Can We Find the Positive in Negative Clinical Trials?

The management of cancer patients is most often underpinned by the highest level of evidence, namely, the randomized clinical trial (RCT). The RCT was an evolutionary step from the anecdotes and experiences of individual practitioners, introducing the application of sound scientific methods to inform the practice of oncology. Since the mid-twentieth century, results from RCTs have become the standard by which oncology is practiced and by which new oncologic agents are approved for use in patients (1). The development and conduct of a well-designed RCT requires enormous effort, time, and expense on the part of many individuals with varied areas of expertise, including clinical scientists, statisticians, and individuals knowledgeable about pharmaceuticals, clinical trials, and regulatory management, along with the many thousands of patients who entrust their care and safety to those clinicians involved with the trial conduct. The National Clinical Trials Network (NCTN) is funded by the National Cancer Institute (NCI) and represents the largest federally funded clinical trials group in the United States that is focused on the development of new therapeutic agents and strategies for cancer patients. NCTN trials encompass patients of all ages and virtually every cancer type. At any given time, the NCTN is responsible for a clinical trials portfolio of approximately 200 actively accruing studies throughout the United States and Canada. Each of these investigations was painstakingly vetted through a series of rigorous scientific reviews by experts in the NCTN, industry partners, the US Food and Drug Administration, and the NCI to ensure that the trial was based on solid preclinical and clinical evidence that supports the concept under consideration, the relevancy of the clinical question, and the soundness of the trial design. Final approval for the conduct of every NCTN trial ensures that each trial represents the true cutting edge of clinical oncology. The completion and ultimate publication of these trials are absolutely critical, regardless of whether the outcome of the trial was positive or negative, in showing the benefit or lack thereof for a new agent or strategy. Trials that show a positive benefit are met with the greatest enthusiasm by investigators, stakeholders, journals, and patients, because they improve outcomes and change the standard of care for all patients who share the clinical and pathologic characteristics of those patients who were trial participants. However, a well-designed and fully executed trial with negative findings also represents a critically important investigation that deserves scientific scrutiny in order to understand why the concept did not meet its expected endpoints. Publication is essential so that the oncologic community can further its collective knowledge concerning the agent or strategy, along with the biology of the cancer under investigation, and therefore avoid a similar mistake in the future for the same or a different cancer. As is true in any other field of endeavor, the oncology community must also evolve its reasoning to account for the ever-growing and more complicated knowledge base that embodies the field of oncology. Such evolution can only be accomplished by embracing a thorough recognition and understanding of all clinical trial results, both positive and negative. It is often the case that journals, particularly high-impact journals, as well as investigators are reluctant to publish clinical trials with negative outcomes, because these are generally not expected to garner the attention and citations typically associated with a positive study. Because the more prominent journals are more often read and referenced, the reluctance to publish negative trials creates an enormous bias in the literature—negative trials may either never be published or may be published in journals that are far less prominent (2). Publication bias is harmful to the scientific process and knowledge acquisition, because both positive and negative results must be available to refine and evolve scientific theories and hypotheses. Failure to publish negative results from clinical trials devalues the enormous effort contributed by the many members of the investigative team and violates the trust that was shared by participating patients. Because of the recognized importance of well-designed, well-conducted, and fully executed clinical trials, the JNCI and JNCI Cancer Spectrum have committed to review any and all such trials regardless of their outcomes. To this end, we strongly encourage NCTN investigators and other leaders of well-conducted trials, with both positive and negative results, to consider the JNCI journals for their publication needs. Department of Medicine/Oncology, University of Florida Health, Gainesville, FL (CJA); Department of Medicine/Division of Medical Oncology and Hematology, Division of Clinical Epidemiology, Lunenfeld Tanenbaum Research Institute at Mount Sinai Hospital, University of Toronto, Toronto, Canada (PJG); Department of Health Policy and Management and Department of Medicine/Hematology-Oncology, UCLA Jonsson Comprehensive Cancer Center, UCLA Fielding School of Public Health, David Geffen School of Medicine at UCLA, Los Angeles, CA (PAG). CJA is Deputy Editor of JNCI. He joined the NCI Clinical Investigations Branch, Cancer Therapy Evaluation Program, Division of Cancer Treatment & Diagnosis in April 2016 to provide expertise and oversight of the gastrointestinal cancers clinical trials portfolio under an interagency agreement between NCI and the University of Florida Health where he is a professor in oncology. PJG is Editor-in-Chief of JNCI Cancer Spectrum. She is an active member of the Canadian Clinical Trials Groups. PAG is Editor-in-Chief of the JNCI. She has been a clinical trials investigator for almost 30 years at SWOG, National Surgical Breast and Bowel Project, and currently NRG Oncology. The authors have no conflicts related to this editorial to declare.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.