BackgroundEarly detection of systemic envenoming is critical for early antivenom therapy to minimise morbidity and mortality from snakebite. We assessed the diagnostic utility of three serum biomarkers in the early detection of systemic envenoming in viper bites in rural Sri Lanka. MethodsAll confirmed snakebite patients admitted to Teaching Hospital Anuradhapura from July 2020 to June 2021 were included. On admission, blood was collected for venom concentrations, prothrombin time (PT)/international normalised ratio (INR), fibrinogen, serum creatinine (sCr), and three serum biomarkers - secretory phospholipase A2 (sPLA2), neutrophil gelatinase-associated lipocalin (sNGAL) and clusterin (sClu). Systemic envenoming was defined by the presence of venom-induced consumption coagulopathy, neurotoxicity, acute kidney injury or the presence of non-specific clinical effects. ResultsA total of 237 confirmed snakebite patients (Russell’s viper,72; Hump-nosed viper,80; non-venomous snakes,31; and unidentified bites,54) with sufficient pre-antivenom serum samples were recruited [median age: 42y (interquartile range: 29-53y); males: 173 (73%)]. Systemic envenoming occurred in 68 (94%) Russell’s viper bites, 48 (60%) hump-nosed viper bites and 45 (83%) unidentified bites. The median sPLA2 activity was 72nmol/ml/min (IQR: 30–164) for Russell’s viper envenoming, 12nmol/ml/min (IQR:9–16) for hump-nosed viper envenoming and 11nmol/ml/min (IQR:9-14) for non-venomous bites. There was no difference in sNGAL and sCLu between the three groups. The median sPLA2 activity of patients with systemic envenoming was 16nmol/min/mL (IQR:11–59nmol/min/mL) compared to 11 nmol/min/mL(IQR:9–14nmol/min/mL) in patients without systemic envenoming; difference between medians of 5nmol/min/mL (95% CI:4–12nmol/min/mL). The AUC-ROC of admission sPLA2 was the best predictor of systemic envenoming in all snakebites (AUC-ROC 0.72, 95% CI: 0.66 to 0.79), whereas sNGAL and sClu were poor predictors. sPLA2 was a better predictor of systemic envenoming in Russell’s viper bites (AUC-ROC 0.90, 95% CI:0.76-1.00) and in those presenting within 2h of bite. A sPLA2 activity >23.5nmol/min/ml had a sensitivity of 97% (95% CI:91-99.5%), but a low specificity of 41% (95% CI:34%-49%) in predicting systemic envenoming. A sPLA2 concentration of >46 nmol/min/ml on admission had a sensitivity of 100% (95% CI:51%-100%) and a specificity of 67% (95% CI:55%-77%) in predicting systemic envenoming in Russell’s viper bites. ConclusionsPLA2 is an early predictor of systemic envenoming following snakebite, particularly in Russell’s viper bites and in those who present early.