%. RDT do not have the technology and staff needed fordetected 41.6% and 31% more positive clinical specimens bacteriological identification methods, ELISA, and PCRthan did bactel;i°logical methods and EUSA, respectively. The for diagnosis of plague during the acute phase of theagreement rate between tests dane at remate centres and in disease.9,lOthe laboratory was 89.8%. With the combination of In such endemic foci, the development of simple,bacteriological methods and F1 ELlSA as reference standard, quick, and specific methods by which local healththe positive and negative predictive values of the RDT were workers could test biological samples from patients with90.6% and 86.7%, respectively. symptoms of plague, or dead rodents, is a key priority.

Indeed, such a test would eliminare doubt in suspectedInterpretation Our RDT is a specific, sensitive, and reliable test cases of plague, facilitating the immediate treatmentthat can easily be dane by health workers at the patient's of patients and contacts, and implementation ofbedside, for the rapid diagnosis of pneumonic and bubonic compulsory flea control measures. Such action is crucialplague. This test will be of key importance for the contrai of ifplague morbidity and lethality are to be reduced.plague in endemic countries. Very early diagnosis of plague would algo be essential

in cases of bioterrorism. Yersinia pestis is one of theLancet 2003; 361: 211-16 bacteriological agents that could potentially be used for

i See Commentary page 191 biological warfare.11 If transmitted in the forro ofI dro?lets that are t?en inhaled ~r spread by ae~osols, the

bacIllus causes hlghly contaglous pneumornc plague,which is always fatal if treatment is not started within24 h óf onset of symptoms.'2

., .FI antigen is specific to Y pestis and is present in., :.. large amounts in blood and bubo samples from

infected people.'~" It is stable in tropical climates, andcan be detected even after the patient has been treatedfor several days.15,16 An RDT for plague based on FI

Institut Pasteur de Madagascar, WHO Collaborating Centre for antigen has been tested in the laboratory, and gavePlague, Antananarivo, Madagascar (S Chanteau PhO, promising results.16 However, since this test is notL Rahalison PhO, M Ratsitorahina MO); Plague National Contrai readily available, another rapid test is needed for use in

I Programme, Ministry of Health, Antananarivo, Madagascar the plague foci of African, Asian, and North and Southi (L Ralafiarisoa, L Ratsifasoamanana MO); and Institut Pasteur, American countries, and in non-endemic countries, in! WHO Collaborating Centre for Yers/nia, Paris, France cases of suspected imported plague or of biological\ (J Foulon, E Carniel MO, F Nato PhD) att~ck. T~ere!ore, we aime~ to d~velop a sensitive FI

i Correspondence to: Dr Suzanne Chanteau, CERMES, antlgen dIPStlck, and to validare It both at the plagueI PO Box 10887, Niamey, Niger reference laboratory and at 26 remote gires in

(e-mail: schanteau@cermes.ne) Madagascar.

,

THE LANCET. Vol361 .January 18, 2003. www.thelancet.com 211

ARllCLES

Methods being plague when the culture was positive, presumptiveOevelapment af plague ROT when culture was negative but rnicroscopy was positive, andTo develop monoclonal antibodies (Mab) against Fi negative when both tests were negative. Confinned andantigen, we immunised rnice (Biozzi BP/2 strain) with Fi presumptive cases were notified to WHO. The secondantigen that was purified as Chen and Meyer describedl7 method was the immunocapture EUSA for detection offrom Y pestis strain 185S/97 (Madagascar). The fusion and FI antigen with F1--O4-A-G1 anti-F1 Mab.16 The lowestscreening methods used have been described.18 concentration of antigen detectable with this test was67 hybridomas with strong IgG responses against F1were 2 ng/rnL.obtained, 16 of which were cloned and characterised. For To assess the sensitivity of the plague RDT we useddevelopment of the plague RDT, two monoclonal 85 strains of Y pestis that had been cultured in brain-heartantibodies (IgG1 and K chain isotype) that bound to two broth at 37°C for 72 h. 55 of these were isolated indifIerent epitopes of the FI antigen19 were selected on the Madagascar from 1996 to 2000 (33 from patients withbasisof their high affinity, and were purified by ammonium bubonic plague, four from those with pneumonic plague,sulphate precipitation and fast protein liquid seven from rats [Rattus norvegicus], three from tleaschromatography (Pharmacia, Bois d'Arcy, France) [Xenopsylla cheopis], seven from shrews [Suncus murinus],according to the column manufacturer's instructions. Mab and one from a hedgehog [Setifer setosus]. 30 strains carneB18-1 (kD <5X1Q-1I moI) was conjugated to gold particles from 14 other countries (collection of the WHOas the mobile phase, and Mab G6-18 (kD=lX10~10 moI) Collaborating Centre for Yersinia, Institut Pasteur, Paris,was deposited on nitrocellulose as the capture antibody. France): tive of biotype Antiqua (four from Kenya, oneThe hybridomas Bl8-1 and G6-18 were stored at the former USSR), tive ofbiotype Medievalis (Kurdistan [four],Collection National des Microorganismes (Paris, France). Turkey [anel), and 20 of biotype Orientalis (Germany

