The emerging therapies are reshaping the first-line treatment for advanced hepatocellular carcinoma: a systematic review and network meta-analysis

Abstract

Background:

Given the superior performance of various therapies over sorafenib in advanced hepatocellular carcinoma (HCC) and the absence of direct comparisons, it is crucial to explore the efficacy of these treatments in phase III randomized clinical trials.

Objectives:

The goal is to identify which patients are most likely to benefit significantly from these emerging therapies, contributing to more personalized and informed clinical decision-making.

Design:

Systematic review and network meta-analysis

Data sources and methods:

PubMed, Embase, ClinicalTrials.gov, and international conference databases have been searched from 1 January 2010 to 1 December 2023.

Results:

After screening, 17 phase III trials encompassing 18 treatments were included. In the whole-population network meta-analysis, the newly first-line tremelimumab plus durvalumab (Tre + Du) was found to be comparable with atezolizumab plus bevacizumab (Atezo + Beva) in providing the best overall survival (OS) benefit [hazard ratio (HR) 1.35, 95% confidence interval (CI): 0.93–1.92]. Concerning OS benefits, sintilimab plus bevacizumab biosimilar (Sint + Beva), camrelizumab plus rivoceranib (Camre + Rivo), and lenvatinib plus pembrolizumab (Lenva + Pemb) appear to exhibit similar effects to Tre + Du and Atezo + Beva. In the context of progression-free survival, Atezo + Beva seemed to outperform Tre + Du (HR: 0.66 CI: 0.49–0.87), while the effects are comparable to Sint + Beva, Camre + Rivo, and Lenva + Pemb. Upon comparison between Asia-Pacific and non-Asia-Pacific cohorts, as well as between hepatitis B virus (HBV)-infected and non-HBV-infected populations, immune checkpoint inhibitor (ICI)-based treatments seemed to exhibit heightened efficacy in the Asia-Pacific group and among individuals with HBV infection. However, combined ICI-based therapies did not show more effectiveness than molecular-targeted drugs in patients without macrovascular invasion and/or extrahepatic spread. As for grades 3–5 adverse events, combined therapies showed comparable safety to sorafenib and lenvatinib.

Conclusion:

Compared with sorafenib and lenvatinib, combination therapies based on ICIs significantly improved the prognosis of advanced HCC and demonstrated similar safety. At the same time, the optimal treatment approach should be tailored to individual patient characteristics, such as etiology, tumor staging, and serum alpha-fetoprotein levels. With lower incidence rates of treatment-related adverse events and non-inferior efficacy compared to sorafenib, ICI monotherapies should be prioritized as a first-line treatment approach for patients who are not suitable candidates for ICI-combined therapies.

Trial registration:

PROSPERO, CRD42022288172.

Plain language summary

Lay summary/Key points

The efficiency of various systemic therapies in advanced HCC patients with specific characteristics remains to be explored. This study revealed that the efficacy of ICI combined therapies is influenced by factors such as tumor staging, etiology, patient demographics, and more. Additionally, ICI monotherapies should be prioritized as a first-line treatment approach for patients who are not suitable candidates for ICI combined therapies. Complementing to recent guidelines, this study indicated that several critical factors needed to be took into consideration for patients with advanced HCC.

Keywords

first-line hepatocellular carcinoma immune checkpoint inhibitors molecular-targeted therapy network meta-analysis

Introduction

Hepatocellular carcinoma (HCC), representing approximately 90% of primary liver cancers, stands as the third leading cause of cancer-related deaths globally.¹ Curative interventions, including resection, transplantation, and ablation, significantly enhance the prognosis for early-stage HCC patients.² Unfortunately, 25–70% of those diagnosed at median-to-advanced stages do not have access to curative treatments, rendering their condition regarded as incurable.^3,4

With the efficacy of reducing the risk of cancer-related death by 31–32%, sorafenib, a tyrosine kinase inhibitor, became the first-line recommended therapy for advanced HCC in 2007.^5,6 Even after lenvatinib, which shows non-inferior efficacy to sorafenib, molecular-targeted therapies (MTDs) remained the first-line therapies for advanced HCC.^7,8 With the advent of immune checkpoint inhibitors (ICIs) and the promising outcomes of the combinations of MTDs and ICIs, this hard nut begins to crack.⁹ The IMbrave150 study found that atezolizumab plus bevacizumab (Atezo + Beva), a combination of programmed cell death ligand 1 (PD-L1) inhibitor and vascular endothelial growth factor (VEGF) inhibitor, surpassed sorafenib in achieving prolonged overall survival (OS). Due to the benefits observed in advanced HCC patients, Atezo + Beva was recommended as the preferred first-line option.^9–11 Moreover, combinations such as sintilimab plus bevacizumab biosimilar (Sint + Beva), tremelimumab plus durvalumab (Tre + Du), camrelizumab plus rivoceranib (Camre + Rivo), and donafenib have shown favorable prognostic benefits and low rates of adverse events (AEs) in advanced HCC patients.^12–15

Nowadays, the majority of guidelines highly recommend combination therapies, including MTDs and ICIs (Atezo + Beva and Sint + Beva) or combinations of ICIs (Tre + Du), as the first-line treatment for advanced HCC patients.^9,11,12,14 There is currently a lack of direct head-to-head studies comparing the merits of these first-line treatment strategies. In addition, in comparison to MTDs, the extent of benefits from combined therapies within various subgroups remains unclear.^16–18 Furthermore, some studies suggested that the etiology of HCC can exert a substantial influence in modulating the response to these therapies.^16,19,20 Some patients with specific characteristics may benefit from MTDs alone rather than the combinations in the view of long-term and economic perspective, and which characteristics should be taken into account to make the optimal decisions is largely unknown.

