|
|
||||||||
| ABSTRACT |
|
|
|---|
| INTRODUCTION |
|
|
|---|
The ultimate confirmation of visceral leishmaniasis is by demonstration of Leishmania donovani amastigotes in stained smears from spleen, lymph glands, or bone marrow aspirates. Splenic aspirates are more sensitive (96%) than aspirates of bone marrow (70%) or lymph nodes (58%).1 During the last 15 years, Médecins Sans Frontières (MSF)Holland has treated more than 58,000 cases of VL in Sudan under harsh field conditions. The use of aspirates in these epidemic and highly endemic settings with large numbers of patients is limited because of the time and skills required in preparing and reading the slides. Due to the small but definite risk of internal bleeding, the use of splenic aspirates in this setting is further restricted. Bone marrow and lymph node aspirates are safer, but the material obtained is more dilute than in a splenic aspirate and thus less sensitive diagnostically. Moreover, bone marrow sampling is inconvenient and often painful for the patient. Many patients would be denied lifesaving treatment on the basis of false-negative aspirates if bone marrow or lymph node aspirations were the only diagnostic tool. These characteristics make diagnostic aspiration of any tissue unsuitable as a routine screening test for VL for large numbers of patients in a field setting.
Serologic tests have been developed for VL, and the direct agglutination test (DAT) is used in the Sudan as the first-line diagnostic tool.2 DAT titers
1:3,200 have a sensitivity of 94% and a specificity of 72% for active disease.1 Lower titers are less specific because of non-specific agglutination and high prevalence of antibodies to Leishmania because of subclinical and/or past infections with Leishmania species in this area. Longitudinal studies in a highly endemic village in eastern Sudan between 1991 and 1996 show a ratio of clinical to subclinical infection between 1.2:1 and 2.6:1, in which Leishmania antibodies are present without illness.3,4 When large numbers of patients are being assessed for kala-azar, e.g., during epidemics, high DAT titers (
1:6,400) plus a typical clinical picture may be used as a surrogate method for diagnosing kala-azar without aspiration, giving a high predictive value. Conversely, a low DAT titer (
1:400), effectively rules out kala-azar.5 Thus, the DAT can be used at a high titer as a diagnostic test, and at a low titer as a screening test. In the situation where the patient has a DAT titer >1:400 but <1:6,400, aspirations are done to confirm a diagnosis of leishmaniasis. The MSF uses this diagnostic protocol, which reduces the number of aspirates by approximately 80%.6
Although the DAT was specifically developed for epidemic field situations, it is still difficult to use in remote conditions. Performance of this test requires a proper laboratory with skilled technicians to perform correct titer setting, cross-checking with controls, meticulous implementation, and overnight incubation. In the field where a laboratory cannot be established, samples have to be transported to a central laboratory, typically resulting in a delay in starting treatment of 12 weeks. Improvements have been made by the introduction of freeze-dried antigen, which replaces the unstable aqueous antigen.7 However, there is an urgent need for a simpler, faster, non-invasive, but reliable test to screen for and diagnose kala-azar at the bedside.
Serologic tests have been developed using the cloned antigen rK39 instead of whole Leishmania parasites. The rK39 antigen, which consists of 39 amino acid repeats of a kinesin-like gene found in L. chagasi, is used in an enzyme-linked immunosorbent assay (ELISA) format and has shown satisfactory results in India, Brazil, and the Mediterranean.813 Several recombinant K39 antigen-based dipstick formats have been developed for rapid field diagnostics by different companies (InBios Inc., Seattle, WA; Arista Biologicals Inc., Allentown, PA; and Amrad ICT, Richmond, Victoria, Australia). However, the latter two companies no longer produce these tests. These dipstick formats showed high sensitivity and specificity in India, Nepal, Bangladesh, and Brazil.1420 In Sudan, previous versions of rK39 antigen-based tests showed less satisfactory results. An rK39-based ELISA had a sensitivity of 93% in Sudanese VL patients.21 However, different dipstick formats showed either good specificity but poor sensitivity,22 or good sensitivity but poor specificity,23 making these rapid tests unreliable for diagnosis of kala-azar under field conditions. The performance of different K39-based tests in different regions is summarized in Table 1
. We evaluated a commercially produced, rK39 antigen-based, rapid diagnostic test in highly endemic and epidemic regions in Sudan.
|
| MATERIALS AND METHODS |
|
|
|---|
Patients.
