Dose-Effect Analysis of Radiation-Induced Liver Disease During Preoperative Chemoradiotherapy for Locally Advanced Gastric Cancer.

Abstract

PURPOSE/OBJECTIVE(S): to investigate the clinical and dose-volume factors of radiation-induced liver disease (RILD) during preoperative chemoradiotherapy for local advanced gastric adenocarcinoma, and provide a reference for optimization of the radiation plans.\n\n\nMATERIALS/METHODS\nWe retrospectively observed the clinical characteristics as well as dose-volume factors in a cohort of 190 patients with local advanced gastric adenocarcinoma in our hospital, including gender, age, treatment group, clinical TNM classification, primary tumor site, hepatitis, body mass index (BMI), alanine aminotransferase (ALT), alkaline phosphatase (ALP), serum albumin (ALB), total bilirubin (TBIL), activated partial thromboplastin time (APTT) from June 2013 to November 2020, based on an ongoing multicentric randomized phase III clinical trial (NCT01815853). There are 98 patients receiving chemotherapy (ChT, 3 cycles of XELOX regimen), and 99 patients receiving chemoradiation therapy (CRT, 1 cycle of induced XELOX regimen and 4500cGy/25f IMRT radiotherapy plus concurrent extenuated 2 cycles of XELOX regimen) before radical gastrectomy. The evaluation of RILD was performed based on clinical symptoms, Child-Pugh score system, Common Terminology Criteria of Adverse Event (CTCAE) 5.0 during preoperative treatment. ROC curves were used for appropriate cut-off values. Univariant and multivariant analysis were performed in dose-volume factors to find significant variables.\n\n\nRESULTS\nIn all of the patients, the incidence rate of RILD is 7.3%. With aid of supportive and symptom-eliminating treatments, all but one patient completed the assigned preoperative treatment. The RILD rate in the CRT group was higher than in ChT group (12.1% vs 2.2%, P\u202f=\u202f0.009). As for the patients in CRT group, there were significant discrepancies in V3.5Gy (P\u202f=\u202f0.017), V5Gy (P\u202f=\u202f0.018), V10Gy (P\u202f=\u202f0.033), V15Gy (P\u202f=\u202f0.019), and borderline significance in the maximum dose (Dmax, P\u202f=\u202f0.078) of liver between RILD patients and non-RILD patients in CRT group. Subsequent analysis showed that patients with the liver V3.5Gy (%) > 99.975%, liver V15Gy (%) > 66.470% had higher risk of RILD (OR: 5.797, P\u202f=\u202f0.016; OR: 6.601, P\u202f=\u202f0.023; respectively). However, V5Gy and V10Gy couldn t be considered as independent factors based on multivariant analysis (P\u202f=\u202f0.757, P\u202f=\u202f0.219, respectively), and the value of Dmax as independent prognostic factor was borderline significant (P\u202f=\u202f0.050). There were no significant association between RILD and other clinical characteristics.\n\n\nCONCLUSION\nCompared to chemotherapy, chemoradiation can increase the risk of RILD during preoperative treatment, but it s safe with regularly monitoring and appropriate treatment. Under the condition of the present liver-dose limitation, reduction of liver radiated volume in low dose (<\u202f=\u202f15Gy) can eliminate the incidence rate of RILD. Besides, further studies are in need for the improvement for radiation plans of preoperative chemoradiation.