Toward determining an appropriate indication for completion pancreatectomy after recurrence of pancreatic adenocarcinoma in the remnant pancreas

Abstract

Background: The effectiveness of operating on local recurrence of adenocarcinoma within the pancreatic parenchyma is poorly understood, because recurrence is more typically systemic. Under the latter circumstance, care is limited to palliative chemotherapy or best supportive care. When recurrence is localized to the pancreas, the benefit of resection is swayed by concerns for endocrine and exocrine insufficiency related to completion pancreatectomy and quiescent disease. This study aimed to define survival probabilities based on interval between index pancreatectomy and completion pancreatectomy. Methods: All completion pancreatectomies for isolated recurrence in the remnant pancreas at our institution from May 2003 with follow-up through August 2018 were identified from our institutional database. All patients had confirmed pancreatic adenocarcinoma at both the index and completion pancreatectomies, and none had distant metastatic disease. Perioperative outcome, survival, and prognostic parameters were assessed. The data naturally fit three inflection points that were used to compare overall survival from time of partial pancreatectomy using Kaplan-Meier methodology. Results: Fifteen males and 13 females were studied; median age at time of index pancreatectomy was 61.6 (IQR: 57.3, 67.7) years. Median follow-up for the sample was 34.7 (IQR 16.75, 84.9) months. 24% received neoadjuvant therapy prior to their index pancreatectomy, and 36% received adjuvant therapy; 24% received radiation therapy. 79% of index pancreatectomies were pancreaticoduodenectomies. Tumors were primarily high grade (80%) and node positive (57.7%). The overall median time to completion pancreatectomy was 23.59 (IQR: 5.9, 58.5) months. 12% received neoadjuvant therapy prior to their completion pancreatectomy, and 50% received adjuvant therapy; 12% received radiation therapy. Tumors were again primarily high grade (80%), but were majority node negative (53.9%). The median overall survival estimate was 64.43 (95% CI 30.9e113.2) months. When the data was trichotomized at 12 (n = 11), 12-36 (n = 6) and >36 (n = 11) months, a significant difference in overall survival probabilities emerged (p = 0.0028): the median overall survival probabilities were 17.98, 30.91 and 97.52 months from index pancreatectomy, respectively (Figure 1). The median time to recurrence by interval was 2.44, 20.67 and 75.38 months, respectively. At 24 months after completion pancreatectomy, 23%, 56% and 32% of patients remained alive, if the interval to completion pancreatectomy was less than 12 months, 12e36 months, or >36 months, respectively. This suggests individuals with isolated local recurrence undergoing completion pancreatectomy 12e36