Currently, most network meta-analyses primarily focus on comparing the advantages of existing therapies with sorafenib or emphasize the comparison of A + T with other regimens.^21–23 However, there is limited coverage of comparative analyses between current first-line therapies and emerging treatments. Therefore, our meta-analysis is specifically dedicated to examining the strengths and weaknesses among first-line therapies and emerging therapies, concurrently exploring potential variations in responsiveness to emerging therapies across different subgroups. This study aims to address existing gaps in the literature, providing a comprehensive perspective on the performance of various first-line treatment strategies within specific patient subgroups.

Methods

This network meta-analysis was reported in line with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement (Supplemental Table S1) for network meta-analysis.²⁴ We have applied the frequentist network meta-analysis for the present analysis. The protocol was registered in the Prospective Register of Systematic Reviews (PROSPERO, CRD42022288172).

Data sources and searches

We conducted a comprehensive search among the online databases, including PubMed, Embase, Cochrane Central Register of Controlled Trials, and ClinicalTrials.gov. The related studies were identified from 1 January 2010 to 1 December 2023 in English by using combinations of the following terms, ‘hepatocellular carcinoma’, ‘tyrosine kinase inhibitors’, ‘immunotherapy’, ‘immune checkpoint inhibitors’, ‘first-line’, within the restrictive limit of ‘randomized controlled trials’. We also scanned the abstracts and presentations of ongoing randomized controlled trials on HCC from several international conferences (American Society of Clinical Oncology, European Society of Medical Oncology) from 2014 to 2023. Finally, the candidate studies were perused for details.

Study selection

Phase III randomized controlled trials that met the following criteria were included: (1) Participants: patients with advanced HCC who were diagnosed histopathologically, cytologically, or clinically (adapted from the Guidelines of the European Association for the Study of the Liver²⁵ or Guidelines for Diagnosis and Treatment of Primary Liver Cancer in China²⁶), (2) Intervention: systemic regimens as first-line treatment, (3) Comparison: sorafenib or lenvatinib as the control group, (4) Outcomes: the primary outcomes are OS, progression-free survival (PFS), and the secondary outcomes are overall response rate (ORR) and Grade 3 or higher treatment-related adverse events (TRAEs).

Exclusion criteria included: (1) incomplete reported outcomes, (2) repeated reports, and (3) sample size of fewer than 100 patients.

Data extraction and risk of bias assessment

Two investigators conducted study selection and data extraction independently, with the disagreements adjudicated by a third consultant. To avoid potential assessment bias caused by investigators, clinical outcomes of interests assessed by the independent review facility were primarily extracted. The TRAEs were employed, and if TRAEs were not reported, general grade 3–5 AEs were used for the network meta-analysis instead. ClinicalTrials.gov and other available sources were evaluated based on the up-to-date data. The qualitative assessment of individual studies was performed using the Cochrane Risk of Bias Tool.²⁷

Data synthesis and statistical analysis

Frequentist framework network meta-analysis (including subgroup analysis based on several standards and integrated analysis) was performed with the ‘netmeta’ package (version 2.0-1, R Foundation, https://cran.r-project.org/web/packages/netmeta/netmeta.pdf) of R software (version 4.0.2, R Project for Statistical Computing, https://www.r-project.org/) to compare and rank the treatment effects directly or indirectly across the regimens. A random-effects model was used to explain the substantial heterogeneity across the studies. The global heterogeneity between treatment effects among the included studies was assessed using Chi² and I² statistics and p values of Chi² > 0.05 or I² < 50% were considered as low heterogeneity, respectively.

Results

Literature search

The process of study selection has been shown in the flow diagram (Figure 1). Initially, we identified 2748 studies from the online database mentioned above. After screening the title/abstract and full text for eligibility, 17 studies involving 18 regimens for advanced HCC were included for further analysis.^{9,13–15,28–37}

Figure 1.

PRISMA flow diagram showing the study selection process.

Study characteristics and data processing

The characteristics of 17 studies are shown in Supplemental Table S2. To summarize, there were 11,922 patients included in these studies, and the number of participants ranged from 323 to 1155. The age range was between 18 and 88 years. The correlation network figures in terms of OS and PFS are presented in Figure 2.

Figure 2.

Network diagrams of comparisons on OS and PFS of treatments: (a) overall survival and (b) progression-free survival.

The primary outcomes were OS and PFS, and the secondary outcome was ORR and grade 3 or higher AEs. As a result, both OS and PFS of 17 regimens, ORR of 15 regimens, and AEs of 14 regimens were included in the network meta-analysis. For integrated analysis, Sint + Beva, Atezo + Beva, Camre + Rivo, lenvatinib plus pembrolizumab (Lenva + Pemb), Tre + Du, and cabozantinib plus atezolizumab (Cabo + Atezo) were integrated as the ICI-combined therapies, patients who received lenvatinib or donafenib were pooled as the MTDs group, and patients who received sorafenib were set as placebo. At the same time, nivolumab, durvalumab, and tislelizumab were integrated into the ICI-monotherapies group.

Efficacy and safety among whole-population analysis

As for the overall network analysis [Figure 3(a)], Sint + Beva, Atezo + Beva, Camre + Rivo, Lenva + Pemb, and Tre + Du stand out as the top five regimens in conferring OS benefits, and all of them exhibited greater OS benefits compared to sorafenib and lenvatinib. Even if no significant difference were observed among these five treatment regimens in terms of providing OS benefits, Sint + Beva, Atezo + Beva, and Camre + Rivo outperformed not only all kinds of MTDs but also ICI monotherapies, while Lenva + Pemb and Tre + Du failed to do so. In the meanwhile, three ICI monotherapies failed to offer marked benefits in comparison with MTDs, which is similar within their corresponding trials. To comprehend the overall effects of ICI-combined therapies, ICI monotherapies, and MTDs, we conducted an integrated analysis. In our integrated analysis, ICI-based combined regimens showed OS benefits compared to various MTD and ICI monotherapies. Interestingly, while ICI monotherapies did not exhibit a significant OS benefit compared to MTDs, this group demonstrated a notable OS advantage over sorafenib [hazard ratio (HR): 0.85; 95% confidence interval (CI): 0.76–0.96] (Supplemental Figure 1A).

