Sorafenib plus drug-eluting bead transarterial chemoembolization for early intrahepatic stage-progressed advanced hepatocellular carcinoma refractory to conventional transarterial chemoembolization

To investigate the effectiveness and safety of the combination of sorafenib and drug-eluting bead transarterial chemoembolization (DEB-TACE) in the treatment of early intrahepatic stage-progressed advanced hepatocellular carcinoma (ISPA-HCC). This study was approved by the ethics committees of six tertiary medical centers in China. Between October 2017 and October 2020, 213 patients with advanced HCC received either sorafenib combined with on-demand DEB-TACE (DTS group, n = 103) or sorafenib monotherapy (S group, n = 110). Overall survival (OS), time to progression (TTP), local tumor response, and adverse events (AEs) were compared between the two groups. The incidences of nause/vomiting, abdonimal pain, hyperbilirubinemia, fever and ALT/AST increasing were higher in the DTS group. The post-treatment partial response, objective response, and disease control rates were significantly higher in the DTS group than in the S group (51.5% vs. 23.6%; 56.3% vs. 25.5%; 77.7% vs. 56.4%, respectively). The median OS was significantly longer in the DTS group than in the S group [16.3 vs. 10.0 months; hazard ratio (HR) = 0.43; P < 0.001], as was the TTP (6.7 vs. 4.3 months; HR = 0.60; P = 0.001). In the DTS group, patients who received ≥ 2 sessions of DEB-TACE benefited more than those who received two sessions of DEB-TACE. Multivariate analysis revealed that the α-fetoprotein level and treatment allocation were independent predictors of OS and TTP. The combination of sorafenib and DEB-TACE is safe and effective for the treatment of early ISPA-HCC.


Introduction
Hepatocellular carcinoma (HCC) is the most frequently occurring and third-deadliest primary liver malignancy worldwide (Sung et al. 2021). Transarterial Wenzhe Fan, Bowen Zhu and Xinlin Zheng contributed equally to this work.
1 3 chemoembolization (TACE) is an effective therapeutic option for unresectable HCC (Sieghart et al. 2015;Lencioni et al. 2016a). Conventional TACE (cTACE) with oilbased chemoembolization provides survival benefits, but the 5-year tumor recurrence rate is 70% because of TACE refractoriness, which means an ineffective response after two or more consecutive TACE procedures, continuous elevation of tumor markers, the appearance of vascular invasion, or extrahepatic spread (Kudo et al. 2014). After cTACE resistance, most intermediate HCC develops to stage progression (SP) HCC. SP serves as the surrogate end-point for TACE refractoriness and a crucial turning point in the survival time of patients with intermediate HCC ). Compared to primary advanced HCC, early intrahepatic stage progression advanced HCC (ISPA-HCC)-which is SP-HCC with portal vein tumor thrombus (PVTT) with/ without extrahepatic metastasis-has reduced tumor burden, poor hepatic artery condition, impaired liver function from repeated TACE, and high probability of unresponsive to TACE Peck-Radosavljevic et al. 2019). Hence, in theory, the treatment of ISPA-HCC should have a difference from primary advanced HCC. However, no standard therapeutic regimen has been suggested for patients with ISPA-HCC.
Sorafenib is one of the standard therapies for patients with advanced-stage HCC and is recommended for managing cTACE-refractory HCC by the Japan Society of Hepatology's consensus-based clinical practice guidelines (Kudo et al. 2014;Ikeda et al. 2014). The noninterventional, prospective, observational OPTIMIS study reported that patients who switched to sorafenib immediately after developing refractoriness to TACE had a better prognosis than those who continued TACE (Peck-Radosavljevic et al. 2018). However, studies reporting the survival outcomes of sorafenib treating patients with ISPA-HCC are limited. And an observational study indicated that the median overall survival (OS) of ISPA-HCC patients who suffered progressive disease (PD) and SP was unsatisfactory and far less than patients who did not . Therefore, further investigations are required for the improvement of sorafenib efficacy.
Drug-eluting bead TACE (DEB-TACE) selectively delivers high doses of chemotherapeutic agents over an extended period, thereby minimizing the blood drug concentration, related systemic effects, and need for embolic agents (Brown et al. 2012). Previous studies have shown that DEB-TACE is superior to cTACE considering its high local response rate, reduced implications, and low liver toxicity (Lammer et al. 2010), and has a local effect on cTACE-resistant HCC (Xiao et al. 2019;Song et al. 2013). For advanced HCC, DEB-TACE has been demonstrated as safe and yields an objective response rate of approximately 69.5% (Liu et al. 2020), which can rapidly reduce tumor burden and enhance the anti-tumor effect of sorafenib (Pawlik et al. 2011). Both the TACE-2 and SPACE studies indicate that DEB-TACE combined with sorafenib is safe and has an increased survival effect in patients with unresectable HCC, with progression-free survival rates of 238 and 169 days, respectively (Meyer et al. 2017;Lencioni et al. 2016b), providing evidence for the continued use of DEB-TACE in ISPA-HCC patients. DEB-TACE is also a potential synergistic agent with sorafenib for the treatment of early ISPA-HCC patients after cTACE resistance.
This study compared the efficacy and safety of this combined treatment with sorafenib monotherapy. We hypothesized that sorafenib plus DEB-TACE is safe and effective in early ISPA-HCC patients with good liver function.

