Impact of intraoperative radiation on postoperative and disease-specific outcome after pancreatoduodenectomy for adenocarcinoma: a propensity score analysis.

For periampullary cancer,intraoperative radiation therapy (IORT) administered to the site with the highest locoregional recurrence risk carries the rationale to improve tumor control. An IORT effect on survival remains unclear. IORT impact on postoperative outcomes after pancreatectomy for adenocarcinoma was analyzed, with a specific attempt to correct for the nonrandom IORT treatment assignment, and to account for treatment group imbalances in the interpretation of outcome differences. A propensity-score-adjusted analysis, based on variable selection by logistic regression, was used to rebalance treatments. Between 1989 and 1999, 61 patients underwent partial or total pancreatectomy for a primary periampullary adenocarcinoma at the City of Hope National Medical Center. Diagnoses included pancreatic (n = 36), duodenal (n = 11), ampullary (n = 10), and bile duct cancer (n = 4). Thirty patients received IORT to the resection area, with a median dose of 15 Gy (range: 10-20), followed by postoperative external beam radiation (n = 24). Mortality was 0%, the complication rate 61%. Of 33 patients with a documented recurrence, 6 had an isolated locoregional recurrence only (1 IORT versus 5 no IORT, = 0.05); the systemic recurrence pattern differed as well (IORT 94%, no IORT 67%; = 0.04). IORT had no significant impact on hospital stay (overall median: 17 days), disease-free survival (16 months), and overall survival (23 months) when adjusted for those most relevant variables reflecting IORT treatment group assignment propensity. After adjustment for relevant propensity factors, IORT was not linked to a significantly increased risk for complications, hospital stay, or survival hazard. The recurrence pattern may be affected in some patients, but systemic recurrences predominate. We continue to explore IORT in combination with systemic chemotherapy.