Patients fulfilling the clinical case definition for suspected kala-azar (fever for >2 weeks with malaria excluded, wasting, and either splenomegaly or lymphadenopathy) were further evaluated. In this study, we only considered patients with no history of previous treatment of kala-azar. All patients were tested by the DAT. A DAT titer
1:6,400 in patients satisfying the clinical case definition was regarded as diagnostic of VL. Those with borderline DAT titers (1: 8001:3,200) underwent further parasitologic examination by aspiration of spleen or lymph nodes. In addition, extra measures were taken for patients who appeared severely ill to make a diagnosis as quickly as possible: they underwent aspiration on the same day as the DAT, and if the initial aspirate was negative, a repeat aspiration was performed the following day with or without a repeat DAT after one week. Kala-azar was diagnosed after a positive aspirate and/or a positive DAT (
1:6,400). Patients were treated according to our standard protocol.6,25 Patients with a negative DAT titer (
1:400) suspected of having VL were evaluated for alternative diagnoses.
A parasitologically confirmed case of kala-azar was defined as a patient with positive parasitologic results on either a spleen or lymph node aspiration smear. A control was defined as a patient admitted to the MSF health centers for other reasons than kala-azar, and who had no clinical signs and symptoms of VL, no history of previous treatment of kala-azar, and a negative DAT test result. The studies were performed after review and approval from the regional health authorities. Informed consent was obtained from the study subjects.
RK39 rapid dipstick test. The rK39 rapid diagnostic test kit (DiaMed-IT Leish®; DiaMed AG, Cressier sur Morat, Switzerland) is composed of a nitrocellulose membrane coated with a line of recombinant antigen K39 across the strip. Antibodies to Leishmania present in a blood sample react with K39 antigen, and their presence is shown by mouse anti-human antibody conjugated to an indicator. A finger prick sample (20 µL) of blood is added to a well and mixed with a drop of buffer. A DiaMed-IT Leish test strip is placed vertically in the well, and the diluted blood rises up the nitrocellulose strip. After the blood is completely wicked up, the strip is transferred to the next well, which contains a few drops of wash buffer, and allowed to clear. The entire process takes approximately 20 minutes and results are read visually. A control line at the top of the strip verifies that the test strip is functional. If this is the only line that appears, then the test result is considered negative; two lines are positive for VL.
Parasitologic diagnosis. Parasitologic confirmation of VL was established by microscopic demonstration of Leishmania amastigotes in Giemsa-stained smears from either spleen or lymph node aspirates. A parasite density score was determined using a scale ranging from 0 (no parasites per 1,000 oil-immersion fields) to +6 (>100 parasites per field).26
Direct agglutination test. The DAT was performed as previously described using freeze-dried antigen (L. donovani promastigotes) and control sera with known titers from the Royal Tropical Institute (Amsterdam, The Netherlands).7 Samples with known DAT titers were included as standards. The titer of the sample is expressed as the highest dilution at which agglutination is still visible.
Statistical analysis. Data entry and statistical analysis was done using Epi-Info (Centers for Disease Control and Prevention, Atlanta, GA). Proportions were compared using the chi-square test with Yates correction, or Fishers exact test, where appropriate. Sensitivity, specificity, and positive predictive values were estimated with exact binomial 95% confidence intervals (CIs).
| RESULTS |
|
|
|---|
Among the 201 parasitologically confirmed cases, 197 had a positive DAT test result (titer
1:6,400). Thus, the sensitivity of the DAT at high titer was 98.0% (95% CI = 95.099.5%), which was significantly higher than the rK39 dipstick (P = 0.001). Using a titer > 1:400, the sensitivity of the DAT was 99.0% (199 of 201).
A total of 133 endemic controls were recruited. Only one control had a positive rK39 dipstick test result, but this control also had a negative lymph node aspirate and a negative DAT result. Thus, the specificity was 99.2% (95% CI = 95.9100%) (Table 2
).
|
Performance of rK39 dipstick in suspected VL cases.
A total of 356 consecutive, unselected clinically suspected kala-azar cases in Lankien and Latjor sites were prospectively evaluated. Fifteen suspected cases with borderline positive DAT results (titers = 1:8001:3,200) were excluded from the analysis because their diagnosis remained uncertain; no aspiration could be done under prevailing field conditions. Among the remaining 341 individuals the rK39 dipstick was compared with diagnosis by high-titer DAT, positive aspirate, or both. The dipstick had a sensitivity of 80.7% (95% CI = 75.285.6%), a specificity of 97.3% (95% CI = 92.499.4%), a positive predictive value of 98.4% (95% CI = 95.499.7%), and a negative predictive value of 71.4% (95% CI = 63.678.4%) (Table 3
).
|
| DISCUSSION |
|
|
|---|
The rK39 dipstick has high specificity (99% in endemic controls and 97% in clinically suspected patients), making overdiagnosis of kala-azar unlikely, despite the high background prevalence of antibodies against L. donovani in this region.3,4 The lower specificity found in aspirate-negative clinical suspects may be attributable to the low sensitivity of lymph node aspirates,1 which results in possible false-negative aspirate results.
