Under-Reporting of Patient Reported Outcome (PRO) in Clinical Trials

Abstract

Introduction: Clinically relevant outcomes as improvements in overall survival (OS) or quality of life (QOL) should guide decision-making. The FDA encourages the implementation of patient-centric PRO measures in clinical trials. In the recent decade there have been a growing number of protocols including PROs in their outcome measures, and an increase in pre-market submissions including those measures. We aimed to evaluate the frequency at which PRO measures, incorporated into clinical trials, are made publicly available, when trial results are published.Methods: We searched Citeline® Trialtrove database, a registry of clinical trials, for randomized phase 2/3 and 3 clinical trials including patients with chronic lymphocytic leukemia (CLL) and non-Hodgkin lymphoma (NHL), initiated between the years 2005-2015, with PRO endpoints listed. We excluded trials involving patients with acute leukemia or involving stem cell transplantation. For these trials we recorded the following data: indication, treatment and comparator, phase, primary endpoint, type of scale or questionnaire used for PRO endpoint. We then identified all available publications associated with the trial, and recoded type of publication (abstract or full text), year, number of randomized patients, reported outcomes.Results: 120 CLL and NHL trials were identified through our search. 44 trials (23 CLL, 21 NHL) were listed with at least one PRO endpoint. Eleven trials were excluded (1 extension of an included trial, 4 not completed, 4 not published, 2 cancelled with little or no recruitment). PRO endpoints were assessed utilizing - EORTC-QLQ30 QoL, QLQ-CLL16, FACIT-fatigue, FACT-leu, FACT-lym, EQ-5D, Skindex-29, and Rituximab Administration Satisfaction Questionnaire as relevant per indication. PRO was the primary outcome in two trials, and secondary in the other 31. Study sponsor was industry-only in 27 trials, and industry-academic in five. Of the 33 trials analyzed, 14 (42%) published results of the PRO endpoint - 6/17 (35%) of CLL trials, and 8/16 (50%) of NHL trials. 26/33 of the trials were published as full-text, 6/33 as abstract, and 1 published on clinicaltrials.gov. Of the 26 full text publications, 12 (46%) reported PRO. In 8 of full-text PRO were reported as part of the primary publication or published within 6 months. Nine of all trials provided a comprehensive report of PRO endpoints. Of the 33 trials, only 2 provided information about missing PRO data. 54% of trials in which primary endpoint was reached, and 12.5% of trials in which it was not reached, were published listing PRO results. Of the trials that showed PFS benefit, with no OS benefit, 9/15 (60%) published the results of the PRO endpoints.Conclusions: Despite a growing emphasis on QOL and use of PROs in oncology clinical trials, and despite patient and health provider efforts to record PRO data, most CLL and NHL randomized trials still do not report the PRO endpoints. In several cases they were published later and in a partial manner, minimizing their impact on treatment decisions. This may indicate a disregard to PRO data collected, incomplete collection or methodological flaws in data analysis. Regardless of the reason, PRO data should be routinely included in study publications to allow for a complete assessment of investigational treatment outcome - including disease related outcomes, as well as those reported by the patients themselves. [Display omitted] DisclosuresNo relevant conflicts of interest to declare.