Evaluation of performance of two monoplex quantitative real time PCR assays in comparison to microscopy for identification of Malaria parasites.

ABSTRACT

Background: Microscopy is the gold standard for Malaria diagnosis with shortcomings such as false positives, false negatives, errors in species identification, and errors in enumeration of parasites. Quantitative real-time PCR (qPCR) has improved submicroscopic malaria diagnosis. This study evaluated the performance, concordance, correlation and methods agreement of two monoplex qPCR assays against expert malaria microscopy for the detection and enumeration of malaria parasites. Methods: This was a cross sectional study utilizing 127 archived blood samples collected from five provinces in Kenya. Malaria microscopy was conducted by two independent microscopists then 18S-rRNA-qPCR and non-18S-rRNA-qPCR assays were done to identify and quantify the infecting species. The sensitivity, specificity, and predictive values. Cohen Kappa value was used to quantify the method agreement and Bland Altman test was used to assess the bias and limits of agreement. Correlation between microscopy and qPCR parasite densities was determined by the Spearman's rank test. Statistical significance was taken at p<0.05. Results: A higher sensitivity and a lower specificity were observed in all the three plasmodium species in non 18S- rRNA-qPCR compared to 18S-rRNA-qPCR. The sensitivity and specificity of 18S-rRNA-qPCR was 91.3% and 75% in detection of P. falciparum, 67.6% and 88.1% in detection of P. malariae, and 55.8% and 91.4% in detection of P. ovale. The sensitivity and specificity of non 18S-rRNA-qPCR was 99.1% and 66.7% in detection of P. falciparum, 77.9% and 88.1% in detection of P. malariae, and 79.4% and 90.3% in detection of P. ovale. All the positive and negative predictive values were above 70% except the negative predictive value for 18S-rRNA-qPCR (47.4%). Kappa of more than 0.5 was observed between microscopy and both18S-rRNA-qPCR and non-18S-rRNA-qPCR in the detection of all three malaria parasites. The non-18S-rRNA-qPCR method had higher kappa > 0.65, in all the three species compared to 18S-rRNA-qPCR method (kappa < 0.55). There was a clear positive correlation between microscopy parasite density and the parasite densities estimated by the 18S-rRNA-qPCR and Non-18S-rRNA-qPCR (P<0.001). Conclusion: The results showed that both monoplex realtime PCR methods demonstrated a high performance compared to microscopy proving to be better methods in the identification and speciation of malaria parasites especially of low parasitemia. The realtime PCR methods also had a positive correlation with parasite density and hence can be used in accurate determination of parasite densities when compared to microscopy. Therefore, this study recommends the utilization of realtime PCR methods in the detection, speciation and quantification of both microscopic and submicroscopic malaria parasites.