According to GBD 2021 data, metabolic dysfunction-associated steatohepatitis (MASH)-associated liver cancer made up 8% of all liver cancer cases and 9% of liver cancer-related deaths (Fig. 1E, 1F). Due to the rising prevalence of metabolic syndrome, MASH has become the fastest-growing cause of HCC in many regions of the world [5,6,25,26], with disproportionate increases seen in low and low-middle-income countries [27,28]. MASLD is the most commonly occurring chronic liver disease, affecting one-fourth to one-third of the global population [29-32]. In addition to being a risk factor for chronic liver disease, MASLD can accelerate cardiovascular risk even in the absence of metabolic comorbidities [33,34]. Over time, hepatic steatosis causes a disruption in hepatic metabolism, increasing lipotoxicity and oxidative stress, and triggering immune system activation [35]. MASH, the inflammatory subtype of MASLD, involves both steatosis and signs of hepatocyte injury and inflammation [36]. It is currently estimated that approximately 25% of patients with MASLD progress to MASH [37-39]. Approximately 40% of MASH patients develop fibrosis, with 15% progressing to cirrhosis [35]. Interestingly, MASH-related HCC is more frequently observed in patients without cirrhosis compared to HCC from other causes [40]. Emerging data also suggest that MASH-HCC patients may have distinct characteristics compared to those with HCC from other etiologies. Patients with MASH-HCC tend to be female and older and has more metabolic comorbidities [41,42]. Additionally, patients with MASH-HCC are often diagnosed at more advanced stages compared to HCC from other causes [43,44].

In 2021, ALD-associated liver cancer represented 19% of both total liver cancer incidence and liver cancer-related deaths (Fig. 1E, 1F). The increase in global per-capita alcohol consumption has resulted in a surge in ALD and associated complications, particularly during the COVID-19 pandemic [45-49]. Another meta-analysis by the World Cancer Research Fund found that liver cancer risk increased by 4% for each 10 g of alcohol consumption per day [50]. A population-based study in Sweden, which utilized liver biopsy-proven ALD, found that patients with ALD had a significantly elevated risk of HCC, with a hazard ratio (HR) of 12.8 (vs. general population).51 Another systematic review and meta-analysis of over 148,000 patients determined that the cumulative incidence of HCC in patients with ALD cirrhosis at 5- and 10-year follow-ups was 3% and 9%, respectively [52]. Alcohol-associated HCC is linked to more advanced Barcelona Clinic Liver Cancer (BCLC) staging at diagnosis, reduced chances of receiving curative treatment, and worse survival outcomes [53]. Chronic alcohol consumption also increases the risk of HCC in patients with other etiologies of chronic liver diseases [54-57]. Therefore, strict abstinence from alcohol is recommended in patients with chronic liver diseases to prevent the progression of liver diseases and ultimately, HCC development.

Chronic HBV infection was responsible for 39% of total liver cancer cases and 37% of liver cancer-related deaths (Fig. 1E, 1F). Patients with chronic HBV infection are at increased risk of developing HCC [58], with risk significantly elevated by factors such as cirrhosis, high HBV DNA levels, specific genotypes, and co-infections (e.g., hepatitis D) [59-62]. Although HBV remains the leading etiology of HCC globally, expanded vaccination coverage has led to a gradual decline in HBV-related HCC [63]. A recent study revealed a nonlinear parabolic association between HBV DNA titer and HCC risk, with HBV DNA around 6 log10 IU/mL (moderate viral loads) corresponded to the highest non-invasive test scores for hepatic inflammation and fibrosis [64,65]. More recently, a large retrospective multinational study confirmed the nonlinear parabolic association between HBV DNA and HCC risk in patients with normal alanine aminotransferase (ALT) [65]. Among patients with HBV infection, antiviral therapy, particularly with nucleos(t)ide analogs, significantly reduces HCC risk by 40–60% by suppressing viral replication and reducing inflammation, contributing to the global reduction of HBV-related HCC [66-69].

