Oncological Outcomes Following Different TACE-Based Conversion Therapies for Intermediate-Advanced Hepatocellular Carcinoma.

1
Introduction
As a common malignant liver tumor with a high mortality rate, hepatocellular carcinoma (HCC) has always been a focus of clinical research [1]. Radical liver resection is not recommended as the first choice for the treatment of intermediate to advanced HCC in both Eastern and Western countries due to its low radical resection rate and poor prognosis [2, 3]. Consequently, improving treatment modalities to enhance long‐term survival for these patients remains an urgent challenge in clinical practice.
Transcatheter arterial chemoembolization (TACE), owing to its advantages of less trauma, faster recovery, and its ability to prolong survival to some extent, has become one of the main treatment strategies for intermediate to advanced HCC [2, 3, 4]. However, the therapeutic efficacy of a single TACE intervention is relatively limited, and disease progression rates are generally high in these HCC patients [4]. Fortunately, with the advent of targeted and immunotherapy in recent years, local TACE combined with systemic therapies has shown great potential in some of these HCC patients, and some may even regain the opportunity for radical liver resection after successfully downstaging during this process [5, 6, 7]. Research in this area has also reported that in patients with intermediate to advanced HCC who were successfully downstaged and subsequently underwent radical resection, long‐term outcomes were significantly improved compared to those who underwent systemic therapies alone [8, 9]. Based on these advances, the concept of conversion therapy for HCC has emerged. Through a series of local treatments combined with or without systemic therapies, patients with initially unresectable HCC may transform into a state where radical liver resection becomes feasible, and this strategy is gradually becoming a novel alternative for intermediate to advanced HCC [10].
However, various systemic therapy options are currently available, and the therapeutic effects of their combinations with TACE are not always consistent [11, 12, 13]. Additionally, in real‐world clinical practice, patients' disease characteristics and general conditions are complex and diverse, making the identification of optimal regimens for different populations highly challenging [14, 15]. Therefore, this study summarized the clinical characteristics of TACE‐based systemic conversion therapies for intermediate to advanced HCC and compared the efficacy and safety of different TACE‐based combination regimens in a real‐world setting, aiming to provide more evidence for the management of HCC and the design of future clinical research.

2
Methods
2.1
Study Design and Patient Selection
Data of HCC patients who were admitted to West China Hospital of Sichuan University between September 2018 and September 2022 were retrospectively collected. Patients meeting the following inclusion criteria were enrolled: (1) aged 18–85 years; (2) diagnosed at Barcelona Clinic Liver Cancer (BCLC) stage B or C; (3) received TACE with or without systemic therapies; (4) had normal liver function (Child‐Pugh A) and well general condition (Eastern Cooperative Oncology Group performance status score of 0–1). Patients meeting any of the following criteria were excluded: (1) incomplete clinical or survival data; (2) received any kind of treatment before admission; (3) diagnosed with recurrent HCC; (4) received liver transplantation or radiofrequency ablation during the treatment period.
Based on the differences of systemic treatments, patients were further divided into four groups: (1) TACE group; (2) TACE combined with tyrosine kinase inhibitors (TACE+TKI group); (3) TACE combined with TKI and immune checkpoint inhibitors (TACE+TKI+ICI group); and (4) TACE combined with Bevacizumab and ICI (TACE+Bev+ICI group). The TKIs included Sorafenib, Donafenib, and Lenvatinib, while the ICIs included Camrelizumab, Sintilimab, Pembrolizumab, Durvalumab, Atezolizumab, and Tislelizumab. The specific regimen used for each patient depended on the availability of certain drugs at the time of diagnosis and the patient's affordability with health insurance. This study was approved by the Institutional Review Board/Ethics Committee of West China Hospital of Sichuan University (Approval number: 2022 Annual Review No. 690) and was conducted in compliance with the Declaration of Helsinki. The requirement for informed consent was waived.

