Presenter(s)
E. Lavrova1, P. Mavroidis1, A. Jones1, P. Hazelrigg2, V. Kalaria3, C. Iloabachie4, M. Lund2, A. K. Juneja4, P. Pazooki4, S. Power5, G. V. Lupu5, J. Badal2, A. M. Moon5,6, and T. K. Yanagihara1,6; 1University of North Carolina, Department of Radiation Oncology, Chapel Hill, NC, 2University of North Carolina, Department of Medicine, Chapel Hill, NC, 3University North Carolina School of Medicine, Chapel Hill, NC, 4University of North Carolina School of Medicine, Chapel Hill, NC, 5University of North Carolina, Division of Gastroenterology/Hepatology, Chapel Hill, NC, 6Lineberger Comprehensive Cancer Center, Chapel Hill, NC
Purpose/Objective(s): External beam radiotherapy (EBRT) for hepatocellular carcinoma (HCC) is associated with low rates of severe toxicity, with most treatment planning goals intending to minimize the risk of late hepatotoxicity. Less is known about the predictors of bothersome early toxicities, which may increase acute care utilization. We evaluated clinical and dosimetric predictors of acute and subacute toxicities leading to the need for prescription medications, emergency room visits and hospitalizations.
Materials/Methods: Patients with HCC treated at a single institution from 2014-2025 with EBRT were included in this study. Clinical endpoints included 1) patients with new ED visit or hospitalization within 3 months of treatment (excluding planned admissions), and 2) patients with new onset nausea or vomiting after EBRT requiring new anti-emetic medication, or pain requiring opioid medication. Dose-volume histogram (DVH) analysis included healthy liver minus gross tumor volume (GTV), stomach, and duodenum as relevant organs at risk. Correlations were calculated using area under the curve (AUC) and relative risk analyses between the clinical endpoints and DVH metrics from EBRT plans as well as clinical characteristics (ECOG, Child-Pugh score, and BCLC stage).
Results:
Complete data were available for 193 patients, of whom 29 (15%) reported acute nausea or vomiting requiring new administration of anti-emetic (28 patients) or pain requiring oxycodone administration (1 patient). DVH analysis showed that for healthy liver minus GTV, D820cc > 8Gy was associated with an increased likelihood of nausea or vomiting or pain requiring new antiemetic or opioid medication (HR 2.3, 95% CI 1.1-4.5, p = 0.02). Of 34 ED visits/hospitalizations, 15 (44%) were for upper GI signs/symptoms, including decompensated cirrhosis (5), abdominal pain (7), GI bleed (2), and dysphagia (1). There was no significant relationship between dose to healthy liver minus GTV and ED/hospital visits. No significant DVH correlations were observed for stomach or duodenum for either clinical endpoint. BCLC stage, CP score and ECOG performance status alone were not significantly associated with either clinical endpoint (AUCs ranging 0.48-0.61); however, a combination of ECOG > 0, Child-Pugh score > CP-A6, and BCLC stage > BCLC-B was associated with increased ED/hospital visits, HR > 2.3 (95%CI: 1.3-4.2), p = 0.01.Conclusion: The dosimetric analysis revealed that dose to liver minus GTV (D820cc > 8Gy), but not stomach or duodenum, was correlated with both a need for prescription medications for nausea/vomiting/pain, but not with acute care utilization. A combination of baseline Child-Pugh score > CP-A6, BCLC stage higher than BCLC-B, and ECOG > 0 predicted increased likelihood of ED/hospitalization within 3 months of EBRT treatment. Future work will evaluate planning guidelines that are tailored to the risk of acute toxicities based on both clinical and dosimetric predictors.