Study design
This retrospective multicenter study collected data from six tertiary medical centers. The study protocol was approved by the institutional review boards. As this was a retrospective study, the requirement for obtaining informed consent from the patients was waived.
The eligibility criteria for enrollment in this study were: (a) age 18-75 years; (b) HCC diagnosed according to the European Association for the Study of the Liver/American Association for the Study of Liver Diseases (Bruix et al. 2005; European Association for the Study of the Liver, 2018); (c) patients with ≥ 1 typical enhanced measurable target lesion based on the modified response evaluation criteria in solid tumors (mRECIST) criteria (Lencioni and Llovet 2010), who had not received DEB-TACE or systemic therapy; (d) PVTT initially observed during follow-up of cTACE-treated intermediate HCC; (e) Child-Pugh class A liver function after stage progression; (f) Eastern Cooperative Oncology Group performance status score of 0; and (g) adequate renal function (serum creatinine concentration of 1.5-fold the upper limit of the normal range or less).
The exclusion criteria were: (a) PVTT in the main portal vein, (b) curative treatment including resection and ablation after stage progression, and (c) severe dysfunction of the heart, kidney, or other organs.
Electronic medical records of 309 patients with ISPA-HCC from the intermediate stage after cTACE-who had been treated with either sorafenib monotherapy or DEB-TACE combination therapy between October 2017 and October 2020-were reviewed. Overall, 213 patients met the inclusion criteria. DEB-TACE combined with sorafenib was administered to 103 patients (DTS group), while the remaining 110 patients received sorafenib monotherapy (S group) (Fig. 1).

Treatments
Sorafenib treatment (400 mg) was administered orally twice a day after detection of stage progression. The patients received continuous sorafenib with no breaks before or after repeated DEB-TACE. The dose was reduced to 400 mg once daily with grade 3 or 4 hematologic toxicity, skin toxicity, gastrointestinal toxicity, hypertension, or hepatic dysfunction, until the adverse events (AEs) were alleviated or eliminated. If these continued after dose adjustment, sorafenib treatment was halted until their disappearance.
On-demand TACE was performed according to the tumor condition. All TACE procedures were performed by interventional radiologists. Standard angiographic facilities and protocols were used for hepatic angiography and catheterization. Imaging of the celiac and superior mesenteric arteries was performed in all patients to evaluate the circulation of the liver vasculature prior to treatment. Super-selective catheterizations were performed in every embolization of DEB-TACE procedures, if possible. However, in patients with bilobar multinodular disease, the lobar artery was selectively catheterized.
The doxorubicin-capable beads (DC Beads™; Biocompatibles, Farnham, UK) used were either 100-300 or 300-500 μm. Each vial of DC Beads mL beads) was loaded with 75 mg of doxorubicin dissolved in sterilized water. After loading for 30 min, at least 5-10 mL of nonionic isotonic contrast [270 mg/mL Visipaque (iodixanol); GE Healthcare, Princeton, NJ, USA] was injected into the vial per 1 mL doxorubicin-DEB. A 10 mL suspension of doxorubicin-DEB was aspirated into a syringe and injected as recommended previously (Lencioni et al. 2012). The embolization protocols rarely necessitated the use of supplementary embolic materials to avoid doxorubicin-DEB overdose, in which case 300-500-μm Embosphere Microspheres (Merit Medical, Salt Lake City, UT, USA) were required for complete devascularization ( Fig. S1 shows a typical patient).
Doxorubicin-DEB doses were adjusted according to the tumor diameter (based on the ellipsoid volume, i.e., height × width × length × π/6). The end-point of primary chemoembolization was complete devascularization of HCC observed on angiograms (Lencioni et al. 2012).