2.2
Treatment Strategies
For TACE treatment, patients underwent femoral artery puncture imaging under local anesthesia. After the tumor‐supplying artery was identified, 5‐fluorouracil combined with epirubicin was slowly infused over 20 min. Embolization was then performed using iodinated oil, iodized oil combined with gelatin sponge particles, or gelatin sponge combined with drug‐eluting microspheres. Re‐imaging was performed for final confirmation. All patients received standard analgesic and antiemetic supportive therapy after the TACE procedure.
If systemic therapy was planned for patients after TACE, it was usually administered at least one week later, provided that liver and thyroid function were normal. The specific dosage of TKIs and ICIs was determined according to the recommendations of medicine instructions.
For patients who achieved at least disease control rate (DCR) status after a series of TACE‐based therapies, a multidisciplinary team (MDT) discussion, including hepatobiliary surgeons, interventional radiologists, oncologists, and pathologists, was conducted to finalize the decision on conversion surgery. In general, well‐preserved liver function (indocyanine green retention rate at 15 min < 10%), the possibility of R0 resection assessed by surgeons, and patients' willingness for resection were taken into consideration. During the subsequent hepatectomy, liver parenchyma dissection was routinely performed while maintaining low central venous pressure and applying the Pringle maneuver. Intraoperative ultrasound was also used to confirm lesions and guide the surgical margin.
Following hospital discharge, patients underwent outpatient follow‐up every three months. The follow‐up protocol included routine blood tests, liver function tests, tumor markers such as alpha‐fetoprotein (AFP) and protein induced by vitamin K absence‐II (PIVKA‐II), as well as radiologic evaluations. Adjuvant treatments were the same as the preoperative strategies and were continued for at least six months or until recurrence was noted. Treatment options for recurrence included re‐resection, radiofrequency ablation, TACE, liver transplantation, and systemic therapies, depending on patients' recurrence patterns and overall physical condition.

2.3
Definitions
The assessment of treatment response was based on the Modified Response Evaluation Criteria in Solid Tumors (mRECIST) [16]. Specifically, responses were graded as complete response (CR), partial response (PR), stable disease (SD), and progressive disease (PD). The objective response rate (ORR) was calculated as the sum of CR and PR, while the DCR was defined as the sum of CR, PR, and SD lasting for at least four weeks. The Common Terminology Criteria for Adverse Events (CTCAE) version 5.0 was used to evaluate the adverse events (AEs) during the treatment period. Overall survival (OS) was measured as the time from the date of treatment to death or the last follow‐up, while progression‐free survival (PFS) was defined as the time from the date of treatment to PD, recurrence after surgery, death, or the last follow‐up. The primary outcomes of this study were PFS and OS following different TACE‐based treatments, while the secondary outcomes were treatment responses and AEs. The last follow‐up date was March 10, 2024.

2.4
Statistical Analysis
All data were processed using SPSS software (version 26.0, IBM, USA) and GraphPad Prism (version 10.0, GraphPad Software, USA). Statistical tests were performed using a two‐sided approach, with a two‐tailed p < 0.05 indicating statistical significance. Categorical variables were expressed as frequencies with percentages and compared using chi‐square tests or Fisher's exact tests. Continuous variables were expressed as median with range and tested for normality using the Kolmogorov–Smirnov or Shapiro–Wilk test. The comparison of continuous data was based on the t‐test or one‐way ANOVA if the data were normally distributed; otherwise, the Mann–Whitney U test or the Kruskal‐Wallis H test was applied. Differences between each two groups were compared using the Z test. Survival‐related risk factors were evaluated using Cox proportional hazards regression models and presented as hazard ratios (HR) with 95% confidence intervals (CI). Since this was a retrospective study with unavoidable selection bias, propensity score matching (PSM) analysis was used to balance confounding factors between the Resection group and the Non‐resection group. AFP, PIVKA‐II, and tumor number were selected as covariates to conduct 1:1 PSM with a caliper value of 0.02. OS and PFS between groups were calculated using Kaplan–Meier curves and compared with log‐rank tests.

