Assessing tolerability with the Functional Assessment of Cancer Therapy item GP5: psychometric evidence from LIBRETTO-531, a phase 3 trial of selpercatinib in medullary thyroid cancer

This psychometric analysis generated evidence to support the use of the Functional Assessment of Cancer Therapy item GP5 (GP5) as a measure of tolerability and confirms the appropriateness of categorizing “high side-effect burden” using a rating of 3 or 4 (score ranges 0–4) in patients with advanced/metastatic RET-mutant medullary thyroid cancer (MTC).

Blinded, pooled interim data from the safety population (n=290) enrolled in the phase 3 LIBRETTO-531 trial (NCT04211337) were used. Intraclass correlation coefficients (ICC) were calculated for test-retest reliability using data from cycles 1-2 post-baseline. Construct validity was evaluated by examining the correlations of GP5 ratings with (a) symptomatic adverse events (AEs; measured by the PRO-CTCAE), and (b) functioning scores of EORTC QLQ-C30. The ability to detect change over time was examined by Cochrane-Mantel-Haenszel tests for GP5 ratings and PRO-CTCAE. The relationship of “high side-effect burden” categories with QLQ-C30 functioning scores was examined.

ICCs for the GP5 ratings after cycle 1 ranged between 0.80 and 0.85, indicating good reliability. Correlations between GP5 and PRO-CTCAE items ranged from 0.18 to 0.62 and ranged from -0.37 to -0.50 for QLQ-C30 functioning scores, consistent with study assumptions. Post-baseline GP5 ratings showed significant associations with PRO-CTCAE scores (p<0.001). Participants with GP5 ratings of 3 or 4 had worse physical function than those with GP5 ratings of 0 to 2 (p<0.0001).

This analysis generated evidence supportive of the psychometric properties of the GP5 as a fit-for-purpose measure to assess treatment tolerability in patients with advanced/metastatic MTC. The definition of “high side-effect burden” was associated with the clinical feature of tolerability.