Factors Predicting Response, Perioperative Outcomes, and Survival Following Total Neoadjuvant Therapy for Borderline/Locally Advanced Pancreatic Cancer

Objective: To identify predictive factors associated with operative morbidity, mortality, and survival outcomes in patients with borderline resectable (BR) or locally advanced (LA) pancreatic ductal adenocarcinoma (PDAC) undergoing total neoadjuvant therapy (TNT). Background: The optimal preoperative treatment sequencing for BR/LA PDA is unknown. TNT, or systemic chemotherapy followed by chemoradiation (CRT), addresses both occult metastases and positive margin risks and thus is a potentially optimal strategy; however, factors predictive of perioperative and survival outcomes are currently undefined. Methods: We reviewed our experience in BR/LA patients undergoing resection from 2010 to 2017 following TNT assessing operative morbidity, mortality, and survival in order to define outcome predictors and response endpoints. Results: One hundred ninety-four patients underwent resection after TNT, including 123 (63%) BR and 71 (37%) LA PDAC. FOLFIRINOX or gemcitabine along with nab-paclitaxel were used in 165 (85%) and 65 (34%) patients, with 36 (19%) requiring chemotherapeutic switch before long-course CRT and subsequent resection. Radiologic anatomical downstaging was uncommon (28%). En bloc venous and/or arterial resection was required in 125 (65%) patients with 94% of patients achieving R0 margins. The 90-day major morbidity and mortality was 36% and 6.7%, respectively. Excluding operative mortalities, the median, 1-year, 2-year, and 3-year recurrence-free survival (RFS) [overall survival (OS)] rates were 23.5 (58.8) months, 65 (96)%, 48 (78)%, and 32 (62)%, respectively. Radiologic downstaging, vascular resection, and chemotherapy regimen/switch were not associated with survival. Only 3 factors independently associated with prolonged survival, including extended duration (≥6 cycles) chemotherapy, optimal post-chemotherapy CA19-9 response, and major pathologic response. Patients achieving all 3 factors had superior survival outcomes with a survival detriment for each failing factor. In a subset of patients with interval metabolic (PET) imaging after initial chemotherapy, complete metabolic response highly correlated with major pathologic response. Conclusion: Our TNT experience in resected BR/LA PDAC revealed high negative margin rates despite low radiologic downstaging. Extended duration chemotherapy with associated biochemical and pathologic responses highly predicted postoperative survival. Potential modifications of initial chemotherapy treatment include extending cycle duration to normalize CA19-9 or achieve complete metabolic response, or consideration of chemotherapeutic switch in order to achieve these factors may improve survival before moving forward with CRT and subsequent resection.