Multi-Institutional Comparison of Ablative 5-Fraction Magnetic Resonance-Guided Online Adaptive Versus 15/25-Fraction Computed Tomography-Guided Moderately Hypofractionated Offline Adapted Radiation Therapy for Locally Advanced Pancreatic Cancer

Background: Radiation dose escalation for locally advanced pancreatic cancer (LAPC) using stereotactic magnetic resonance (MR)-guided online adaptive radiation therapy (SMART) or computed tomography (CT)-guided moderately hypofractionated ablative radiation therapy (HART) can achieve favorable outcomes although have not previously been compared. Methods: We performed a multi-center retrospective analysis of SMART (50 Gy/5 fractions) vs. HART (75 Gy/25 fractions or 67.5 Gy/15 fractions with concur-rent capecitabine) for LAPC. Gray’s test and Cox proportional regression analyses were performed to identify factors associated with local failure (LF) and overall survival (OS). Re-sults: A total of 211 patients (SMART, n = 91; HART, n = 120) were evaluated, and none had surgery. Median follow-up after SMART and HART was 27.0 and 40.0 months, respectively (p < 0.0002). SMART achieved higher gross tumor volume (GTV) coverage and greater hotspots. Two-year LF after SMART and HART was 6.5% and 32.9% (p < 0.001), while two-year OS was 31.0% vs. 35.3% (p = 0.056), respectively. LF was associated with SMART vs. HART (HR 5.389, 95% CI: 1.298–21.975; p = 0.021) and induction mFOLFIRINOX vs. non-mFOLFIRINOX (HR 2.067, 95% CI 1.038–4.052; p = 0.047), while OS was associated with CA19-9 decrease > 40% (HR 0.725, 95% CI 0.515–0.996; p = 0.046) and GTV V120%(HR 1.022, 95% CI 1.006–1.037; p = 0.015). Acute grade > 3 toxicity was similar (3.3% vs. 5.8%; p = 0.390), while late grade > 3 toxicity was less common after SMART (2.2% vs. 9.2%; p = 0.037). Conclusions: Ablative SMART and HART both achieve favorable oncologic outcomes for LAPC with minimal toxicity. We did not observe an OS difference, although technical advantages of SMART might improve target coverage and reduce LF.