Association of Radiation Dose with Local Failure in Hepatocellular Carcinoma

External beam radiation therapy is a well-established treatment for patients with unresectable hepatocellular carcinoma (HCC), however there remains some controversy regarding the relationship between radiation dose delivered and its impact on local control. We hypothesized that a dose-response relationship exists in patients treated with radiation therapy. Patients with HCC treated at a single institution using either stereotactic body radiation therapy (SBRT) or fractionated regimens with a biologically equivalent dose (BED) ≥ 60 Gy were included in this retrospective study. Both photon and proton therapy were included. We excluded patients who were treated with regimens with a BED < 60 Gy to limit the analysis to patients treated with curative, rather than palliative, intent. The primary objective was to assess whether local failure was improved with increased BED. Competing risk modeling was used to investigate the association between BED and local failure (LF), with death or liver transplant as a competing risk. Local progression-free survival (LPFS) was also examined with a Cox-proportional hazard model fit using a penalized cubic spline with two degrees of freedom to capture the impact of BED. 136 patients were included, with a median follow-up of 2.4 years. Median patient age was 63 (Range: 12-93). The proportion of patients with Child-Pugh Group A, B, and C at baseline were 61.7%, 25.8%, and 3.9% respectively. The proportion of patients with ECOG scores of 0, 1, 2, and 3 at baseline were 48.9%, 39.7%, 9.2%, and 2.3%. Median tumor size was 4.5 cm (Range: 1.1 – 19). Median BED among SBRT and non-SBRT patients was 100 Gy and 69 Gy, respectively. Among the patients in the fractionated group, about 90% received a hypofractionated regimen (dose per fraction >2 Gy). Local failure at 2 years was 14.9% for SBRT and 30.1% for fractionated patients, respectively. In a multivariable competing risk model with BED and tumor size as independent variables, increased BED correlated with decreased LF with a hazard ratio of 0.79 per 10 Gy increase (95%CI: 0.61-1.03, p = 0.08). In a retrospective cohort study of 136 patients, patients treated with higher biologically equivalent doses achieved decreased local failure, suggesting that HCC patients should be prescribed BED doses of 90-100 Gy.Abstract 2504; Table 12-Year local failureHazard ratio [95% confidence interval]PUnivariable modelBED (per 10 Gy)0.78 [0.61-1.0].05BED 90-100Gy vs. <90 Gy0.52 [0.24-1.16].11Tumor size (per cm)1.02 [0.94-1.1].60Age (per 10 years)0.86 [0.57-1.3].45Child PughA vs. C0.71 [0.09-5.69].75B vs. C0.94 [0.11-8.0].95ECOG0 vs. 21.69 [0.4-7.1].481 vs. 20.78 [0.16-3.8].76Multivariable model 1BED, continuous (per 10Gy)0.79 [0.61-1.0].08Tumor size (per cm)0.99 [0.91-1.09].88Multivariable model 2BED 90-100Gy vs. <90 Gy0.54 [0.24-1.25].15Tumor size (per cm)0.99 [0.91-1.09].98 Open table in a new tab