Outcomes and assessments
OS was measured from the first administration of sorafenib until death or the last follow-up. Time to progression (TTP) was defined as the time from the first sorafenib treatment until the detection of progressive disease. Patients followed up once per month. Within 1 week prior to the treatment, all patients underwent triphasic contrast-enhanced CT or MRI, and all parameters, including serum α-fetoprotein (AFP) level and hepatic function, were documented. Tumor response and safety were assessed at 1 month intervals until death or progression. Once any residual tumor or recurrence was observed, additional DEB-TACE was performed on demand according to the criteria described above. The mRECIST criteria were used to assess the efficacy of local tumor response according to images acquired 1 month after TACE (EASL 2018). In each institution, measurements were performed by two independent radiologists from the Department of Medical Imaging, with a consensus review performed for equivocal cases by a third experienced radiologist. Triphasic contrast-enhanced CT and MRI were used for follow-up imaging. The best overall response during treatment was considered the final response. The last follow-up date was June 31, 2021.
AEs were classified according to the Common Terminology Criteria for Adverse Events v. 5.0 (NIH 2017). Abdominal pain, nausea, and vomiting that occurred Fig. 1 Flowchart showing the procedure for the selection of patients. BCLC Barcelona Clinic Liver Cancer, DTS group patients treated with drug-eluting bead transarterial chemoembolization plus sorafenib, ECOG Eastern Cooperative Oncology Group (performance status), S group patients treated with sorafenib monotherapy within 14 days after the TACE procedure were considered post-TACE syndrome (Vogl et al. 2009) and were not recorded as sorafenib-related AEs. Hepatic reserve function was evaluated using the albumin-bilirubin (ALBI) score. The ALBI score was calculated based on the total bilirubin (TBil) and serum albumin (ALB) levels using the following formula: ALBI score = [− 0.085 × ALB (g/L)] + [0.66 × log10 TBil (µmol/L)], and it was categorized into three grades based on the following scores: ≤ − 2.60 = grade 1, > − 2.60 to ≤ − 1.39 = grade 2, and > − 1.39 = grade 3 (Hiraoka et al. 2017).

Statistical analysis
Continuous variables and categorical data were expressed as mean ± standard deviation and frequencies, respectively, whereas between-group comparisons were evaluated using either Student's t test, Pearson's Chi-squared test or Fisher's exact test. Survival curves were estimated using the Kaplan-Meier analysis, and univariate analysis was performed using the log-rank test. Multivariate analysis was performed using Cox regression analysis for the significant variables identified in the univariate analysis. All statistical tests were two-sided, and between-group differences were considered significant at P < 0.05.

Patient characteristics
A retrospective review of the records of 309 consecutive patients with stage-progressed advanced HCC between October 2017 and October 2020 was performed. A total of 213 patients were enrolled in the study. Of these, 103 patients consented to receive the combination of sorafenib and DEB-TACE (DTS group), whereas the remaining 110 patients consented to receive sorafenib monotherapy (S group) ( Table 1). The median follow-up duration was 11.3 months (2.4-36.6 months) and 9.5 months (2.4-27.5 months) in the DTS and S groups, respectively. The average pre-cTACE times were 2.6 and 2.4 in the DTS and S groups, respectively. In the DTS group, all patients received repeated on-demand DEB-TACE; 56 (54.4%) patients received two DEB-TACE sessions and 47 (45.6%) patients received ≥ 3 DEB-TACE sessions. No significant differences were observed in the baseline characteristics between the two groups. Subsequent treatment after discontinuation was shown in Supplemental Table 1.