Our RDT was derived from a prototype developed by the [two], Morocco [anel, Turkey [anel, India [anel, NarnibiaNaval Medical Research Institute (Bethesda, Maryland, [two], Vietnam [seven], Brazil [two], Argentina [anel,USA) and assessed in Madagascar.16 For this new test, we Senegal [anel, South Africa [anel, Congo [anel).used a combination of B 18-1 and G6-18 produced at the Additionally, we used 198 clinical specimens from plagueInstitut Pasteur, whereas the original RDT combined patients that had tested positive by both bacteriology andF1--O4-A-G1 Mab with a polyclonal rabbit antiserum, and FI EUSA: 17 sputum samples, 35 serum samples, 43 urinewas based on one-step, vertical-tlow immunochro- samples, and 103 bubo aspirates.matography.'o The colloidal gold particles (40 nrn diameter) To assess the specificity of the test we used 189 cultureswere conjugated to anti-F1 Mab B18-1 (British Biocell of yersinia and other enterobacteria grown in brain-heartIntemational CardifI, UK) and lyophilised (A54Qnrn=3) onto broth, at 37°C for 72 h. These samples were: 71 cultures ofpolyester release pads (Filtreri, Ambérieu en Bugey, enteropathogenic yersinia, consisting of Y pseudotuberculosisFrance). An automatic thin layer chromatography sampler (44 strains, serotypes I-V from 17 difIerent animal species(CAMAG, Muttenz, Switzerland) was used to spray the and environrnental samples isolated in 22 difIerentanti-F1 capture Mab G6-18 in a tine, at a concentration of countries), and Yenterocolitica (27 strains isolated in seven2 ~cm, onto nitrocellulose membrane (Irnmunopore FP, countries and belonging to six biotypes [lA-5] and toWh}tman Intemational, Chateau Giron, France); together various serotypes); 52 non-pathogenic yersinia of variouswith the contraI capture tine, which was affinity-purified serotypes belonging to eight species and isolated in variousgoat anti-mouse IgG (ICN Biomedicals, Aurora, Orno, countries (Y kristenseni [nine), Y intermedia [ten],USA), sprayed as a liDe higher up the strip at a Y mollaretii [ten], Y frederiksenii [ten], Y bercovieri [ten],concentration of 0,8 Ilg/cm. Cellulose filter paper was used Y rhodei [anel, Y ruckeri [anel, and Yaldovae [anel; andfor the wicking pad and sample pads (Schull and Schleicher, 66 enterobacteria belonging to 19 genera and 55 speciesEcquevilly, France). (Escherichia spp [eight], Enterobacter spp [ten], Erwinia sp

The immunostrips were trimmed to a width of 5 mm and [anel, Hafnia sp [anel, Serraria spp [ten], Klebsiella sppstored in a waterproof bag at 4°C, or in 5 rnL disposable [ten], Salmonella spp [four], Shigella spp [four], Levinea spplastic tubes at roam temperature at non-central sites. The [anel, Citrobacter spp [two], Edwa~iella spp [two], Proteustest was dane in the plastic tube with a sample volume of spp [two], Providencia spp [two], Morganella sp [anel,about 200 j.l.L. After 10-15 min, a positive result appeared Kluyvera spp [two], Budvicia sp [anel, Cedecea spp [two],as two pink tines (upper controltine and lower F1-positive Leminorella spp [two], and Obesumbacterium sp [one]).liDe), and a negative result as a single upper pink contraI Additionally we used 137 clinical specimens that weretine. Semiquantitative analysis was possible, with the negative for plague: 47 sputum samples from suspectedintensity afilie test tine scored from 0'5 to 4 (indicating the tuberculosis patients (14 positive for Mycobacteriumconcentration ofF1 antigen in the test sample). The cutof! tuberculosis), 66 normal serum samples diluted 1/10, andand the range of FI concentration that could be detected 24 undiluted normal urine samples. For ethical reasons,was measured with tenfold dilutions (50 j.l.g/rnL to bubo aspirates were not taken from patients who were not0,05 ng/rnL) of purified FI antigen. The reliablility of the thoUght to have plague.RDT was assessed with two concentrations of purified FI The negative and positive predictive values (NPV andantigen (0'5 ng/rnL and 1 j.l.g/rnL). PPV, respectively) of the FI dipstick were calculated by

