The Journal of Health, Population and Nutrition icddr,b ISSN: 1606-0997 EISSN: 2072-1315 Vol. 28, Num. 3, 2010, pp. 281-285

Journal of Health Population and Nutrition, Vol. 28, No. 3, May-June, 2010, pp. 281-285

Article

In search of an ideal test for diagnosis and prognosis of kala-azar

Dharmendra P. Singh¹ , Rahul K. Goyal ¹ , Rahul K. Singh¹ , S. Sundar² , and Tribhuban M. Mohapatra¹

¹ Department of Microbiology and
² Department of Medicine, Institute of Medical Sciences, Banaras Hindu University, Varanasi 221005, India

Correspondence Address: Prof. T.M. Mohapatra, Department of Microbiology Institute of Medical Sciences Banaras Hindu University Varanasi - 221005, India

tmmohapatra2000@yahoo.com

Code Number: hn10036

Abstract

Keywords: Diagnosis, Laboratory; Direct agglutination test; KAtex; rK39 strip test; Visceral leishmaniasis; India

Introduction

Leishmaniasis is endemic in 88 countries spread over five continents, with an estimated yearly inci-dence of 1-1.5 million cases of cutaneous leishman-iasis and 500,000 cases of visceral leishmaniasis (VL). The number of people at risk is estimated to be 350 million, with an overall prevalence among 12 million people ^[1] . VL is a potentially-fatal dis-ease, caused by Leishmania donovani complex. Of all the VL cases, more than 90% are from India, Nepal, Bangladesh, southern Sudan, and northeast Brazil. For decades, the detection of the parasite has been considered to be the definitive diagnosis of this disease. It requires invasive procedure to find the parasite in the tissue, i.e. spleen, bone-marrow, and lymph-node. Although it is the gold standard for the diagnosis of VL, the procedure is cumber-some, time-consuming, technically demanding, risky, and very difficult to apply in field condi-tions or remote places/primary health centres. To obviate these procedures, various serological tests have been developed, evaluated, and tried. These tests have the advantage of being safe, are less in-vasive, and can be carried out in large numbers of samples. Conventional methods of antibody de-tection are immunodiffusion, complement fixa-tion test, indirect haemagglutination test, and countercurrent immunoelectrophoresis ^[2],[3],[4] . However, aside from practical difficulties at pe-ripheral laboratories, the sensitivities and speci-ficities of most of the above tests have been the limiting factors.

Recently, quest for a simple diagnostic test for VL has led to the development of several novel tests, such as direct agglutination test (DAT), rK39 dip-stick assay, and latex agglutination test (KAtex) of urine. In particular, the development of different formats of DAT and its subsequent improvement has been an impetus to the improvement of di-agnostic procedures of VL under field conditions ^[5],[6],[7] . DAT is a sensitive test and well-adapted to screening larger populations, especially when using its variant referred to as fast agglutination screening test ^[8] . However, data regarding its specificity in different populations and patient-groups are vari-able, and many users of the test find it complicated to perform in remotely-located health facilities.

The rK39 dipstick assay-which is based on a cloned recombinant antigen of 39 amino acid repeats of a kinesin-like gene ^[9],[10] -and the KAtex of urine latex agglutination test ^[11] have emerged as attrac-tive options owing either to their simplicity (in the case of rK39 dipstick) or their high specificity (in the case of KAtex). A comparative evaluation of DAT, KAtex, and rK39 dipstick test was conducted aiming at obtaining more reliable estimates of their sensi-tivity, specificity, and reproducibility in the field for the diagnosis and prognosis of VL in India.

Materials and Methods

Patients were recruited at the Kala-azar Medical Research and Training Centre (KAMRC) in Muzaf-farpur and Rajendra Memorial Research Institute (RMRI), Patna, Bihar, India. All clinically-suspected VL cases were subjected to parasitological examina-tion. In total, 455 subjects-comprising 150 con-firmed VL cases, 160 endemic controls, and 100 non-endemic controls, and 45 other febrile diseas-es-were included in the study. Sera and urine sam-ples were collected. The cases were bled on the first encounter. The serum was separated and stored at -20 °C till further use. However, the KAtex was done in fresh urine samples. The tests were carried out in duplicate in each patient's sample with known positive and negative controls by a technical expert who had no knowledge about the patient's data.