Chronic HCV infection accounted for 29% of total liver cancer cases and 30% of liver cancer-related deaths (Fig. 1E, 1F). Chronic HCV infection is a well-established risk factor for HCC, with the highest risk observed in patients who have developed cirrhosis [70]. Increased implementation of HCV treatment programs using direct-acting antivirals has significantly reduced HCV-related HCC, although increased efforts are needed to meet World Health Organization (WHO) viral hepatitis elimination goals [71-74]. Further, HCC risk is significantly reduced but not eliminated after viral eradication, with studies showing persistent annual HCC incidence rates exceeding 1% more than 5 years after sustained virologic response, underscoring the importance of ongoing surveillance [75-77].

Aflatoxins are potent carcinogenic mycotoxins that commonly contaminate foods such as maize and groundnuts, exposing approximately 4.5 billion people globally [78,79]. The majority of liver cancer cases linked to aflatoxin exposure are concentrated in Sub-Saharan Africa, Southeast Asia, and China [80,81]. The primary aflatoxin involved in liver carcinogenesis is aflatoxin B1 (AFB1), produced by Aspergillus species. AFB1 leads to the formation of DNA adducts, which interact with guanine bases in hepatocyte DNA, causing a mutation in the P53 tumor suppressor gene at the codon 249 hotspot in exon 7, ultimately resulting in HCC [82,83]. The risk of HCC from aflatoxin is not evenly distributed, with a higher risk observed in children and individuals with HBV infection [80,84]. The increased risk observed with aflatoxin and HBV may be due to chronic HBV infection-inducing cytochrome P450 enzymes, which convert inactive AFB1 into the mutagenic AFB1–8,9-epoxide, thereby causing mutations in hepatocytes and ultimately hepatocarcinogenesis [85].

Current guidelines recommend semi-annual HCC surveillance in at-risk populations [86-88]. Traditionally, surveillance was deemed cost-effective in cirrhosis patients with an annual HCC incidence rate of 1.5% or greater [89]. More recent cost-effectiveness analyses demonstrated an HCC incidence of more than 0.4% per year and surveillance adherence of more than 19.5% biannually for ultrasound (US) with alpha-fetoprotein (AFP) to be cost-effective in patients with compensated cirrhosis [90]. Cirrhosis patients with Child-Pugh class A or B, regardless of etiology, should undergo surveillance, while this should be restricted to transplant-eligible patients in the setting of Child-Pugh class C cirrhosis [86,91]. Non-cirrhotic, chronic hepatitis B patients are also at high risk to warrant surveillance based on their risk factors such as age, sex, endemic region, family history, HBV DNA titer, or the PAGE-B score [86]. Although 25–30% of HCC cases in patients with MASLD occur in the absence of cirrhosis, the annual risk of HCC in noncirrhotic MASLD is low. Universal HCC surveillance is not cost-effective, underscoring the need for risk stratification in this patient population [40]. Precision surveillance based on risk group using four patient factors (sex, etiology of liver disease, Child-Pugh class, and body mass index) was able to detect a higher proportion of HCC at an early stage and was more cost-effective than the universal approach, although it requires further validation before implementation in clinical practice [92].

The best data for HCC surveillance comes from a large randomized controlled trial (RCT) in patients with chronic HBV infection [93]. Although there is not a similar RCT among patients with cirrhosis, several cohort studies show associations between HCC surveillance and early-stage HCC detection, curative treatment receipt, and prolonged survival [94]. The benefits of HCC surveillance must be weighed against potential physical, psychological, and financial harms. There are fewer studies enumerating surveillance-related harms, although available data suggest these are of mild severity in most patients [94-98]. The balance of benefits vs. harms may be particularly relevant when expanding surveillance to patients with advanced fibrosis but no cirrhosis, in whom the annual incidence of HCC is lower, and the number needed to screen will inevitably become larger, further escalating the extent of unnecessary harm [99,100].

