Stereotactic body radiation therapy (SBRT) for definitive treatment and as a bridge to liver transplantation in early stage inoperable Hepatocellular carcinoma

The study included all consecutive HCC patients who were not candidates for primary resection and/or local therapy and who were treated with intensity-modulated radiation therapy (IMRT)/image-guided radiation therapy (IGRT)-based SBRT at the Institute of Oncology between November 2011 and January 2016. All cases were discussed in a multidisciplinary tumor board including hepatologists, hepatobiliary/transplant surgeons, medical oncologists, radiation oncologists and radiologists as well as invasive radiologists.

This retrospective study was approved by the medical center institutional Helsinki review board. No informed consent was required. Data were collected from medical records and included demographics, viral status, history of cirrhosis and underlying liver disease, tumor location and size, performance status, response to treatment, survival, and cause of death. Liver enzyme, albumin, creatinine and total bilirubin levels, as well as coagulation studies (international normalized ratio [INR]) were recorded before and every 2 weeks after completion of radiotherapy up to 12 weeks. Treatment outcome was assessed by liver magnetic resonance imaging (MRI) or triphasic computed tomography (CT) scan. Tumor response in patients who underwent liver transplantation was assessed by explant pathology. Progression-free survival (PFS) was defined as the time between SBRT and the first imaging study that indicated disease progression – local or distant. Local progression was defined as an increase of 20% or more of the primary tumor volume. Local control was defined as the lack of local progression. Overall survival (OS) was calculated from the time of SBRT. Treatment-related death was defined as death within 30 days of treatment. The surgical report and post-op charts were reviewed for remarks on surgical difficulty and vascular complications.

Patients were immobilized for simulation with a customized vacuum cushion and abdominal compression was applied. Patients were simulated using a multiphase 4-dimentional CT simulation to monitor breathing-related liver motion. Images were reconstructed on the Advantage Workstation (GE Healthcare, Chicago, IL). When anatomically feasible, fiducial markers were placed. The internal target volumes (ITVs) were created using all 10 phases to account for maximal tumor excursion.

The gross tumor volume (GTV) was defined as the contrast enhancing tumor volume on a triphasic CT or MRI scan. The planning treatment volume (PTV) was defined as a 3-mm margin around the GTV, after expansion for ITV was made. The PTV was reduced in case of proximity to vital normal tissue. Patients were treated with IMRT using dynamic sliding window multileaf collimator (MLC) or volumetric modulated arc therapy (VMAT). Specification of the dose-volume histogram (DVH) constraints is available in Table 1.

Dose calculations were performed using the Eclipse™ treatment planning system (Varian, Palo Alto, CA), AAA algorithm version 8. Treatment was prescribed to the 95% isodose line with PTV tolerance of ±5%. Quality assurance verification plans were performed with the ArcCHECK™ dosimeter (Sun Nuclear Corporation, Melbourne, FL).

Before each treatment, cone beam CT (CBCT) was used to position the patient appropriately. Doses of 30, 48 and 54 Gy in 5, 4 and 3 fractions respectively were delivered every other day. Dose of radiotherapy was recommended to be 54 Gy for patients who were CP-A and 30 Gy in 5 fractions for CP-B. Some patients also received 48 Gy in 4 fractions when normal tissue constraints could not be met.

The treated tumors were assessed by MRI or triphasic CT eight weeks from completion of SBRT. Following imaging studies were scheduled at the treating physician’s discretion. For the patients in whom alpha-fetoprotein level was elevated before SBRT, repeat tests were done monthly. In cases where the tumor marker indicated progression, an imaging study was performed.

The liver explants were cut in slides in an attempt to reproduce the sections obtained in imaging and to receive correlation with the site of the nodules. Sections from each major tumor nodule and representative samples of small nodules were taken. Satellites nodules, multifocal HCC and intrahepatic metastasis were not distinguished and were considered multiple tumors. Tumor grade and extension were evaluated, including extension into the hilar vessels, inferior vena cava, the hepatic capsule, and margins.

The liver sampling consisted of: (1) Hepatic hilus including hepatic artery, portal vein, and bile duct; (2) Inferior vena cava; (3) Sections from the nodules with the detail of the location; and (4) Section of the cirrhotic liver from the left and right lobes. Conventional hematoxylin-and-eosin stain was performed, as well as special stains, including: Periodic acid–Schiff, Periodic acid–Schiff–diastase, Masson’s trichrome, Reticulum, Ferrum, Orcein, Keratin 19 and 7.

