Early phase trials: the perils of nonpublication

In a recent article published in The Oncologist, a group of Phase I trialists dissect the problem of publication (or lack thereof) for early phase clinical trials.1 A well-known problem in oncology drug development is that clinical trials are often not published if they were not completed, did not meet their endpoints, or were overtaken by events in the field. Lucassen et al. used a qualitative semi-structured in-depth interview process to evaluate the barriers to publishing early phase clinical trials. Investigators defined meaningful themes that were then distilled, reported, and discussed in the paper. A decade ago, we reviewed trials reported at ASCO for their final publication in the medical literature. Reviewing 1075 abstracts describing 378 randomized and 697 nonrandomized clinical trials reported between 2009 and 2011, we found that 39% of oncology clinical trials remained unpublished; notably, 25% for randomized trials.2 Several previous studies had noted similar results. In response to this concern, The Oncologist launched a special section in 2012, Clinical Trial Results (CTR), with the goal to provide a simple, straightforward venue in which to publish clinical trials that were considered “difficult to publish,” as well as complete successful clinical trials.3 It was designed to be a templated method of publishing that would allow entry of key facts, and not require the generation of a full manuscript. An extended abstract would appear in print, with the remaining study details found online. Since its launch in 2012, the section has published over 325 CTRs, many trials that could not complete accrual and were discontinued early. But in every trial, there were lessons learned and lessons discussed. These publications meant—most of all—that every patient’s clinical trial enrollment would “count.”3,4 Patients enroll on clinical trials in part with the hope that they will individually benefit and in part with the hope that others will benefit. Indeed, the altruistic motive behind clinical trial enrollment has been well documented.5,6 Thus, the CTR section has offered investigators the opportunity to meet the goals highlighted by Lucassen et al., that there is an ethical and moral responsibility to publish, that there should be no loss of knowledge, and that the resources spent on the trial should not be wasted (See Figure 1 in Lucassen et al1). Most of all, CTRs afford an opportunity to ensure that the altruistic motives of patients participating in clinical trials are honored and that trial data are indeed shared in the hope of helping others. In their paper, the trialists discuss barriers to publication. These include the cost of preparing a full manuscript after the Phase 1 study ends; the problem of multiple clinical trial sites enrolling only a handful of patients each such that no one investigator is deeply invested in completing the report; and that the sponsor’s study team may be rapidly disbanded once the study ends. This raises the question as to whether publication should be considered as a required aspect of proper study close-out. Phase I trials are particularly susceptible to the lack of reporting if an early efficacy signal is not apparent. With further study, an agent may meet with success but the initial Phase I dose-finding study may not be the study that generates the enthusiasm that will later take an agent into registration. The trialists’ recommendations include strategies to simplify the publication process, as well as increasing the regulatory requirement to publish. Their recommendations should be carefully considered. The trialists note that the desire for high-impact papers often lead to the rejection of early phase studies, especially. One interesting recommendation is that all journals should set aside a certain space for publishing these Phase 1 and 2 trials. In publishing the CTRs, The Oncologist implemented this over a decade ago. The trialists are right that negative early phase trial reports will be unlikely to raise a journal’s impact factor and that innovative thinking around this is needed. Perhaps clinical trials of any type should not be counted in the impact factor. That would neutralize both positive and negative influences on the impact factor and underlying biases.7 This would also end the inherent conflict of interest that exists in which journals prefer to be venues for publishing income-producing Phase III trials, rather than early phase studies. Another option would be to deposit clinical trials in medRxiv or similar archive—medRxiv being a preprint archive of clinical trials with over 1675 oncology papers at this time. Comments are allowed and posted, such that the oncologic community can effectively provide peer review on the research in a public forum. These papers can still be published in a journal with peer review. Physicists began using this type of prepublishing format, in arXiv.org, with the archive reaching 2.4 million scholarly articles to date. The ClinicalTrials.gov result posting was meant to provide a publicly available dataset but doesn’t include the explanation and discussion found in a proper manuscript. The authors of “Barriers to publishing early phase clinical trials: The oncologists’ perspective” are to be applauded for bringing the problem of publication to light again. Perhaps the time is right to identify innovative solutions for the future. In our 2016 analysis of the subject, we noted that similar findings from ASCO annual abstracts had previously appeared in the literature, and with regularity, in 1992, 2003, 2005, 2008, 2009, and 2011.8-13 With greater attention, this may be changing; a Swiss cooperative group analyzing 261 of their own trials, reported a 95.8% publication rate, including for the 29% of trials that were closed prematurely.14 However, this seems aspirational for most. Recent reports include one for pancreatic cancer trials noting 54% “published” and 70% with “available results”, the latter referring to both publication and ClinicalTrials.gov postings15; another in GI cancers (77% for negative studies16); and in GYN cancers (55.5% in 5 years17), suggesting that this remains a gap still in need of closing. Adrian Sacher (Writing—original draft; review and editing). Susan E. Bates (Writing—original draft; review and editing). The authors indicated no financial relationships. Adrian Sacher is the Section Editor for Clinical Trial Results in The Oncologist; Susan E. Bates is the Editor-in-Chief of The Oncologist.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.