Individualized Adaptive Radiation Therapy Allows for Safe Treatment of Hepatocellular Carcinoma in Patients with Child-Pugh B Liver Disease

Abstract

Liver stereotactic body radiation therapy (SBRT) is an effective treatment for hepatocellular carcinoma (HCC). However, previous reports suggest unacceptably high rates of treatment related toxicity in patients with Child Pugh B (CPB) liver disease. Our approach to limit toxicity in these patients has been to treat with an individualized adaptive approach, based on baseline and mid-treatment liver function. We hypothesized that this approach would maintain good local control while limiting toxicity in CPB patients receiving SBRT for HCC. Eighty patients with CPB liver disease and HCC were treated on prospective clinical trials of individualized adaptive radiation therapy at a single institution from 2006-2018. Patients underwent pre and mid-treatment liver function assessment using indocyanine green, a water soluble compound cleared exclusively by the liver, thereby providing a dynamic assessment of liver function. The mid-treatment assessment was typically performed one month after the delivery of the first 60% of the total intended dose, with the potential to receive 0-40% of the remaining dose after the 1 month break. Treatment related toxicity was defined as a ≥2 point rise in CP score from pre-treatment. Logistic regression models were used to assess predictors of toxicity and linear mixed models were used to assess longitudinal changes in CP score over time. Overall survival (OS) was assessed using Cox proportional hazard models. In total, 80 CPB patients with HCC were treated. Median follow-up was 13.1 months and median survival was 17.1 months. 37 patients were CPB-7, 28 CPB-8, and 15 CPB-9. Median tumor size was 2.6 cm. Median treatment dose was 36 Gy delivered in 3 fractions with a median mean liver dose (MLD) of 11.0 Gy. 2-year local control was 89.3%. Overall, 18 patients (22.5 %) had a ≥2 point rise in CP score within 6 months of SBRT. In a multivariate model assessing predictors of toxicity, baseline CP, tumor size, and MLD were not associated with toxicity (all p>0.35). In a longitudinal model assessing changes in CP score over 12 months post-SBRT, controlling for baseline CP and tumor size, increasing MLD was associated with an increased risk of eventual increase in CP score (p=0.04). In a multivariate model for OS controlling for tumor size and baseline CP score, increasing MLD was associated with decreased OS (HR: 1.04, 95% CI: 1.00-1.12, p=0.04). Tumor size and baseline CP score were not significantly associated with OS. Our individualized adaptive treatment approach allows for safe and effective treatment of patients with CPB liver disease. Our rates of treatment related toxicity in these high-risk patients is comparable to reported series in patients with Child-Pugh A liver dysfunction, and substantially less than those reported to date in CPB patients.