Industry Sponsorship Bias in Collagenase Clinical Trials for Dupuytren Disease: A Meta-analysis.

Abstract

<h4>Background</h4>Collagenase clostridium histolyticum (collagenase) was introduced in 2010 creating a nonoperative treatment option for Dupuytren disease with promising results in sponsored clinical trials. A meta-analysis was performed to investigate industry sponsorship bias.<h4>Methods</h4>A systematic review of collagenase treatment of Dupuytren contracture was conducted. Articles containing mesh terms including "microbial collagenase" and "Dupuytren's contracture" were searched and limited to only clinical trials with similar protocols for inclusion. Meta-analysis of treatment endpoints of correction of contracture to 0-5 degrees after first and last injection was conducted comparing sponsored versus nonsponsored studies.<h4>Results</h4>Sixteen of the 29 identified articles met criteria for inclusion. Nonsponsored studies reported a significantly higher rate of meeting the primary treatment endpoint compared to sponsored studies after single injection for all joints (69.6% vs 56% P < 0.01), metacarpophalangeal joint (96% vs 64% P < 0.01), and proximal interphalangeal joint (67% vs 36% P = 0.011). The correction in contracture rates was similar between groups with studies evaluating more than one injection.<h4>Conclusions</h4>Nonsponsored studies published higher success rates in meeting the primary endpoint of full correction after single injection than sponsored studies; however, similar results with multiple injections. This study demonstrated that sponsored studies of collagenase produced highly powered studies that may be reliably depended on for evidence-based clinical application.