The shelf life of the strips in the laboratory was assessed Bayes' theorem, and 95% CI by Fleiss' quadraticby testing twice per week for 3 weeks after storage at 60°C, approximation method (Sp=sensitivity, Se=sensitivity):4°C, -20°C, and :-80°C.'0 The re~iability of the RDT was Sp (l-prevalence)assessed by repeanng the test ten nInes on the same sample NPV=and in ten difIerent series oftests. Sp (l-prevalence) + (l-Se) prevalence

Reference tests and samples Se (prevalence)We used two reference standard tests for the diagnosis of PPV = Se (prevalence)+ (l-Sp) (1- prevalence)

plague. The first was gram staining and isolation of Y pestis,either direcrly from the patients' samples or after mouse The agreement between RDT and reference tests resultsinoculation. A specimen was judged to be confinned as (bacteriology or FI EUSA, and bacteriology combined

212 THE LANCET. Vol361 .January 18, 2003' www.thelancet.com

---

ARTICLES

Assessment of ROrWe undertook a prospective study to assess rlle effectivenessof rlle FI dipstick during rlle 200G-Ql plague season inMadagascar, as part of rlle national plague controlprogramme. We obtained a biological sample (buboaspirate, sputum, or post-mortem organ puncture, as

O Plague endemic areas appropriate) from patients wirll suspected plague, and rllePilot sites patients were treated wirll streptomycin. Samples were sent

on a swab in Cary Blair agar, at room temperature, to rlleCentral Laboratory at rlle Institut Pasteurde Madagascar.On arrival, rlle specimen was washed out of rlle swab byincubation in 1 mL phosphate-buffered saline, and wastested wirll rlle two reference merllods and wirll RDT. Themean transport times of rlle positive and negative samples torlle Central Laboratory were compared wirll rlle Kruska11-Wallis non-parametric testo

To test rlle RDT in remote areas, a network of 26 pilotsites was set up from Dec 1, 2000, to Jan 25, 2001,consisting of six district hospitaIs and 20 healrll care centresin remote areas (figure). After rlle efficiency of rlle plagueRDT had been assessed at rlle Central Laboratory, kits-each composed of one syringe, one needIe, PBS, one plagueRDT in a disposable test tube wirll silica gel, one Cary Blairtube, one swab, a user's manual, and a questionnaire-weredistributed to rllese field sites. The robustness of rlle testunder field conditions was assessed, based on severalcriteria: reliability of rlle results obtained by non-biologiststaff, presentation of rlle kit, ease of manipulation andinterpretation, shelf life under real field conditions, lengili

I I and content ofrlle training programme, and intelligibility of200 km rlle illustrated manual.

The RDTs were stored at room temperature during rllestudy (2o-30°C). 29 medical doctors, 19 nurses, and nine

I healrll workers participated in rlle assessment. They weretrained onsite for 3 h as to how to obtain clinical samplesand to use, read, and archive rlle dipsticks. An illustrated

Location of 26 pilot sites in plague endemic districts of instruction guidebook, in French and Malagasy, was ~venMadagasc to each centre. The plague RDT was done at rlle patlent'sar bedside. The trained personnel were rllen asked to send a

wirll FI ELISA) was assessed wirll rlle percentage of sample of rlle specimen to rlle Central Laboratory for aconcordant results and rlle Kappa coefficient wirll dipstick test wirll rlle same batch of strips, bacteriologicalconfidence intervals. The correlation between FI ELISA identification, and FI ELISA assay. The technicians whoand RDT results was assessed by Kendall's correlation did rlle tests at rlle Central Laboratory were not aware of rllecoefficient. Finally, we used rlle RDT to test 283 bubo results obtained in rlle pilot centres.aspirates, which were stored at rlle Institut Pasteur, rllat had From May to June, 2001 (end of rlle plague season), rllebeen isolated from patients who were suspected on clínical 26 sites were visited again, RDT results were compared,grounds to have plague but tested negative wirll rlle two and rlle registers were checked. The healrll personnel werereference standard assays. asked to fil1 in an individual user's questionnaire, to provide

