Integrating Molecular Classification Based on FGFR3 , KRAS , HRAS , and TP53 Mutation With TNM Staging Modestly Improves Prediction of Prognosis of Upper Tract Urothelial Carcinoma After Surgery

PURPOSE: To evaluate whether a simplified mutation-based classification using archival formalin-fixed paraffin-embedded (FFPE) specimens provides prognostic information beyond pathological stage in patients with upper tract urothelial carcinoma (UTUC) undergoing nephroureterectomy. MATERIALS AND METHODS: We retrospectively identified 107 UTUC patients who underwent nephroureterectomy between 2005 and 2020. DNA was extracted from FFPE specimens, and representative exons of FGFR3, KRAS, HRAS, and TP53 were analyzed by PCR followed by Sanger sequencing. Patients were classified into TP53, RAS, FGFR3, and triple-negative groups according to mutation status. Non-urothelial-tract recurrence-free survival (nutRFS) and overall survival (OS) were evaluated. To assess the incremental prognostic value of mutation-based classification beyond pathological information, parsimonious Cox models including pT, pN, and nuclear grade were constructed, and model discrimination was assessed using the concordance index. RESULTS: Mutation status could not be fully determined in 17 patients; thus, 90 patients comprised the final cohort. Mutations in TP53, KRAS, HRAS, and FGFR3 were identified in 20 (22%), 4 (4%), 3 (3%), and 32 (36%) patients, respectively. In general, alterations in TP53, KRAS/HRAS, and FGFR3 were mutually exclusive. In exploratory multivariable analyses, the FGFR3 group showed more favorable nutRFS and OS than the TP53 group. Addition of mutation-based classification to a model including pT, pN, and nuclear grade modestly improved discrimination for nutRFS (c-index, 0.782 to 0.804) and OS (0.726 to 0.754). CONCLUSIONS: Although the improvement in discrimination was modest, a simplified mutation-based classification using archival FFPE specimens may provide additional prognostic information beyond pathological stage and contribute to postoperative risk stratification in UTUC.