Data were analyzed using the Statistical Package for the Social Sciences 22.0 (SPSS) at a significance level of 0.05. Survival was estimated using the Kaplan-Meier method and Cox’s regression analysis.

The study included 23 patients. Median age at RT was 62.2 (range, 43.0–83.9) years, 69.6% were male and 30.4% female, 70% were viral hepatitis carriers, all had liver cirrhosis. All patients were either CP-A (56.5%) or CP-B (43.5%) and ECOG performance status ≤2. All patients had early stage disease (BCLC stage A). CP-B score was determined based on biochemical findings, no patient had encephalopathy or ascites before being treated with SBRT. Six of the CP-B patients were CP-B score 7 (CP-B7), and 4 were CP-B score 8 (CP-B8). Two patients were previously treated with invasive locoregional methods. In 21 patients, SBRT was the first treatment modality (Table 2).

The median number of lesions was 1 (range, 1–4), the median tumor diameter was 2.5 (range 1.0–4.8) cm, and the median tumor volume was 12.7 (range 2.2–53.6) cm³. The median prescribed dose to the tumor was 54 (range, 30–54) Gy. The median of the mean dose to the uninvolved liver (mean liver dose, MLD) was 6.0 (range, 1.6–12.6) Gy. The median V5/V7/V15 to the uninvolved liver was 33.4% (range, 6.2–48.4%), 27.2% (range, 5.7–44.9%), and 14.6% (range, 2.9–27.0%) respectively. The median mean dose to the right kidney and to 1 cm³ bowel was 0.4 (range, 0.05–8.29) Gy and 7.8 (range, 0.1–17) Gy, respectively.

After a median follow-up time following RT of 12 (range, 2.0–48.7) months, the median OS for the entire cohort was 34.2 (range, 2.0–53.7) months. The median OS for patients who were not transplanted was 23.3 (range, 2.3–34.3) months and was not reached for the transplanted patients (Fig. 1). There was no 30 day radiation-related mortality.

The median PFS for the entire cohort and for the transplanted patients was not reached (as there were too few events of disease progression), and was over 54 months (Fig. 2). The median PFS for patients who were not transplanted was 14.0 (range, 2.3–32.6) months (Fig. 2).

Two patients had grade 1upper gastrointestinal toxicity with abdominal discomfort and grade 1 weakness and fatigue. One patient (CP-B8, treated to 30 Gy) developed RILD two weeks after SBRT that manifested in rising serum total bilirubin levels and INR, as well as new ascites. This patient’s hepatic decompensation was determined to be RILD be a multidisciplinary team. The V15 and mean dose to the uninvolved liver were 9.16 Gy and 5.08 Gy respectively. This patient subsequently underwent urgent and successful transplantation. No other patient had significant changes in 12 weeks of laboratory follow-up.

Eleven of 16 eligible patients (68.7%) underwent orthotopic liver transplantation. The surgical teams reported no significant surgical difficulties. All tumors conformed to the Milan Criteria. The median time to transplantation was 4.8 (range, 0.16–8.5) months. Of these 11 patients, 3 (27.3%) achieved pathological complete response (pCR), 6 (54.5%) achieved pathological partial response (pPR), and 2 (18.2%) achieved pathological stable disease (pSD) (Table 3). Local control was achieved in all patients. A patient who was transplanted 5 weeks after SBRT died of sepsis 4 weeks after transplantation. This patient did not experience any surgical/radiation-related complications and died from antibiotic-resistant sepsis while in the intensive care unit (Fig. 3 and Table 4).

Of the 12 patients who were not transplanted, 7 (58.3%) were not candidates for transplantation (elderly patients or beyond the Milan Criteria) and 5 (41.7%) are awaiting transplantation. All 5 patients awaiting transplantation are without progression with a median time from SBRT completion of 6.7 (range, 2.7–32.9) months. Of the 7 non-transplant candidates, 5 (71.4%) had progression within a median time of 6.8 (range, 4.1–14.1) months. Two patients developed a new lesion in the liver - one was re-irradiated, was progression-free for 9 months after re-irradiation, and then died of hepatic decompensation. Another was treated with TACE. Two patients had systemic progression, and one had local progression that was not treated due to poor PS. The 2 remaining patients died of hepatic decompensation and variceal bleeding (10 and 12 months after RT) without tumor progression.