We did PCR with any cultures of Y pestis rllat tested criticisms, comments, and observations about rlle plaguenegative on RDT, to find out wherller cafl gene, which RDT and rlle instruction book.encodes for FI antigen, was present. The cafl gene was

i amplified by PCR, wirll heat-killed bacteria as templates Role of the funding sourceand two oligonucleotide primers: forward 5'-CAGTTC The sponsors ofrlle study had no role in study design, dataCGTTATCGCCATTGC-3', and reverse 5/-TATTGG collection, data analysis, data interpretation, or writing ofTTAGATACGGTTACGGT-3'. PCR amplification was rlle reportodone in a final volume of 50 ILL, in a PTC 1 00 rllermalcycler (MJ Research, Basel, Switzerland). We mixed 7 ILL Resultsof boiled bacterial suspension (target DNA), 200 ltInol of The lower detection threshold of rlle newly developedeach deoxyribonucleotide triphosphate, 2 mmol MgCI. plague RDT was 0'5 ng/mL, and rlle range of FI0.1 ILmol of each primer, and 1 U of rllermostable DNÀ concentrations extended from 0'5 ng/mL to 50 lLg/mLpolymerase (Roche, Meylan, France) in the corresponding (5 log) , wirllout any hook effect (in which no signal isIX polymerase buffer. The gene was amplified over detected for high concentrations). Results obtained were25 cycles, each consisting of denaturation for 1 min at much rlle same after storage of RDT for 21 days at 60°C,95°C, annealing for 1 min at 55°C, and polymerisation for 4°C, -20°C, and -80°C.2 min at 72°C; followed by one cycle at 55°C for 1 min and The RDT had a sensitivity of 100% on fresh1y isolated Y72°C for 3 mino Plasmid DNA was extracted21 and digested pestis strains and positive control clínical samples: alI 55wirll EcoRV. PCR products and digested plasmid DNA bacterial cultures and all198 positive specimens tested werewere electrophoresed in agarose gels and stained wirll positive wirll rlle RDT. Positive RDT results were obtainedethidium bromide. for 28 of 30 Y pestis strains from various continents. PCR

THE LANCET. Vol361 .January 18, 2003' www.thelancet.com 213

ARTlCLES

Plague RDT Bacteriology Fi ELlSA Comblnatlon of reference testsConfirmed Presumptive Negative Positive Negative Positive Negative

Number (%) samplesPositive 279 (40%) 151 (83%) 12 (80%) 116 (24%) 212 (100%) 67 (14%) 232 (87%) 47 (11%)Negative 412 (60%) 31 (17%) 3 (20%) 378 (77%) 0(0%) 412 (86%) 34 (13%) 378 (89%)Total 691 182 15 494 212 479 266 425

Table 1: Comparison between plague RDT, bacteriology, and Fj. ELISA for specimens tested at Central Laboratory

assays done on the two RDT -negative strains with primers At the 26 remote sites, 128 patients (122 survivors andinternal to the cafl gene yielded no amplification product, six patients who died) were suspected, on clinicalindicating that this gene (or the entire pFra plasmid) was grounds, to have plague (123 bubonic and tivedeleted in these two strains. Comparison of the plasmid pneumonic fonns) and were treated. Results of RDTsrestriction profiles of these strains with that of wild-type done by health workers at lhe patient's bedside, and bystrain 6/69 from Madagascar confirmed the absence of the skilled technicians at the Central Laboratory, on therestriction fragments corresponding to pFra (data not garoe specimens (table 2) were concordant in 115 of 128shown). The specificity of the plague RDT was algO 100% patients (90%). Four test~ done at remote sites (3%)for alI bacterial cultures and alI negative control clinical were not valid since the controlline was absent (table 3).specimens. Of the 283 bubo aspirates from suspected plague With bacteriology and FI EUSA as reference standards,patients thar tested negative in bacteriology and EUSA, 256 58 samples were identified as positive among the 128(90.5%) algo tested negative by RDT, but 27 (9'5%) gave tested, whereas RDTs done at remote sites and at lheweak positive results with this test (score of 0'5). Central Laboratory identified 53 and 55 positive samples,

