Subtype-specific representation of thyroid cancers in contemporary basket clinical trials (2018–2024): A pooled analysis.

Abstract

e18021 Background: Basket clinical trials (BCTs) are designed to overcome histology-based drug development by enrolling patients according to shared molecular alterations. A pooled analysis by Hazim and Prasad (2018) focused on biomarker (BM)-driven BCTs and reported an under-representation of thyroid cancers (TCs) as a single entity. Since then, BCT methodology has evolved and multiple actionable alterations have been identified in TCs. We assessed the representation of TC subtypes in contemporary BCTs, contextualizing enrollment patterns against population-level cancer incidence. Methods: We conducted a systematic review of BCTs published between April 2018 and December 2024. BCTs were categorized according to prior BM selection (none BM, single BM, multiple BM). Cancer-specific crude incidence rates per 100,000 persons in 2022 were derived from the Global Cancer Observatory across the EU and US for both sexes aged >20 years. Association between incidence and BCT enrollment was assessed using semi-log and log-log linear models. Incidence estimates for TC subtypes were literature-derived. An observed-to-expected (O/E) enrollment ratio (ER) was calculated for anaplastic, differentiated, and medullary TCs (ATC/DTC/MTC), comparing observed BCT accrual with expected numbers based on subtype-specific incidence-weighted counts. Results: Among 137 screened publications, 26 BCTs enrolling 2,448 patients in 372 centers (58.6% US, 27.3% EU) met inclusion criteria. Across 25 cancer types, incidence was associated with BCT enrollment in both semi-log (β ≈ 69.7, 95%CI 26.8–112.5, p=0.003, R² ≈ 0.34) and log–log models (β ≈ 0.39, 95%CI 0.08–0.69, p=0.015, R² ≈ 0.24). Overall, 83 patients with TC (3.4%) were included: 57 ATC, 12 DTC, 3 MTC, 11 unspecified. TC inclusion was concentrated in 7 BCTs (26.9%): 97.2% of ATC patients were enrolled in 2 BCTs with BRAF inhibitors or immune checkpoint inhibitors. MTC patients were mainly enrolled in a RET-targeted BCT, and DTC patients in a PD-L1 inhibitor trial. TC macro-category was well represented, lying above the regression line in both models. However, O/E analyses revealed marked subtype-specific heterogeneity, with over-representation of ATC (ER 133.7), and under-representation of MTC (ER 9.6) and DTC (ER 0.15), indicating a strong deviation from incidence-driven representation. Conclusions: Overall TCs inclusion in BCTs has increased since 2018, driven by few trials on a restricted number of targets or ICI. TC’s apparent adequate representation in contemporary BCTs reflects aggregation effects that mask clinically relevant subtype-specific disparities. These findings highlight structural barriers in translating molecular actionability into equitable access to novel therapies for the majority of patients with advanced TCs, and support the need for integrated precision oncology pathways and centralized referral strategies.