Evidence To Support the Use of the Functional Assessment of Cancer Therapy – General Item GP5 (FACT-GP5) To Assess Comparative Tolerability Endpoint: Results From the LIBRETTO-531 Trial

OBJECTIVES: To generate psychometric evidence supporting the use of FACT-GP5 to measure tolerability and the categorization of “high side-effect bother” using FACT-GP5 rating of 3 or 4 in patients with advanced or metastatic medullary thyroid cancer.

METHODS: This pre-specified psychometric analysis of FACT-GP5 used blinded, pooled data from the LIBRETTO-531 (NCT04211337) safety population (N=290) (cutoff: 22May2023). Participants completed FACT-GP5 electronically at baseline (Cycle1,Day1) and then weekly post-baseline during on-treatment period. Intraclass correlation coefficients (ICC) were calculated for test-retest reliability using data ≤8 weeks post-baseline. Convergent validity was assessed using Spearman correlations of FACT-GP5 ratings with symptomatic adverse events (AEs) using the PRO version of Common Terminology Criteria for Adverse Events (PRO-CTCAE) and functioning scores of EORTC QLQ-C30. Known-group validity was tested between FACT-GP5 ratings and treatment discontinuation, and hospitalization. Ability to detect change over time was examined between FACT-GP5 ratings and worsening of symptomatic AEs using baseline-adjusted Cochrane-Mantel-Haenszel tests. The appropriateness of categorizing “high side-effect bother” using FACT-GP5 ratings of 3 or 4 were examined.

RESULTS: In this subgroup of participants with stable symptomatic AEs, ICCs ranged between 0.80-0.85 in weeks 5-8 post-baseline. Correlation coefficients between FACT-GP5 and a) PRO-CTCAE items ranged from 0.18 (vomiting) to 0.62 (fatigue); b) QLQ-C30 functioning scores ranged from -0.39 (emotional) to -0.50 (social). The findings were per priori hypotheses and supported convergent validity of FACT-GP5. Known-group validity was supported by an increase in proportion of participants who reported higher levels of side-effect bother at assessments closer to treatment discontinuation and hospitalization events. Post-baseline FACT-GP5 ratings showed significant association with worsening of symptomatic AEs (p<.001). Participants with FACT-GP5 ratings of 3 or 4 had poorer physical functioning than those who reported “low side-effect bother” (mean±SD: 60.2±20.8 vs. 84.7±16.2; p<.0001).

CONCLUSIONS: FACT-GP5 is fit-for-purpose to assess treatment tolerability and to categorize “high side-effect bother” in LIBRETTO-531.