Acute Toxicities and Short-Term Follow-up of 64 Patients Treated with Intensity-Modulated Proton Beam Radiotherapy (IMPT) or Intensity-Modulated Photon Radiotherapy (IMRT) for Esophageal Carcinoma (EC): A Single-Center Experience

Abstract

IMPT allows for a significant dosimetric advantage over IMRT in patients (pts) with EC, particularly for sparing the heart and lungs. However, increased uncertainties are associated with IMPT, such as organ motion and robustness. We present a clinical comparison of acute RT-related toxicities and short-term outcomes of pts treated with IMPT and IMRT for EC at our institution. A retrospective review was conducted on consecutive pts who underwent concurrent chemoradiation using IMPT or IMRT for EC with definitive or neoadjuvant intent from 2014-2018. Patient, treatment, toxicity, and disease outcome characteristics were extracted from medical records. Univariate and multivariate (MVA) analyses were performed (categorical variables by Fischer’s exact test and continuous variables by Wilcoxon-rank sum test). Survival outcomes were estimated by the Kaplan-Meier method. Sixty-four (32 per group) pts were included, with 10 months median follow-up for IMPT and 14 months for IMRT. Baseline characteristics were similar between groups except for (IMPT vs IMRT, respectively): dysphagia on presentation (85 vs. 63%, P=0.04) and histology (adenoma. 63 vs. 91%, P=0.02). Pts were more commonly male, with median age 72, ECOG 0-1, and distal tumor location (none statistically significant between IMPT and IMRT). Treatment intent was primarily neoadjuvant (72% IMPT, 81% IMRT) with the most common concurrent chemotherapy regimen being weekly carboplatin/paclitaxel (95% per group). The most common radiation regimen was 50 Gy in 25 fractions (75% IMPT, 78% IMRT). There was no significant difference in number of pts who underwent surgery (47% IMPT, 56% IMRT) and of those, no difference in pathologic complete response rates (33% IMPT, 39% IMRT). Post-operative complications were also similar between groups including pneumonia, anastomotic leak, anastomotic stricture, and cardiac arrhythmia rates. At one year, clinical outcomes were (for IMPT vs IMRT, respectively): local control 92 vs 84%, local-regional control 92 vs 80%, distant metastasis-free survival 87 vs 65%, and overall survival 74 vs 71% (none statistically significant between groups). IMPT pts had statistically significantly improved progression free survival (PFS) (75% vs 47%, P=0.02), and it remained significant for improved PFS in the final MVA model (P=0.04). No statistically significant association was seen between the two groups for acute treatment-related grade 3 toxicities (P=0.47). We report early, promising retrospective evidence for safety and efficacy of IMPT compared to IMRT for EC. Direct comparative studies of IMPT vs. IMRT utilizing prospective trials accruing larger, balanced patient cohorts are warranted in the future.