A scoping review of thresholds for responder and time‐to‐event analysis of patient‐reported outcomes in breast cancer trials

Le résumé fourni par la source

Patient-reported outcomes (PROs) are essential for assessing clinical benefit and tolerability from the patient's perspective. In oncology trials, PRO data analysis often involves responder and time-to-event analyses, which depend on predefined thresholds to determine clinically meaningful change. There is considerable variability in how these thresholds are selected, applied and reported. This scoping review with systematic methods aimed to explore how thresholds for responder and time-to-event analysis are defined, applied and reported in randomized controlled trials (RCTs) involving breast cancer patients with PRO-based endpoints. A systematic PubMed search was conducted to identify eligible RCTs in breast cancer patients published between 2020 and 2024 that reported PRO data analysed with responder or time-to-event analyses. Details on trial design, PRO measures, threshold selection and reporting were extracted. Of the 53 eligible RCTs identified in the search, about half (49.1%) were phase III trials. Only 47.2% of studies explicitly reported the use of thresholds. In 56.6%, thresholds were defined ad hoc without justification, 39.6% cited external sources, and 5.7% provided an explicit rationale. The most cited source for a threshold was an article by Osoba et al. (1998). Differentiating thresholds by PRO domain or by improvement/deterioration was uncommon. Reporting on threshold selection and its relevance to the trial population was very limited. There is significant variability and limited transparency in how PRO thresholds are used in breast cancer RCTs. Greater standardization in methodology and reporting is necessary to improve the interpretability and clinical usefulness of PRO analyses in oncology.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.