A Comparison of the Efficiency of Three Sampling Methods for use in the Molecular and Conventional Diagnosis of Cutaneous Leishmaniasis

In human cutaneous leishmaniasis (CL), the success of positive diagnoses and species identifications depends, primarily, on how biopsies are taken and then processed and examined. The efficiency of three methods of taking skin biopsies from suspect cases of CL was compared using the classical methods of microscopy of stained smears, in vitro culture of tissue aspirate, and internal transcribed spacer region 1 (ITS1)-polymerase chain reaction in diagnosing positive cases and identifying the species of Leishmania causing them. From 1994 to 2014, biopsy samples from the skin lesions of 2232 CL-suspected patients were collected as unstained smears, as smears stained with Giemsa's stain and on filter paper, and compared in the diagnostic tests employed. Matched comparison based on testing biopsy samples from 100 patients, microscopy, in vitro culture and ITS1-PCR were also conducted to assess the most suitable combination of methods for diagnosing leishmaniases. In the 100-case-matched comparison, the three different types of sample proved to be equally good with no significant difference (P > 0.05). However, skin tissue imprints on filter paper revealed most cases of CL. The kappa statistic for measuring the degree of agreement among the three samples was 89%, which is considered good. Agreement was highest between imprints on filter paper and unstained smears, and lowest was for stained smears. In the overall comparison between the ITS1-PCR and conventional methods, the ITS1-PCR using samples from filter papers was the most sensitive method but the difference was insignificant (P = 0.32). The combination of microscopy together with ITS1-PCR on samples from filter papers increased the sensitivity significantly to 46%, compared to using the methods individually (P = 0.003-0.0008). On comparing the results of the tests done on the samples from the 2232 patients after applying ITS1-PCRs to their samples from filter papers, unstained smears, in vitro culture, microscopy, and stained smears showed, respectively, test sensitivities of 81, 69, 64, 57 and 48%. Of the tests and samples adjudicated, ITS1-PCRs run on skin tissue samples from filter papers proved best for the routine laboratory diagnosis of CL. Adding microscopy of stained smears to it, improved its diagnostic value significantly.