Despite the advancement in surveillance modalities for early HCC detection, HCC surveillance in clinical practice is underused. Among 1,014 patients with cirrhosis with HCC, only 37.2% of the patients had regular outpatient care the year before HCC presentation, and only 24.7% of those under regular outpatient care were under surveillance. Nearly half lacked surveillance orders despite knowing the patient had cirrhosis [101]. Underuse of HCC surveillance is related to a combination of patient and provider-reported barriers [102,103]. Common patient-reported barriers include cost, difficulty scheduling, uncertainty of where to receive an ultrasound exam, or difficulty with transportation [102]. Underuse of surveillance is particularly high in patients with non-viral liver diseases, including those with ALD or MASLD, related to a combination of difficulty recognizing at-risk patients as well as increased social and financial barriers [53,104]. HCC surveillance is also underused in racial and ethnic minorities, populations disproportionately impacted by HCC mortality [105]. Several interventions are efficacious in increasing surveillance, but studies are needed to see their effectiveness when implemented in routine clinical practice [106].

Biannual surveillance for HCC using abdominal ultrasound with/without AFP is recommended by multiple professional society guidelines with few notable similarities and differences (Table 1) [86-88,107]. Biannual ultrasound has been the standard imaging of choice due to its cost-effectiveness, low risk of adverse events, and detection of other complications of cirrhosis [87]. However, the sensitivity of ultrasound for early-stage HCC is operator-dependent as well as varies widely based on the patient’s body habitus, liver disease etiology, and presence of cirrhosis (Table 2) [108-110]. Therefore, there are concerns that ultrasound performance will worsen over time given the shift from viral-related liver disease to increasing proportions due to MASLD and ALD. Although ultrasound in combination with AFP increases sensitivity for early-stage HCC detection, this strategy still misses over one-third of HCC at an early stage [90].

There are multiple emerging blood- and imaging-based strategies for HCC surveillance (Table 2). AFP-L3, which measures a subfraction of the AFP, and des gamma carboxy prothrombin (DCP), also known as protein induced by vitamin K absence/antagonist-II (PIVKA-II) are two serum protein biomarkers extensively studied and currently approved by the Food and Drug Administration (FDA) for risk stratification [111]. Although each biomarker has insufficient sensitivity for early-stage when used alone, biomarker panels including multiple biomarkers appear to have higher accuracy. The original hepatocellular carcinoma early detection screening (HES) score combining AFP with age, ALT, and platelet count has shown to be superior to AFP alone for detection of early HCC, and a more recent version of HES V2.0 including AFP-L3 and des-gamma-carboxy prothrombin was proposed with excellent performance [112,113]. The GALAD (gender, age, AFP-L3, AFP, des-gamma-carboxy prothrombin) model combined the three biomarkers and was recently evaluated in a national phase 3 biomarker study demonstrating an increase in accuracy: GALAD had the highest true positive rate (63.6%, 73.8%, and 71.4% for all HCC, and 53.8%, 63.3%, and 61.8% for early HCC within 6, 12, and 24 months, respectively) [111,114,115]. Another phase 3 study revealed that longitudinal GALAD further improved the sensitivity up to 66.7% (vs. 54.8% for single point GALAD) [116]. Given these promising results, GALAD is now undergoing prospective validation compared to ultrasound in a large RCT in the US enrolling over 5,000 patients with cirrhosis (NCT06084234).

Liquid biopsy is also emerging as a supplementary approach to imaging for the early detection of HCC, which consists of analysis of circulating tumor DNA, tumor cells, or tumor-derived extracellular vesicles (EV) [117]. Methylation profiling of circulating tumor DNA has evolved as a promising surveillance tool for detecting early HCC in at-risk populations, although phase 3 biomarker validation studies are yet to be done [117,118]. A recent phase 2 study on a six-marker methylated DNA marker panel (HOXA1, EMX1, AK055957, ECE1, PFKP, CLEC11A) yielded an area under the curve (AUC) of 0.96 with a sensitivity of 95% and specificity of 92% [119]. An algorithm incorporating sex and AFP levels with the methylation biomarkers (HOXA1, TSPYL5, and B3GALT6) was also developed, showing a higher early-stage sensitivity compared to AFP level (≥20 ng/mL) and GALAD (≥–0.63), although specificity was lower [118]. In addition, the HCC EV surface protein assay based on four HCC-associated surface protein markers and two EV markers is also a promising circulating biomarker for early-stage HCC detection and was validated in a phase 2 study [120]. These strategies are promising, although most data are limited to case-control studies, which can overestimate performance, and prospective validation in cohort studies is needed prior to adoption.