3
Results
3.1
Patient Characteristics
A total of 518 HCC patients treated with TACE‐based therapies were finally enrolled according to the inclusion and exclusion criteria (Figure S1). Among them, 117 (22.6%) patients underwent subsequent liver resection. Baseline characteristics of the entire cohort are summarized in Table 1. The median age of all patients was 55 (18–85) years, with 465 (89.8%) male patients. The hepatitis B virus infection was diagnosed in 450 (86.9%) patients, and liver cirrhosis in 365 (70.5%) patients. Regarding tumor distribution, 398 (76.8%) patients had multiple liver lesions, and 283 (35.3%) had tumors in both right and left lobes. The median tumor size was 8.5 cm (1.0–20.0), and 40.2% of tumors were larger than 10 cm. Additionally, 467 (90.2%) patients had significantly elevated tumor markers (either AFP or PIVKA‐II), and macrovascular invasion was present in 288 (55.6%) patients. According to BCLC staging criteria, 196 (37.8%) of these cases were stage B and 322 (62.2%) were stage C.
In terms of specific treatment regimens, the 518 patients were further divided into the TACE group (n = 143), TACE+TKI group (n = 158), TACE+TKI+ICI group (n = 180), and TACE+Bev+ICI group (n = 37). As shown in Table 1, the proportions of cirrhosis (p = 0.002), macrovascular invasion (p = 0.011), BCLC stage (p = 0.006), and extrahepatic metastases (p = 0.001) differed significantly among the four groups. Further pairwise comparisons showed that the proportion of cirrhotic patients was lower in the TACE group compared to the TACE+TKI group or the TACE+TKI+ICI group (p < 0.05). Compared to the TACE group, patients in the TACE+TKI+ICI group had a higher rate of macrovascular invasion. Regarding extrahepatic metastases, the TACE+Bev+ICI group showed a significantly higher rate compared to any of the other three groups (p < 0.05).

3.2
Treatment Responses and AEs
During a median follow‐up time of 13.6 (0.1–67.5) months, 10 (1.9%) patients achieved CR, 99 (19.1%) achieved PR, 221 (42.7%) had SD, and 188 (36.3%) had PD status (Table 2). The ORR for all patients was 21.0% (109/518) and the DCR was 63.7% (330/518). Comparison of treatment responses among the four groups showed statistically significant differences in PR (p = 0.012), PD (p = 0.018), ORR (p = 0.009), and DCR (p = 0.018). Further pairwise comparisons showed that the TACE+TKI+ICI group had a significantly higher ORR and DCR compared to the TACE group (p < 0.05), while no statistical differences were noted among the other groups. The overall conversion surgery rate was 22.6%, which differed significantly among the four groups (p = 0.010). Specifically, only 21 (13.3%) patients underwent surgical resection after TACE+TKI treatment compared to the other groups (p < 0.05).
In terms of AEs, 147 (28.4%) patients experienced at least one treatment‐related AE during the treatment period. 34 patients experienced high‐grade AEs (≥ grade 3) and 1 patient in the TACE+Bev+ICI group died due to severe abdominal hemorrhage. 13 patients (9.1%) in the TACE group had AEs documented during follow‐up, which was significantly lower compared to the other three groups (p < 0.001). In contrast, the highest rate of grade 3–5 AEs was observed in the TACE+Bev+ICI group, with 8 (21.6%) patients experiencing treatment‐related severe AEs during follow‐up, which was significantly higher than that in the TACE and TACE+TKI groups (p < 0.05). Specifically, statistically significant differences were found in the incidence of increased transaminases (p = 0.001), hypertension (p = 0.034), hypothyroidism (p = 0.001), thrombocytopenia (p = 0.002), myelosuppression (p = 0.001), and dermatologic toxicity (p = 0.007) among the four groups.