Figure 3.

Pooled estimates of the network meta-analysis. (a) Pooled hazard ratios (95% credible intervals) for progression-free survival (upper triangle) and overall survival (lower triangle). Data in each cell are hazard ratios (95% credible intervals) for the comparison of row-defining treatment versus column-defining treatment. And hazard ratio of less than 1 favors row-defining treatment. (b) Pooled odds ratios (95% credible intervals) of objective response rate (lower triangle). Data in each cell are odds ratios (95% credible intervals) for the comparison of row-defining treatment versus column-defining treatment. An odds ratio of more than one favors row-defining treatment.

With regard to PFS [Figure 3(a)], with comparable efficacy among themselves, Camre + Rivo exhibited the most significant benefit, followed by Sint + Beva, Lenva + Pemb, and Atezo + Beva – all of which outperformed sorafenib and ICI monotherapies. Notably, Tre + Du was inferior to each of the four therapies and lenvatinib. And, in contrast to its performance in OS, Tre + Du failed to show PFS benefits compared to sorafenib and ICI monotherapies. Interestingly, none of the ICI-combined therapies demonstrated their superiority in PFS improvements when compared to lenvatinib. In integrated analysis, ICI-based combined regimens showed PFS benefits versus sorafenib and ICI monotherapies; however, such superiority was absent when compared to MTDs (HR: 0.84; 95% CI: 0.67–1.07) (Supplemental Figure 1A). ICI monotherapies did not demonstrate PFS benefit over sorafenib in this analysis.

In terms of ORRs [Figure 3(b)], all the ICI-based combinational therapies, ICI monotherapies, and lenvatinib demonstrated ORR benefits over sorafenib. Notably, in accordance with the outcomes in PFS, our analysis revealed that none of these therapies exhibited ORR superiority over lenvatinib. Nevertheless, when contrasted with donafenib, an alternative first-line therapy in China, Sint + Beva, Camre + Rivo, and Tre + Du exhibited an ORR advantage. The duration of response could not be evaluated due to the lack of data from the included trials.

As for safety analysis (Supplemental Table S3), the ICI monotherapies, durvalumab (52.1%), nivolumab (73.3%), and tislelizumab (76.6%), demonstrated superiority in safety compared to other regimens with fewer TRAEs of all grades. The addition of anti-CTLA-4 (Cytotoxic T-lymphocyte-associated protein 4) tremelimumab to Durvalumab significantly increased the incidence of TRAEs of all grades (52.1–75.8%) and grade 3–5 (12.9–25.8%). Despite this, Tre + Du induced TRAEs of all grades and grade 3–5 TRAEs comparable to the other two ICI monotherapies. Despite significantly increasing the incidence of TRAEs compared to ICI monotherapies, the occurrence rate of TRAEs with the addition of MTDs to ICIs is similar to that of sorafenib and lenvatinib, except for Camre + Rivo. Notably, compared to other combinations of MTDs and ICIs, Camre + Rivo led to a significantly higher incidence of grade 3–5 TRAEs and TRAEs that led to discontinuation. For the ICI therapies, the most common grade 3–5 TRAEs were lipase increased and aspartate aminotransferase increased in Tre + Du groups, and aspartate aminotransferase increased in nivolumab or durvalumab groups. The combinations of ICIs and MTDs, including Atezo + Beva, Sint + Beva, Cabo + Atezo, Camre + Rivo, and shared similar types of grades 3–5 AEs, such as hypertension and aspartate aminotransferase increase.

Subgroup analysis

Subgroup analyses were performed to answer the question of whether specific patients would benefit from particular regimens, hoping to provide preliminary evidence for decision-making in future clinical practice.

It is generally believed that the etiology of HCC differs between the Western and Eastern populations and subgroup analyses were performed according to the geographical region [Figure 4(a)]. In subgroup analyses for the Asia-Pacific and non-Asia-Pacific regions, there were no significant differences among the five treatment regimens – Sint + Beva, Atezo + Beva, Camre + Rivo, Lenva + Pemb, and Tre + Du. In the Asia-Pacific region, all five regimens demonstrated superiority over sorafenib. However, in the non-Asia-Pacific region, Atezo + Beva emerged as the exclusive regimen exhibiting superiority over sorafenib and simultaneously demonstrating advantages over lenvatinib. In our subgroup integrated analysis (Supplemental Figure 1C), in both the Asia-Pacific and non-Asia-Pacific populations, ICI-based combined therapies exhibited superior efficacy compared to MTDs, including sorafenib. Notably, in the Asia-Pacific region, the HR values were even smaller, underscoring the heightened effectiveness of these combined therapies in this specific demographic. Interestingly, while ICI monotherapies did not exhibit superiority over the MTDs group, they demonstrated OS benefits compared to sorafenib in both subgroups.

Figure 4.

Pooled estimates of the subgroup network meta-analysis in terms of OS. Pooled estimates of subgroup analysis: (a) Pooled hazard ratios of OS for other regions (upper triangle) and Asia-Pacific (lower triangle). (b) Pooled hazard ratios of OS for non-HBV (upper triangle) and HBV positive (lower triangle). (c) Pooled hazard ratios of OS for VI and/or EHS negative (upper triangle) and VI and/or EHS positive (lower triangle). (d) Pooled hazard ratios of OS for AFP low (upper triangle) and AFP high (lower triangle). Data in each cell are hazard ratios for the comparison of row-defining treatment versus column-defining treatment. And hazard ratio of less than one favors row-defining treatment.

