ABSTRACT: Stereotactic body radiotherapy (SBRT) is a novel noninvasive treatment for hepatocellular carcinoma. SBRT can achieve effective local control, but it requires a relatively high input of resources; this systematic review was performed to assess the cost effectiveness of SBRT in the treatment of hepatocellular carcinoma to provide a basis for government pricing and medical insurance decision-making. The PubMed, EMBASE, Cochrane Library, CNKI, Wanfang and SinoMed databases were searched to collect economic evaluations of SBRT for the treatment of hepatocellular carcinoma from the date of database inception through December 31, 2018. Two reviewers independently screened the studies, extracted the data and performed a descriptive analysis of the basic characteristics, methods of economic evaluation and main results, as well as the quality and heterogeneity of the reports. A total of 5 studies were included. Among them, the level of heterogeneity was relatively acceptable, with a median score of 90%. Four studies were cost-utility analyses (CUAs), and 1 was a cost-effectiveness analysis (CEA). The incremental cost effectiveness ratio (ICER) for sorafenib compared to SBRT was US $114,795 per quality-adjusted life year gained (cost/QALY) in patients with advanced hepatocellular carcinoma. The ICER for proton beam therapy compared to SBRT was US $6465 in patients with inoperable advanced hepatocellular carcinoma. The ICER for SBRT compared to RFA was US $164,660 for patients with unresectable colorectal cancer (CRC) with liver metastases and US $56,301 for patients with early-stage hepatocellular carcinoma. For patients with inoperable localized hepatocellular carcinoma, compared with RFA-SBRT therapy, the ICERs for SBRT-SBRT and SBRT-RFA were US $558,679 and US $2197,000, respectively; RFA-RFA was dominated. In conclusion, there is limited evidence suggesting that SBRT could be cost-effective for highly specific subpopulations of HCC patients, and further economic evaluations based on randomized controlled trials (RCTs) or cohort studies are needed.
