FDA Officials Look to Tackle PRO Bias in Cancer Trials
Thanks to the 21st Century Cures Act, the US Food and Drug Administration (FDA) is expecting to see more patient-reported outcome (PRO) assessments, and in light of that anticipated growth, three officials in FDA’s Center for Drug Evaluation and Research are explaining the ways in which open-label bias may arise and how to address some of the challenges.A potential effect of PRO bias noted in FDA’s guidance is overly optimistic reports of outcomes among patients in the investigational arm and overly pessimistic reports for patients in the control arm. Indeed, the authors say that a frequent concern with PROs in open-label cancer trials is that a patient’s perception of their symptoms or function may be influenced by knowledge of their assigned treatment, they write in the JAMA Oncology Viewpoint.
They also cite a meta-epidemiological analysis including over 1000 trials across several therapeutic areas that found treatment effects for “subjective” outcomes, including clinician-reported outcomes and PROs, were larger in open-label trials than in blinded trials, suggesting bias. But, the officials note, there are “surprisingly few data to suggest open-label bias regarding PRO measures in cancer clinical trials.”
And there are ways researchers can combat bias in some circumstances.
“If researchers are concerned that completion and/or dropout bias may be present in their study, the extent of this bias can be explored by comprehensive reporting of PRO completion, patient enrollment between arms throughout the trial, and careful sensitivity analyses that test the robustness of the PRO findings,” they write.
The officials also point to other ways open-label trial designs may bias PROs and how to explore bias further.
And FDA’s Oncology Center of Excellence (OCE) is beginning to investigate these issues based on the treatment arm of the study, they note. OCE is looking into comparing change from baseline of PRO results in matched investigational and control arms where the same treatment was evaluated for the same cancer in both double-blind and open-label trials.
Safety data is also a concern, they add. “For instance, the same forces that may lead patients to report fewer or milder toxic effects on a PRO measure when they are aware of their assignment to the investigational arm could lead them to underreport adverse effects to the physician in the clinic visits that generate clinician-reported safety data.”
Moving forward, the FDA officials call for more targeted research on the existence and magnitude of biases.
JAMA Oncology