Several studies have observed that hepatitis B virus (HBV)-related HCC tends to exhibit better OS outcomes when treated with immune-based therapies, in contrast to cases without HBV etiologies. In light of these findings, we conducted a subgroup analysis based on the HBV status to further explore the nuanced treatment responses within specific etiological subgroups. We observed significant OS benefits over sorafenib in the HBV-positive group across five ICI-combined therapies, while none of them performed better than sorafenib in the non-HBV group [Figure 4(b)]. At the same time, while no discernible differences in efficacy were observed among these five treatments, it is noteworthy that only Lenva + Pemb and Sint + Beva retained their association with superior OS benefits when compared to lenvatinib in the HBV-positive subgroup. Given the challenging nature of the results presented above, making analysis and summarization difficult, we conducted an integrated subgroup analysis. We found that, in comparison to sorafenib, ICI-combined therapies provide significant survival benefits in both HBV-positive groups and non-HBV groups. Notably, in the HBV-positive group, the HR values were significantly smaller. In addition, these therapies demonstrated OS benefits compared to MTD in the HBV-positive group, while no such benefits were observed in the non-HBV group (Supplemental Figure 1D). Interestingly, similar to the performance in a demographical subgroup, though ICI monotherapies did not exhibit superiority over the MTDs group, they demonstrated OS benefits compared to sorafenib regardless of HBV status.

Given tumor staging may influence the treatment decision, subgroup analysis was performed according to macrovascular invasion (MVI) and/or extrahepatic spread (EHS) status [Figure 4(c)]. Regardless of tumor stage, there were no significant differences among the five treatment regimens – Sint + Beva, Atezo + Beva, Camre + Rivo, Lenva + Pemb, and Tre + Du. In HCC patients diagnosed with MVI/EHS, Sint + Beva, Atezo + Beva, Camre + Rivo, and Lenva + Pemb consistently outperform sorafenib and lenvatinib in terms of OS. However, Tre + Du exhibited superiority only compared to sorafenib and lacked an advantage over lenvatinib. While in the MVI/EHS negative subgroup, only Sint + Beva and donafenib demonstrated OS benefits over sorafenib. In addition, an integrated subgroup analysis was conducted. We found that in the MVI/EHS-positive group, ICI-combined therapies demonstrated advantages over ICI monotherapies and MTDs, while in the absence of MVI/EHS, no such advantages were observed (Supplemental Figure 1E). Besides, ICI monotherapies also showed a better OS benefit over sorafenib in the MVI/EHS-positive group.

As tumor biomarkers may indicate treatment efficacy according to previous studies,^38,39 serum alpha-fetoprotein (AFP) was used as a useful biomarker for treatment decisions under second-line treatment of HCC.⁴⁰ Subgroup analyses were performed based on AFP status. In Figure 4(d), no significant differences were observed among these five regimens, including Sint + Beva, Atezo + Beva, Camre + Rivo, Lenva + Pemb, and Tre + Du, in both the AFP high and low subgroups. However, only Sint + Beva and Camre + Rivo demonstrated OS benefits over sorafenib in patients, irrespective of whether AFP was high or low. Furthermore, Atezo + Beva exhibited OS benefits exclusively in the AFP low subgroup, whereas Tre + Du and Lenva + Pemb demonstrated OS benefits solely in the AFP high subgroup [Figure 4(d)]. Nevertheless, in our integrated subgroup analysis, we observed that ICI-combined therapies consistently outperformed both MTDs and sorafenib, regardless of AFP status, while ICI monotherapies exhibited greater benefit over sorafenib but not over MTDs (Supplemental Figure 1F).

Risk of bias

The Cochrane tool was used to assess the risk bias of included studies. After the assessment, most of the studies we included were considered low risk in terms of most aspects. Nine studies were ranked as high risk with respect to blinding of participants and personnel.^7,32,34,35 Due to the lack of a detailed description of the blinding of outcome assessment of PFS, the interpretation of outcomes regarding PFS should be cautious (Supplemental Figure 2).

Discussion

Recently, emerging drugs and regimens have continuously challenged sorafenib’s status as the first-line treatment. As the landscape of advanced HCC treatment evolves, there is an urgent need for a comprehensive review and summarization of the efficiency and safety of both current first-line therapies and emerging schemes. Such an analysis could provide crucial insights for clinical decisions in the ever-changing landscape of HCC treatment. Sonbol et al.⁴¹ and Fulgenzi et al.²³ conducted separate network meta-analyses on first-line treatments for advanced HCC, but it is worth mentioning that their studies did not include several high-quality studies reporting impressive results in recent years. At the same time, Liu et al.²¹ and Fulgenzi et al.²³ conducted separate meta-analyses, assessing the comparative effectiveness of current therapies against sorafenib. In addition, they explored the efficacy of Atezo + Beva in comparison to other treatments. By contrast, we also analyzed the distinctions between emerging combination therapies and novel treatments in comparison to other MTDs, such as lenvatinib and donafenib. Furthermore, comprehensive subgroup analyses were also valuable to fully understand the optimal combination for advanced HCC patients with specific characteristics, which might help us figure out the population who might benefit from the specific regimens. Our preliminary results indicated that (1) ICI-combined regimens, including Sint + Beva, Atezo + Beva, Tre + Du, Lenva + Pemb, and Camre + Rivo, demonstrated OS benefits compared to both sorafenib and lenvatinib, except for Tre + Du, which exhibited OS benefits only compared to sorafenib. (2) The OS benefits of first-line regimens were affected by patient characteristics, including etiologies, tumor burdens, AFP levels, and so on. (3) ICI not only demonstrated comparable efficacy to sorafenib but sometimes surpassed it, exhibiting a lower incidence of AEs. (4) While the ICI-based combined regimens did not significantly increase grade 3–5 TRAEs compared to sorafenib and lenvatinib, a notable increase was observed when compared to ICI monotherapies.

