Central Hepatobiliary Tree As an Organ at Risk for Stereotactic Body Radiation Therapy (SBRT) of the Liver: Is It High Priority for Hepatocellular Carcinoma (HCC)?

Abstract

SBRT dose objectives for the central hepatobiliary tree (cHBT) when treating primary liver cancers (PLCs) have been reported and were derived from HCC and cholangiocarcinoma (CCA) patients. It has been our practice to only define pre-planning objectives for mean liver dose (MLD) modified per Child Pugh Score (CP). We retrospectively reviewed dosimetric metrics and toxicity in a pure HCC dataset to assess whether the cHBT should be considered for the general application of HCC SBRT. Sixty consecutive HCC patients treated to 74 tumors from 2012-2018 were reviewed. Target volumes were defined on a 4D-CT with fusion of diagnostic contrast enhanced imaging. Treatment was delivered using volumetric modulated arc therapy with daily image guidance, 30-60 Gy in 3-6 fractions over 1-2 weeks. The MLD was the only planning goal (<13 Gy for Child Pugh A (CPA), <9 Gy for CPB/C). The cHBT was retrospectively contoured with a uniform 15-mm expansion of the portal vein (PV) from the splenic confluence to the first bifurcation of the left and right PV. Dosimetric data (VBED1030 < 37cc, VBED1040 < 45cc, DmeanBED10 < 25Gy) and tumor characteristics (location, size, cHBT overlap) as well as toxicity endpoints were analyzed using statistical software. Fifty-seven patients (CPA =35, CPB= 17, CPC = 5) with follow-up >3 months were included (median age=62). The median tumor size was 3.2cm (0.7-10cm) with 88% located in the right lobe of the liver. The MLD was 8.7Gy (1.4-15Gy, CPA), 7.8Gy (3.1-10.4Gy, CPB), and 4.8Gy (4-5.9Gy, CPC). Although 20 treatment targets overlapped with the cHBT with 9 overlap regions > 1.5cm (indicating direct overlap with the portal vein) , only 9% of patients did not meet published cHBT dose-volume metrics. Of these, 5 exceeded the VBED1030 and 3 the VBED1040 thresholds. 1 patient met criteria for hepatobiliary toxicity. Interestingly, this patient only had 0.3cc of overlap with the cHBT and was also the only patient who exceeded the recommended MLD of <13Gy (15Gy). The patient died 16 weeks post-treatment with toxicities manifesting as hyperbilirubinemia and large volume ascites. No other patients with overlapping target volumes or who exceeded published cHBT dose objectives met criteria for cHBT toxicity. We have not routinely placed a planning priority on the cHBT for treatment planning. Despite this, just 10% of treated tumors in this pure HCC dataset exceeded reported dose objectives. In the few patients where these were exceeded, only 1 met criteria for cHBT toxicity, but the patient had minimal overlap and exceeded the MLD. Given these findings, we have continued to place highest planning priority on the MLD with lesser priority on the cHBT threshold until additional reports justify the merits for conformal avoidance of this structure.