Radiation-induced liver disease (RILD) is an adverse effect that is pathologically characterized as a veno-occlusive disease (VOD) [22], and is categorized into “classic” and “nonclassic” RILD [23]. “Classic” RILD involves anicteric hepatomegaly and ascites, typically occurring within 4 months, and often exhibits severe or fatal complications following conventional radiotherapy for large hepatic volumes [23, 24]. In contrast, “nonclassic” RILD, typically occurring between 1 week and 3 months after therapy, involves liver transaminases elevated more than 5 times the normal upper limit or Common Terminology Criteria for Adverse Events (CTCAE) grade 4 levels in patients with baseline values more than 5 times the normal upper limit within 3 months after completion of radiotherapy, or a decline in liver function [measured by a worsening of Child-Pugh (CP) score by 2 or more], in the absence of classic RILD [23]. Recent advances in imaging and radiation techniques provide high radiation doses to conform to focal HCC, such as SBRT, and several studies listed in Table 14.1 have described good treatment results for SBRT without severe clinical signs of “classic” RILD. However, toxicities of more than grade 3 CTCAE criteria (so called “nonclassic” RILD) is frequently observed, shown in Table 14.1. Particularly, patients with worse baseline liver function are at higher risk for developing “nonclassic” RILD even in SBRT. Lee et al. reported that in their experience of 131 patients with HCC who received 3-dimensional conformal radiotherapy, the incidence of liver complications was significantly increased in patients with CP class B (p = 0.044). They concluded that indicators of liver function status such as CP class may be important and useful parameters for predicting radiation-related liver disease [25]. In addition, in our unpublished data of 65 patients (56 patients with CP class A and 9 patients with CP class B) who underwent SBRT (median 48 Gy/4 fr), 15 patients (23.1 %) developed grade 3 toxicity including 5 patients with decreased platelet counts before SBRT. CP class B was the only significant factor associated with more than grade 3 (CTCAE ver 4.0.) hepatic toxicity. Irradiated liver dose is also risk factor for severe hepatic toxicity. Bujold et al. reported a significantly higher median liver mean dose was observed in patients who developed grade 5 toxicity (CTCAE ver 4.0.) compared with those who did not (18.1 Gy versus 15.4 Gy; p = 0.02) in their phase I/II study [15]. Son et al. reported in their experience of 47 patients with HCC who received SBRT using CyberKnife, 4 (11 %) showed progression of CP class, and multivariate analysis showed that the only significant parameter associated with the progression of CP class was liver volume after receiving a dose of less than 18 Gy [26]. To avoid hepatic toxicity, the dose- volume limits guidelines recommended by the Quantitative Analyses of Normal Tissue Effects in the Clinic (QUANTEC) for normal liver dose constraints of 3–6 fractions of SBRT, have been offered. For example, the MLD (liver minus GTV) should receive <13–18 Gy, or that an MLD of ≥700 mL of the normal liver should receive ≤15 Gy [23].

Gastrointestinal toxicity is another severe problem associated with SBRT for HCC. Kang et al. reported that 5 (10.5 %) of 47 patients experienced more than grade 3 gastrointestinal toxicity including grade 4 gastric ulcer perforation in 2 patients (4.3 %) [18]. In patients with liver cirrhosis, portal hypertension probably affects the gastrointestinal mucosal defensive and healing mechanisms, whereas liver cirrhosis increases gastrointestinal toxicity [27]. It is recommended that target proximity to luminal gastrointestinal tract should be far from tumor more than, i. e., 2 cm [20].