Our standard diagnostic work-up has been based on microscopy and the DAT, both of which require laboratory services with well-trained laboratory technicians. In situations where field laboratories cannot be established (for reasons of security or capacity), transportation of filter papers containing blood samples for DAT testing from peripheral units to centralized laboratories often results in an unacceptable treatment delay. Simple, rapid, and cheap serologic dipstick tests are therefore urgently required. The DiaMed-IT Leish can be used with whole capillary blood, does not require a cold chain, and can be stored for least one year under the high field temperatures in the Sudan (2540°C). Community health workers in peripheral health units can easily perform the test. A positive dipstick test result in a clinically suspected case of kala-azar can be considered confirmation of the disease with a high level of certainty, and the patient can be treated immediately. Especially in southern Sudan, where access to kala-azar care is very poor, kala-azar patients often come for treatment after a prolonged symptomatic period in a critical condition.30 Decentralized diagnosis will allow active case-finding, reduce delays, and improve chances of survival. Because the sensitivity of the rK39 dipstick is 90%, 10% of kala-azar patients (dipstick false-negative results) may die without treatment if the disease is not otherwise confirmed. Therefore, a backup protocol (using the DAT and/or aspiration) remains essential for clinically suspect VL cases with a negative rK39 dipstick test results. Therefore, rapid parasite antigen tests or Leishmania DNA/RNA detection tests might be an important improvement in serodiagnosis. This is particularly important in view of increasing human immunodeficiency virus/VL co-infection rates in eastern Africa, which are expected to reduce the validity of serologic tests.11
Received December 17, 2004. Accepted for publication September 1, 2005.
Acknowledgments: We thank the MSF teams in the field in Sudan for their enthusiastic co-operation in this study, Dr. Patrick Jacquier (DiaMed AG) for providing the DiaMed-IT-Leish® tests, and Professor Piet Kager (University of Amsterdam) for the critical review of this manuscript. DiaMed AG provided 500 free VL diagnostic tests to MSF-Holland for this study, but had no involvement with the study design, data gathering, or analysis, and have provided no financial contribution to anyone associated with the study.
Financial support: This study was supported by Médecins sans Frontières-Holland.
* Address correspondence to Koert Ritmeijer, Médecins sans Frontiéres-Holland, P.O. Box 10014, 1001 EA, Amsterdam, The Netherlands. E-mail: koert.ritmeijer{at}amsterdam.msf.org ![]()
Authors addresses: Koert Ritmeijer, Médecins sans Frontières-Holland, P.O. Box 10014, 1001 EA, Amsterdam, The Netherlands, Telephone: 31-20-5208700, Fax: 31-20-6205170, E-mail: koert.ritmeijer{at}amsterdam.msf.org. Yoseph Melaku and Sammy Kipngetich, Médecins sans Frontières-Holland, Fourth Floor, Chaka Place off Argwings Khodhek Road, Hurlingham, P.O. Box 40643, Nairobi, Kenya. Marius Mueller and Caroline OKeeffe, Médecins sans Frontières-Holland, Steet 33, Amarat, P.O. Box 3319, Khartoum, Sudan. Robert N. Davidson, Infection and Tropical Medicine, Northwick Park Hospital, Harrow HA1 3UJ, United Kingdom.
| REFERENCES |
|
|
|---|
This article has been cited by other articles:
![]() |
D. Mansour, E. M. Abass, M. el Mutasim, A. Mahamoud, and A. el Harith Use of a Newly Developed -Mercaptoethanol Enzyme-Linked Immunosorbent Assay To Diagnose Visceral Leishmaniasis in Patients in Eastern Sudan Clin. Vaccine Immunol., December 1, 2007; 14(12): 1592 - 1595. [Abstract] [Full Text] [PDF] |
||||
![]() |
F. Chappuis, S. Rijal, A. Soto, J. Menten, and M. Boelaert A meta-analysis of the diagnostic performance of the direct agglutination test and rK39 dipstick for visceral leishmaniasis BMJ, October 7, 2006; 333(7571): 723. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. Mueller, M. Balasegaram, Y. Koummuki, K. Ritmeijer, M. R. Santana, and R. Davidson A comparison of liposomal amphotericin B with sodium stibogluconate for the treatment of visceral leishmaniasis in pregnancy in Sudan J. Antimicrob. Chemother., October 1, 2006; 58(4): 811 - 815. [Abstract] [Full Text] [PDF] |
||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |