Customize your JAMA Network experience by selecting one or more topics from the list below.
Shepshelovich D, Tibau A, Molto C, et al. Assessment of Frequency and Reporting of Changes in Cancer Trial Design After Initiation of Patient Accrual. JAMA Oncol. 2019;5(1):107–109. doi:10.1001/jamaoncol.2018.5877
Changes in trial design after the onset of patient accrual can affect the scientific basis of evidence-based medicine and should be reported and justified.1 Data on the prevalence of these changes in oncology drug development are scarce. We aimed to quantify these changes in trials that supported cancer drug approval by the US Food and Drug Administration (FDA) and to explore the associations of such changes with study characteristics and drug-approval processes.
The FDA website2 was searched in April 2017 for trials supporting approval of new drugs for treatment of solid tumors from January 1, 2010, to December 31, 2016. ClinicalTrials.gov entries at trial initiation were compared with entries for the same trials effective on April 1, 2018, thereby identifying modifications in trial design. We then identified the publication of the registered studies and assessed the reporting of these changes. The association of changes in the trial design with trial characteristics and regulatory approval pathways were explored using logistic regression. Statistical analysis was performed using SPSS version 21 (IBM, Inc). Statistical significance was defined as 2-sided P < .05.
The trial characteristics and regulatory approval pathways are detailed in Table 1. Changes in ClinicalTrials.gov entries are shown in Table 2. Of 101 included trials, 56 (55.4%) had modifications in planned sample size, 34 (33.7%) in inclusion criteria, and 27 (26.7%) in primary outcome. Thirty-nine of 56 studies (69.6%) with changes in the preplanned sample size, 19 of 34 studies (55.9%) with changes in the inclusion criteria, and 10 of 27 studies (37.0%) with changes in the primary outcome definitions were reported in the matching publication. A breakthrough therapy designation was associated with more changes in inclusion criteria (odds ratio [OR], 2.73; 95% CI, 1.01-7.39; P = .048) and primary outcome definitions (OR, 3.95; 95% CI, 1.11-14.13; P = .03). Accelerated approval was associated with more changes in inclusion criteria (OR, 2.84; 95% CI, 1.12-7.21, P = .03) and sample size (OR 2.71; 95% CI, 1.00-7.34; P = .049). A single-arm trial design was associated with more changes in sample size (OR, 4.00; 95% CI, 1.33-12.50; P = .01). Tumor and drug types and other regulatory pathways were not associated with modifications in study design.
Changes in cancer trial design after the onset of patient enrollment are common and often unreported in publications. Previous research has shown that changes in outcome definitions and premature data analysis can influence the interpretation of trial results.3,4 Scientific validity may be compromised even when these changes are reported and justified transparently and more so when transparency is compromised and the rationale behind the changes and their timing are unknown. Most of the trials included in our cohort were published in influential journals, where readers might assume that trial methodology has undergone thorough peer review and the results can be relied on to support changes in clinical practice.
Breakthrough therapy designation and accelerated approval are FDA programs aimed at speeding up the availability of drugs for life-threatening conditions. These programs potentially allow the evidentiary standards that support marketing approval to be lowered.5 The greater frequency of modifications made to clinical trials supporting drug registration when regulatory review and approval processes are shortened may undermine credibility.
Our study has limitations. Because most journals do not make serial versions of protocols available publicly,6 we were not able to assess whether changes in ClinicalTrials.gov entries are accurate representations of modifications in trial protocol or corrections of clerical errors in previous registry entries, as in the case of changing ambiguous outcome definitions to specific ones. In addition, multivariable analyses of the factors associated with changes in ClinicalTrials.gov entries were not feasible owing to the small number of studies included in the analysis.
Among trials supporting FDA approval of cancer drugs, modifications in study design after patient accrual has begun are common, often unreported, and associated with breakthrough therapy designation, accelerated approval, and single-arm trials. Health care professionals, reviewers, journal editors, and regulators should demand more transparent justification for changes to the trial design after commencement of patient accrual.
Accepted for Publication: October 9, 2018.
Corresponding Author: Eitan Amir, MD, PhD, Division of Medical Oncology & Hematology, Princess Margaret Cancer Centre, 700 University Ave, Floor 7, Ste 721, Toronto, ON M5G 1Z5, Canada (email@example.com).
Published Online: December 6, 2018. doi:10.1001/jamaoncol.2018.5877
Author Contributions: Drs Shepshelovich and Amir had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Concept and design: Tibau, Ocana, Amir.
Acquisition, analysis, or interpretation of data: Shepshelovich, Tibau, Molto, Goldvaser, Šeruga, Amir.
Drafting of the manuscript: Shepshelovich.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: Shepshelovich, Molto, Amir.
Administrative, technical, or material support: Molto, Goldvaser, Amir.
Supervision: Tibau, Ocana, Amir.
Conflict of Interest Disclosures: Dr Ocana reported receiving personal fees from Daiichi Sankyo, Servier, and Entrechem outside the submitted work. Dr Amir reported receiving personal fees from Genentech/Roche, Myriad Genetics, Agendia, and Apobiologix outside the submitted work. No other disclosures were reported.
Create a personal account or sign in to: