Pathologic Response and Locoregional Control After Preoperative Pancreatic Stereotactic Body Radiation Therapy. International journal of radiation oncology, biology, physics Dworkin, M. L., Miller, J. A., Toesca, D. A., Baclay, J. R., Pollom, E., Chang, D. T. 2021; 111 (3S): e37

Abstract

PURPOSE/OBJECTIVE(S): In light of ALLIANCE02150, the role of preoperative stereotactic body radiotherapy (SBRT) for pancreatic cancer is controversial. We studied patients who had surgery after preoperative SBRT to assess pathologic and clinical outcomes, and evaluate if elective nodal irradiation (ENI) decreases locoregional failure (LRF).MATERIALS/METHODS: Patients with pancreatic cancer who received SBRT at one center from 2007 to 2020 followed by oncologic surgery were reviewed. Local (primary tumor), regional (peripancreatic or perivascular per RTOG consensus), and distant failures were coded. Pathologic treatment response per College of American Pathologists was reviewed. Incidence of LRF with death as competing risk was assessed. Time to overall (OS) and progression-free survival (PFS) from date of pathologic diagnosis was assessed via Kaplan-Meier. Association testing via Cox analysis for OS/PFS and Gray test for LRF was performed.RESULTS: Twenty-eight patients were evaluable. Median (range) ECOG was 1 (0-2). Tumor stage was cT4 in 19 (67.9%) patients. Nodal stage was cN1 in 10 (35.7 %) patients. There were 7 and 21 patients with initially unresectable and borderline resectable disease respectively. Chemotherapy (CT) was given prior to SBRT in 27 (96.4%) patients, with FOLFIRINOX and gemcitabine-paclitaxel in 21 and 4 patients, respectively. Median (range) duration of CT was 5 (0-12) cycles. Median (range) dose, BED10, and fractionation were 40 Gy (25 - 50 Gy), 72 Gy (54.78 - 100 Gy), and 5 (1-5) respectively. The median (range) time between SBRT and surgery was 6.7 (2.6 - 21.9) weeks. Whipple, Appleby, and distal pancreatectomy was performed in 22, 4, and 2 patients respectively. The R0 rate was 23/28 (82.1%). Pathologic complete, near complete, partial, and poor/no treatment response was seen in 1 (3.6%), 10 (35.7%), 15 (53.6%), and 2 (7.1%) patients, respectively. The pN1 rate was 12/28 (42.3%). Median (range) follow up was 21.5 months (6.9 - 67.2 months). The 18-month (95% CI) overall survival (OS) and LRF were 66.6% (50.0 - 88.6%), and 7.8% (0.0 - 48.4%) respectively. Complete or near complete pathologic response was associated with improved OS (HR?=?0.21, P?=?0.022) and PFS (HR?=?0.25, P?=?0.021), but not LRC (P?=?0.780). Longer (=8 weeks) time between end of SBRT and surgery was associated with improved complete or near complete response rate (P?=?0.024) but not OS, PFS, or LRF. BED10 did not predict for pathologic treatment response or margin status. Sixteen patients received ENI. There were no other statistically significant differences in the above baseline characteristics, OS, PFS, or pathology outcomes, with vs without ENI. The 18-month (95% CI) LRF with vs without ENI was 0.0% (0.0 - 0.0%) vs 12.8% (0.0 - 38.8%), P?=?0.29.CONCLUSION: Our cohort of patients with pancreatic cancer treated with preoperative SBRT and surgery showed good pathologic response and R0 rate. ENI was associated with numerically lower LRF. Increased BED10 was not associated with improved pathologic treatment response.

View details for DOI 10.1016/j.ijrobp.2021.07.356

View details for PubMedID 34701293