Outcomes of Carbon-Ion Radiation Therapy Versus Photon Therapy for Isolated Paraortic Lymph Node Recurrence From Colorectal Cancer

Abstract

Purpose: To assess the effectiveness of carbon-ion radiation therapy (CIRT) versus that of photon radiation therapy (RT) in patients with isolated paraortic lymph node metastasis from colorectal cancer postcurative surgery. Methods and Materials: A retrospective analysis was conducted on 116 patients, including 53 treated with CIRT and 63 with photon RT. The primary endpoint was local control (LC), with progression-free survival (PFS) and overall survival (OS) as secondary endpoints. Kaplan-Meier and Cox regression analyses were used to evaluate outcomes. A propensity score matching analysis was also performed to adjust for baseline imbalances. Results: At a median follow-up of 54.8 months for surviving patients, CIRT was associated with higher 2-year LC than photon RT (78.9% vs 68.1%, P = .028). Cox regression analysis identified CIRT (hazard ratio, 2.152; 95% confidence interval, 1.082-4.279) and a smaller gross tumor volume at RT as significant independent predictors of improved LC. The difference in LC remained significant after propensity score matching (2-year 78.9% vs 63.3%, P = .036). No significant differences in PFS or OS were observed between the groups. Grade 2 or higher acute gastrointestinal toxicity occurred in 17.4% of the photon RT group but was absent in the CIRT group. Acute hematologic toxicity of grade 2 or higher occurred in 22.6% of the CIRT group and 17.5% of the photon RT group. Late toxicity, reported only in the photon RT group (11.1%), consisted of mild gastrointestinal symptoms. Conclusions: CIRT demonstrated superior LC to photon RT in the treatment of isolated paraortic lymph node metastases, with a more favorable toxicity profile. Although no significant differences were observed in PFS or OS, careful patient selection and integration of systemic therapies may further optimize outcomes. (c) 2025 Elsevier Inc. All rights are reserved, including those for text and data mining, AI training, and similar technologies.