From Dec 1, 2000, to May 30, 2001, 671 cases of respectively. Agreement between results from RDTssuspected plague were declared to the national surveillance done at the field sites and bacteriology was moderatesystem: 598 (89%) survived, 61 (9%)died, 12 (2%) had an (K=0.62, 95% CI 0'57-0.66), but agreement was veryunknown outcome. 642 (96%) patients had bubonic plague, good when RDT in the field was compared with FI20 (3%) had pneumonic plague, and nine (1 %) had an EUSA (0.83, 0'78-0.87). The false negatives with RDTunknown clipical status. In total, 691 clinical specimens corresponded with samplesthat tested positive in culturewere sent to the Central Laboratory in Cary Blair agar: 643 after amplification in two mice (low bacterial load) but(93%) were bubo aspirates, 13 (12%) sputa, and 35 (5%) nega tive with FI EUSA. Five patients who werepost mortem lung or tiver puncture samples. The median identified at remote sites as plague-infected were negativetransport time to the Central Laboratory of the 691 samples for plague with the combination of reference tests.was 8 days, ranging from O to 66 days (25th percentile On the assumption that the combination of lhe two5 days, 75th percentile 14 days). The mean transport time reference tests is lhe gold standard (which is notwas 10.6 days (SD 9.6) for bacteriologically confirmed lhe case), lhe calculated PPV and NPV would bespecimens and 12.1 days (11'9) for negative specimens 90.6% (95% CI 78.6-96.5) and 86.7% (76.4-93.1),(non-significant difference, Kruskall-Wallis non-parametric respectively, at lhe pilot sites, and 92.7% (81.6-97.6)test). ~ and 90.4% (80.7-95.7), respectively, at lhe Central

After recovery of lhe samples from Cary Blair media, Laboratory. There was 85% concordance between lheY pestis was cultured from 182 (26%) samples (classified results of the combined reference methods and those foras confinned specimens), whereas 15 (2%) were negative the plague RDT in the field (K=0'76, 95% CIby culture but positive by microscopy (classified as 0'71-0.80), and 91 % concordance between lhepresumptive specimens; table 1). Thus, the total number combined and RDT methods at lhe Central Laboratoryof plague patients identified by bacteriology was 197 of (0.82, 0'77-0.86).691 suspected. With RDT, 279 (40%) specimens were The 57 individual user's questionnaires gave a meanpositive for plague, 116 (42%) of which tested negative overall satisfaction score of 17'6 of 20 [range 14--20}.both by microscopy and culture. The concordance Most (51 of 57,90%) ofthe health staffthought that lhebetween lhe results of bacteriological and plague RDT RDT would help to reduce plague morbidity andresults was 78,3% (K=0'52, 95% CI 0.50-0'54; moderate mortality in Madagascar.agreement). Therefore, use of plague RDT alone led to Sputum obtained from 16 patients with suspectedidentification of 41.6% more cases than when only pneumonic plague were tested with lhe plague RDT asbacteriology was used (279 vs 197 positive tests). well as with lhe two reference tests for comparisonHowever, 31 (17%) samples tested nega tive with the (table 4): 13 sputa were tested only at lhe Centralplague RDT were positive by bacteriology. Laboratory; lhe remaining three sputum samples were

When FI EUSA was used as the reference method, algo tested at lhe field sites. The eight patients for whomresults obtained with this technique and with RDT were the diagnosis of plague was confinned by bacteriologyhighly concordant (concordance 90'3%, results ofthe two algo tested positive by RDT (one tested negative bytests agreed for 624 of691 specimens tested) and strongly ELISA). However, two patients who tested negative bycorrelated (Kendall's correlation coefficient r=0.834, bacteriological methods gave a weak positive result withp<10-3). K was 0.79 (95% CI 0'77-0.81; good agreement). lhe plague RDT, one of whom algo tested positive byAlI EUSA-positive samples algo tested positive on RDT. EUSA.Moreover, RDT detected 67 weak1y positive samples(score of 0'5) among lhe 479 samples that tested negative ROT, remate sites Tot~1by EUSA (table 1). Therefore, use of RDT resulted in Positive Negative Invalid patlents31 % (67 of212) more positive results than were obtained RDT, Central Laboratorywith FI EUSA alone. When we used a combination of Positive (n=55) 49 (92.5%%) 5 (7%) 1 (25%) 55 (43%)bacteriological tests and FI EUSA as lhe reference Negative (n=73) 4 (7'5%) 66 (93%) 3 (75%) 73 (57%)standard (table 1), 266 of the 691 specimens tested Total patients (%) 53 71 4 128positive, whereas with RDT alone, 279 tested positive Table 2: Comparison between plague RDTs done at remote(K=0.75, 95% CI 0.73-0'77). sites and at Central Laboratory on the same specimens