An abbreviated MRI has been a suggested alternative imaging modality to improve sensitivity since a complete MRI is not feasible as a surveillance tool due to prolonged scanning time. A pooled estimate of 15 studies demonstrated 86% sensitivity and 94% specificity per patient, comparable in sensitivity and specificity between non-contrast and contrast-enhanced MRI [121]. Abbreviated MRI is undergoing prospective validation in large trials in the US, France, and South Korea (NCT05486572, NCT05095714, NCT06312826) [122]. Although these studies show high test performance, issues including radiological capacity, costs, and patient acceptance would need to be considered for implementation.

Follow-up of surveillance test results for subsequent procedures or diagnostic testing is recommended based on the LI-RADS (Liver Imaging Reporting and Data System) visualization score, size of the lesion, and AFP levels [86]. Patients with well-visualized sonography (visualization A) with no lesions and normal AFP levels can repeat surveillance in 6 months.

For patients with subcentimeter liver lesions in ultrasound, a short-interval repeat test in 3–4 months is reasonable when the initial ultrasound demonstrated optimal visualization [123]. However, patients with suboptimal ultrasound visualization or increased odds of poor visualization, such as obesity, alcohol-related liver disease, or MASLD cirrhosis, should be considered for further computed tomography (CT) or MRI [124]. For patients with lesions that are more than 1 cm on the ultrasound, AFP levels over 20 ng/mL, or increasing AFP levels, diagnostic contrast-enhanced multiphase MRI or CT should be performed.

CT/MRI LI‐RADS is used to further characterize abnormal imaging findings [125]. For LR‐1 (definitely benign) and LR‐2 (probably benign) lesions, return to ultrasound surveillance at a routine 6‐month interval is recommended for LR‐1 observations, while follow‐up CT or MRI in 6 months or less may be considered for LR‐2 observations. For LR‐3 (intermediate probability of HCC) lesions, close monitoring with follow‐up CT or MRI in 3 to 6 months is warranted because approximately 30% to 40% of LR‐3 observations may represent HCC. For LR-4 (probably HCC) lesions, multidisciplinary discussion for tailored workup is recommended with options including a biopsy or repeated imaging in a short interval of around 3 months. LR‐5 is diagnostic for HCC, as discussed further in the following section.

Patients with localized HCC without evidence of vascular invasion or extrahepatic metastases are suitable candidates for liver resection. The underlying liver function and presence of portal hypertension (defined as HVPG ≤10 mmHg) play a major role in determining resectability [138]. Although most data are limited to BCLC 0 or A stages, studies on extending resection criteria to select patients with limited multifocal HCC or those with locally advanced HCC have suggested survival benefits compared to TACE or radiofrequency ablation (RFA). It suggests that, for early multinodular HCC patients ineligible for transplant, resection should be prioritized [139-145]. Most of the data for these extended indications are derived from Asian cohorts, with fewer data from Western cohorts in non-HBV populations. Minimally invasive liver resection including laparoscopic and robotic approaches is applied more to reduce morbidity and mortality [146]. Recurrence occurs in more than half of the patients undergoing surgical resection, mostly within 2 years. Nevertheless, surgical resection can achieve long-term survival by detecting and treating recurrent tumor early [147-149].

Local ablation is highly effective for tumors up to 3 cm in diameter. RFA and microwave ablation (MWA) are commonly used ablation treatments, although other ablative modalities are available with similar outcomes [150]. RFA induces radiofrequency-generated heat to destroy tumoral tissue and its efficacy has been demonstrated in early HCC patients [151]. MWA uses high-frequency electromagnetic energy to induce tissue death with large tumor ablation volumes, faster ablation time, and lower risk of heat-sink effect [152,153]. Although the local recurrence rate and disease-free survival rate were similar between RFA versus MWA, MWA had a lower distant recurrence rate, potentially decreasing the rate of long-term recurrences [154]. Both are a valuable option for patients who are not a candidate for surgery due to tumor size, number, central location, impaired liver function, or comorbidities and are a cost-effective option compared to surgery [155,156]. Depending on the location of the tumor, RFA can be adopted as first-line due to excellent local control of small HCC [87,157]. It has also been proven effective as downstaging and bridging therapy in the pre-transplant setting [158]. However, its efficacy diminishes in patients with large lesions (more than 3 cm) or lesions adjacent to large vessels [86]. Alternative treatment should be considered in patients with HCC adjacent to other critical structures, including central bile ducts, or if the area is not accessible via a percutaneous method. Similar to surgical resection, the high rate of tumor recurrence and/or distant metastasis remains a challenge [159]. Furthermore, there are alternative ablative modalities such as cryoablation which demonstrated lower local tumor progression and equivalent efficacy and safety compared to RFA in a multicenter RCT [160], or irreversible electroporation which utilizes short pulses of direct current at high voltage through multiple electrodes and recently demonstrated efficacy in a clinical setting on HCC patients [161].

