Urine metabolites for the identification of Onchocerca volvulus infections in patients from Cameroon

Abstract

Abstract Background The tropical disease onchocerciasis (river blindness), caused by Onchocerca volvulus filarial nematodes, is targeted for elimination by mass treatment with nematocidal and antimicrobial drugs. Diagnosis of O. volvulus infections is based on counts of skin-borne microfilariae, but additional diagnostic tools, e.g. worm- or host-derived small RNAs, proteins or metabolites, are required for high-throughput screening. N-acetyltyramine-O,β-glucuronide (NATOG) was suggested as a biomarker for onchocerciasis but its viability as diagnostic tool has been challenged. Methods We performed a screening program of urine samples from individuals from Cameroon infected with O. volvulus, Loa loa, Mansonella perstans or a combination thereof. Urine metabolites were measured by liquid chromatography–mass spectrometry (LC-MS). Principle component analysis (PCA) revealed that onchocerciasis causes complex changes of the urine metabolome. Results The mean NATOG content was elevated in urine of O. volvulus-infected compared with non-infected individuals, but NATOG levels showed considerable variation. However, 13.8% of all O. volvulus-infected individuals had high NATOG levels never reached by individuals without filarial infections or only infected with L. loa or M. perstans. Therefore, the identification of individuals with high NATOG levels might be used to screen for the elimination of onchocerciasis after mass drug application. Additional metabolites, including a compound identified as cinnamoylglycine, had high PC1/PC2 loadings in the data set. Mean levels of cinnamoylglycine were increased in O. volvulus-infected individuals, and 17.2% of all O. volvulus individuals had elevated cinnamoylglycine levels not reached by the controls. Conclusions On an individual level, NATOG alone had poor discriminative power distinguishing infected from non-infected individuals. However, 13.8% of all O. volvulus-infected individuals had NATOG levels never reached by individuals without filarial infections or infected with only L. loa or M. perstans. Discrimination of O. volvulus infections from controls or individuals suffering from multiple infections was improved by the measurement of additional metabolites, e.g. cinnamoylglycine. Thus, measuring a combination of urine metabolites may provide a way to assess onchocerciasis on the population level. This provides the possibility to design a strategy for large-scale onchocerciasis epidemiological screening programs based on urine rather than invasive techniques. Graphical abstract