Discussion
In this phase II clinical trial of patients with SBE, the addition of varespladib to standard of care, including the use of antivenom did not result in a difference in the SSS measured as a change from baseline to 6 and 9 hours postbaseline (primary outcome). Key secondary outcomes slightly favoured varespladib but were not statistically significant. Importantly, however, within the prespecified subgroup of patients treated within 5 hours of the snakebite, clinically important differences and nominal statistical significance were observed for patients receiving varespladib versus placebo for the SSS AUC day 7 and for measures of physical function and recovery.
The primary results from the BRAVO study should be considered within the context of widely reproduced preclinical data and the promising additional subgroup analyses.17 18 Factors that may have hindered the ability to identify benefit in the primary outcome in this trial include (1) limitations of the outcome measure, (2) initiation of study drug an average of 6 hours after the bite, (3) efficacy signal obscured by concurrent administration of antivenom as well as resuscitative and symptomatic treatments, (4) random but important baseline differences in time to treatment and severity favouring the placebo group with the potential for residual confounding and (5) heterogeneity in the severity and clinical manifestations of SBE.
There is not an accepted single outcome measure for global SBE clinical trials.23 SBE is a heterogeneous disease with diverse clinical syndromes driven by variation in venom composition. To address this heterogeneity, a global consortium of SBE researchers currently recommend collecting a wide variety of ‘core outcome’ measures.23 These ‘core outcomes’ include key clinical signs, laboratory values and clinician-reported and patient-reported outcomes that evaluate important endpoints ranging from organ-specific toxicity to disability and mortality.
The SSS is an existing embodiment of the ‘core outcomes’ recommendation and addresses the challenge of diverse toxicities from SBE by using a composite measure summing subscores across multiple organ systems. To begin updating the SSS for international use, we modified it by removing the gastrointestinal subscore and adding a renal subscore which better aligns the SSS with toxicities that contribute most to morbidity and mortality globally. Nonetheless, the SSS still has limitations. Rapid death from venom neurotoxicity is generally caused by neuromuscular blockade resulting in weakness or total paralysis, but the existing neurologic subscore of the SSS focuses mostly on central nervous system effects. Objective measures of neuromuscular weakness, such as head-lift and grip strength, may better capture the most common patient conditions that put them at risk for death or reliance on advanced life-preserving measures.24
The timing of outcome assessments is also important. Based on findings from studies of varespladib in animal models with lethal doses of venoms, we selected a primary outcome measured 6 and 9 hours after randomisation. The results of this trial confirm that, regardless of treatment arm, some toxicities resolve within 9 hours (eg, cardiovascular) and some are unchanged over the first 48 hours (eg, local wound and blood components that do not normally regenerate in 6–9 hours). The challenge of identifying a time frame for recovery that captures benefit for different toxicities is a study design issue that will need to be addressed in future global SBE studies. A measure that integrates differences observed over time, such as AUC, may provide a practical solution to this challenge.
Patient-reported physical function during recovery from SBE is an important patient-centred aspect of SBE that has been incorporated into prior studies. More specifically, the PSFS is an extensively validated, patient-reported measure that has had an in-depth psychometric evaluation in SBE.25 In patients treated within 5 hours, the point estimate for the increase in PSFS at day 3 for varespladib versus placebo exceeded the established minimum clinically important difference of 1.0 on the 11-point PSFS scale. We augmented this analysis of recovery by performing a post-hoc evaluation of the proportion of patients achieving complete recovery, defined as an SSS of 0. This outcome provides an intuitive and unambiguous clinical endpoint that approximates the condition under which a patient has had resolution of symptoms. The strength of the complete recovery data is supported by the size of the treatment effect as well as the consistency of directional benefit for different snake types. One possible mechanism by which varespladib may promote recovery is by reducing the total injury experienced by the patient, thus reducing the time required for physiologic homeostasis and tissue repair.
The concurrent administration of standard of care, including antivenom, likely influenced the outcomes in this trial. A distinguishing characteristic of an oral direct toxin inhibitor is the potential for use as a field or prehospital treatment of SBE prior to antivenom administration. Studies conducted where antivenom is not available, such as in the field, community health clinics or ambulances, would be informative, but the logistics of conducting rigorous trials in these settings are daunting. Withholding antivenom when available in the hospital would be unethical without further evidence of the efficacy of varespladib alone.
Nonetheless, in this study of hospitalised patients receiving concurrent standard of care, a signal of benefit from varespladib was seen for the subgroup of patients receiving study drug within 5 hours of bite or symptom onset when evaluating multiple toxicities over time using the SSS AUC day 7 and when evaluating function and recovery. This signal is biologically plausible because SBE is a time-dependent disease for which early treatment is essential to optimise outcomes.26 27 Because sPLA2s are relatively small and highly active toxins that diffuse rapidly to produce systemic effects, early treatment may be particularly important when using an sPLA2 inhibitor.8 17 It is encouraging to see consistent benefits for recovery across the major snake types included in the study. As such, these findings are supportive of the use of varespladib as an oral treatment that can be carried in remote and austere environments and taken shortly after envenoming to minimise morbidity and mortality.
The BRAVO trial results suggest that varespladib is safe and well tolerated for the treatment of SBE. These results add to a large body of safety data from earlier trials of varespladib for other indications. Two notable safety findings were observed in this earlier work. First, a 1.2% increase in recurrent myocardial infarction was observed in a study of patients hospitalised with acute coronary syndrome who had risk factors for recurrent myocardial infarction.28 No increase in cardiac events was observed in two large studies of patients with stable coronary artery disease or any of the trials of varespladib for other indications.29 30 In the BRAVO trial, there were no cardiac events and no patients in either treatment group had ECG findings suggestive of ischaemia. Second, increases in hepatic enzymes equal to or exceeding 3× the upper limit of normal were observed in 4% (15 of 396) of patients treated with varespladib for a period of 8 weeks in two studies of stable coronary artery disease.29 30 All patients were asymptomatic and there was no accompanying increase in bilirubin or alkaline phosphatase. The duration of treatment tested in BRAVO is substantially shorter than these earlier trials. Furthermore, in the BRAVO trial, among patients with normal liver enzymes at baseline, no patient randomised to varespladib had an elevation in hepatic enzymes equal to or exceeding 3× the upper limit of normal through 28 days of follow-up.
Because of their small size, children are disproportionately affected by SBE.25 This is even more evident in low-income settings.31 Even as the number of paediatric patients treated in this study was small, encouraging safety data for the paediatric patients adds to a large body of safety data and should inform benefit–risk assessments about appropriate use of varespladib in children bitten by venomous snakes.
The BRAVO trial lays an important foundation for additional clinical studies of direct toxin inhibitors such as varespladib and other treatments for SBE. The signal consistent with benefit in patients treated within 5 hours of the bite is encouraging because it is biologically plausible, consistent with antivenom studies and points to the potential value of varespladib as a field or prehospital treatment. Future studies should consider focusing on patients initiating treatment within 5 or 6 hours and consider outcomes such as the SSS AUC day 7, the PSFS or measures of recovery.
This study has several limitations. Imbalances between the treatment groups included a longer time from bite to initiation of study drug and a higher baseline SSS in the varespladib group, creating the possibility of residual imbalances that may have impacted comparisons. There was no control over the timing or amount of antivenom given to patients. Long-term physical and psychological disability are important components of the total burden of SBE but were not within the scope of this study. This trial does not provide direct information regarding the value of varespladib administered without antivenom, for example, as a stand-alone or initial field therapy.