Using Hydrogel to Create Spatial Separation between the Pancreas and Duodenum in Patients with Pancreatic Cancer: A Multi-Institutional Safety and Feasibility Study

Abstract

<h3>Purpose/Objective(s)</h3> Emerging data supports the use of dose-escalated radiation therapy (RT) for pancreatic adenocarcinoma (PDAC), but administration of ablative dosing with good target coverage is often limited by the proximity of the duodenum. Endoscopic injection of a hydrogel spacer into the pancreatico-duodenal (PD) groove may increase spatial separation between pancreatic tumors and the duodenum, thereby better enabling ablative RT, and previous preclinical data have supported this approach. Herein, we report results from a multi-institutional study exploring the safety and feasibility of hydrogel injections into the PD groove in patients with PDAC. <h3>Materials/Methods</h3> Patients undergoing RT for localized PDAC centered in the pancreatic head were eligible across three institutions. Patients with evidence of duodenal invasion were excluded. Hydrogel was injected in 1 mL increments along the PD groove under endoscopic ultrasound (EUS) guidance. The primary endpoint was safety, defined by procedure-related adverse events (AEs) that resulted in a delay in RT initiation, and feasibility, defined as evidence of hydrogel in the PD groove on simulation CT. Additional endpoints included the amount of space created between the pancreas and the duodenum and hydrogel visibility on cone-beam CT (CBCT). Pathologic evidence of hydrogel-related duodenal injury was also assessed in patients who underwent Whipple resection. <h3>Results</h3> Six patients were enrolled with a median age of 70 years (range 60-80 years). Disease stage at presentation included borderline resectable (n=4) and locally advanced (n=2) disease, with a median tumor diameter of 3.2 cm (range 2.3-3.5 cm). All patients underwent at least four months of systemic therapy before RT. Radiation modality included five-fraction stereotactic body RT (n=5, median dose 33 Gy, range 33-40 Gy) and intensity modulated RT (n=1, 58.8 Gy). All patients successfully underwent EUS-guided hydrogel injection into the PD groove with a median injected volume of 3.4 cc (range 2.5-10 cc) over a median of 4 injections per procedure (range 3-12). No patients experienced a procedure-related AE that delayed RT initiation. Additionally, over a median follow-up of 14.8 months (range 5.5-18.9 months), no device-related AEs occurred. On post-injection simulation CT, hydrogel was present in the PD groove in all six patients. Median space creation at the level of the hydrogel was 7.7 +/- 2.4 mm. Hydrogel was visible on CBCT in all patients. No evidence of pathologic duodenal damage was present in three patients who underwent resection. <h3>Conclusion</h3> Endoscopic injection of hydrogel into the PD groove appears safe and feasible. Continued efforts are needed to optimize the ability to inject larger volumes of hydrogel along the PD interface and to develop methods that can predict specific locations along the PD groove that would benefit most from hydrogel placement.