-

214 THE LANCET' Vo1361' January 18, 2003' www.thelancet.com

~

AR11CLES

Total patients (%) RDT, remote sites RDT, Central Laboratory

Positive Negative Invalid Positive Negative

Combination bacteriology/ELlSAPositive 58 (100%) 48 (83%) 10 (17%) O (0%) 51 (88%) 7 (12%)Negative 70 (100%) 5 (7%) 61 (87%) 4 (6%) 4 (6%) 66 (94%)Total 128 (100%) 53 (41%) 71 (56%) 4 (11%) 55 (43%) 73 (57%)

Table 3: Comparison between plague RDT done at remote sites and at Central Laboratory and combination of bacteriology andFi ELiSA

Discussion the reference tests (and positive with the initial plagueWe have shown that our RDT is as specific as, and at dipstick) subsequently underwent FI antibody sero-least as sensitive as, the two available standard methods. conversion.16The excellent specificity of the RDT, its low detection Conversely, 116 samples that tested positive by plaguethreshold, and the higher number of positive specimens RDT gave negative results in bacteriological tests,detected among samples from patients with suspected probably for the reasons given above. These reasonsplague, suggest a greater sensitivity than bacteriology and explain the fair concordance between RDT andELISA. Thus, in the operational conditions ofthe plague bacteriology. Similarly, 67 samples that tested negativecontrol programme at the Central Laboratory, we with FI ELISA gave weak positive results with RDT,recommend a combination of bacteriological tests and probably because the detection threshold of RDT wasRDT for diagnosis and surveillance of plague in lower than that of EilSA. Finally, 27 of the 283 clinicalMadagascar. specimens classified as negative with the combined

Our new test detected FI antigen at a wider range of bacteriological and ELISA tests tested positive by RDT.concentrations than other tests, in 10-15 min, without Since the RDT was highly specific and had a lowprozone effect. Absence of prozone phenomenon is detection threshold, we suggest that these 27 specimensessential to avoid false-negative results, especially in post- were true plague-positive specimens, and that the plaguemortem samples or sputum samples that contain very RDT is thus more sensitive than the two referencehigh concentrations of FI antigen. The highest FI assays.concentration ever noted in clinical samples was Of the 30 Y pestis strains of the three biotypes and30 J.lg/mL (S Chanteau, unpublished). from different continents, most tested positive by RDT,

The existing reference tests, which can only be done in indicating that this test is potentially applicable to plaguethe laboratory-the isolation of Y pestis and the detection foci worldwide. The two strains that tested negativeof FI antigen by ELISA-are specific, but lack sensitivity proved to have total or part deletions of the cal] gene;in ali countries in which plague is endemic. The failure of this deficiency seems to be due to the loss of the 110 kbthese standard assays is due to the deterioration, plasmid, probably during the long storage time of thecontaminati~, or both, of specimens during their long strains in the laboratory.transport time to the laboratory, antibiotic treatment of The invalid tests in the conditions of the remote sitespatients before sampling, and the diffusion of FI antigen (no control liDe in 3% of the specimens) were due toin the Cary Blair transport medium. The difference moistening of the dipsticks, caused by the humidity ofbetween the numbers of suspected plague samples found the air. This drawback can be overcome by improvementpositive by RDT and culture could algo be accounted for of the RDT packaging, with waterproof bags instead of aby diffusion ofthe FI antigen in the Cary Blair transport disposable test tube. We expect the shelflife ofthe RDTmedium, a difficulty that is not encountered if these tests in such conditions to be about 2 years at roomare directly done on clinical samples by health staff. temperature, on the basis of stability after 21 days atFurthermore, other investigators have shown that some 60°C.21 The test kit should therefore tolerate storage atpatients who were originally diagnosed as negative with room temperature for long periods in plague endemic

zones. The other discordant results were due either toweak positive bands that were not seen in the field, or to

Plague RDT F:1 ELlSA Bacteriology (degree of positivity) (ngjmL) a mlsmterpretatlon ofthe control and pOSltlve lIDes. ThlS

difficulty could be solved by a second training sessionPatlent's number focusing on interpretation of the test results.230/01 Neg Neg Neg ...

