Impact of Open-Label Design on Patient-Reported Outcomes (PROs) Data in Randomized Clinical Trials of Immuno-Oncology (IO) Agents in Patients with Advanced or Metastatic Cancer: A 10-year systematic literature review (SLR)

Abstract

Abstract Background The unblinded nature of open-label clinical trials may bias patient assessments of treatment efficacy and PRO measures. In this context, we conducted a SLR of IO trials across indications, to assess and quantify the potential impact of open-label vs. double-blind designs on treatment effects of PROs. Methods Independent reviewers conducted a systematic search of indexed literature published from January 2009 to May 2019, using PubMed/MEDLINE, Cochrane Library, EMBASE databases and manual search. A detailed search algorithm identified all randomized clinical trials of IO therapies focused on advanced or metastatic cancer patients reporting PRO data. We conducted descriptive analyses quantifying differences at baseline and over time according to the type of studies (open vs. blinded) in terms of PRO questionnaire completion rates and on obtained PRO measures. We determined frequencies, percentages and 95% confidence intervals of the estimated differences when comparing open-label vs. blinded trials. Complementary analyses also assessed the concordance between PRO results and efficacy. Results Overall, we identified 8,285 references. After duplicates were removed, 5,888 papers were screened and finally 28 papers were identified according to our predefined criteria, corresponding to a total of 24 studies. Among them, 15 (62.5%) were open-label studies. The main cancer sites were melanoma (N = 10, 41.7%) and non-small cell lung cancer (N = 7, 29.2%). The principal IO-drugs investigated, alone or as part of a combination therapy regimen, were nivolumab (N = 12, 50.0%), ipilimumab (N = 8, 33.3%), and pembrolizumab (N = 5, 20.8%). Descriptive analyses are ongoing and will be presented at the conference. Conclusion To our knowledge, this SLR is the first one to focus on the impact of open-label design on PROs in randomized clinical trials of IO agents using a large search algorithm in the 3 most relevant scientific literature databases. Future research should use patient-level data to investigate potential differences in PROs between open-label and blinded trials. Legal entity responsible for the study Bristol-Myers Squibb. Funding Bristol-Myers Squibb. Disclosure A. Anota: Honoraria (self), Research grant / Funding (institution), Travel / Accommodation / Expenses: Bristol-Myers Squibb. C. Lefevre: Full / Part-time employment: Bristol-Myers Squibb. H. Lemasson: Full / Part-time employment: Bristol-Myers Squibb. F-E. Cotte: Full / Part-time employment: Bristol-Myers Squibb. G. Mouillet: Honoraria (self): Bristol-Myers Squibb. All other authors have declared no conflicts of interest.