SBRT with Simultaneous Integrated Protection: Incidence and Pattern of Recurrence in Borderline Resectable and Locally Advanced Pancreatic Cancer

Abstract

To analyze the incidence and pattern of recurrence and to define a dosimetric map of loco-regional failure (LRF) after Stereotactic Body Radiation Therapy (SBRT) with Simultaneous Integrated Protection (SIP) for pancreatic cancer (PC) patients (pts). We retrospectively reviewed 50 consecutive pts (25 borderline resectable and 25 locally advanced disease) treated in our Department from November 2016 to August 2018. All the pts received induction chemotherapy prior to SBRT for at least 4 months (17 FOLFIRINOX, 30 Gemcitabine-Abraxane, 3 others) and 27 pts underwent surgical tumor resection after SBRT. SBRT was delivered in 5 consecutive daily fractions, prescribing 30 Gy to the Planning Target Volume (PTV) while simultaneously delivering 50 Gy to the region of vessel abutment/encasement. SIP was generated prescribing 25 Gy on the overlap area between the PTV and the Planning Organ at Risk Volume (PRV corresponding to OAR + 3 mm). Elective Nodal Irradiation (ENI) was not performed. For the analysis we defined 3 different site of failure based on the predominant area of first recurrence: in field (inside the isodose 30 Gy), out field (area corresponding to the ENI) and distant failure. We realized for each location of LRF a dosimetric evaluation matching the computed tomography (CT) that showed LRF with the SBRT treatment plan CT. A deformable image registration (DIR) was performed to better compare the CT anatomy. Median follow up time after SBRT was 9.3 months (range 4.2-24.1 months). Overall, a Local Control Rate (LCR) of 80% was obtained (85%, 23/27 in surgical vs 74%, 17/23 in nonsurgical pts; p = 0.32). Tumor relapse occurred in 29 pts (58%): distant alone in 19 pts, LRF only in 6 pts and a combination of distant and LRF in 4 pts. Concerning LRFs, 7 were exclusively in field and 3 a combination of in and out field failure. Median Local Progression Free Survival (LPFS) from SBRT was 8.6 months (8.5 vs 9.7 months in surgical vs nonsurgical pts; p = 0.78). The dosimetric evaluation showed that all in field failures occurred inside the area covered by 95% of the prescribing isodose, with extent in the area immediately close to the PTV. Interestingly, we didn’t observe isolated nodes recurrences in the ENI area. At the last follow-up 34 pts (68%) were alive, with a median Progression Free Survival of 16.2 months and a median Overall Survival of 17.4 months from the time of diagnosis. In our experience, SBRT-SIP for PC patients pretreated with systemic therapy was able to provide an acceptable local control rate and the lack of ENI didn’t significantly increase the rate of failure in the regional nodes area. However, the predominant incidence of in field recurrence support a clinical trial with SBRT-SIP at higher doses to further improve disease local control.