3096 - Dosimetric Outcomes and Clinical Toxicity from CT-Based Online Adaptive Hypofractionated Radiotherapy for Non-Operable Pancreatic Cancer

Purpose/Objective(s): CT-based online adaptive radiotherapy (oART) offers isotoxic treatment while maximizing therapeutic dose for non-operative pancreatic tumors. As CT-based oART pancreas programs expand, a hypofractionated approach (60 Gy/10 Fx) can initially be used (prior to transitioning to 50 Gy/5 Fx regimen) to obtain favorable therapeutic dose (BED₁₀=96 Gy) while accounting for the inherent uncertainties of surface guidance radiation therapy (SGRT). We report dosimetric and toxicity outcomes of a 10-fraction (Fx) oART scheme.

Materials/Methods: We retrospectively reviewed 110 CT-guided oART Fx given to 11 non-operable patients with pancreatic tumors. All received 60 Gy/10 Fx to gross tumor volume (GTV) with optional 35 Gy/10 Fx to clinical target volume (CTV). Three breath hold (BH) scans were acquired at simulation with patient surface tolerance of ±2 mm. A 3-5 mm planning organ at risk volume (PRV) was added to GI luminal structures to account for internal motion seen across the BH scans. Daily contoured bowel, duodenum, and stomach PRVs were combined (BDS_PRV) for adaptive dose optimization. All 10 Fx of 350 cGy to 35opti = ([CTV + 5mm] – BDS_PRV) and simultaneous integrated boost of 600 cGy to 60opti ([GTV + 5mm] – [BDS_PRV + 5mm]) were given daily with BH via SGRT after 30 minutes supine bowel rest. Daily contouring done by physician trainees and physicists underwent final review by physician faculty. Clinical variables, constraints, and toxicity data (CTCAE V6) were recorded. Paired t-tests with a Bonferroni-adjusted threshold (p <.0083) determined significance of dose changes between scheduled and adapted plans.

Results: Of the 11 patients, 10 had pancreatic adenocarcinoma tumors, 1 had a grade 2 well-differentiated neuroendocrine tumor; median age: 71 (Interquartile [IQR] 67-78). Mean treatment time (CBCT to treatment completion): 57.5 minutes (SD + 14.1). Adaptation significantly reduced D_0.03cc for bowel, duodenum, and stomach PRVs, by 26.0%, 28.0%, 26.8% respectively (all p <.0001), while maintaining GTV coverage, V600 and V570 (Table 1). Median PRV expansion was 3 mm (IQR: 3-4.5). Mean BDS_PRV D_0.03cc reduced 37.2% with adaptation (p <.0001). No treatment interruptions occurred. At a median follow-up of 4 months (IQR: 3-5.5), there was no treatment toxicity Grade >3.

Conclusion: CT-based oART significantly improved critical OARs sparing while maintaining target dose in the delivery of 60 Gy/10 Fx. This regimen had low toxicity for pancreatic patients and may be an effective approach for centers starting oART pancreas programs.