Abstract
Background The scientific and ethical justification for early phase clinical trials relies on evidence produced in preclinical experiments. Numerous studies have raised concerns about the rigor and reproducibility of such experiments. Little is known about how clinician investigators navigate this contested terrain. Methods We conducted semi-structured interviews with 17 clinician investigators involved in early phase neuro-oncology research. Interviews explored their approaches to evaluating supporting evidence during trial development. Data were analyzed using a systematic process based in grounded theory. Results Interviews surfaced three themes. First, clinician investigators framed trials as extensions of clinical care. They viewed trials as ways to address the lack of therapeutic alternatives and uphold their role as caregivers. Second, clinician investigators approached review of supporting evidence with an awareness of common evidential vulnerabilities. Assessments relied on heuristics, like the existence of plausible molecular mechanisms or replication. Third, clinician investigators maintained equanimity about trialing in the face of doubts about supporting evidence and unfavorable prospect of therapeutic success. This equanimity was secured by an understanding of themselves as caregivers and by a concern about missing important clinical advances. Conclusion Concerns about rigor and reproducibility for preclinical research were understood as tolerable imperfections when trials were supported by plausible claims relating to measured effective sizes and molecular mechanisms. Evidential thresholds for trial initiation were intertwined with clinical objectives of providing care in the face of severe disease. Clinician investigators thus shielded trials from the contestation arising from reproducibility. Ultimately, doubts about preclinical evidence were approached with an equanimity that seldom justified slowing trial efforts. Trial Registration Not applicable
Full Text
The Full Text of this preprint is available as a PDF (256.1 KB). The Web version will be available soon.