ICIs targeted the molecular markers on the immune cells or tumor cells and were able to activate or block the immune cell response or tumor cell escapes, respectively. Although ICIs have achieved unprecedented clinical benefits, there is only a subset of patients reaching limited benefits.⁴² On the other hand, MTDs targeted intracellular signaling pathways associated with angiogenesis, cancer cell survival, or growth have been proven to be effective in improving disease control rates, including HCC. Moreover, some studies indicate that MTDs also possess the capability to regulate immune cells and the immune microenvironment,^43,44 which suggests that their combination with ICIs may have a synergistic effect. And mounting evidence has shown that the addition of MTDs to ICIs was able to enhance anti-tumor immunity by blocking the intracellular pathways associated with immune evasion.^45,46 Promising results in other cancers^47,48 encouraged the following clinicians to practice the novel MTD and ICI regimens clinically and basically in HCC. In this network meta-analysis, we observed that most ICI-combined therapy regimens were basically successful. Particularly, Sint + Beva, Atezo + Beva, and Camre + Rivo exhibited superiority not only over current MTDs but also over emerging ICI monotherapies. Though Tre + Du did not show superiority over lenvatinib and donafenib, the recently released exploratory 4-year follow-up data from the HIMALAYA trial show that it exhibits unprecedented 3- and 4-year OS rates (30.7% and 25.2%, respectively).⁴⁹ In LEAP002, Lenva + Pemb did not demonstrate an OS benefit over lenvatinib. However, our network meta-analysis revealed an OS advantage when comparing Lenva + Pemb against both lenvatinib and sorafenib. This phenomenon may be explained by the longest-ever observed OS in the lenvatinib arm (19.0 months; 95% CI: 17.2–21.7) in LEAP002. When conducting the network meta-analysis, all studies related to lenvatinib were integrated into a single overall effect for comparison. Besides, the failure of the Cabo + Atezo combination suggests that the role of MTDs independent of their anti-VEGF action in antitumor effects requires further illustration.

Though nivolumab, anti-PD-1 therapy, did not achieve what they expected, patients who received nivolumab had marginally higher objective response rates and more durable disease control (median 7.5 months versus 5.7 months) with relatively low rates of grade 3–5 AEs (22% versus 49%).²⁸ And responders had higher PD-L1 expression in tumor tissue, which indicated that nivolumab might function well in a specific group.^28,50 Besides, durvalumab and tislelizumab, the other two ICI monotherapies, demonstrated non-inferiority to sorafenib in extending OS and exhibited comparatively lower rates of TRAEs. Moreover, our integrated analysis indicates that ICI monotherapies exhibit OS benefits over sorafenib. As mentioned earlier, considering the significantly lower incidence of TRAEs in ICI monotherapies, they may emerge as a preferred first-line treatment option over sorafenib for patients who are not suitable for ICI-combined therapies.

As we know, HCC primarily develops in the context of chronic liver inflammation, and the immune dysfunction varies from different hepatitis viruses to non-alcoholic steatohepatitis (NASH).^16,51 Our subgroup and integrated analysis provides further indication that the efficacy of ICIs-based combination therapies shrank among the non-HBV subgroup compared with the HBV-positive subgroup. This might result from that HBV-related HCC increased immunogenicity provided by viral nonpeptides,⁵² which facilitated ICI-combined therapies to exert greater efficacy. Besides, Cheng et al. found that tumor-resident memory (Trm) CD8⁺ T cells were associated with a favorable outcome in patients with HBV-related HCC.^20,53 However, Pfister et al.¹⁶ reported that a subset of protumorigenic cells in NASH favors the development of HCC and hamper response to ICIs. In the meanwhile, a former study found that hepatitis C virus (HCV)-related HCC tended to induce more exhausted and dysfunctional CD8⁺ T cells, which might impair the efficacy of ICI-based combination therapies.^19,20 Interestingly, our subgroup analysis based on different regions showed that patients from Asia-Pacific tend to benefit more from regimens containing ICIs than those from the rest of the world [Figure 4(a)]. It is generally believed that the types of viral hepatitis between the Asia-Pacific region and the non-Asia-Pacific region were dramatically different. Patients from Asia-Pacific, especially China, were usually diagnosed with HBV hepatitis. By contrast, in the European and American regions, patients tended to be diagnosed with HCV hepatitis or alcoholic steatohepatitis.⁵⁴ This medical phenomenon was consistent with our subgroup analysis associated with HBV status [Figure 4(b) and Supplemental Figure 1D]. Certainly, the discrepancy of efficacy that ICIs exerted on patients from different regions could not be simply accounted for the status of hepatitis infection, as such phenomenon has also been observed in other kinds of cancer, such as lung cancer and renal cell carcinoma.^47,55 Thus, the factors that affected the efficacy of ICI-based therapy in patients from the western world should be identified, which might help us comprehensively understand more mechanisms of ICI therapy and become a target in the future.

A study has suggested that customized dosing strategies can contribute to an extended OS with sorafenib.⁵⁶ At the same time, we found that ICI-combined therapies did not always improve OS among patients without MVI and EHS, which indicated relatively low risk [Figure 4(c) and Supplemental Figure 1E]. This suggests that the benefit level for intermediate-stage patients receiving ICI-combined therapies is likely comparable to those receiving MTDs. However, it is crucial to note that clinical trials are typically designed with relatively short follow-up periods and a higher proportion of participants in the advanced stage. This circumstance makes it challenging to demonstrate clear advantages in the intermediate-stage subgroup. Therefore, when formulating recommendations for intermediate-stage HCC, it is essential to consider the broader clinical context. Decisions should not solely rely on subgroup analyses with acknowledged limitations. In renal cell carcinoma, given ICI-based therapies did not significantly improve OS compared with MTDs, MTDs alone are listed as the preferred option in the low-risk group according to International Metastatic Renal Cell Carcinoma Database Consortium (IMDC) by NCCN kidney cancer guidelines.^48,55,57 However, both ICIs combined with targeted therapy and the combinations of ICIs provided significant OS benefits versus targeted monotherapies in the high-risk group, according to IMDC, and NCCN kidney cancer guidelines listed combination regimens as the preferred option in first-line therapies.^48,55,57 Thus, patients classified as low risk were advised to use MTD monotherapy first, which might ease the economic burden of patients and be suitable for follow-up treatment decisions after progression. So, there remains uncertainty regarding whether ICI-combined therapies are superior to sorafenib across all subgroups. Inspired by these results mentioned above, there is an urgent need to identify and stratify HCC patients who would receive ICI-based regimens into different risk subgroups according to some characteristics.