Safety
AEs associated with sorafenib treatment are shown in Table 2. Only events that occurred in ≥ 10% of the patients in each group were recorded. The median duration of sorafenib administration was 5.9 and 4.0 months in the DTS and S groups, respectively. The median daily dose of sorafenib was 600 mg for both groups. The rates of dose reduction and interruption due to AEs were 57.3% and 47.3% in the DTS group and 47.6% and 34.5% in the S group, respectively. No unexpected AEs or drug-related deaths occurred within 30 days of final dose administration.
The following AEs of any grade were more common in the The mean changes in the ALBI score 7 days after chemoembolization were consistent between the two groups (P = 0.407) (Fig. S2).
The forest plots show the results of the subgroup analysis of OS and TTP (shown in Figs. 4 and S4). The DTS group was superior to the S group for the OS in most subgroups except without HBV-infection, main tumor size ≤ 5 cm, PVTT in the first portal vein branch, and with extrahepatic spread. Similar advantages of the DTS group were observed in the subgroup analysis of TTP. In the DTS group, the patients who received more sessions of DEB-TACE (≥ 2 sessions) benefited more from the combined therapy, regarding OS (HR = 0.53, P = 0.027), than those who received two sessions of DEB-TACE (Fig. S5). Univariate analysis revealed that the AFP level (P < 0.001), intrahepatic tumor number (P = 0.038), and treatment type (P < 0.001) were significant prognostic factors of OS, whereas the AFP level (P < 0.001) and treatment type (P = 0.001) were significant determinants of TTP. Multivariate analysis showed that the AFP level (HR = 1.96; P < 0.001) and treatment type (HR = 0.44; P < 0.001) were independent predictors of OS, whereas the AFP level (HR = 1.97; P < 0.001) and treatment type (HR = 0.60; P = 0.001) were independent predictors of TTP (Table 3).

