Sample‐Related Factors Associated With FGFR Testing Validity in Urothelial Cancer: A Japanese THOR Post Hoc Analysis

Timely access to valid fibroblast growth factor receptor (FGFR) testing is essential for treatment selection in metastatic urothelial carcinoma; however, invalid results remain a barrier to timely treatment selection. We conducted a post hoc analysis of the Japanese subgroup of the Phase 3 THOR trial to describe FGFR testing validity across sample‐related factors and explore its predictors. Among 612 patients with FGFR testing records, 563 (92.0%) had valid results, and FGFR alterations were detected in 82 of 563 (14.6%). The proportion of valid results tended to be lower in biopsy‐derived samples than in surgical‐resection samples (76.9%–83.3% vs. 92.3%–97.7%). Samples that provided invalid testing results had smaller median total tissue surface area (16.60 vs. 388.85 mm 2 ) and target tumor surface area (2.00 vs. 85.10 mm 2 ). In contrast, for upper tract urothelial carcinoma biopsies, even valid samples had small median total tissue and target tumor surface areas (22.60 and 7.55 mm 2 ), underscoring that specimen source and collection method must be accounted for when interpreting surface area. In multivariable logistic regression, larger target tumor surface area was associated with higher test validity (adjusted odds ratio 1.10 per 10 mm 2 increase; 95% confidence interval 1.05–1.16), whereas neither archival sample age nor any of the defined fixation‐duration categories was associated with validity. Most invalid results reflected insufficient tumor cells rather than RNA quality. Specimen source, collection method, and target tumor surface area should therefore be considered together when selecting samples to improve FGFR testing validity, leading to greater access to targeted therapy. Trial Registration: ClinicalTrials.gov identifier NCT03390504