Although various treatments may exhibit differences in performance based on AFP levels, our integrated analysis revealed that ICI-based therapies consistently outperformed both MTDs and sorafenib in both AFP high and AFP low subgroups [Figure 4(d) and Supplemental Figure 1F]. However, a study reported that AFP combined with C-reactive protein was able to predict the response to ICI therapy, independent of Child-Pugh class and performance status.⁵⁸ In the meantime, our former study indicated that the early reductions in AFP and prothrombin induced by vitamin K deficiency or antagonist-II (PIVKA-II) could be predictors of the efficacy of ICI therapy in HCC patients.³⁸ The difference between our analysis and former studies may derive from the different standards of high AFP levels and low AFP levels in each study. And some information might be covered up after pooling the data. Therefore, the follow-up studies should aim to explore better-predicting combinations that can provide the immune status of HCC patients for clinical decisions.

In the present study, we first included all the phase III RCTs that were published or up-to-date recently and comprehensively analyzed the major outcomes of efficacy and toxicity. By synthesizing evidence with the most extensive data, our results will assist clinical decisions for evaluating different regimens in treating patients with advanced HCC. For example, though there is no significant difference among the first-line or potential first-line regimens, including Sint + Beva, Atezo + Beva, Camre + Rivo, Lenva + Pemb, and Tre + Du, these three regimens (Sint + Beva, Atezo + Beva, and Camre + Rivo) demonstrate superior efficacy compared to other treatments, such as lenvatinib and ICI monotherapies. We were the first to propose that ICI monotherapies may emerge as a preferred first-line treatment option over sorafenib for patients who are not suitable for ICI-combined therapies. Moreover, this study separately analyzed the efficacy of treatments based on several factors that might play critical roles in affecting antitumor effectiveness, which has never spotted before. For instance, though the efficacy of ICI-combined regimens showed no significant differences in various subgroups, unlike the Asia-Pacific population, Atezo + Beva is the only one demonstrating superiority over sorafenib and lenvatinib in the non-Asia-Pacific population. Just as described earlier, the performance of ICI-combined therapies varies among different groups. These findings could provide valuable insights to assist clinicians in customizing the most suitable treatment plan based on the individual characteristics of the patients.

This network meta-analysis has several limitations. First, the estimations of this study were based on observational data from clinical trials, which contain unavoidable confounding factors. Second, each comparison between treatments included only one or two studies, which may cause lower statistical power and uncertainty in estimation due to heterogeneity. Third, patients were not stratified according to the volume and number of tumors which might influence the treatment benefits. For example, our analysis involved post hoc stratification based on factors such as MVI/EHS, HBV status, AFP levels, and geographical regions. This post hoc nature introduces the risk of potential biases, and caution is warranted in interpreting the results. Fourth, this network meta-analysis was conducted with summary aggregated data rather than individual patient data, which would impair the power of this analysis. For example, in the HBV-negative subgroup, we pooled the effect of HCV-infected or non-viral infection patients, which might impair their efficacy.

Conclusion

The ICI-combined therapies significantly improved the prognosis of advanced HCC and demonstrated similar safety compared with sorafenib and lenvatinib. However, the optimum treatment regimens should be shaped by patient characteristics, such as etiologies, tumor stage, and AFP. It is important to note that ICI monotherapies, with lower incidence rates of TRAEs, not only demonstrated comparable efficacy to sorafenib but sometimes surpassed it.

Supplemental Material

sj-docx-1-tag-10.1177_17562848241237631 – Supplemental material for The emerging therapies are reshaping the first-line treatment for advanced hepatocellular carcinoma: a systematic review and network meta-analysis

Supplemental material, sj-docx-1-tag-10.1177_17562848241237631 for The emerging therapies are reshaping the first-line treatment for advanced hepatocellular carcinoma: a systematic review and network meta-analysis by Wei Peng, Yangxun Pan, Lan Xie, Zhoutian Yang, Zhiwei Ye, Jinbin Chen, Juncheng Wang, Dandan Hu, Li Xu, Zhongguo Zhou, Minshan Chen, Aiping Fang and Yaojun Zhang in Therapeutic Advances in Gastroenterology

Footnotes

Acknowledgements

AcknowledgementsSincere thanks to Yizhen Fu for the valuable insights provided during the revision process of the manuscript.

Declarations

ORCID iD

Wei Peng

Supplemental material

Supplemental material for this article is available online.

References

Asrani

Devarbhavi

Eaton

, et al. Burden of liver diseases in the world. J Hepatol 2019; 70: 151–171.

Llovet

Bustamante

Castells

, et al. Natural history of untreated nonsurgical hepatocellular carcinoma: rationale for the design and evaluation of therapeutic trials. Hepatology 1999; 29: 62–67.

Park

Chen

Colombo

, et al. Global patterns of hepatocellular carcinoma management from diagnosis to death: the BRIDGE Study. Liver Int 2015; 35: 2155–2166.

Altekruse

McGlynn

Reichman

ME.

Hepatocellular carcinoma incidence, mortality, and survival trends in the United States from 1975 to 2005. J Clin Oncol 2009; 27: 1485–1491.

Llovet

Ricci

Mazzaferro

, et al. Sorafenib in advanced hepatocellular carcinoma. N Engl J Med 2008; 359: 378–390.

Cheng

Kang

Chen

, et al. Efficacy and safety of sorafenib in patients in the Asia-Pacific region with advanced hepatocellular carcinoma: a phase III randomised, double-blind, placebo-controlled trial. Lancet Oncol 2009; 10: 25–34.