Systemic therapies in the adjuvant and neoadjuvant settings for early-stage HCC have been investigated recently. In a phase 3 trial of adjuvant treatment with sorafenib after HCC surgical resection or local ablation, there was no benefit compared to the placebo in recurrence-free survival [162]. The IMbrave 050 study was the first phase 3 study to report positive results. There was a significantly improved recurrence-free survival (HR 0.72, 95% confidence interval [CI] 0.53–0.98) in the adjuvant therapy group (atezolizumab plus bevacizumab) vs. control group of high-risk HCC patients after curative-intent surgical resection or ablation [163]. However, more recently, longer follow-up results suggested a lack of improvement in progression-free survival [164]. Neoadjuvant therapies are also promising in priming the immune system to prevent recurrence and downstaging of locally advanced tumors for surgical resection or liver transplants [165]. HCC develops in a setting of chronic inflammation of the liver in which the microenvironment is altered and immune cells are altered; thus, several synergistic mechanisms for the efficacy of neoadjuvant immunotherapy have been proposed [165]. Neoadjuvant cabozantinib and nivolumab converted locally advanced HCC into resectable disease (80% with margin negative resection, 42% with major pathologic response) in a total of 15 patients [166]. A phase 2 study on perioperative nivolumab and nivolumab plus ipilimumab was safe and tolerable. 33% of the patients achieved major pathological response (MPR) and 5/6 MPR patients achieved complete response [167]. Multiple phase 1–2 clinical trials are ongoing to investigate the role of neoadjuvant immunotherapy for HCC (NCT05908786, NCT03299946, NCT05471674, NCT04224480, NCT03337841, NCT05440864, NCT03682276, NCT03510871, NCT04658147, NCT04123379).

LT is an optimal treatment option for early-stage, unresectable HCC in patients with cirrhosis as it can not only remove intrahepatic tumors but also cure the underlying cirrhosis. About 15% of patients experience HCC recurrence after LT with a mortality rate of 19.8% at 5 years and 35.7% at 10 years in patients receiving deceased donor LT (DDLT). There has been a significant improvement in transplant outcomes over the past few decades [168,169]. The selection of patients for LT has been historically guided by the Milan criteria (one tumor that is less than 5 cm in diameter; up to three tumors, each less than 3 cm in diameter) while there were ongoing efforts on extending the patient population eligible for LT [170]. Additional criteria have also been proposed such as the UCSF criteria where the overall size limitation has expanded (one tumor ≤6.5 cm in diameter or up to three tumors ≤4.5 cm with total tumor diameter ≤8 cm) with no significant difference in 5-year follow-up [171-173]. Successful downstaging, defined as a reduction of viable tumor burden by locoregional therapy within the Milan criteria, demonstrates good post-LT outcomes and is currently adopted in the United Network for Organ Sharing (UNOS) in the U.S [174,175]. Over 80% of the cases had a successful down-staging rate with HCC within the UNOS-DS criteria (a single lesion greater than 5 cm and up to 8 cm; two to three lesions, with at least one lesion less than 3 cm and each 5 cm, with a total tumor diameter of 8 cm; four to five lesions, each less than 3 cm, with a total tumor diameter of 8 cm) [176]. The outcome of “all-comers,” HCC exceeding UNOS-DS criteria in regard to primary tumor burden, had a slightly lower successful downstaging but acceptable rate (67%), suggesting that it is a realistic goal in selected all-comer patients [177].

