Serum and urinary biomarkers for early detection of acute kidney injury following Hypnale spp. envenoming

Abstract

Background: Hump-nosed pit viper (HNV; Hypnale spp.) bites account for most venomous snakebites in Sri Lanka. Acute kidney injury (AKI) is the most serious systemic manifestation (1-10%) following HNV envenoming. We aimed to identify the value of functional and injury biomarkers in predicting the development of AKI early following HNV bites.

Methods: We conducted a prospective cohort study of patients with confirmed HNV envenoming presenting to two large tertiary care hospitals in Sri Lanka. Demographics, bite details, clinical effects, complications and treatment data were collected prospectively. Blood and urine samples were collected from patients for coagulation and renal biomarker assays on admission, at 0-4h, 4-8h, 8-16h and 16-24h post-bite and daily until discharge. Follow-up samples were obtained 1 and 3 months post-discharge. Creatinine (sCr) and Cystatin C (sCysC) were measured in serum and kidney injury molecule-1 (uKIM-1), clusterin (uClu), albumin (uAlb), β2-microglobulin (uβ2M), cystatin C (uCysC), neutrophil gelatinase associated lipocalin (uNGAL), osteopontin (uOPN) and trefoil factor-3 (uTFF-3) were measured in urine. Definite HNV bites were based on serum venom specific enzyme immunoassay. Kidney Disease: Improving Global Outcomes (KDIGO) criteria were used to stage AKI. Two patients had chronic kidney disease at 3 month follow-up, both with pre-existing abnormal sCr, and one developed AKI following HNV envenoming.

Results: There were 52 patients with confirmed HNV envenoming; median age 48y (Interquartile range [IQR]:40-59y) and 29 (56%) were male. Median time to admission was 1.87h (IQR:1-2.75h). Twelve patients (23%) developed AKI (AKI stage 1 = 7, AKI stage 2 = 1, AKI stage 3 = 4). Levels of five novel biomarkers, the functional marker serum Cystatin C and the damage markers urinary NGAL, cystatin C, β2-microglobulin and clusterin, were elevated in patients who developed moderate/severe acute kidney injury. sCysC performed the best at 0-4 h post-bite in predicting moderate to severe AKI (AUC-ROC 0.95;95%CI:0.85-1.0) and no biomarker performed better than sCr at later time points.

Conclusions: sCysC appears to be a better marker than sCr for early prediction of moderate to severe AKI following HNV envenoming.

Fig 1. Flow chart of included patients and grading of acute kidney injury.

Fig 2. Scatter plots of the peak HNV venom concentrations in serum within 24 h post bite (median and interquartile range) for each of the acute kidney injury groups.

Fig 3

Maximum renal biomarker concentration within 24 h of bite. Scatter plots of the maximum biomarker concentrations reached within 24 h post-bite for patients with no acute kidney injury (No AKI; green), mild AKI (AKI 1; blue) and moderate to severe AKI (AKI 2–3; red), for serum creatinine (sCr; Panel A), serum cystatin C (sCysC; Panel B), urinary neutrophil gelatinase associated lipocalin (uNGAL; Panel C), urinary cystatin C (uCysC; Panel D), urinary beta-2 microglobulin (uβ2M; Panel E), and urinary clusterin (uClu; Panel F). The grey shaded is the normal range based on respective biomarkers measured in healthy individuals.

Fig 4

Time course of the median biomarker concentrations (with interquartile ranges) for each of the three patient groups ([No AKI; green], [Mild AKI; blue] and [moderate to severe AKI; red]) post-bite for 48 h including serum creatinine (sCr; Panel A), serum cystatin C (sCysC; Panel B), urinary neutrophil gelatinase associated lipocalin (uNGAL; Panel C), urinary cystatin C (uCysC; Panel D), urinary beta-2 microglobulin (uβ2M; Panel E), and urinary clusterin (uClu; Panel F). The grey shaded is the normal range based on respective biomarkers measured in healthy individuals.

Fig 5

Plots of the AUC-ROCs versus time for the best biomarkers in detecting moderate/severe AKI versus no AKI/mild AKI, including serum creatinine (sCr; red line), serum cystatin C (sCysC; blue line), and urinary neutrophil gelatinase-associated lipocalin (uNGAL; green line).

Fig 6

A. Scatter plots of the estimated GFR at 1 month or 3 months* in patients with no acute kidney injury (No AKI; green), mild AKI (AKI 1; blue) and moderate/severe AKI (AKI 2–3; red). The dotted line indicates the normal eGFR of 60ml/min/1.73m². B. Time course of sCr concentrations for the two patients with CKD. (Patient A; green line, patient B; red line) The grey shaded is the normal range for sCr measured in healthy individuals. *In all patients eGFR at 3 months was considered if available. In the absence of an eGFR at 3 months eGFR at 1 month was considered.