According to a recent report by the International Agency for Research on Cancer, among different types of cancer, liver cancer has the sixth highest incidence in the world, but its mortality is ranked the third highest
[1]. Recently, both the incidence and mortality of liver cancer have decreased in many high-risk countries, including those in East- and Southeast Asia, because of vaccination against the hepatitis B virus (HBV) and the development of treatment against the hepatitis C virus (HCV)
[1]. However, infections with viruses, such as HBV or HCV, are still major causes of cirrhosis and multifocal hepatocellular carcinoma (HCC); therefore, resection was selected as the initial treatment in only 38.3% of patients in Japan
[2]. Recent technical advances in radiation therapy have made it possible to deliver high radiation doses to focal tumors; thus stereotactic body radiotherapy (SBRT) has been widely used in various tumors, such as lung, liver and bone tumors. Promising retrospective results of SBRT in early-stage HCC have reported that local control (LC) rates and overall survival (OS) rates have generally ranged from 66–100% and from 60–70% at 2–3 years, respectively
[3][4][5][6]. However, due to a lack of specific evidence, SBRT should not be considered as a first-line treatment of localized HCC according to several guidelines
[7][8][9]. The 2022 updated Barcelona Clinic Liver Cancer (BCLC) strategy showed that surgery, liver transplantation (LT), and ablative therapies, that is, radiofrequency ablation (RFA), are considered the first treatment options in early-stage HCC
[7]. Although a randomized phase III study comparing SBRT with other modalities has not been reported until now, several prospective phase II studies have shown promising data
[10][11][12][13][14][15][16][17].
2. SBRT in Early-Stage HCC
2.1. Treatment Outcomes
At this moment, SBRT could be regarded as a substitute therapy to surgery, LT, RFA, transarterial infusion chemotherapy or chemoembolization (TACE)
[7][8][9]. Recent prospective studies showed comparable OS to the other modalities, almost 70% in 3 years, with an excellent LC, almost 90% in 2 or 3 years for small HCC
[10][11][12][13][14][15][16][17].
Table 1 summarizes the prospective studies on SBRT in early-stage HCC. The toxicities of SBRT are relatively low, and the incidence of grade III or more toxicities ranged from 2% to 38% in
Table 1. The most frequent adverse effects were associated with liver injury, such as the elevation of total bilirubin and transaminase and the decrease of platelets and ascites. Rim et al. reported a meta-analysis of 32 published studies involving 1950 HCC patients; OS and LC in 3 years were 48.3% and 83.9%, respectively, and grade III or more hepatic and gastrointestinal (GI) toxicities were 4.7% and 3.9%, respectively
[18]. In these series, the Child–Pugh classification (CPC) was a significant prognostic factor for OS and toxicities
[5][6][13][18]. Lasley et al. reported the results of SBRT in 38 patients with CPC-A and 21 patients with CPC-B; the 3-year OS was significant lower in patients with CPC-B (61% for CPC-A, 26% for CPC-B,
p = 0.03). In addition, the incidence of grade III or more toxicities was higher in patients with CPC-B (11% for CPC-A, 38% for CPC-B); patients with CPC-B experiencing grade II or more liver toxicity had significantly higher dosimetric parameters of the normal liver
[13]. GI toxicities have been reported
[11]. Kang et al. reported that 5 (10.5%) of 47 patients experienced more than G3 GI toxicity, including grade IV gastric ulcer perforation in two patients (4.3%). They concluded that preexisting gastro-duodenal disease with cirrhosis was a significant risk factor, because in patients with liver cirrhosis, portal hypertension probably affects the gastrointestinal mucosal defensive and healing mechanisms, whereas liver cirrhosis increases GI toxicity. In general, it is recommended that the target proximity to the luminal GI tract should be more than 2 cm from the tumor. The incidence of central liver toxicities, such as central biliary tract (CBT) stenosis and portal vein (PV) thrombosis, are not so high
[19][20][21]. Eriguchi et al. reported that only two patients (3.6% in 55 patients) experienced asymptomatic bile and concluded that SBRT for liver tumors adjacent to the CBT was feasible with minimal biliary toxicity
[19]. However, Toesca et al. reported that grade III ≥ CBT stenoses were observed in seven patients (17.5% of 40 patients)
[20] They recommended the limiting dose of CBT to be V
BED1040 < 37 cc and V
BED1030 < 45 cc. Takahashi et al. reported that grade III ≥ PV thrombi were observed in three patients (4.8% in 63 patients)
[21]. They concluded that PV thrombosis may be needed to be considered in patients with a higher Child–Pugh class, with higher doses received to 2% of the PV volume.