Salvage LT is also a strategy for patients with HCC who have undergone resection and developed tumor recurrence within the LT criteria or liver decompensation [178,179]. It has shown better disease-free survival compared to second resection for recurrent HCC [180]. A major limitation of LT has been organ shortage, and living donor LT (LDLT) has been a viable option, especially in Asia [181]. LDLT recipients showed comparable long-term outcomes compared to DDLT recipients. LDLT may provide better survival benefits, especially in regions that had low deceased organ availability [182].

TACE is the standard of care treatment for intermediate-stage HCC with well-defined nodules, preserved portal flow, and selective access [86,183]. It is a method of injecting chemotherapy into tumor-feeding arteries while hindering the arterial blood supply to the tumor, inducing tumor necrosis [184]. Two RCTs confirmed the survival benefit of TACE for unresectable HCC [185,186]. It is often used as a bridging or downstaging therapy for the patient to be a candidate for LT or resection [184]. However, challenges remain as patients may develop TACE-refractory tumors or experience failure to achieve objective response after two or more treatments [187]. Furthermore, transarterial embolization using microspheres alone when compared to TACE demonstrated no difference in RECIST response in a randomized trial in a single tertiary center, challenging the role of chemotherapy [188].

TARE delivers radioactive microspheres loaded with Y90 leading to tumor destruction with excellent efficacy in solitary HCC [189]. A phase 2 study randomized BCLC stage B or C, Child-Pugh stage A patients with unablatable/unresectable HCC to either Y90 therapy or TACE and was able to demonstrate that the Y90 radioembolization group had a significantly longer median time to progression (TTP) (>26 versus 6.8 months) with better tumor control and increased likelihood to receive transplantation and slightly favorable side effect profile [190]. The superiority of TARE in terms of TTP was insufficient to prolong the overall survival compared to TACE [191]. However, prolonging TTP through TARE can provide higher rates of successful bridging to transplantation, especially as Y90 is an intra-arterial therapy that eliminates the risk of tract seeding. TARE is better tolerated than TACE [190]. The TRACE phase 2 RCT compared Y90 glass TARE with doxorubicin drug-eluting bead TACE and demonstrated superiority in tumor control and overall survival [192]. In addition, a phase 2 trial (DOSISPHERE-01) reported a personalized dosimetry approach delivering at least 205 Gy to the index lesion improved overall survival, especially in patients downstaged for resection compared to the prior standardized approach, highlighting the importance of using personalized dosimetry [193].

Stereotactic body radiation therapy (SBRT) is an emerging treatment option across early, intermediate and advanced stages of HCC [194]. It is a modality that delivers high doses of conformal radiation to the tumors in 1–5 fractions, enabling a precise steep dose gradient without involving the surrounding structures including the liver [194-196]. It is a promising alternative to RFA or TACE in terms of local control but also demonstrated superior outcomes when combined with TACE plus RT compared to TACE alone [197,198]. Patients with advanced HCC and macroscopic vascular invasion may also benefit from first-line treatment with TACE plus RT vs sorafenib, although superiority of this combination approach to systemic treatment needs to be revisited given the improved efficacy of systemic treatment with the introduction of ICI [199].

Proton beam therapy is also an emerging option that is non-invasive and used in locations that are unsuitable for other local treatments such as RFA or surgery [200]. In a phase 3 clinical trial on recurrent/residual HCC (size <3 cm, number ≤2), patients were randomized to proton beam and RFA. Proton beam therapy was non-inferior to RFA with 2-year local progression-free survival being 92.8% vs. 83.2% [201]. Of note, crossover was permitted and PBT demonstrated greater feasibility than RFA, with a notably lower crossover rate (8.3 versus 26.4%) [201]. Overall, external beam radiation treatment is a viable option for early-to-intermediate stage HCC in achieving local control and is promising in combination therapies with TACE in intermediate-stage, and combination with systemic therapy in advanced HCC, especially with vascular invasion [199,200].

