Despite statistically negative results, a large proportion of presenters of phase 3 trials at major oncology conferences provide positive conclusions or emphasize positive subgroups or secondary endpoints, according to results of a research letter published in JAMA Oncology.
“When the primary endpoint is not met, the word negative should be explicitly used,” the authors said.
“Oral presentations are not subject to peer review, and some authors’ conclusions may not be completely justified by the results,” the authors wrote.
Results from phase 3 clinical trials presented at oncology conferences are widely reported by the media, and resonate with the scientific community. The goal of this study was to determine the frequency of not-negative conclusions given by presenters of formally negative trials.
The study included 208 randomized, phase 3 clinical trials presented as oral abstracts at the American Society for Clinical Oncology’s Annual Meeting and the European Society for Medical Oncology’s Congress held in 2017 through 2019. Trials with a prespecified noninferiority design were excluded. The authors classified the trials as either positive (ie, not-negative) or negative based on the rejection of the null hypothesis for the primary endpoint. Conclusions were also classified as positive or negative based on the conclusions section of the oral presentation.
Formally negative trials that were given positive conclusions were then classified according to the reason why the conclusion was considered positive, including 1) numerically higher, but statistically nonsignificant, outcome in the experimental group; 2) emphasis on a positive subgroup; 3) emphasis on a positive secondary endpoint; and 4) noninferiority conclusion despite a superiority design.
The trials were formally negative in 44% of cases. Of these, 29% were given a positive conclusion. This varied by year of the conference; in 2017, 22% of formally negative trials were given a positive conclusion, 13% in 2018, and 47% in 2019. Positive conclusions were given for formally negative nonprofit studies in 30% of cases and for 26% of for-profit studies.
For the formally negative trials, a numerically higher outcome that was statistically nonsignificant was emphasized in 50% of cases, a positive subgroup was highlighted in 46% of cases, and a positive secondary endpoint result was highlighted in 38% of cases.
“Positive subgroup analysis of negative trials can be misleading, should not support treatment adoption, and, at best, should be hypothesis generating,” the authors wrote.
However, the authors did note that there were instances in which “methodological strengths and weaknesses of the subgroup analysis were correctly discussed by the study presenter.”
A post hoc noninferiority conclusion was made in 27% of the formally negative trials. “The noninferiority hypothesis should be prospectively planned, with a clear definition of the margin acceptable to define noninferiority,” the authors said.
“We showed that the risk of wrongly emphasizing borderline significance, already described in the oncology literature, is present also in meeting presentations,” the authors wrote.
The authors highlighted that, “the primary endpoint is the measure used for the study hypothesis and should condition study interpretation.” They noted that the role of the discussant during oral abstract sessions is critical to ensure that misinterpretation is not missed.
Di Maio M, Audisio M, Cardone C, et al. The use of not-negative conclusions to describe results of formally negative trials presented at oncology meetings [published online April 16, 2020]. JAMA Oncol. doi: 10.1001/jamaoncol.2020.0475