Diagnostic laboratory tests for visceral leishmaniasis and exploration of genotype and genetic diversity of Leishmania donovani isolates from kalaazar patients

Abstract

Visceral leishmaniasis (VL) or kala-azar is a neglected chronic systemic illness caused by Leishmania donovani. Clinical diagnosis of VL is inaccurate even in patients meeting the clinical case definition, so laboratory confirmation is essential. Current diagnostic tests include parasitological, immunological and PCR-based molecular methods. Except serology all other methods are not suitable as point-of-care test in remote endemic areas. Serological test by rK39 antigen is being used in the Indian subcontinent with very high sensitivity but it lacks discriminatory power between active VL and past or asymptomatic cases. So, always there is search for new diagnostic innovation which will be highly accurate, cheap, simple and rapid for successful VL management and to achieve the goal of current elimination programme. The present study was designed to formulate an applicable diagnostic algorithm based on comparative evaluation of available VL diagnostic tests aiming to facilitate more case detection from different levels of health care facilities in our country. The work also focused on exploration of genotype and genetic diversity of Leishmania strain for its taxonomical status. Study population comprised of 200 subjects including 100 confirmed VL patients (spleen smear positive for LD body) admitted in the Rajshahi Medical College Hospital (RMCH), Bangladesh and 100 controls (30 endemic, 30 non-endemic and 40 disease controls) from July, 2010 to January, 2013. Parasitological (spleen and buffy coat smear microscopy), serological (RDT and ELISA against both rK39 & rK28) and molecular diagnostic tests like Loop-mediated isothermal amplification (LAMP) and different PCR (Ln-PCR, Mini-exon, ITS1 and ITS2) methods were carried. Buffy coat was separated by concentration gradient separation technique to be utilized for smear preparation and DNA extraction. Parasite load in spleen smear was graded from 1+ to 6+ and performances of tools other than serology were compared with parasitic load. Diagnostic sensitivity, specificity and accuracy were calculated using online clinical calculator. Performances of different diagnostic tools against gold standard (spleen smear) were compared by McNemar’s test and correlations between tools were done by Spearman correlation. The kappa value of different tools against spleen smear was used to construct a receiver-operator characteristic (ROC) curve. Diagnostic tools were ranked (1 to 10) based on kappa value, cost, interpretation, availability, user friendliness, test type and potential for field use. Diagnostic algorithm for VL in Bangladesh was formulated based on the diagnostic indices and index scoring of tools. Genotyping by ITS1-RFLP and DNA sequence analysis of ITS1 and ITS2 PCR products was done for polymorphism. Phylogenetic tree was constructed using maximum parsimony by the MEGA (Molecular Evolutionary Genetics Analysis) computer program, Version 5.05 and compared with reference strains. Parasitic load was demonstrated as grade 1+ (8%), 2+ (41%), 3+ (34%), 4+ (10%), 5+ (6%) and 6+ (1%). Diagnostic sensitivity, specificity and accuracy at 95% CI respectively of rK39 ICT (99%, 96%, 95%), rK39 ELISA (98%, 97%, 95%), rK28 ICT (99%, 90%, 89%), rK28 ELISA (98%, 94%, 92%), Buffy coat smear (93%, 100%, 93%), Ln-PCR (94%, 99%, 93%), LAMP (89%, 100%, 89%), Mini-exon PCR (86%, 100%, 86%), ITS1 PCR (85%, 100%, 85%) and ITS2 PCR (80%, 100%, 80%) were noted. All diagnostic tools correlated significantly (P=0.0001) with gold standard. Correlation between rK39 and rK28 ICT was 0.942 (P=0.0001) and between rK39 and rK28 ELISA was 0.930 (P=0.0001). Pair-wise comparison of Buffy coat smear, Ln-PCR, LAMP, Mini-exon, ITS1 and ITS2 PCR against gold standard was found significant (P<0.05). Different PCR methods were also found significantly (P<0.05) comparable to Ln-PCR. Buffy coat smears was found 100% positive among grades ≥3+ parasite load cases but other tools did not show significant correlation with parasite load. ROC analysis in respect to gold standard revealed AUC (at 95% CI) of 0.965 (SE=0.0129), 0.945 (SE=0.0157), 0.930 (SE=0.0174), 0.925 (SE=0.0179) and 0.900 (SE=0.0201) respectively for Ln-PCR, LAMP, Mini-exon, ITS1 and ITS2 PCR. While, AUC (at 95% CI) of rK39 ICT, rK39 ELISA, rK28 ICT, rK28 ELISA and buffy coat smear was 0.975 (SE=0.0110), 0.975 (SE=0.0111), 0.945 (SE=0.0159), 0.960 (SE=0.0139), and 0.965 (SE=0.0128) respectively. Index score (IS) of diagnostic tools showed rK39 ICT as the best diagnostic option (Rank=1, IS=1.65335) followed by rK28 ICT (Rank=2, IS=1.48776), Buffy coat smear (Rank=3, IS=0.50764), rK39 ELISA (Rank=4, IS=0.2963), rK28 ELISA (Rank=5, IS=0.2963), Ln-PCR (Rank=6, IS= -0.7158), LAMP (Rank=7, IS= -0.88139), Mini-exon PCR (Rank=8, IS= -0.88139) ITS1 PCR (Rank=9, IS= -0.88139), and ITS2 PCR (Rank=10, IS= -0.88139) for kala-azar. In the diagnostic algorithm, RDT (rK39 or rK28) was proposed as the most preferred test. Genotyping revealed L. donovani as the sole agent for kala-azar and DNA sequencing explored no polymorphism. All 14 Bangladeshi strains including 4 from present study (Raj 21, Raj 64, Raj 68 and Raj 85) were found exactly identical to strain ID, DON 39, with Zymodeme MON-2, isolated from Indian VL patient (gene accession No. AJ634376). Kala-azar patients’ characteristics showed male to female ratio of 2.13:1 with mean age, monthly income and duration of fever as 20.66±15.863 years, Tk. 4005±2631.871 and 20.61±12.518 weeks respectively. Patients had splenomegaly (98%), hepatomegaly (76%), anaemia (99%), wasting (100%), blackening (96%), weight loss (100%), family history of VL (18%), past history of VL (13%), jaundice (2%), bleeding (9%) and pancytopenia (40%) as presenting features. All diagnostic tests showed good to excellent sensitivity (80% to 99%) and specificity (90% to 100%). rK39 ICT ranked #1 and was proposed as the most preferred diagnostic test in the diagnostic algorithm. Buffy coat smear may be suitable alternative to conventional invasive procedures and is preferred parasitological confirmation for asymptomatic, treatment failure, relapse or immunodeficient VL patients. Ln-PCR or LAMP can be good adjunct for confirmation of kala-azar. Genotyping revealed L. donovani as the agent of kala-azar in Bangladesh, identical to Indian strain (DON 39, Zymodeme MON-2) and there is no strain polymorphism.