SBRT is considered an alternative to other modalities, and most patients who undergo SBRT are non-naïve, therefore, previous treatments could affect the treatment results of SBRT, especially the OS in previous reports. Only a few reports have been published on SBRT in naïve patients
[16][17][22]. Durand-Labrunie et al. reported on a prospective phase II study of SBRT using 45 Gy in 3 fractions in 43 naïve patients. The 2-year LC and OS were 98% and 69%, respectively, with 31% toxicities being greater than grade III
[16]. Kimura et al. also reported on a prospective phase II study of SBRT using 40 Gy in 5 fractions in 36 naïve patients and showed that the OS and LC in 3 years were 78%, 90%, respectively, with 11% of toxicities being greater than grade III
[17]. Considering that the eligible patients of these prospective studies were not suitable for resection, liver transplantation and RFA, the results of SBRT were comparable to those of the first treatment options in naïve patients
[23][24].
Table 1. Prospective studies of SBRT in early-stage HCC.
Abbreviations: * BCLC: Barcelona Clinic Liver Cancer, ** N.A.: not available, # CPC-A/B: Child-Pugh class A/B.
There are several unresolved issues in SBRT in early-stage HCC. The first issue is the timing of response evaluation. The evaluation method for SBRT is usually judged by whether there is early arterial enhancement of the tumor using dynamic enhanced CT or MRI
[25][26]. According to the Modified Response Evaluation Criteria in Solid Tumors (mRECIST), a complete response (CR) was defined as the “Disappearance of any intratumoral arterial enhancement in all target lesions”
[25]. However, this intratumoral arterial enhancement may be prolonged over 6 or more months in several cases, and this fact confuses the timing of response evaluation. In a previous report, Kimura et al. evaluated the patterns of dynamic enhanced CT appearance of tumor responses after the completions of SBRT. They observed residual early arterial enhancement in 19 lesions (28.4%) more than 3 months after SBRT in 59 patients with 67 tumors
[27]. Based on this result, researchers propose that the response evaluation at 6 months, not at 3 months, after the completion of SBRT is the appropriate time point, because in most cases there was ab observed disappearance of residual early arterial enhancement within 6 months.
Figure 1 shows a typical case of CR at 3.5 months after the completion of SBRT. The second issue is the optimal dose-fraction schedule. The several dose-fraction schedules are shown in
Table 1. Kim et al. analyzed the dose–response relationship in a multi-institutional retrospective cohort that included 510 patients treated with SBRT
[28]. Patients treated with a biological effective dose (BED) ≥ 100 Gy showed a better 2-year freedom from local progression (FFLP) and OS than did patients treated with a BED < 100 Gy (FFLP, 89% vs. 69%; OS, 80% vs. 67%;
p < 0.001). In addition, a multivariate analysis before and after propensity score-matching (PSM) in 198 selected patients between BED ≥ 100 Gy and BED < 100 Gy, identified BED ≥ 100 Gy as the main prognostic factor for both FFLP and OS (
p < 0.01). Higher dose-fraction schedules may improve LC and OS. The third issue is the combination with TACE. Kimura et al. compared SBRT alone (28 patients) with SBRT+TACE (122 patients) in small HCC, the median tumor size was < 20 mm, retrospectively
[29]. The 2 year OS and local progression-free survivals (LPFS) for SBRT alone and SBRT+TACE groups were 78.6% and 80.3% (
p = 0.6583) and 71.4% and 80.8% (
p = 0.9661), respectively. On the other hand, Su et al. also compared SBRT alone (50 patients) with SBRT+TACE (77 patients) in large HCC, with median tumor size being 85 mm, retrospectively
[30]. The 5 year OS was significantly higher in the SBRT + TAE/TACE group (46.9%) than that in the SBRT alone group (32.9%;
p = 0.049). The LPFS did not differ significantly between the two groups. These opposite results suggested that the combination of SBRT and TACE has the potential to improve treatment results, compared with SBRT alone, especially in patients with larger HCC, such as those of >5 cm, but SBRT alone could be a significant treatment option for patients with small HCC, such as those of <2 cm. To resolve these issues, further prospective studies are warranted.
Figure 1. A typical case of complete response at 3.5 months after the completion of SBRT. (A) Dynamic MRI appearance (arterial phase) before SBRT; the early arterial enhancement is obvious (red arrow). (B) Dynamic MRI appearance (arterial phase) after 3.5 months; the early arterial enhancement has disappeared (red arrow). (C) Dose distribution of SBRT: the prescribed dose of 40 Gy covered 95% of PTV with 125% maximum dose of 40 Gy (80% isodose) in 4 fractions.
2.2. Comparison to the Other Treatments
Recently, several PSM studies, which compared SBRT with the different modalities of local therapy, have been published. Table 2 summarizes these PSM studies in early-stage HCC.
