Patient-reported outcomes corresponding to most common symptomatic adverse events in lung cancer clinical trials.

Abstract

11103 Background: FDA recommends assessing the most relevant expected symptomatic adverse events (AEs) with patient-reported outcomes (PROs) in cancer clinical trials (CTs) to complement traditional safety data. To inform selection of symptomatic AEs, we previously identified the most common (i.e., all-grade, clinician-reported in ≥20% of patients) symptomatic AEs from FDA-approved lung cancer product approvals from 2015 through 2021 based on CT monotherapy experimental arm safety data from US Prescribing Information (USPI). The present research aims to assess whether these CTs included PROs which adequately captured common symptomatic AEs. Methods: We assessed 30 CTs that supported lung cancer approvals from 2015 through 2021. We reviewed the study protocol, statistical analysis plan, and/or clinical study report to determine the PRO measures included. We matched symptoms included within PRO measures and items to the most common clinician-reported symptomatic AEs per the USPI to determine the extent to which the PROs and symptomatic AEs corresponded. Results: Twenty-three out of 30 (77%) CTs supporting non-small cell lung cancer (NSCLC) approvals included PRO(s). Nineteen included more than one PRO measure. The most frequently used PRO measures were the European Organisation for Research and Treatment of Cancer-Core Questionnaire (EORTC QLQ-C30; n=19), EORTC-Lung Cancer (EORTC QLQ-L13; n=17), EuroQoL 5-Dimension (EQ-5D; n=12), and selected items from the Patient-reported Outcomes version of the Common Terminology Criteria for Adverse Events item library (PRO-CTCAE; n=4). Nearly all (20 of 23) CTs included PRO items corresponding to more than half of the common symptomatic AEs reported. Symptomatic AEs for types of pain corresponded to general pain PRO items (e.g., “musculoskeletal pain” matched to EORTC item for “pain”). However, PRO items for specific types of pain were not often used. While important symptomatic AEs such as fatigue, diarrhea, nausea, and decreased appetite were often assessed by PROs, others including rash and edema were not usually captured. Ultimately, only 4 of 23 (17%) CTs with PROs included items that assessed all symptomatic AEs that occurred in ≥20% of patients. Of the 4 CTs which included an item library, 3 (75%) comprehensively assessed patient-reported symptomatic AEs, compared to 3 of 19 (16%) CTs that did not employ an item library for patient-reported symptom collection. Conclusions: Based on USPI safety data and PRO measures from CT monotherapy experimental arms of recently approved NSCLC cancer drugs, we determined that while the majority of CTs include PRO items for some relevant symptomatic AEs, other important expected symptomatic AEs (e.g., edema, rash, specific types of pain) were not measured. Future NSCLC CTs should include PRO items for common expected symptomatic AEs which is feasible using well-established PRO item libraries.