Background
The primary treatment for hepatocellular carcinoma (HCC) is surgery, including hepatic resection and liver transplantation, which results in 5-year survival rates of 30–70 % [
1]. However, <20 % of HCC patients are eligible for surgery. In unresectable cases, local therapies such as radiofrequency ablation (RFA) or percutaneous ethanol injection (PEI) are reportedly effective, but not all patients are eligible for these treatments due to the tumor’s percutaneous inaccessibility, its invisibility via ultrasonography, or bleeding risks. Transcatheter arterial chemoembolization (TACE) has been widely used as the first-line non-curative therapy for HCC cases that are non-surgical or unsuitable for local ablative therapies [
2]. However, TACE alone rarely produces a complete response, and additional treatments are often required. Multiple repetitive sessions of TACE have been widely performed in Korea, but such treatments deteriorate liver function, increase TACE-related adverse effects, and offer less therapeutic efficacy due to vascularity decrease [
3]. Therefore, various modalities such as RFA, PEI, sorafenib, conventional radiotherapy (RT), and stereotactic ablative radiotherapy (SABR) have been suggested in addition to TACE, but a definitive guideline has not been established [
4,
5].
The role of RT in HCC is limited owing to the liver’s low tolerance to radiation and the risk of radiation-induced liver damage [
6,
7]. However, recent radiotherapeutic developments have gradually expanded the indications for external beam radiotherapy from palliative to curative with high doses of radiation safely delivered to the tumor while avoiding adverse effects to the liver function. In several studies on three-dimensional conformal radiotherapy (3D-CRT) to primary HCC, substantial effects of RT have been observed [
8‐
10]. With the introduction of SABR, it is now possible to accurately deliver more radiation doses to tumors using fewer fractions while sparing the normal liver tissue [
11,
12]. Recent clinical data have demonstrated the feasibility of SABR for HCC treatment with high local control (LC) and overall survival (OS) rates and low treatment-related severe toxicities [
13‐
21].
However, there remains a lack of randomized studies comparing the effects of TACE combined with RT and those of TACE alone. Although several phase II trials of SABR have reported outstanding results, including LC rates of > 90 % and 5-year OS of > 50 % in well selected cases of HCC, RT is still not endorsed as a curative treatment option for HCC in most guidelines or consensus strategies [
1,
22,
23]. Therefore, this retrospective study aimed to provide the basis for initiating a randomized trial to investigate whether a combination of SABR and TACE would improve the long-term OS compared with TACE alone. The survival outcomes of TACE plus SABR were compared with those of other treatment modalities combined with TACE or TACE alone.
Discussion
In previous phase I and phase II studies conducted at our institute [
4,
19], the feasibility of SABR for the treatment of primary HCC with incomplete response to TACE has been demonstrated with a 2-year LC rate of 94.6 % and a 2-year OS rate of 68.7 %. In the long term follow up study [
20], LC and OS rates at 5 years were 82 % and 39 %, respectively. Furthermore, at the last follow-up (4.5 years), patients receiving high-dose SABR (>54 Gy) reported a 100 % LC rate and a 68 % OS rate, which were comparable to RFA outcomes. Based on the results of these studies of SABR, HCC cases with a CTP score of 5–7, total sum of tumor size < 10 cm, and ≤ 3 nodules seem to be feasible for curative SABR. Therefore, to investigate whether the addition of SABR to TACE offers a survival benefit to HCC patients with incomplete response to TACE, we enrolled those meeting the above-mentioned eligibility criteria. In this study, the addition of SABR after incomplete TACE demonstrated significant survival benefits compared with additional non-curative treatments such as repeated TACE, sorafenib, and chemotherapy. Moreover, it also showed similar survival outcomes to those of the good prognostic groups, such as patients with complete response to TACE or those undergoing surgery with incomplete TACE.
TACE is widely adopted as the current standard of care for HCC patients at BCLC intermediate stage. Additionally, it might be indicated for patients with early stage HCCs that are unresectable or ineligible for local ablative therapies such as RFA or PEI due to tumor location or other medical conditions. Although complete response to TACE has not yet been defined in radiological findings, compact lipiodolisation is an important goal of TACE. Cabibbo et al. [
25] reported that a complete radiological response after TACE significantly increases OS, and thus proposed utilizing the observation as a surrogate treatment endpoint. Kim et al. [
26] indicated that compact lipiodol uptake following TACE predicted favorable survival outcomes for unresectable HCC. The 1-year and 5-year OS rates of patients with compact lipiodolisation were 93 % and 52 %, respectively, versus 61 % and 17 %, respectively, in patients with noncompact lipiodolisation. Similarly, in our current study, the OS rates of group 1 patients with complete response to TACE at 1 and 5 years were 96 % and 53 %, respectively. In addition, only 14 % of the enrolled patients in this study achieved a complete response after 1–4 sessions of repeated TACE. Furthermore, approximately 40 % of the patients were in stage 0, and almost 80 % had tumors <3 cm, suggesting that group 1 patients had relatively lower BCLC stages and smaller tumors than those in the incomplete response groups. Therefore, although group 3 patients, who received SABR after incomplete response to TACE, might have had poorer prognostic factors than group 1 patients, similar OS rates were observed between the 2 groups.
