Radiation-induced liver disease in three-dimensional conformal radiotherapy for primary liver carcinoma

Abstract

To identify the risk factors of radiation-induced liver disease (RILD) in three-dimensional conformal radiotherapy (3DCRT) for primary liver carcinoma (PLC) and find a dosimetric threshold for RILD. Between April 1999 and August 2003, 128 patients with PLC were treated by 3DCRT at Cancer Hospital, Guangxi Medical University. All of the patients were technical unresectable or medical inoperable due to poor liver function or cardiovascular diseases. The clinical characteristics of these patients were as follows: 113male, 15 female; median age of 48.2(27–72); with portal vein thrombosis (PVT) in 34 cases, without in 94 cases; liver cirrhosis of Child-Pugh grade A in 108 cases, Child-Pugh grade B 20 cases. 3DCRT was carried out by 8MV x-ray with Topslane treatment planning system. In 48 patients transarterial chemoembolization (TACE) was performed prior to 3DCRT with DDP, ADM and MMC. 3DCRT was delivered by4.88 ± 0.47Gy(4-8Gy)/fraction, three fractions per week (Mon., Wed. Fri.) with a median total dose of 53.6Gy. The mean value of gross target volume (GTV) was 458.92 ± 429.8 cm3. RILD was defined as either anicteric elevation of alkaline phosphatase level of at least twofold and non-malignant ascites (classic RILD), or elevated transaminases of at least fivefold the upper limit of normal or of pre-treatment level(non-classic RILD). The diagnosis of RILD should be distinguished from the progression of PLC, which occurred within 4 months. Parameters evaluated for the occurrence of RILD included: gender, age, GTV, AFP level, HBV, PVT, TACE, Child-Pugh grade of liver cirrhosis. Among 128 patients, 84 patients had complete 3-dimensional dose-volume data, and these dosimetric parameters, including DVH were also analyzed. During a median follow-up time of 17.2 months (4–56) after 3DCRT, 19 patients were diagnosed as RILD with the incidence of 14.8%(19/128). Classic RILD 13cases,non-classic RILD 6 cases. 16 cases died of hepatic failure, 8 of them died within 4 months after completion of radiation with the median survival time of 5.2 (1–14) months. GTV, PVT and Child-Pugh grade of liver cirrhosis had correlation with occurrence of RILD, in favor of small GTV, without PVT and Child-Pugh grade A (p = 0.000, 0.002, 0.001, respectively). The Grade of hepatic toxicity of Common Toxicity Criteria (Version 2.0) during 3DCRT also related to RILD (p = 0.000). Ten of 84 patients, who had 3DCRT dosimetric data developed RILD after treatments. When whole liver volume was taken as a single organ, which included normal liver and tumors, the mean whole liver dose was significantly higher in patients with RILD than those without (27.5Gy ± .8.5Gy vs.33.9Gy ± 5.9Gy,p = 0.027). Multivariate analysis demonstrated that only Child-Pugh grade of liver cirrhosis played an independent factor (p = 0.000, RR = 29.7) for occurrence of RILD with much high incidence for Child-Pugh B patients. In Child-Pugh grade A patients, the threshold of dosimetric parameters, which would not produce >5% of RILD was 81% for V5, 69% for V10, 51% for V15, or 42% for V20.For Child-Pugh A patients, when mean normal liver (excluding tumors) dose was ≤19Gy(33 cases), there was no case of RILD (0%),whereas, when mean normal liver dose was over 19Gy,the incidence of RILD was 2/36(5.6%).The incidence of RILD were 0/38(0%) and 2/31(6.5%), respectively for mean whole liver dose of ≤28Gy and >28Gy. In Child-Pugh grade B patient, the probability of developing RILD was 53% (8/15),which implied that the doses used in this study was not tolerable. Liver cirrhosis was the most critical risk factor for occurrence of RILD when PLC was treated by irradiation. For Child-Pugh grade A patients, the safe threshold dose could be 81% for V5, 69% for V10, 51% for V15, or 42% for V20, and mean normal liver dose of 19Gy or mean whole liver dose of 28Gy when 3DCRT was carried out by fractionation used in this study. The fractionation and total doses implemented in this study was not tolerable for PLC patients with liver cirrhosis of Child-Pugh grade B, and not recommended for the further clinical trials