Utilizing Intensity Modulated Proton Therapy with a Single Posterior-Anterior Beam for Esophageal Chemoradiation: Dosimetry and Long-Term Clinical Outcomes

Abstract

<h3>Purpose/Objective(s)</h3> Optimal techniques for intensity-modulated proton therapy (IMPT) in the treatment of locally advanced esophageal cancer remain unresolved. We previously demonstrated dosimetric advantages of a single posterior-anterior (sPA) beam technique and our initial experience employing it for these cancers. Herein, we present an update of our institution's experience with long-term clinical outcomes using this technique. <h3>Materials/Methods</h3> Retrospective review was performed of patients with stage II-IVA (AJCC 8^th ed.) esophageal or gastroesophageal junction cancer treated with chemoradiation (CRT) from 2015-2020. Inclusion criteria: 1) age ≥ 18-years-old, 2) pathologic diagnosis of squamous cell carcinoma or adenocarcinoma 3) neoadjuvant or definitive CRT utilizing IMPT prescribed to 50.4 Gy(RBE) with volumetric rescanning for plan robustness, 6) concurrent chemotherapy utilizing carboplatin and paclitaxel. Clinical outcomes were gathered on overall survival (OS), disease-free survival (DFS), pathologic complete response (pCR) for patients receiving trimodality therapy (TMT), and acute and late toxicities. Comparisons were made with patients treated with non-single PA beam IMPT techniques. <h3>Results</h3> Seventy patients met inclusion criteria of which 54 were treated with a sPA beam technique and 16 were treated with multiple proton beams or a single non-PA beam. The median follow-up time for the sPA beam cohort was 4.1 years and 2.3 years for the non-sPA beam cohort. The 2-yr OS and DFS rates (95% CI) for the sPA beam and non-sPA beam cohorts were 74.4% (63.0-86.8%) vs 60.6% (35.1-86.0%), and 57.3% (43.7-70.9%) vs 57.1% (31.2-83.0%), respectively. The pCR rates were not significantly different for patients receiving TMT in the sPA beam cohort (n=44) and the non-sPA beam cohort (n=11) at 15.9% vs 36.3%, respectively (p=0.20). The rate of late grade 3+ toxicity was not significantly different for the sPA beam and non-sPA beam cohorts at 25.9% vs 18.8% (p=0.49), respectively. Esophageal stenosis was the driver of late grade 3+ toxicity accounting for 70.5% (n=12) events, and 91.6% (n=11) of these events occurred in TMT patients. There were 2 patients (3.7%) with asymptomatic vertebral compression fracture (VCF) in the sPA beam cohort and 1 patient (6.3%) with symptomatic VCF in the non-sPA beam cohort. The PTV was significantly larger in the non-sPA beam cohort (384 cc vs 206 cc, p=0.015). The heart V5 was significantly lower in the sPA beam cohort (30% vs 37%, p=0.015). The mean lung (4.0 vs 4.4 Gy), lung V20 (8.2 vs 8.9%), and lung V5 (13.3 vs 17.1%) dosimetry was not significantly different between cohorts (p>0.14). <h3>Conclusion</h3> CRT for locally advanced esophageal cancer utilizing IMPT with a single PA beam has disease outcomes and toxicity rates comparable to the photon literature and other IMPT beam arrangements used at our institution.