3.3
Survival Outcomes
The survival outcomes showed that the median PFS (mPFS) was reached in all groups. Specifically, mPFS was 10.9 months in the TACE group, 15.6 months in the TACE+TKI group, 20.7 months in the TACE+TKI+ICI group, and 15.6 months in the TACE+Bev+ICI group. There was a statistically significant difference in PFS among the groups (p < 0.001, Figure 1a). Further pairwise comparisons showed that PFS in the TACE+TKI+ICI group was significantly better than in the TACE group (p < 0.001) and the TACE+TKI group (p = 0.011).
On the other hand, median OS (mOS) was also achieved in all treatment groups. In specific, mOS was 13.3 months in the TACE group, 25.0 months in the TACE+TKI group, 44.0 months in the TACE+TKI+ICI group, and 32.3 months in the TACE+Bev+ICI group. The difference in OS among four groups was also statistically significant (p < 0.001, Figure 1b). Further pairwise comparisons showed that OS in the TACE+TKI+ICI group was significantly better than in the TACE group (p < 0.001) and the TACE+TKI group (p < 0.001).
The Cox regression analyses were performed on all patients to identify potential risk factors for long‐term survival. As shown in Table 3, multivariate analysis revealed that age < 65 years [HR = 1.648, 95% CI (1.257–2.162), p = 0.001] and single TACE therapy [HR = 1.583, 95% CI (1.233–2.033), p < 0.001] were risk factors for increased disease progression, whereas low AFP levels [HR = 0.795, 95% CI (0.637–0.993), p = 0.043], tumor size < 5 cm [HR = 0.721, 95% CI (0.553–0.939), p = 0.015], conversion resection [HR = 0.649, 95% CI (0.498–0.846), p = 0.001], and TACE+TKI+ICI therapy [HR = 0.713, 95% CI (0.551–0.923), p = 0.010] significantly reduced the risk of disease progression.

3.4
Subgroup Analysis of Patients Received Liver Resections
Based on whether patients received subsequent liver resection, we analyzed the impact of conversion resection on long‐term survival. Of these, 117 patients in the Resection group and 401 patients in the Non‐resection group were identified (Table 4). Comparative analysis of baseline characteristics between the two groups showed that patients in the Resection group generally had a lower incidence of multiple tumor lesions (59.8% vs. 81.8%, p < 0.001), higher rates of AFP > 400 ng/mL (47.9% vs. 32.9%, p = 0.003), and PIVKA‐II > 400 mAU/ml (29.1% vs. 15.7%, p = 0.001) compared to those in the Non‐resection group before PSM. Therefore, a PSM processing of the two groups was carried out to balance the baseline characteristics. As a result, 112 patients in the matched Non‐resection group and 117 patients in the Resection group were generated with well‐balanced and comparable characteristics after PSM (Table 4).
Based on this, PFS analyses were performed (Figure 2a,b). The results showed that the mPFS was 15.0 months in the Non‐resection group and 20.6 months in the Resection group after PSM. The 1‐, 3‐, and 5‐year PFS rates were 58.9%, 23.2%, and 12.7% in the Non‐resection group, and 70.1%, 34.2%, and 17.1% in the Resection group, respectively (p = 0.016, Figure 2b). The log rank test showed that conversion surgery significantly prolonged PFS both before and after PSM (p < 0.05).
Similarly, OS analyses were conducted for both groups (Figure 2c,d). The mOS was 26.8 months in the Non‐resection group, while the Resection group had not yet reached mOS at the last follow‐up. The OS rates at 1‐, 3‐, and 5‐year were 74.1%, 41.8%, and 21.3% in the Non‐resection group, and 91.5%, 70.2%, and 62.2% in the Resection group, respectively (p < 0.001, Figure 2d). The log rank test showed that conversion surgery significantly improved OS both before and after PSM (p < 0.05).

