Automated detection of Loa loa: a field trial in Cameroon

Abstract

Onchocerciasis (river blindness) is targeted for elimination through mass administration (MDA) of ivermectin (IVM) to endemic populations. In areas where loiasis, caused by the blood-borne filarial parasite Loa loa, is co-endemic, IVM MDA faced significant challenges because individuals harboring more than 30,000 L. loa microfilaria (mf)/mL of blood are at high risk of developing serious adverse events (SAEs) that can sometimes be fatal. An alternative strategy was developed for safe IVM distribution: the so-called 'Test and Not Treat' (TaNT), which identifies and excludes individuals with high mf density from IVM MDA and treats only those with minimal risk. TaNT requires a point-of-care diagnostic device to rapidly quantify L. loa parasites in field conditions. The NTDscope is a handheld device which captures bright-field videos of whole blood in capillaries. An onboard algorithm detects and counts the live microfilaria of L. loa via movement of red blood cells. This paper describes (i) a new detection algorithm; and (ii) a May 2025 field trial in Cameroon using the NTDscope with new algorithm (550 patients). Results for the TaNT use case (i.e. flagging cases with L. loa mf densities > 30,000 mf/mL): 94% to 97% sensitivity, and 94% to 97% specificity. The results indicate that the NTDscope and new algorithm provide a greater margin of safety than the predecessor device, and can potentially offer rapid, effective field detection of high mf infections to enable scalability of the TaNT strategy.