I 307/01 Neg Neg Neg Dunng non-central evaluatlon of the RDT, thlS test369/01 Neg Neg Neg was applied to the diagnosis of pneumonic plague as well382/01 Neg Neg Neg as bubonic plague. Although sputum contains high967/00 Neg Neg Neg concentrations of FI antigen, we thought that its density1058/00* Neg Neg Neg and viscosity might inhibit antigen-antibody interactions23:1/01 Pos (0.5+) Pos (5) Neg .:1052/00 Pos (0.5+) Neg Neg and obstruct the flow of lmmune complexes through the969/00 Pos (0.5+) Pos (:13) Confirmed dipstick. This difficulty was overcome by diluting sputum:1027/00 Pos(1+) Pos(44) Confirmed samples with saline or PBS (1 in 2 to 1 in 10 dilution).999/00 Pos (2+) Neg Confirmed Since the sensitivity and specificity of the test with1157/00 Pos (3+) Pos (125) Confirmed samples from patients with confirmed and suspected:1:157/00* Pos (3+) Pos (125) Confirmed . 1 h.974/00 Pos (4+) Pos (>:125) Confirmed pneumornc pague was ~gh, the RDT seems ~o b~ a1:158/00 Pos (4+) Pos (>125) Confirmed useful method for rapld, easy, and non-mvaSlve1:158/00* Fios (4+) Pos (>125) Confirmed confirmation of this disease. The availability of a reliableTotal positive/tested 10/16 8/16 8/16 onsite diagnostic test should result in rapid treatment ofConfirmed;Y pestis isolated. Neg;negative. Pos;positive. 0.5+ to 10+;score patients and efficient administration of chemoprophylaxis(see text). 'Patient algo tested at remote sites. to the contact population; these measures should, in turn,Table 4: Diagnosis of pneumonic plague in sputum with plague help to prevent the spread of pneumonic plague inRDT, Fi ELiSA, and bacteriology endemic countries or in cases of bioterrorist use.

TOE LANCET. Vol361 .January 18, 2003. www.thelancet.com 215

ARTlCLES

Assessment of the RDT results and of the users' Referencesquestionnaires showed that the technical transfer of the 1 Cole RA, Lu HM, Shi YZ, WangJ, De-Hua T, Zhou AT. Clinicaltest to remote pilot sites was successful and allowed the evaluation of a rapid immunochromatographic assay based on me 38kDareliable diagnosis of bubonic and pneumonic plague. antigen of Mycobacterium tuberculosis on patients wim pulmonaryTh find' I d I fr h Ith ff tuberculosis in China. Tuber Lung Dis 1996, 77: 363--{í8.

ese mgs e to a genera request om ea sta 2 W iIGJ Lammi PT W . N Th ICTFiI ..' d "k.. I fi . fi li . b d .1 bl e, e J' elSS .e anaS1S test: a rapl .onnatwor mg m pague OCl or lS test to e ma e aval a e at antigen test for diagnosis ofbancroftian filariasis. Parasito! Today 1997their centres. Bacteriological and EUSA tests are between 13: 401-04. '

tive and 50 times more expensive than the RDT, and they 3 Makler MT, Palmer C], Ager AL. A review ofpractical techniques forcannot be used in remQte locations. me diagnosis of malaria. Annals Trop Med Parasito11998; 92: 419-33.

Overall, we have sh~wn that with our test, the rapid and 4 Per;ry~' r:ertherstonlD. Yersiniapestis, etiologic agent ofplague.ffi . d.. fb b . d . I ChnMlCrobiolRev 1997; 10: 35-66.

cost-e ecttve lagnosls o U ornc an pneumornc Pague 5 Th W Id H l-L O .. O b ak 'UTL' E","Á-' I Re..e or ea Ul rganlZanon. ut re news. w"'y "~ff'W Ccould easlly be achleVed by health workers in remote sites. 2000, 75: 337-44.Use of the test could help to reduce mortality (through 6 Chanteau S, Ratsifasoamanana L, Rasoamanana B, et ai. Plague, arapid and efficient treatment of patients), morbidity (by reemerging disease in Madagascar. Emerg lnfect Dis 1998, 4: 101-04.rapid implementation of preventive measures), and 7 Chanteau S, Rats~torahina M, Rahalison L, et ai. Current epidemiologyinsecticide resistance of fleas (through rational use of ofh~an plagu: m Madagascar. MlCrobes lnfect 2000; 2: 25-~1.