4
Discussion
Given the widespread use of TACE in clinical practice, this study explored the characteristics of different TACE‐based combination strategies. It presents a real‐world analysis of the current status of these regimens as conversion therapies for intermediate to advanced HCC. Comparative analyses of the efficacy and safety of different TACE‐based combinations showed that TACE combined with TKI and ICI significantly improved long‐term survival with manageable AEs profiles. For patients who successfully underwent subsequent liver resection, both PFS and OS were significantly better compared to those without resection.
From the perspective of the overall patient cohort, the ORR of TACE‐based systemic therapies was 21%, and the DCR was 63.7%, with the TACE+TKI+ICI regimen showing the highest therapeutic effectiveness. Indeed, the selection of a specific treatment strategy varied according to patient characteristics. As presented in this study, patients with cirrhosis or macrovascular invasion were less common in the TACE group compared to other groups. This reflects real‐world practice, where clinicians tailor therapy to disease severity. The presence of cirrhosis or large vessel invasion has been identified in the literature as key factors affecting the long‐term prognosis of HCC [17, 18]. Consequently, clinicians tend to choose more aggressive treatment strategies to prolong survival in these patients as much as possible. Furthermore, in cases with extrahepatic metastases, more patients were treated with TACE+Bev+ICI. The reason for selecting such a regimen could be complex, but reported evidence on the effectiveness of anti‐VEGF strategies for lymph node or lung metastasis may support this combination in this setting [19]. On the other hand, the overall conversion to surgical resection rate was 22.6%, which was comparable to previously reported evidence focusing on TACE‐based treatments [8, 20]. It should be noted that the factors determining a patient's acceptance of subsequent hepatic resection do not fully correspond to preoperative treatment effectiveness. In clinical practice, the possibility of receiving subsequent liver surgery could be considered once patients achieved ORR or at least DCR after systemic treatments. However, a patient's liver function reserve, future liver remnant, financial support, and willingness to undergo surgical resection all influence the decision for surgery to varying degrees. Therefore, the conversion surgery rate alone does not represent the true effectiveness of conversion therapy. Instead, achieving optimal ORR during systemic therapy may better indicate the potential for subsequent resection. This also emphasizes the clinical significance of conversion therapy for intermediate to advanced HCC.
Regarding treatment‐related AEs, 28.4% of patients experienced at least one AE of varying severity during treatment. Based on case distributions, significantly more patients were treated with TACE combined with TKI or TKI+ICI than with TACE+Bev+ICI. Although this may introduce selection bias affecting the reliability of results for the TACE+Bev+ICI group, it reflects the actual characteristics of liver cancer populations in Asia, which have a high prevalence of hepatitis B and liver cirrhosis, limiting the choice of Bevacizumab due to safety concerns regarding hemorrhage. This was also reflected in the comparison of AEs, showing the highest rate in the TACE+Bev+ICI group. The number of patients included in the TACE+Bev+ICI group was relatively small, but only one treatment‐related death was observed. This underscores the critical need for careful patient selection when considering this regimen. Therefore, this combination should be used with particular caution in patients with compromised liver function, active hepatitis, or a history of significant bleeding. More actively monitoring strategies are needed, including regular assessment for hypertension, proteinuria, signs of bleeding, and liver function deterioration. Overall, all TACE‐based systemic treatments increased the incidence of complications to varying degrees, with triple therapy resulting in significantly more AEs than double therapy. Analysis based on different treatment regimens showed that any form of systemic therapy could increase the likelihood of hepatic impairment, which may be related to the metabolic and detoxification processes of drugs in the liver [21]. For TACE combined with TKI therapies, the probability of hypertension was significantly higher compared to TACE therapy alone, regardless of whether ICI was applied. Consistently, Piscaglia et al. analyzed 476 HCC patients treated with lenvatinib and found that 44.5% of patients developed treatment‐emergent hypertension, but it did not impair long‐term outcomes when detected early and managed appropriately [22]. Therefore, monitoring blood pressure and the timely use of antihypertensive agents are necessary during TACE+TKI therapy. In addition, we observed that the use of ICI significantly increased the incidence of thyroid insufficiency. The similar phenomenon was reported by Wu et al., who collected data on 195 HCC patients treated with PD‐1 monoclonal antibodies and found a high rate of 57.9% for thyroid dysfunction, but most cases with timely intervention did not interrupt PD‐1 treatment [23].
