Pancreatic Stereotactic Body Radiation Therapy (SBRT) in the Era of Multi-Agent Chemotherapy for Locally Advanced and Borderline Resectable Pancreatic Cancer: A Prospective Single Institution Study

Abstract

Few studies have examined the role of stereotactic body radiation (SBRT) after multiagent chemotherapy (ChT) in pancreatic ductal adenocarcinoma (PDAC). This was a prospective study of SBRT after ChT for PDAC. To prospectively assess safety and efficacy of pancreatic SBRT following neoadjuvant multiagent chemotherapy. Eligible pts had locally advanced (LA), borderline resectable (BR), or locally recurrent (LR) PDAC and received either gemcitabine/nab-paclitaxel (GA) or FOLFIRINOX. SBRT was delivered to tumor and tumor-vessel interface and dose escalated from 33 to 36 Gy in 5 fractions. Pts with prior abdominal RT received 25 Gy in 5 fractions. Primary endpoint was severe (≥ grade 3) acute (≤ 6 mos from SBRT) toxicity rate. Kaplan Meier and multivariate Cox regression were used for survival analyses. 35 pts completed treatment on study with an additional 10 treated off study but per protocol (total 45). Median age was 66 yrs; 22, 17 and 6 pts, respectively, had BR, LA, and LR disease. 69% received FOLFIRINOX before SBRT and 24% received GA. Severe acute toxicity rate after SBRT was 4.4% (2/45, both grade 3). 16 patients were resected after SBRT; 12 had partial responses, but none had pathologic complete response. For all pts, at a median follow-up of 15.4 mos, median overall survival (OS) was 22.8 (95% CI 17.6-28.0) mos from diagnosis and 11.8 (10.4-13.1) mos from SBRT. In patients with LA, BR, and LR disease, median OS from diagnosis was 17.6, 25.9, and 36.4 mos, respectively, while median OS from SBRT was 9.6 mos, 19.8 mos, and not reached. Median progression-free survival (PFS) from SBRT was 8.4, 16.1, and 20.3 mos in patients with LA, BR, and LR disease, respectively. Actuarial 1-year local PFS (LPFS) rate was 78.5%, with 1-year LPFS rates of 71.6% and 100% in pts treated with ≤33 Gy and >33 Gy, respectively (p = 0.13). Cox regression showed that resection (p = 0.023), and pre-SBRT CA 19-9 (p = 0.005) were significant predictors of post-SBRT survival, but type of ChT was not (p = 0.75). Median OS from diagnosis was 25.2 mos in resected and 18.4 mos in unresected pts (log-rank p=0.047). SBRT can be performed safely after multiagent ChT for PDAC. There was a trend towards improved local control in patients treated with dose-escalated (36 Gy/5 fractions) SBRT. Patients with BR/LR disease and low baseline CA19-9 benefited the most from SBRT. Further study is needed to define the role of SBRT in PDAC.