TACE is considered an alternative to resection or RFA, and its eligibility is regardless of tumor conditions, such as its size, number and location in several guidelines
[7][8][9]. Spair et al. reported the comparison between TACE in 84 patients with 114 HCCs and SBRT in 125 patients with 173 HCCs using PSM
[31]. SBRT produced significantly better LC (2 year: 91% for SBRT and 23% for TACE,
p < 0.001) with lower grade III (and lower) toxicity (8% for SBRT and 13% for TACE,
p = 0.05) despite no difference in OS. A meta-analysis, which compared TACE alone with radiation therapy (RT) including SBRT with TACE, showed that the median survival for TACE with RT (22.7 months) was significantly better than that for TACE alone (13.5 months) (
p < 0.001)
[32].
Several comparative PSM studies with RFA, which is a first-line treatment, have been reported
[33][34][35][36]. Recent large-size studies showed significantly better LC for SBRT compared to RFA and comparable OS and toxicities
[35][36]. On the other hand, Rajyaguru et al. showed opposite outcomes with RFA producing significantly better OS; the 5 year OS of RFA was 29.8% and that of SBRT was 19.3% (
p = 0.001) using the National Cancer Database
[34]. However, several limitations, such as selection bias, have been pointed out by several investigators. Several meta-analyses, which compared SBRT with RFA, showed that the LC of SBRT was better or comparable to that of RFA, especially for a tumor size of ≥ 2 cm, but the opinions were divided regarding OS
[37][38][39]. Pan et al. concluded that OS of SBRT was inferior to that of RFA because of the tumor burden or liver profiles of the enrolled study participants
[37]; in contrast, Wang et al. concluded that SBRT was well-tolerated with an OS equivalent to that with RFA
[39]. These findings could suggest the limitations of these retrospective studies, therefore further investigations are still needed.
Table 2. Comparison with the other modalities.
In comparisons with resection, which is also a first-line treatment, PSM studies from China showed that SBRT produced a similar OS to that in resection
[40][42]; in addition, Su et al. showed that an advantage of SBRT over resection was its lesser degree of invasiveness
[40]. In contrast, Nakano et al. from Japan reported that the 5 year OS and PFS rates for the resection and SBRT groups were 75.2% vs. 47.8% (
p = 0.0149) and 33.8% vs. 16.4% (
p = 0.0512), respectively, and a multivariate analysis showed that resection was a significant favorable factor for OS and PFS
[41]. They concluded that resection should be considered a first treatment option for potentially resectable patients.
Charged particle therapy (CPT), including carbon ion therapy and proton beam therapy (PBT), provides physical and biological advantages compared with SBRT, which uses a photon beam. In a physical aspect, the “Bragg peak”, which is a rapid energy fall-off at a specific depth, allows for the delivery of a very localized dose distribution that potentially reduces the incidence of hepatic toxicity. In a biological aspect, the greater relative biological effectiveness of CPT, especially carbon ion therapy, compared with SBRT, could be expected to improve LC and OS. Based on this background, CPT is promising for large tumors with relatively less toxicity. Qi et al. reported a meta-analysis of the CPT and SBRT in patients with HCC
[43]. The OS for CPT was similar to that for SBRT, and the toxicity tended to be lower for CPT compared to that for SBRT. From 2022, CPT for HCC (≤4 cm) is covered by the national health insurance in Japan. Now, a prospective non-randomized trial on PBT vs. surgery for operable untreated HCC is ongoing (Japanese Clinical Oncology Group, JCOG1305).
Further studies, including randomized phase III studies to define which patients are more suitable for each curative local treatment, are needed.
2.3. Repeated SBRT
According to the latest Japanese survey, recurrence was reported within two years of diagnosis in 50.5% patients with HCC
[2]. Because of the multifocal nature, intrahepatic recurrence is the one most frequently observed in 80–95% of cases
[44], and Imamura et al. reported that two types of recurrence may be distinguished: early and late recurrences
[45]. Early recurrence is considered a metastatic occurrence and late recurrence a multicentric occurrence of HCC, therefore, the late recurrence shares the same risk factors as primary HCC
[45]. Considering these recurrent patterns and their frequencies, repeated locoregional therapies play an important role. The treatment strategy for patients with intra-hepatic recurrent HCC after initial treatments, such as surgery or RFA, should principally follow the same eligibility which was used for naïve HCC patients
[46]. SBRT could also be considered a substitute therapy to surgery or RFA in this situation. Kimura et al. reported the results of repeated SBRT in 81 patients with 189 tumors of two courses or more (median two times; ranged from two to four times)
[47]. The 5 year local recurrence rate, OS and liver-related death rates from the first SBRT were 6.3%, 60.4% and 32.9%, respectively. The 3 year OS and liver-related death rates from the second SBRT were 61.0% and 34.5%, respectively, with almost the same frequency of grade III toxicity between the first and second SBRT (first: 11%; second: 15%,
p = 0.48). Repeated SBRT for patients with intra-hepatic recurrent HCC achieved satisfactory LC and OS without severe toxicities; therefore, SBRT could be a good treatment option in these recurrent cases.