Long-Term Outcomes of IDH2-Mutant Sinonasal Carcinomas Compared With IDH2-Wild Type Sinonasal Undifferentiated...

Sinonasal carcinomas are aggressive and often locally invasive at presentation, with no universally accepted treatment approach. Histopathologic IDH2-mutant status appears to be associated with improved treatment response and survival. However, the clinical significance of histopathologic classification among IDH2-mutant sinonasal carcinomas, as well as that of IDH2 mutation status within SNUC, remains unclear and long-term clinical outcome data are limited.

Records from a prospectively maintained skull-base database at a large, tertiary care hospital system from 1986 to 2024 were reviewed. Patients with a diagnosis of SNUC or sinonasal tumor of any histopathology positive for IDH2 R172 mutation were identified. SMARCB1-deficient, SMARCA4-deficient, NUT sinonasal carcinomas, and other small round cell mimickers were excluded. After histopathologic confirmation, 22 IDH2-mutant SNUC, 14 IDH2-mutant non-SNUC, and 8 IDH2-wildtype SNUC cases were amenable to comprehensive clinicopathologic data review. A presence of IDH2 R172 and pathogenic TP53 mutations was determined by molecular testing and/or by immunohistochemistry.

The median age at diagnosis was 58 years and 61% of patients were male. The majority of patients presented with T4b disease (41% of IDH2-mutant SNUC; 57% of IDH2-mutant non-SNUC; 75% of IDH2-wildtype SNUC; p = 0.3). There was no significant difference in the proportion of patients presenting with orbital invasion (55% IDH2-mutant SNUC; 64% IDH2-mutant non-SNUC; 75% of IDH2-wildtype SNUC; p = 0.7) or intracranial invasion (82% IDH2-mutant SNUC; 79% IDH2-mutant non-SNUC; 75% of IDH2-wildtype SNUC; p = 0.7). On histopathology, there was no significant difference in perineural invasion and margin status between the three subgroups (p > 0.6). A greater proportion of IDH2-mutant non-SNUCs had TP53 mutations (67%) than IDH2-mutant SNUC (31%) and IDH2-wildtype SNUC (29%) (p = 0.13). Across the cohort, 41% of patients received neoadjuvant chemotherapy, with a partial response by RECIST criteria in 94% of patients across the subgroups. Median follow-up was 28 months (IQR 11-64 months). 5-year DSS of 73% in IDH2-mutant SNUC, 75% in IDH2-mutant non-SNUC and 80% in IDH2-wildtype SNUC (p = 0.78). 5-year RFS was 53% in IDH2-mutant SNUC, 48% in IDH2-mutant non-SNUC and 43% in IDH2-wildtype SNUC (p = 0.78).

Upfront assessment of IDH2 mutation status in sinonasal carcinoma patients may be warranted to select patients for clinical trials with IDH-inhibitors. However, no significant differences in the extent of disease at presentation or response to induction chemotherapy were observed according to histopathologic classification among IDH2-mutant sinonasal carcinomas or IDH2 mutation status within SNUC. Larger series are required to determine the prognostic significance of IDH2 mutation status in sinonasal carcinoma patients.