Kudo

Finn

Qin

, et al. Lenvatinib versus sorafenib in first-line treatment of patients with unresectable hepatocellular carcinoma: a randomised phase 3 non-inferiority trial. Lancet. 2018; 391: 1163–1173.

Kudo

Systemic therapy for hepatocellular carcinoma: latest advances. Cancers (Basel) 2018; 10: 412.

Finn

Qin

Ikeda

, et al. Atezolizumab plus bevacizumab in unresectable hepatocellular carcinoma. N Engl J Med 2020; 382: 1894–1905.

10.

Llovet

Villanueva

Marrero

, et al. Trial design and endpoints in hepatocellular carcinoma: AASLD Consensus Conference. Hepatology 2021;73(Suppl. 1): 158–191.

11.

Singal

Llovet

Yarchoan

, et al. AASLD Practice Guidance on prevention, diagnosis, and treatment of hepatocellular carcinoma. Hepatology 2023; 78: 1922–1965.

12.

Ren

Bai

, et al. Sintilimab plus a bevacizumab biosimilar (IBI305) versus sorafenib in unresectable hepatocellular carcinoma (ORIENT-32): a randomised, open-label, phase 2–3 study. Lancet Oncol 2021; 22: 977–990.

13.

Qin

, et al. Donafenib versus sorafenib in first-line treatment of unresectable or metastatic hepatocellular carcinoma: a randomized, open-label, parallel-controlled phase II–III trial. J Clin Oncol 2021; 39: 3002–3011.

14.

Abou-Alfa

Chan

Kudo

, et al. Phase 3 randomized, open-label, multicenter study of tremelimumab (T) and durvalumab (D) as first-line therapy in patients (pts) with unresectable hepatocellular carcinoma (uHCC): HIMALAYA. American Society of Clinical Oncology; 2022.

15.

Qin

Chan

, et al. Camrelizumab plus rivoceranib versus sorafenib as first-line therapy for unresectable hepatocellular carcinoma (CARES-310): a randomised, open-label, international phase 3 study. Lancet 2023; 402: 1133–1146.

16.

Pfister

Núñez

Pinyol

, et al. NASH limits anti-tumour surveillance in immunotherapy-treated HCC. Nature 2021; 592: 450–456.

17.

Bejjani

Finn

RS.

Hepatocellular carcinoma: pick the winner-tyrosine kinase inhibitor versus immuno-oncology agent-based combinations. J Clin Oncol 2022; 40 :2763–2773.

18.

Casadei-Gardini

Rimini

Tada

, et al. Atezolizumab plus bevacizumab versus lenvatinib for unresectable hepatocellular carcinoma: a large real-life worldwide population. Eur J Cancer 2023; 180: 9–20.

19.

Haber

Puigvehí

Castet

, et al. Evidence-based management of hepatocellular carcinoma: systematic review and meta-analysis of randomized controlled trials (2002–2020). Gastroenterology 2021; 161: 879–898.

20.

Llovet

Castet

Heikenwalder

, et al. Immunotherapies for hepatocellular carcinoma. Nat Rev Clin Oncol 2022; 19: 151–172.

21.

Fulgenzi

CAM

Scheiner

Korolewicz

, et al. Efficacy and safety of frontline systemic therapy for advanced HCC: a network meta-analysis of landmark phase III trials. JHEP Rep 2023; 5: 100702.

22.

Liu

Zhu

Immunotherapy or targeted therapy as the first-line strategies for unresectable hepatocellular carcinoma: a network meta-analysis and cost-effectiveness analysis. Front Immunol 2022; 13: 1103055.

23.

Fulgenzi

CAM

D’Alessio

Airoldi

, et al. Comparative efficacy of novel combination strategies for unresectable hepatocellular carcinoma: a network metanalysis of phase III trials. Eur J Cancer 2022; 174: 57–67.

24.

Page

McKenzie

Bossuyt

, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ 2021; 372: n71.

25.

European Association for the Study of the Liver. EASL Clinical Practice Guidelines: Management of hepatocellular carcinoma. J Hepatol 2018; 69: 182–236.

26.

Zhou

Sun

Wang

, et al. Guidelines for diagnosis and treatment of primary liver cancer in China (2017 edition). Liver Cancer 2018; 7: 235–260.

27.

Cumpston

Page

, et al. Updated guidance for trusted systematic reviews: a new edition of the Cochrane Handbook for Systematic Reviews of Interventions. Cochrane Database Syst Rev 2019; 10: ED000142.

28.

Yau

Park

Finn

, et al. Nivolumab versus sorafenib in advanced hepatocellular carcinoma (CheckMate 459): a randomised, multicentre, open-label, phase 3 trial. Lancet Oncol 2022; 23: 77–90.

29.

Llovet

Kudo

Merle

, et al. Lenvatinib plus pembrolizumab versus lenvatinib plus placebo for advanced hepatocellular carcinoma (LEAP-002): a randomised, double-blind, phase 3 trial. Lancet Oncol 2023; 24: 1399–1410.

30.

Qin

Kudo

Meyer

, et al. Tislelizumab vs sorafenib as first-line treatment for unresectable hepatocellular carcinoma: a phase 3 randomized clinical trial. JAMA Oncol 2023; 9: 1651–1659.

31.

Kelley

Rimassa

Cheng

, et al. Cabozantinib plus atezolizumab versus sorafenib for advanced hepatocellular carcinoma (COSMIC-312): a multicentre, open-label, randomised, phase 3 trial. Lancet Oncol 2022; 23: 995–1008.

32.

Cheng

Kang

Lin

, et al. Sunitinib versus sorafenib in advanced hepatocellular cancer: results of a randomized phase III trial. J Clin Oncol 2013; 31: 4067–4075.

33.