Reference standard

The cases presented with clinical symptoms and signs suggestive of kala-azar, i.e. fever of two weeks or more, splenomegaly (8.4±2.4 cm), and anaemia. Routine investigations were carried out to exclude other aetiologies of fever with hepato-splenomeg-aly. The final diagnosis of VL was established by Karnofsky score, white blood cell (WBC) count, haemoglobin assay, and demonstration of parasites in splenic aspirates.

Non-endemic controls included healthy persons from Europe and North America where kala-azar is not endemic. Endemic controls included healthy persons free of signs and symptoms of any disease, residing in same endemic area and are closely re-lated to VL patients.

Latex agglutination test

The KAtex kit made in the USA was provided by Kalon Biological Limited (Sitting Bourne ME 98 AQ, UK)-lot no. K1-22 and Catalogue no. 5050-001. The test was conducted as per instructions of the manufacturers. Briefly, all reagents were brought to ambient temperature. The test latex was shaken well before use. Fifty mL of urine samples were heated by boiling for five minutes. Heat-treated urine samples were first allowed to cool to ambient temperature and then transferred to a glass-slide. Subsequently, the processed sample was mixed with a drop of test latex and stirred. The glass-slides were continuously rotated for two minutes before final reading of reactions. Positive reactions were determined by the degree of agglutination and re-corded as 1+, 2+, and 3+. Appropriate positive and negative controls, supplied by the manufacturer, were included as test control.

Direct agglutination test

L. donovani (MHOM/IN/80/Dd8) obtained from the World Health Organization reference centre was used as antigen and prepared as per the meth-od of Goyal et al. ^[12] . Serum samples were serially two-fold diluted with saline and 0.2% gelatin so-lution. Fifty mL of parasites suspension was added to an equal volume of diluted serum in each well. After the completion of the test, microtitre plates were shaken on a flat and smooth surface and were incubated at 22 °C overnight. The results were re-corded in titres. The positive result was expressed as the highest dilution of serum that showed definite agglutination of parasite. Uniform mat formation suggested a positive result. A sample was considered positive if it had a titre of 21:3200 (cut-off value).

rK39 dipstick

The kit was provided by In Bios International, Inc. Seattle, Washington, DC, USA. The test was con-ducted according to instructions of the manufac-turer. The strip for VL assay is a qualitative mem-brane-based immunoassay for the detection of antibodies to VL in human serum. In brief, 20 pL of serum was placed on pad. Following this, 2-3 drops of wash buffer (provided) were added. The mixtures were carried up by capillary action of the membrane VL. The result was read within 10 min-utes. The appearance of two pink lines-one con-trol and one test-indicated a positive result while the appearance of one pink line (control) indicates that the test methodology is correct but sample is negative for antibodies.

Followed up cases

All the patients were provided a standard treatment with amphotericin B (Sarabhai Chemicals, Vado-dara, India) in 1 mg/kg infusions on alternate days up to 15 infusions. After the completion of treat-ment, the patients were free of symptoms. The hae-matological parameters were restored. Fifty easily accessible (those who reported) among treated pa-tients (n=150) were further followed up. Serum and urine samples were collected to know the prognos-tic value of the rK39 dipstick test, DAT, and KAtex.

Statistical analysis

The serological data, thus, obtained were ana-lyzed after entry in the Epi Info software (version 6) and the SPSS package (version 16). The Fleiss quadratic method was used in calculating the con-fidence interval of proportions. The sensitivity and specificity were mathematically calculated as fol-lows: sensitivity=(true positive/true positive+false negative)x100; Specificity=(true negative/true negative+false positive)x100.

Ethics

The Ethical Committee of the Institute of Medical Sciences, Banaras Hindu University, approved the study. Written consent was obtained from all the subjects.

Results

Of the 150 parasitologically-proven index cases, 131 (87%), 140 (93.3%), and 148 (98%) were posi-tive for the KAtex, DAT, and rK39 strip test respec-tively. Of the healthy controls (n=160), 2 (1.2%), 5 (3.1%), and 17 (10.6%) showed positive reaction for the KAtex, DAT, and rK39 strip test respectively while none from the non-endemic area showed positive reaction. No cross reactivity was found be-tween the VL cases and the other febrile cases. [Table 1] shows the results.

The sensitivity of the KAtex, DAT, and rK39 strip test was 87% (95% CI 80-96), 93.3% (95% CI 88- -100), and 98% (95% CI 93-100) respectively. The specificity of these tests was 98% (95% CI 93-100), 93% (95% CI 87-100), and 89% (95% CI 82-97) for the KAtex, DAT, and rK39 strip test respectively [Figure - 1]. Fifty clinically-cured cases were followed up and subjected to the KAtex, DAT and rK39 strip test af-ter 30 days of successful treatment. The DAT and rK39 strip test showed positive results in all the 50 (100%) cases whereas the KAtex showed no posi-tive reaction in any case [Figure - 2].