For patients with intermediate‐advanced HCC, the primary goal of any treatment is to prolong survival. Our analyses demonstrated that the TACE+TKI+ICI regimen provided significantly longer PFS and OS than TACE alone or TACE+TKI. Further multivariate regression analysis also revealed that TACE+TKI+ICI as a conversion therapy effectively reduced the risk of disease progression. In contrast, TACE alone, without any form of systemic therapy, impaired long‐term survival to some extent. These results are consistent with previous evidence provided by Liu et al., who reported a significantly higher ORR and PFS in the TACE combined with targeted therapy group compared to TACE alone [24], further emphasizing the importance of combining locoregional therapy with systemic therapy for HCC management. In addition, our results identified better long‐term survival in patients older than 65 years, as well as in those with lower AFP levels and smaller tumors. In elderly patients, tumors often exhibit less aggressive biological behavior, which may contribute to more favorable long‐term outcomes [25]. The levels of AFP also represent the aggressiveness and differentiation degree of tumors. Consequently, for patients with relatively low AFP levels, survival may be greatly improved after a series of conversion therapies. On the other hand, for patients who successfully underwent subsequent hepatectomy, this study demonstrated the survival benefit following conversion resection. Among the 22.6% of patients who successfully underwent liver resection, long‐term survival was significantly better than in those who did not. Similar results have been reported in many studies [26, 27, 28]. In light of these findings, selecting an appropriate approach to conversion hepatectomy emerges as a key strategy to improve outcomes for intermediate to advanced HCC.
The findings from this real‐world analysis offer several practical implications for managing intermediate to advanced HCC. Firstly, for patients eligible for aggressive conversion therapy, initiating a triple combination of TACE+TKI+ICI appears to provide the highest likelihood of tumor downstaging and superior survival outcomes. Secondly, the decision‐making process should be dynamic and involve a MDT discussion. Patients achieving a favorable response should be re‐evaluated for surgical resectability, as conversion surgery conferred a profound survival advantage. Moreover, treatment selection must be personalized, integrating both efficacy and safety profiles. While TACE+TKI+ICI showed promising efficacy, the TACE+Bev+ICI regimen was associated with higher severe toxicity, necessitating stringent patient selection and vigilant monitoring. Besides, our data reinforce that TACE alone is suboptimal for most intermediate‐advanced HCC patients in the era of effective systemic therapies, and its combination with systemic agents should be standard when conversion therapy is the goal.
It should be acknowledged that this study has limitations. First, the retrospective, non‐randomized design inherently introduces selection bias and the potential for unmeasured confounding factors. Despite employing statistical methods like PSM to balance observed variables between subgroups, residual confounding from factors not captured in our database could influence the outcomes and their interpretation. Second, the heterogeneity in patient characteristics across treatment groups, including the complexity of patients' disease backgrounds, differences in affordability, and willingness to pursue different treatment options, could also affect the interpretation of the results to varying extents. Importantly, the sample size for the TACE+Bev+ICI group was notably smaller than for other groups. This limits the robustness of conclusions specific to this regimen and increases the uncertainty of efficacy and safety to some extent. However, this also reflects the current state of real‐world clinical practice and provides a foundation for insights into the management of HCC. Moreover, as numerous options for systemic treatment of HCC are available, there may be differences in efficacy among varying TKIs or ICIs. Due to the lack of head‐to‐head clinical study designs and differences in market accessibility of certain drugs during treatment, combining TKIs and ICIs in a single analysis may further reduce the precision of the current results. However, this study was limited by its relatively small sample size, and subdividing each treatment regimen may further reduce statistical power. Therefore, more detailed and comprehensive clinical trial designs are warranted in the future. Specifically, prospective, randomized controlled trials are needed to validate the efficacy and safety hierarchy of these TACE‐based combination regimens. Furthermore, head‐to‐head comparisons between different systemic therapies (e.g., TACE+TKI+ICI vs. TACE+Bev+ICI) are essential to define the optimal conversion therapy strategy for specific patient subgroups.
In conclusion, this study compared different TACE‐based systemic therapies as conversion strategies and demonstrated that TACE+TKI+ICI had a better ORR, manageable AEs, and significantly improved PFS and OS. Long‐term survival was significantly improved in patients who underwent subsequent liver resection compared to those without resection. Further prospective multicenter studies with larger sample sizes are required to validate these findings.

Author Contributions
Conception and design: Y.W., Y.L., H.X., B.L., K.C., H.W.; Data collection and quality assessment: H.X., Y.L., H.Z., H.W., K.C.; Analysis and interpretation of data: H.X., Y.L., H.Z., Y.W.; Drafting the work: H.X., Y.L., H.Z., Y.W.; Critical review and final approval: H.X., Y.L., H.Z., H.W., B.L., K.C., Y.W.

Funding
This work was supported by a grant from Sichuan Science and Technology Program (No. 2023YFS0230 & 2024NSFSC1941).

Ethics Statement
This study was approved by the Institutional Review Board/Ethics Committee of West China Hospital of Sichuan University (Approval number: 2022 Annual Review No. 690) and was conducted in compliance with the Declaration of Helsinki. The requirement for informed consent was waived.

Conflicts of Interest
The authors declare no conflicts of interest.

Supporting information

Figure S1: Flowchart of patient selection.