Johnson

Qin

Park

, et al. Brivanib versus sorafenib as first-line therapy in patients with unresectable, advanced hepatocellular carcinoma: results from the randomized phase III BRISK-FL study. J Clin Oncol 2013; 31: 3517–3524.

34.

Cainap

Qin

Huang

, et al. Linifanib versus sorafenib in patients with advanced hepatocellular carcinoma: results of a randomized phase III trial. J Clin Oncol 2015; 33: 172–179.

35.

Abou-Alfa

Shi

Knox

, et al. Assessment of treatment with sorafenib plus doxorubicin vs sorafenib alone in patients with advanced hepatocellular carcinoma: phase 3 CALGB 80802 randomized clinical trial. JAMA Oncol 2019; 5: 1582–1588.

36.

Jouve

Lecomte

Bouché

, et al. Pravastatin combination with sorafenib does not improve survival in advanced hepatocellular carcinoma. J Hepatol 2019; 71: 516–522.

37.

Zhu

Rosmorduc

Evans

, et al. SEARCH: a phase III, randomized, double-blind, placebo-controlled trial of sorafenib plus erlotinib in patients with advanced hepatocellular carcinoma. J Clin Oncol 2015; 33: 559–566.

38.

Sun

Mei

Lin

, et al. Reductions in AFP and PIVKA-II can predict the efficiency of anti-PD-1 immunotherapy in HCC patients. BMC Cancer 2021; 21: 775.

39.

Nishino

Ramaiya

Hatabu

, et al. Monitoring immune-checkpoint blockade: response evaluation and biomarker development. Nat Rev Clin Oncol 2017; 14: 655–668.

40.

Bruix

Reig

Merle

, et al. 755P – Alpha-fetoprotein (AFP) response in patients with unresectable hepatocellular carcinoma (HCC) in the phase III RESORCE trial. Ann Oncol 2019; 30: v291.

41.

Sonbol

Riaz

Naqvi

SAA

, et al. Systemic therapy and sequencing options in advanced hepatocellular carcinoma: a systematic review and network meta-analysis. JAMA Oncol 2020; 6: e204930.

42.

Sharma

Hu-Lieskovan

Wargo

, et al. Primary, adaptive, and acquired resistance to cancer immunotherapy. Cell 2017; 168: 707–723.

43.

Granito

Forgione

Marinelli

, et al. Experience with regorafenib in the treatment of hepatocellular carcinoma. Therap Adv Gastroenterol 2021; 14: 17562848211016959.

44.

Kimura

Kato

Ozawa

, et al. Immunomodulatory activity of lenvatinib contributes to antitumor activity in the Hepa1-6 hepatocellular carcinoma model. Cancer Sci 2018; 109: 3993–4002.

45.

Rahma

Hodi

FS.

The Intersection between tumor angiogenesis and immune suppression. Clin Cancer Res 2019; 25: 5449–5457.

46.

Ruiz de Galarreta

Bresnahan

Molina-Sánchez

, et al. β-Catenin activation promotes immune escape and resistance to anti-PD-1 therapy in hepatocellular carcinoma. Cancer Discov 2019; 9: 1124–1141.

47.

Paz-Ares

Ciuleanu

Cobo

, et al. First-line nivolumab plus ipilimumab combined with two cycles of chemotherapy in patients with non-small-cell lung cancer (CheckMate 9LA): an international, randomised, open-label, phase 3 trial. Lancet Oncol 2021; 22: 198–211.

48.

Motzer

Jonasch

Boyle

, et al. NCCN guidelines insights: kidney cancer, version 1.2021: featured updates to the NCCN guidelines. J Natl Compr Cancer Netw 2020; 18: 1160–1170.

49.

Sangro

Chan

Kelley

(eds.). Four-year overall survival update from the phase 3 HIMALAYA study of tremelimumab plus durvalumab in unresectable hepatocellular carcinoma. In: ESMO World Congress on Gastrointestinal Cancer, Barcelona, Spain, 2023.

50.

Sangro

Melero

Wadhawan

, et al. Association of inflammatory biomarkers with clinical outcomes in nivolumab-treated patients with advanced hepatocellular carcinoma. J Hepatol 2020; 73: 1460–1469.

51.

Sangro

Sarobe

Hervás-Stubbs

, et al. Advances in immunotherapy for hepatocellular carcinoma. Nat Rev Gastroenterol Hepatol 2021; 18: 525–543.

52.

Rooney

Shukla

, et al. Molecular and genetic properties of tumors associated with local immune cytolytic activity. Cell 2015; 160: 48–61.

53.

Cheng

Gunasegaran

Singh

, et al. Non-terminally exhausted tumor-resident memory HBV-specific T cell responses correlate with relapse-free survival in hepatocellular carcinoma. Immunity 2021; 54: 1825–1840.e7.

54.

Akinyemiju

Abera

Ahmed

, et al. The burden of primary liver cancer and underlying etiologies from 1990 to 2015 at the global, regional, and national level: results from the Global Burden of Disease Study 2015. JAMA Oncol 2017; 3: 1683–1691.

55.

Rini

Plimack

Stus

, et al. Pembrolizumab plus axitinib versus sunitinib for advanced renal-cell carcinoma. N Engl J Med 2019; 380: 1116–1127.

56.

Tovoli

Ielasi

Casadei-Gardini

, et al. Management of adverse events with tailored sorafenib dosing prolongs survival of hepatocellular carcinoma patients. J Hepatol 2019; 71: 1175–1183.

57.

Cella

Grünwald

Escudier

, et al. Patient-reported outcomes of patients with advanced renal cell carcinoma treated with nivolumab plus ipilimumab versus sunitinib (CheckMate 214): a randomised, phase 3 trial. Lancet Oncol 2019; 20: 297–310.

58.

Scheiner

Pomej

Kirstein

, et al. Prognosis of patients with hepatocellular carcinoma treated with immunotherapy – development and validation of the CRAFITY score. J Hepatol 2022; 76: 353–363.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

1.07 MB