Proposed Model of Dose Escalation in Localized Pancreatic Cancer Patients Eligible for Surgical Exploration: Neoadjuvant Intensity Modulated Radiation Therapy (IMRT) vs. Stereotactic Body Radiation Therapy (SBRT) and Intra-Operative Radiotherapy (IORT)

Abstract

Multi-agent chemotherapy and radiation therapy (RT) has increased the proportion of locally advanced pancreatic cancer (LAPC) patients (pts) able to undergo resection and improve overall survival compared to those who are not resected. However, prior data from our institution (2016-2019) indicate that despite an R0 rate of 88%, LF developed in > 30% of LAPC pts resected after neoadjuvant therapy (NAT). This high LF rate is a rationale to explore dose escalation (DE) in the NAT setting, as data highlighting the value of DE have been limited to unresectable disease. Importantly, the optimal method for DE remains unclear. While conformal external beam RT can ablate part of the target, under-coverage in close proximity to the stomach and bowel remains problematic. For pancreatic head tumors, Whipple resection presents an opportunity for subsequent DE with IORT, as the duodenum is resected and direct access to microscopic peri-vascular residual disease is allowed. It also allows for precise DE based on the surgeons' recommendations. Herein, we model DE approaches, comparing pre-operative DE-IMRT with an approach in which a high-dose rate (HDR)-IORT boost is given after SBRT, with a goal biological effective dose (BED10) of 100 Gy.Ten consecutive LAPC pts were modeled with both DE-IMRT to 75 Gy in 25 fractions and SBRT to 40 Gy in 5 fractions followed by HDR-IORT to 15 Gy, using an 192Ir source. Notably, to minimize LF after surgery, our clinical tumor volume (CTV) has been the "Heidelberg Triangle" of tissue between the celiac, superior mesenteric, and common hepatic arteries and portal vein, which contains perineural and lymphatic tissue at risk of harboring subclinical disease. A 5mm planning tumor volume (PTV) expansion was used. A 3mm expansion on stomach and bowel was constrained to 60 Gy and 40 Gy for DE-IMRT and SBRT, respectively. For IORT modeling, the Freiburg Flap was superimposed on the planning CT, and dose was prescribed to the target volume surface. Given rapid dose fall-off with HDR-IORT, SBRT was optimized to compensate for IORT under-coverage, with higher dose allowed posteriorly but with an overall hotspot less than 50 Gy. Descriptive statistics were generated for coverage to target structures.With DE-IMRT, mean V75Gy for the CTV and PTV was only 79% +/- 14.4% and 52% +/- 13%, respectively. In comparison, posteriorly optimizing the SBRT plan in anticipation of anterior IORT supplementation improved coverage with ablative BED. For the anterior half of the target volume, HDR-IORT created a dose gradient of 10-15 Gy, whereas for the posterior half of the target volume, HDR-IORT created a dose gradient of 5-10 Gy. Optimizing the SBRT plan to complement this with 40 Gy to the anterior half of the volume and 45 Gy to the posterior half of the volume was feasible with > 90% coverage with BED10 of 100Gy in composite.Dose escalation with SBRT/IORT may allow for better target coverage with ablative RT dose as compared to DE-IMRT and should be explored clinically.