Discussion
Survival time is poor once a patient enters ISPA-HCC, and the treatment at this stage seriously affects the OS. This study evaluated the efficacy of the combined treatment with sorafenib plus DEB-TACE in patients with early ISPA-HCC. We found that the combined therapy was superior to sorafenib monotherapy with respect to OS (16.3 months vs. 10.0 months) and TTP (6.7 months vs. 4.3 months). This result was similar to prospective studies of primary unresectable HCC treated with DEB-TACE combined with sorafenib, including TACE-2 (TTP 7.9 months) (Meyer et al. 2017) and SPACE (TTP  (Lencioni et al. 2016a, b), and is superior to the trials of primary advanced HCC with PVTT treated with cTACE combined with sorafenib upon STAH (OS 12.8 months) (Park et al. 2019) or Zhu's study (OS 11.0 months) (Zhu et al. 2014). According to the latest result of IMbrave150 (Finn et al. 2020(Finn et al. , 2021, the median OS of patients treated with atezolizumab plus bevacizumab was 19.2 months, which is longer than our study. The participants in IMbrave150 included BCLC stage A (2%), BCLC stage B (15%), which may account for this difference. The OS and TTP of our study were calculated from the administration of sorafenib instead of the diagnosis of HCC, indicating that all patients had suffered disease progress, while all patients in IMbrave150 were treated since the diagnosis of primary HCC, making it difficult to directly compare the results. Compared to those with primary advanced HCC, patients with early ISPA-HCC normally have less tumor burden and poor liver functional reserve. However, with portal invasion, the intrahepatic tumor burden of these patients is still in the early progression stage, similar to that of HCC patients at intermediate substage B3a, as reported by Kudo et al. (2015). At this stage, local treatment to minimize liver function damage, such as super-selective TACE, could still have a high response rate in intrahepatic tumor control on the premise of systemic treatment. Therefore, early ISPA-HCC patients are most likely to benefit from selective DEB-TACE when they are refractory to cTACE.
Although the STAH trial failed to achieve a survival benefit, the results of post-hoc analysis showed that the subgroups of cTACE ≥ 2 could benefit from sorafenib plus TACE (Park et al. 2019). In the present study, all patients received ≥ 2 DEB-TACE sessions in the DTS group, and the patients who received more than two sessions of DEB-TACE had longer median OS than those who received two sessions. This suggested that for whom repeated DEB-TACE is tolerable, sorafenib combined with DEB-TACE might be more effective.
Except for the characteristics of early ISPA-HCC itself, the survival advantages in our study might be due to the mutual benefits for efficacy improvement of both DEB-TACE and sorafenib treatments. First, intrahepatic tumor burden appears to be an important predictor of survival in patients with advanced HCC. Tumor size, tumor number, and PVTT have been found to be predictors of TACE response (Vogl et al. 2011). DEB-TACE promptly reduces the tumor burden refractory to cTACE, which could further enhance sorafenib efficacy. Conversely, in our study, sorafenib was initiated before DEB-TACE, as previous studies suggested that early administration of sorafenib suppresses the post-TACE angiogenesis resulting from increased levels of vascular endothelial growth factor and other angiogenic factors after TACE (Hatooka et al. 2016;Park et al. 2012). As such, sorafenib shows marked efficacy against local tumors and extrahepatic metastases when combined with TACE. Furthermore, sorafenib can normalize the tumor blood vessels and reduce numerous bilobar HCCs simultaneously, which results in enhancement of TACE efficacy by the dense accumulation of anticancer drugs and embolic agents, which was confirmed by the TACTICS trial (Kudo et al. 2020). Our sub-analyses demonstrated that DEB-TACE was more effective in patients with more advanced-stage HCC, including AFP ≥ 400 ng/mL, intrahepatic tumor number ≥ 3, and main tumor size ≥ 5 cm. On-demand DEB-TACE for patients could quickly reduce tumor burden, thereby protecting liver function from intrahepatic lesion progression.
AE occurrence was routinely checked prior to each TACE session. Although the incidences of nausea/vomiting, abdonimal pain, hyperbilirubinemia, fever and ALT/ AST increase were higher in the DTS group, these adverse events were considered post-embolization syndrome, which were all managble and tolerated, and would alleviate within 14 days after DEB-TACE procedures. Except for the postembolization mentioned above, the most common AEs of all grades in two groups were hand-foot syndrome, diarrhea, and fatigue. The incidence of AEs associated with sorafenib was similar to previous studies (Zhu et al. 2014;Wu et al. 2017). These AE results indicate that sorafenib combined with DEB-TACE is well tolerated by patients with early ISPA-HCC after cTACE-refractory. Fig. 3 Response assessment of patients in the DTS and S groups according to the mRECIST. CR complete response, DCR disease control rate, DCR = CR + PR + SD; DTS group patients treated with drug-eluting bead transarterial chemoembolization plus sorafenib, mRECIST modified response evaluation criteria in solid tumors, ORR objective response rate, ORR = CR + PR; PR partial response, SD stable disease, S group patients treated with sorafenib monotherapy Fig. 4 Forest plots of overall survival in the patient subgroups. Subgroup analyses for overall survival. AFP α-fetoprotein, HBV hepatitis B virus, HR hazard ratio, PVTT portal vein tumor thrombus, Second/Lower PVTT in the second-or lower-order branches, First PVTT in the first-order branches Previous studies have suggested that DEB-TACE might be effective in cTACE refractoriness. The causes of resistance to cTACE include cTACE-induced liver function impairment (Johnson et al. 1991), heterogeneous lipiodol retention (Lee et al. 2002), and arterioportal shunt formation caused by vascular invasion (Seki and Hori 2012). HCC patients with heterogeneous lipiodol retention have poor blood supplies that are unlikely to be effectively perturbed with arterial embolization alone. In contrast, a positive response to DEB-TACE may be associated with the slow release of doxorubicin from the beads in the tumor vascular plexus. Additionally, the lipiodol used during cTACE may pass through the arterioportal shunt, which would not produce embolization of the arterial blood supply to the HCC but would cause extensive ischemia in the normal liver tissue (Oh et al. 2015). However, DEB beads are inherently larger than lipiodol particles and cannot traverse arterioportal shunts; hence, shunts with diameters < 300 µm would be occluded, and continuous doxorubicin delivery would eventually kill the tumors.
There are several limitations to our study. First, as this was a retrospective study, further prospective randomized controlled studies are required to confirm our findings. Second, to date, the first-line treatments for advanced HCC include sorafenib, lenvatinib, and atezolizumab plus bevacizumab. The most suitable systemic therapy for combination with TACE remains to be explored. Last, although early detection of advanced HCC is addressed, it is difficult to define the time limit for early detection from complete cTACE resistance to early progression, and preoperative BCLC B-stage tumor burden and cTACE sessions could affect the degree of ISPA-HCC progression, all of which should be considered in further research.
In conclusion, sorafenib plus DEB-TACE is safe and effective for the treatment of early ISPA-HCC with wellpreserved liver function. Combination therapy improves OS, TTP, and local tumor response and does not increase AEs compared to sorafenib monotherapy.

Supplementary Information
The online version contains supplementary material available at https:// doi. org/ 10. 1007/ s00432-022-04107-w. Data availability The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request.