Field performance of Plasmodium falciparum lactate dehydrogenase rapid diagnostic tests during a large histidine-rich protein 2 deletion survey in Ethiopia

Abstract

Background: Malaria rapid diagnostic tests (RDTs) have expanded diagnostic service to remote endemic communities in Ethiopia, where 70% of malaria services per annum are reliant on them. However, diagnostic strategies are threatened by Plasmodium falciparum parasites with deletions of the histidine-rich protein 2 and/or 3 (pfhrp2/3) genes. Studies have reported pfhrp2/3 gene deletion prevalence in Ethiopia that exceeds the WHO recommended threshold to switch to non-HRP2 targeted RDTs for detection of P. falciparum. Therefore, RDTs that target alternative antigens, such as P. falciparum lactate dehydrogenase (PfLDH) are increasingly in programmatic use. Methods: Malaria suspected patients visiting health facilities of Amhara, Tigray, Gambella, and Oromia regions of Ethiopia were screened by community health workers using Carestart Pf/Pv (HRP2/Pv-LDH) and SD-Bioline Pf (HRP2 for Pf/LDH for Pf) RDTs. Dried blood spot (DBS) samples were collected from selected patients for molecular and serological analysis. The clinical data and RDT results were recorded on standard forms, entered into EpiInfo, and analysed using STATA. The Pf-LDH detecting RDT results were compared with real-time PCR and bead-based immunoassay to determine their diagnostic performance. Results: The 13,172 (56% male and 44% female, median age of 19 years ranging from 1 to 99 year) study participants were enrolled and tested with PfHRP2 and PfLDH detection RDTs; 20.6% (95% CI: 19.6 to 21.6) were P. falciparum RDT positive. A subset of samples (n = 820) were previously tested using P. falciparum lactate dehydrogenase (pfldh) quantitative real-time PCR, and 456 of these further characterized using bead-based immunoassay. The proportion of samples positive for P. falciparum by the PfHRP2 Carestart and SD-Bioline RDTs were 66% (539/820) and 59% (481/820), respectively; 68% (561/820) were positive for the PfLDH band on the SD-Bioline RDT. The sensitivity and specificity of the PfLDH RDT band were 69% and 38%, respectively, versus pfldh qPCR; and 72% and 36%, respectively, versus PfLDH detection by immunoassay. Among samples with results for RDT, qPCR, and immunoassay, higher proportions of P. falciparum were recorded by pfldh qPCR (90%, 411/456) and PfLDH immunoassay (88%, 363/413) compared to the PfLDH band on the SD-Bioline RDT (74.6%, 340/456). Conclusion and recommendation: Both PfHRP2 RDTs detected fewer P. falciparum cases than PfLDH, and fewer cases than qPCR or immunoassay. The poor sensitivity and specificity of the PfLDH RDT compared to qPCR and to immunoassay in this study raises concern. Continuous operator training and RDTs quality assurance programme to ensure quality diagnostic services are recommended.