The Importance of Publishing Negative Data
Publication of all data is critical to ensure the forward movement of research that is needed to create clinically meaningful improvement in patient outcomes.
Publication of study results is critical to ensure forward movement in the field of oncology. Data suggest, however, that the publication of negative data is lacking.
A commentary published in The Oncologist by Georgios Pongas, MD, of the National Cancer Institute in Bethesda, Maryland, and Tito Fojo, MD, of Columbia University Medical Center in New York, New York, lauded the timely and concise publication, also in The Oncologist, of the early termination of a phase 1b trial of the dual pan-class PI3K and mTOR inhibitor, BEZ235, for the treatment of renal cell carcinoma, due to high rates of grade 3-4 adverse events and marginal efficacy.1,2 Overall, 50% of patients experienced grade 3-4 adverse events and 50% discontinued the study due to toxicity. Of 5 evaluable patients, the best response was stable disease in 2 patients and progressive disease in 3.
Dr Fojo, also the section editor for Clinical Trials Results at The Oncologist, told Cancer Therapy Advisor that clinical trial results should be concise and direct. “Many articles are far more nuanced in their presentations. Others, unfortunately, place too much emphasis on ‘statistically significant' but marginally beneficial outcomes.”
Low Publication Rates of Negative Outcomes
Evidence suggests that negative or undesirable data are less likely to be published compared with positive data. A recent study demonstrated that 39% of abstracts of randomized and nonrandomized trials presented at the American Society of Clinical Oncology Annual Meetings between 2009 to 2011 were not published in peer-reviewed journals within 6 years.3 If any outcome was not significant, or if the authors were unable to determine significance due to lack of statistical analyses or defined outcomes, the publication rate was lower (75% and 61%, respectively) compared with the publication rate of studies with statistically significant findings (82%).
“An important problem is that not all statistically positive trials are really positive, or at least, not in the way we would want them to be, and not in a way that the results are meaningful to patients,” said Dr Fojo. Investigators often emphasize positive effects or rationalize negative outcomes.
RELATED: Final Analysis Demonstrates Some Benefit With Everolimus in Papillary mRCC
Possible Reasons for Low Publication Rates
There are several plausible reasons for why the publication of negative data or undesirable outcomes is lacking. “One knows that if the trial is negative, it will very likely not be published by a journal,” Dr Fojo said. The effort involved in publication or resubmission to a journal willing to publish a negative outcome becomes a burden.
Publication of trials requires the interest of multiple people, and data analyses require a substantial dedication of time and resources. The study by Dr Massey and colleagues found that participation in a study with at least 1 investigator with an industry affiliation or a cooperative group participation significantly increased the likelihood of publication.