Since a compact radiological response after TACE significantly increases OS, technical improvements and refinements have been described for transarterial administrative methods [
3,
27]. However, many patients still experience incomplete response after TACE, especially in cases of tumors >3 cm [
28]. Several studies [
29‐
32] have shown that the combination of TACE and other modalities, mainly surgery, RFA, PEI, or RT, was associated with higher survival rates. According to a prospective cohort study by Lee et al. [
30], in which surgical resection after primary TACE was compared with TACE alone, the OS rate was significantly higher for the surgical resection group than for the TACE alone group (5-year OS rate, 56 % vs. 23 %). Studies on TACE plus RFA showed that patients in the TACE plus RFA group had better LC rates than patients in the TACE alone group [
31], and better OS rates than patients in the RFA alone group [
32]. TACE plus RFA have also been reported to provide similar OS rates to those achieved with surgical resection [
33]. Several randomized trials were conducted to compare TACE plus PEI and TACE alone [
34‐
36], suggesting that TACE plus PEI performed better than either TACE or PEI alone, and the reported 3-year OS rates for TACE plus PEI were approximately 35 %–65 %. In our current study, group 2 patients underwent surgery, RFA, or PEI after incomplete TACE. This group only included patients with CTP class A. Our findings suggested that TACE might have been performed as a bridge therapy in patients who were initially eligible for curative treatments and subsequently underwent additional surgery. Also, group 2 patients had significantly lower CTP scores than the non-curative treatment group 4 patients. Therefore, group 2 patients might initially have had good prognostic factors, and our results also indicated a significant survival benefit in this group compared with group 4. Additionally, although not statistically significant, group 2 consisted of patients with relatively lower CTP scores compared to group 3, and smaller tumors compared to groups 3 and 4. Therefore, although group 3 patients might have had poorer prognostic factors than group 2 patients, similar OS rates were observed.
The benefit of additional 3D-CRT to incomplete TACE over TACE alone in unresectable HCC has also been reported. Meng et al. [
10] conducted a meta-analysis of 17 trials, including 5 randomized controlled and 12 non-randomized controlled studies. Although TACE plus RT showed significantly improved survival and tumor response compared with TACE alone in this meta-analysis, none of the 5 randomized trials reported the random allocation sequence in detail, and most of the original studies included were published in Chinese, making general acceptance challenging. On the other hand, in a retrospective study, Shim et al. [
5] reported that a combination of 3D-CRT and incomplete TACE significantly improved survival rates compared with TACE alone. Their study included single tumors ≥5 cm, and the 2-year OS rates of the TACE plus RT and TACE alone groups were 37 % and 14 %, respectively. The combination of SABR and TACE has previously been studied by Honda et al. [
21]. Their results showed that SABR combined with TACE was safe and effective for loco-regional treatments, increasing both LC and OS compared with TACE alone. The 1-year and 3-year OS rates for the SABR group were 100 % and 100 %, respectively, whereas the corresponding rates for the TACE group were 89 %, and 66 %, respectively. Although excellent OS rates were reported, patient selection was very strict, with only solitary tumors and small tumors ≤3 cm included. Our results also indicated a significant survival benefit of TACE plus SABR (group 3) over TACE alone (group 4) with similar baseline characteristics between the 2 groups.
There is not enough evidence to establish any combination therapy as a standard treatment for HCC after incomplete TACE. In our study, combining SABR to TACE offered similar survival outcomes to TACE and curative treatments, and survival benefits over repeated TACE, suggesting that SABR might be recommended as a treatment modality after TACE failure. Furthermore, SABR, being a noninvasive procedure, might be more advantageous than other invasive curative modalities such as surgery, RFA, or PEI. However, since this was a retrospective study, patients were not controlled with respect to variable prognostic factors. Additionally, the dose and fractionation schedules of SABR and the number of previous TACE sessions prior to SABR are still not well defined and are yet to be determined. Therefore, further randomized trials are needed to validate our results before SABR can be recommended routinely. Hence, based on the results of this study, a multicenter randomized controlled trial is being planned to investigate the potential benefits of SABR as an alternative modality in the treatment of HCC after incomplete TACE.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
EKP participated in the design of the study, literature research, data acquisition, data analysis, data interpretation, statistical analysis, manuscript preparation, and manuscript editing. MSK contributed to study concepts, study design, definition of intellectual content, data analysis, data interpretation, manuscript editing, and manuscript review. All authors made substantial intellectual contributions to drafting the article, revising the article, data analysis, and data interpretation. All authors read and approved the final manuscript.