Real-World-Feasible Immunohistochemistry of ATRX, DAXX, and Menin Identifies a Subgroup of Non-Functioning Pancreatic Neuroendocrine Tumors with low Recurrence Risk to Guide De-Escalating Surveillance.

Non-functioning pancreatic neuroendocrine tumors (NF-pNETs) show a variable prognosis. Despite 40-60% of patients remaining recurrence-free after surgery, guidelines recommend ≥ 10 years of follow-up. Recent studies identify prognostic subgroups based on ATRX, DAXX, and MEN1 mutations and chromosomal aneuploidy, highlighting a subgroup with favorable prognosis. We aimed to classify resected NF-pNETs into molecular subgroups using ATRX, DAXX, and Menin immunohistochemistry as surrogate markers for genomic status. To do so, we retrospectively collected resected primary sporadic grade 1 and 2 NF-pNETs without synchronous metastases from five international local pathology archives. ATRX, DAXX, and Menin immunohistochemistry was performed, and tumors were categorized into three groups: ATRX-DAXX-loss-group (ATRX or DAXX loss), isolated-Menin-loss-group (Menin loss without ATRX and DAXX loss), and unspecified-group (retained Menin, ATRX and DAXX). Kaplan-Meier analysis evaluated prognostic differences, and multivariable Cox regression assessed the added value of molecular subgroups beyond established prognostic factors. In total, 139 patients with grade 1 (64%) and 2 (36%) NF-pNETs were included. The median follow-up was 28 months (range 0-195) and recurrences occurred in 20% (28/139). No recurrences occurred in the isolated-Menin-loss-group (0/33), versus 12.5% (8/64) in the unspecified-group and 48% (20/42) in the ATRX-DAXX-loss-group. Kaplan-Meier analysis showed better recurrence-free interval in the isolated-Menin-loss-group than other subgroups (P < 0.001). In univariate analysis, the isolated-Menin-loss-group had a lower recurrence hazard (HR 0.03; 95%CI: 0.00-0.55; P = 0.018), than other subgroups, with little change in the hazard ratio after adjustment, but loss of significance. In conclusion, molecular subclassification by immunohistochemistry identifies NF-pNET patients with a low risk of recurrence and lays the groundwork for future prospective studies assessing genetics-based risk stratification as a tool for guiding clinical management.