Discussion

The present study was conducted on 455 subjects for the field evaluation of the KAtex, DAT and rK39 strip test as a diagnostic and prognostic tool for kala-azar. In this study, the DAT and rK39 strip test were evaluated on serum because the tests were carried out in one-go. In clinical prac-tice, however, these tests may be performed on whole fingerprick blood. In our study, the KAtex showed 87% (95% CI 80-96) sensitivity. These results corroborated well with the results of other researchers ^[13],[14] who showed a sensitivity in the range of 87-95% where fresh urine samples were tested. However, the results do not corroborate with those of other studies ^{[15],[16],[17]} where poor sen-sitivity has been reported, which may be due to low level of parasitaemia at a particular stage of disease or may be attributed to the effect of transport and subsequent storage of urine. Poor-to-moderate re-producibility of the KAtex mentioned by other researchers ^[15],[16] is an additional disadvantage of the test. On the other hand, our results on the specificity of the KAtex are consistent with those of previous studies which ranged from 98% to 100% ^{[14],[15],[16]} . The positive KAtex by 1.2% (2 of 160) in the endemic controls points out to subclinical in-fection, having a low level of parasitaemia. Subclin-ical VL with a few or no symptoms with positive anti-leishmanial serology is documented ^[9] . In this study, cut-off titre of ≥1:3,200 was employed for the DAT. Our results showed the sensitivity and specificity of 93.3% and 93% respectively. These results corroborated well with the results of other researchers ^{[18],[19],[20],[21],[22],[23]} . The rK39 strip test showed 98% sensitivity and 89% specificity. It showed negative reaction with the non-endemic controls whereas 10.6% of the subjects from an endemic area were positive by the rK39 strip test; 10.6% positivity in the endemic controls but negative reaction with the non-endemic controls indicates false-positive reaction. False-positivity with serological tests is significantly higher than with the antigen test (KA-tex). On the contrary, all the febrile cases showed non-reactive results by this method. This result cor-roborated well with the finding of other researchers ^{[24],[25],[26]} . The specificity of the test is lower because of the higher prevalence of subclinical or past infec-tions. On the contrary, the study in Sudan showed lower sensitivity and specificity of this test ^[27] . The reason for this difference in antibody response may be because of the ethnic differences or the dif-ference in antibody response.

In the present study, 50 successfully-treated indi-viduals were followed up to evaluate the prognostic value of the KAtex, DAT, and rK39 strip test after 30 days of the completion of treatment. All showed positive reaction for the rK39 strip test and DAT, which indicates persistence of antibodies whereas the KAtex showed no positive reaction, indicating absence of antigens in urine. These observations were supported in various studies performed in suc-cessfully-treated cases ^{[26],[28],[29],[30]} .

If we compare the overall yield of these diagnostic tests, the DAT and rK39 strip test are comparable. The advantage of DAT includes the availability of titres; however, the procedure for the DAT is cum-bersome and requires incubation, and the storage of antigen also requires a cold-chain. This may not be possible in remote rural India. By contrast, the rK39 strip test procedure is easy, simple, and user-friendly and can be easily performed and inter-preted in most peripheral areas even by an auxil-iary health worker. Thus, the rK39 strip test has the potential to be used as a suitable candidate for the diagnosis of VL whereas the prognostic value of the KAtex seems to be promising.

The study was carried out only in a limited number of post-treated kala-azar cases. For the correct as-sessment of these tests, more number of cases with well-matched controls may be carried out to reach a definite conclusion. However, with these obser-vations, it may be concluded that the rK39 strip test might be used as a diagnostic tool for kala-azar. Since this is a highly-sensitive test, an ideal format for use under field conditions, quick and no special equipment is needed. For monitoring successful treatment, the KAtex may be used because anti-gen becomes undetectable in urine after successful treatment. For sero-epidemiological studies, both rK39 and KAtex may be carried out. Their simplici-ty and detection of both antigen and antibodies may give better information on the disease preva-lent in a particular area.

Acknowledgements

The authors express their gratitude to Dr. S.K. Kar, former Director of KAMRC, Patna, for providing the patients and the facilities to carry out the DAT and KAtex.

References

Copyright 2010 - Journal of Health Population and Nutrition

The following images related to this document are available:

Photo images