We retrospectively reviewed the hospital records, laboratory results, and imaging studies for patients with HCC who received radiotherapy at our institution between January 2008 and January 2013. Patient eligibility criteria included HCC confirmed by clinical or histological examination, inoperability due to underlying disease or technical unresectability, unfeasible percutaneous radiofrequency ablation, receipt of definitive radiotherapy, good general condition with Eastern Cooperative Oncology Group (ECOG) performance status of 2 or less, a Child-Pugh classification of A or B, no extrahepatic metastases, and a follow-up duration of at least 12 months.
Each patient underwent basic laboratory studies and liver function tests including α-fetoprotein (AFP) concentration detection, abdominal ultrasonography, and computed tomography (CT). Most patients also underwent liver magnetic resonance imaging (MRI). A diagnosis of HCC was based on the practice guidelines of the Korean Liver Cancer Study Group. The cancer stage of each patient was assigned based on the American Joint Committee on Cancer staging system (7th edition).
The institutional review board of our institution approved this study, and the research was carried out in compliance with the Helsinki Declaration.
For CT simulation, patients were immobilized supinely with their arms above their heads using posterior vacuum bags and anterior vacuum-sealed cover sheets (BodyFix, Medical Intelligence Medizintechnik GmBH, Schwabmünchen, Germany). To reduce the movement of the liver during respiration, patients were instructed to take shallow breaths. All patients received intravenous contrast agents, and axial CT images were acquired with a 3-mm slice thickness.
The simulation CT data were transferred to the Hi•Art Planning Station (TomoTherapy Inc., Madison, WI, USA) for inverse planning. The gross tumor volume (GTV) was delineated according to all tumors identified on the abdominal CT and MRI scans. Subsequently, a 5-mm margin was added to create the clinical target volume (CTV), and the planning target volume (PTV) was created by adding an additional 10- to 15-mm margin to the CTV, taking into account target movement during respiration.
The prescription dose was determined by the physician according to the patient’s general condition, PTV, and the radiation dose to normal liver. A daily dose of 2 to 4 Gy was delivered at 5 fractions per week, resulting in a total dose of 40 to 60 Gy. The biologically equivalent dose was calculated using a linear quadratic model with respect to acute tumor effects as an α/β ratio of 10.
We evaluated each treatment plan using a dose-volume histogram and visually inspecting isodose curves. In general, we considered plans acceptable if the PTV was covered by 95% isodose curves, inhomogeneity of the PTV ranged from 95% to 107%, and doses to normal structures were limited in their tolerances. The dose constraints for normal liver were as follows: no more than 30% of a normal liver should have received more than 27 Gy, and no more than 50% of a normal liver should have received more than 24 Gy. Additionally, the mean normal liver dose should have been less than 28 Gy. For the spinal cord, the maximum dose constraint was to be less than 45 Gy. The dose constraints for the stomach and small intestine were as follows: no more than 10% of each normal organ should have received more than 50 Gy, and no more than 15% of each normal organ should have received more than 45 Gy. All radiation doses are biologically corrected doses. The biologically equivalent dose was calculated as an α/β ratio of 3.
Radiotherapy was administered using a tomotherapy system (TomoTherapy Inc., Madison, WI, USA). Triangulation marks were used to verify that the patient did not roll and to quickly position the patient correctly. Before each treatment, a 3.5-MV fan beam CT image was acquired using a CT detector mounted on a ring gantry and matched to the planning CT image for comparison. Then, if necessary, the patient’s position was corrected.
Outcome evaluation and statistical analyses
After treatment, the patients were examined monthly. Liver function, blood cell counts, and AFP concentrations were measured with standard laboratory tests. Treatment responses and tumor recurrence were determined by using CT or MRI every 1 to 2 months.
Treatment response was defined according to the Modified Response Evaluation Criteria in Solid Tumors. An objective response was defined as complete response (CR) or partial response (PR). Local recurrence was defined as the appearance of a new enhanced tumor within the PTV after an objective response, and intrahepatic recurrence was defined as the appearance of a new tumor outside the PTV. Tumors indicating progressive disease (PD) or local recurrence received further treatment, such as trans-arterial chemoembolization (TACE) or surgical resection. Tumors with an objective response without local recurrence or stable disease (SD) received no further treatment. Patients with intrahepatic recurrence were treated for the recurrence.
Radiation-induced general and gastrointestinal toxicities were assessed using the Common Terminology Criteria for Adverse Events, version 4.0. Radiation-induced liver disease and hepatitis B virus reactivation were also evaluated; detailed definitions are given elsewhere. Treatment response time was calculated from the date of radiotherapy completion to the date of the imaging study on which a treatment response was determined. In cases of PR, treatment response time was calculated to the date at which the enhanced tumor stopped becoming smaller. The times to local and intrahepatic recurrences were also calculated from the date of radiotherapy completion.
We also sought to identify factors potentially influencing treatment response time: age, sex, ECOG performance status, tumor size, GTV, pre-radiotherapy AFP concentrations, change in AFP concentration after radiotherapy (calculated as post-radiotherapy concentration/pre-radiotherapy concentration), total radiotherapy dose, daily radiotherapy dose, and pre-radiotherapy TACE.
Actuarial rates were estimated using the Kaplan-Meier method, and groups were compared with log-rank tests for univariate analysis. The Cox proportional regression hazard model was used for multivariate analysis. For all analyses, alpha was set at 0.05. All analyses were performed using SPSS version 18.0 (SPCC Inc., Chicago, IL, USA).