The risks and benefits of biopsies done for research need to be better represented in study protocols and in informed consents, concluded investigators who recently explored the matter.
Research biopsies are increasingly being incorporated into clinical trials, acknowledged Michael J. Overman, MD, of the Department of Gastrointestinal Medical Oncology at The University of Texas MD Anderson Cancer Center in Houston, and colleagues, in their report for Journal of Clinical Oncology. However, they noted, limited information is available regarding how study protocols and informed consents integrate and describe the use of biopsies.
For their own study, Overman’s group used an interventional radiology database to identify all therapeutic clinical trials from January 2005 until October 2010 in which image-guided research biopsies were performed. The investigators analyzed data from study protocols and informed consents, and recorded procedural complications.
A total of 745 research biopsies were performed on 576 patients throughout the 57 clinical trials identified. The majority of trials (67%) contained at least one mandatory biopsy from participants. Analysis of the research biopsy tumor tissue was a study endpoint in nearly all of the trials (95%). The primary indication for a research biopsy was for integral biomarker analysis in 32% of the trials and for correlative science in 68%.
A statistical justification for the research biopsy sample size was present in 50% of the studies with mandatory biopsies. A statistical analytic plan for the correlative science research biopsy tumor tissue was mentioned in 26% of the trials, described as exploratory in 51%, and not mentioned in 23%.
Overall and major complication rates were 5.2% (39 of 745 biopsies) and 0.8% (6 of 745 biopsies), respectively. Complication rates for intrathoracic and abdominal/pelvic solid organ biopsies were 17.1% (36 of 211 biopsies) and 1.6% (3 of 189 biopsies), respectively. Site-stratified research biopsy–related risks were discussed in five consents.