CDKN2A homozygous deletion is associated with muscle invasion in FGFR3-mutated urothelial bladder carcinoma.

The gene cyclin-dependent kinase inhibitor 2A (CDKN2A) is frequently inactivated by deletion in bladder carcinoma. However, its role in bladder tumourigenesis remains unclear. We investigated the role of CDKN2A deletion in urothelial carcinogenesis, as a function of FGFR3 mutation status, a marker for one of the two pathways of bladder tumour progression, the Ta pathway. We studied 288 bladder carcinomas: 177 non-muscle-invasive (123 Ta, 54 T1) and 111 muscle-invasive (T2-4) tumours. CDKN2A copy number was determined by multiplex ligation-dependent probe amplification, and FGFR3 mutations by SNaPshot analysis. FGFR3 mutation was detected in 124 tumours (43.1%) and CDKN2A homozygous deletion in 56 tumours (19.4%). CDKN2A homozygous deletion was significantly more frequent in FGFR3-mutated tumours than in wild-type FGFR3 tumours (p = 0.0015). This event was associated with muscle-invasive tumours within the FGFR3-mutated subgroup (p < 0.0001) but not in wild-type FGFR3 tumours. Similar findings were obtained for an independent series of 101 bladder carcinomas. The impact of CDKN2A deletions on recurrence-free and progression-free survival was then analysed in 89 patients with non-muscle-invasive FGFR3-mutated tumours. Kaplan-Meier survival analysis showed that CDKN2A losses (hemizygous and homozygous) were associated with progression (p = 0.0002), but not with recurrence, in these tumours. Multivariate Cox regression analysis identified CDKN2A loss as a predictor of progression independent of stage and grade. These findings highlight the crucial role of CDKN2A loss in the progression of non-muscle-invasive FGFR3-mutated bladder carcinomas and provide a potentially useful clinical marker for adapting the treatment of such tumours, which account for about 50% of cases at initial clinical presentation.