Characterizing Compromised Target Coverage With Hypofractionated Radiation Therapy for Pancreatic Cancer.

Introduction Proximity of organs at risk (OAR) hinders radiation dose escalation for the treatment of pancreatic cancer. To address this limitation, there is interest in protracted-fractionation (PF: 15 to 25 fractions) courses employing moderate hypofractionation (MHF: 3-4 Gy/fraction). However, there persists underdosing where tumor interfaces with OAR. The significance of compromised tumor coverage and dose heterogeneity on tumor control remains unknown. Here, we report our initial planning experience with PF-MHF in pancreatic cancer. Methods We retrospectively reviewed radiation courses for locally advanced or recurrent pancreatic cancer with a PF-MHF approach: 45 Gy in 25 fractions (1.8 Gy/fraction) to PTV with 75 Gy (3 Gy/fraction) as an integrated boost to the GTV. We reviewed dosimetric parameters for the GTV: percentage overlap with planning OAR volume (PRV-GTV overlap), D99.9%, D0.1cc, Dmean, V75Gy, and V60Gy. We also calculated the GTV's generalized equivalent uniform dose (gEUD) value using two different  values (-5 and -15). Lastly, we reoptimized two plans with two approaches: increasing gEUD or relaxing the maximum dose constraint. Results A total of 26 plans were included in our analysis: 14 locally advanced and 12 locally recurrent pancreatic cancer cases. While the D0.1cc median value was 81.7 Gy, target volume coverage was relatively low (V75Gy median 71%). Median gEUD were 71 Gy ( = -5) and 62.8 Gy ( = -15) and inversely correlated with PRV-GTV overlap. On reoptimized plans, both approaches yielded similar results, but an increase in target coverage and gEUD were seen only when there was limited PRV-GTV overlap. Conclusion Although radiation dose can be escalated within the GTV, there continues to be low coverage by the prescription dose, especially with high PRV-GTV overlap. Relaxing the maximum dose constraint in planning allows for meaningful improvement in tumor coverage in limited PRV overlap scenarios. Continued refinement of the PF-MHF approach is needed.