Implications of SMAD4 Status in Pancreatic Carcinoma Treated With Radiation Therapy: A Multi-Institutional Analysis

PURPOSE/OBJECTIVE(S) Loss of the tumor suppressor gene SMAD4 is a critical genetic alteration in pancreatic carcinoma (PC). We hypothesized that SMAD4 status in PC is associated with outcomes in patients who received neoadjuvant (NARx) or adjuvant (ARx) radiotherapy. MATERIALS/METHODS PC patients who underwent surgical resection at two high-volume centers following NARx-or those treated with ARx-between 2008-2019 were identified. SMAD4 status was determined based on immunohistochemical staining and classified as preserved (SMAD4+) or lost (SMAD4-). Kaplan-Meier survival estimates and multivariate analysis were used to analyze correlations between SMAD4 status, radiation therapy, and clinical outcome measures. RESULTS A total of 290 patients (mean age at diagnosis 66 years, 51% female) were identified; 131 (45%) were SMAD4+ and 159 (55%) SMAD4-. Resectable disease was diagnosed in 95 (33%) and borderline-resectable disease in 166 (57%); 29 patients (10%) had locally-advanced disease. NARx was administered in 147 (51%) in combination with chemotherapy while 143 (49%) received ARx; 26 (18%) of which received ARx solely. NARx in SMAD4- PC was associated with a significantly increased incidence of near-complete/complete histopathologic response and reduced incidence of none/poor response compared to SMAD4- PC not treated with NARx (12% vs 2% and 44% vs 19% respectively, P = 0.001). On adjusted analysis, NARx was a significant predictor of histopathologic response in SMAD4- patients (HR: 3.5, 95% CI 1.6-7.6, P < 0.001) while no association was seen for SMAD4+ PC. Neither radiation therapy receipt nor SMAD4 status were associated with overall survival (OS). Yet, SMAD4- PC had a worsened disease-free survival (DFS) compared to SMAD4+ PC (19 vs 16 months, P = 0.03). This difference persisted even among patients who had received NARx (21 vs 16 months, P = 0.04) and those with histopathologic treatment response (24 vs 16 months, P = 0.031). No difference in DFS between SMAD4- and SMAD4+ PC was identified in the ARx group. Lastly, NARx, significantly improved local-recurrence free survival in SMAD4+ PC but not in SMAD4- PC (33 vs 21 months, P = 0.047). CONCLUSION `Outcomes following surgical resection for PC remain primarily driven by SMAD4 status irrespective of radiation therapy timing (NARx vs ARx). However, this analysis suggests that SMAD4 status may help identify a subset of patients who are most likely to benefit from NARx.