" .. d ) Th. ki .11b d .. b d I 8 Bolsler P, Rahallson L, Rasolomaharo M, et ai. Four succeSSlve armualexpe~slv~ msecttcl es .lS t Wl e lstn u~e to a I outbreaks ofbubonic plague in me coastal city of Mahajangathe distnct and health care centres of the endemlc plague (Madagascar): epidemiological features. Emerg lnfect Dis 2002, 8:areas ofMadagascar in 2002. The next step will be, with 311-16.the assistance of international organisations to make the 9 Williams JE, Arntzen L, Tyndal GL, lsaacson M. Application of enzyme

I RDT .1 bl th .' . th d .immunoassays for me confirmation of clinically suspect plague inp ague ~val a e to o er countnes Wl en em1C Namibia 1982. Bull World Health Organ 1986, 64: 745-52.plague worldWlde. 10 Rahalison L, Vololonirina E, Ratsitorahina M, Chanteau S. Diagnosis of

bubonic plague by PCR in Madagascar .J Clin Microbio12000; 38:Contrioutors 260-63.S Chanteau designed me study, developed anti-F1 Mabs and RDT, 11 Bellamy R], Friedlander AR Bioterrorism. QJM 2001, 94: 227-34.planned me user's guide book, trained healm staff in me field, analysed 12 Ratsitorahina M, Chanteau S, Rabalison L, Ratsifasoamanana L, Boisierdata, and wrote me manuscript. L Rahalison supervised assays at Central P. Epidemiological and diagnostic aspects afilie outbreak ofpneumonicLaboratory, trained healm staffin me field, and made me user's guide plague in Madagascar. Lancet 2000; 355: 111-13.book (French and Ma)agasy languages). L Ralafiarisoa produced RDT 13 Baker EE, Sommer H, Foster LE, Meyer E, Meyer KF. The isolationkits, trained healm staffin me field, and did RDT and EUSA tests at me and characterisation afilie soluble antigen of PasteureUa pestis. lnfectCentral Laboratory.] Foulon assessed RDT wim bacterial cultures and lmmun 1952,68: 131-45.did PCR. L Ratsifasoamanana coordinated me plague national contrai 14 Brubaker RR. Interconversion of purine mononucleotides in Pasteurellaprogramme and me network ofhealm centres. E Carniel designed me pestis. lnfect lmmun 1970, 1: 446-54.study, supervised o.f me assessment ~f RDT wim b~cterial cultures, and 15 Chanteau S, Rabarijaona L, Hager ], Boisier P, Burans ], Rasolomaharowro,te me man~scnpt. F Nato coordmated me In~ntut Pasteur .RDT M. F1 antigenemia in bubonic plague patients, a marker of gravity andpro)ect, supervlsed me ~evelopment and producnon afilie ann-F1 Mabs, efficacy oftreatrnent. Trans R Soc TropMed Hyg 1998, 92: 572-73.and wrote me manuscnpt. 16 Ch S Rahli L Ra . ahin M ai E I di . fanteau, a son, tsltor a , et .ar y agnosls o

bubonic plague using F1 antigen capture EUSA assay and rapidConfli~ of interest statement immunogold dipstick. IntJ Med Microbio12000; 290: 279-83.None declared. 17 Chen TH, Meyer KF. An evaluation of Pasteurella pestis fraction

l-specific antibody for me confinnation of plague infection.Acknowledgments Bull World Health Organ 1966; 34: 911-18.We mank me US Naval Medical Research Institute (Bemesda -MD- 18 Ko?ler G, Milstein C. Con~~ous cultures offused cells secretingUSA) which shared wim us meir experience in me development ofRDTs annbody ofpredefined specificlty. Nature 1975,265: 495-97.and provided me immunocapture EUSA assay for evaluation at me 19 Phalipon A, Arondel], Nato F, Rouyre S, Mazie ]C, Sansonetti P].InstituI Pasteur de Madagascar. We mank]ean Luc Guesdon for me ldentification and characterization ofB-cell epitopes afilie IPA Cpurification by FPLC afilie monoclonal antibodies, me technicians afilie invasin of Shigeilaflexneri. lnfect lmmun 1992,60: 1919-26.Central Plague Laboratory, and me healm staffin me pilot sites for meir 20 Paek SH, Lee SH, Cho]H, Kim YS. Development ofrapid one-stepcontribution to me validation afilie test under field conditions. This immuno-chromatographic assay. Methods 2000; 22: 53--{í0.investigation was funded by me InstituI Pasteur, Paris (Grant PTR 21 Bimboim HC, Doly]. A rapid alka1ine extraction procedure for2000-11), InstituI Pasteur de Madagascar, and me MinisIry ofHealm of screening recombinant plasmid DNA. NucleicAcids Res 1979, 7:Madagascar (World Bank IDA 3302 MAG). 1513-23.

.j216 THE LANCET. Vol361 .]anuary 18, 2003. www.melancet.com