Systemic treatment is indicated for unresectable advanced-stage HCC, intermediate-stage HCC with extensive, infiltrative liver involvement, and those who had disease progression despite locoregional therapy or are poor candidates for locoregional treatment [86]. Tyrosine kinase inhibitors (TKI) and ICI with or without anti-vascular endothelial growth factor (VEGF) inhibitors have become the mainstay of systemic treatment for HCC (Table 3) [163,202-213]. When deciding on the appropriate systemic therapy, several factors need to be considered. For example, history of autoimmune disease or prior adverse events needs to be considered when starting an ICI, whereas a history of cardiac comorbidities or risk of bleeding/thrombosis should be accounted for before initiation of an anti-angiogenic treatment [214]. Screening for esophageal varices is recommended before initiating therapy with bevacizumab due to the potential adverse effects of gastrointestinal bleeding associated with the treatment [215].

Atezolizumab (anti-PDL1) plus bevacizumab, an antiangiogenic agent, has become the standard of care and the first-line regimen for patients with advanced HCC after the IMbrave 150 clinical trial in 2020 [202,215]. This has also been validated by several studies regarding its efficacy and safety profile/tolerance in Child-Pugh class B patients, although the presence of main portal vein tumor thrombus and a higher albumin-bilirubin (ALBI) grade was associated with a poorer prognosis [216,217]. The combination treatment of durvalumab and tremelimumab is also currently approved as the first-line regimen for patients with advanced HCC [214]. A phase 3 trial demonstrated that STRIDE (Single Tremelimumab Regular Interval Durvalumab) significantly improved overall survival (HR 0.78, 95% CI 0.65–0.93) compared to sorafenib, and durvalumab monotherapy was non-inferior to sorafenib for patients with unresectable HCC [203]. The main factor to consider in selecting first-line combination regimens between the two is whether the patient can tolerate an anti-VEGF inhibitor [218]. While nivolumab plus ipilimumab was tested before as second-line (CheckMate-040), a recent phase 3 study (CheckMate-9DW) on first-line nivolumab plus ipilimumab compared to sorafenib or lenvatinib reported improved median overall survival (23.7 versus 20.6 months, HR 0.79, 95% CI 0.65–0.96) suggesting that this combination ICI therapy will likely be approved as a first-line regimen for unresectable HCC soon [204,205].

Sorafenib is the first TKI that showed improved survival and TTP in untreated advanced HCC with Child-Pugh class A [206]. It is also the treatment of choice for advanced HCC recurrence after LT, although drug-to-drug interactions with immunosuppressants should be closely monitored [219]. Lenvatinib, another TKI, has subsequently demonstrated non-inferiority (vs. sorafenib) in overall survival in the REFLECT trial [207].

Regorafenib was shown to improve overall survival in BCLC stage B or C HCC patients not eligible for local treatment who previously progressed on sorafenib in the RESOURCE study [208] and cabozantinib demonstrated improved survival in HCC not amenable to curative treatment with up to 2 lines of prior treatment including sorafenib [209]. Ramucirumab, a monoclonal antibody against VEGF2, was studied in BCLC stage B or C HCC patients previously treated with sorafenib and with an AFP above 400 ng/mL and showed improved median overall survival compared to placebo [210]. Pembrolizumab showed efficacy in a phase 3 trial in advanced HCC, previously treated with sorafenib compared to placebo [211]. These treatments were U.S. FDA-approved as subsequent line treatments for advanced HCC.

Combining systemic treatment with other modalities such as ablation, TACE, TARE or SBRT is expected to improve outcomes and is under active investigation in intermediate and advanced-stage HCC. For example, local ablation in combination with tremelimumab showed additional increased immunogenicity effects, leading to increased intratumoral CD8+ cells and reduction in HCV viral load [220]. The EMERALD-1 is a phase 3 trial that demonstrated improvement in PFS with the combination of ICI plus VEGF inhibitor and TACE compared to only TACE in patients with embolization-eligible unresectable HCC, Child-Pugh A to B7 liver function, ECOG 0–1, and no evidence of extrahepatic disease, though there needs to be caution as further studies need to explore overall survival and the safety profile [221]. There was also a recent study comparing the addition of a locoregional therapy to systemic therapy in advanced HCC. TACE-ICI-VEGF treatment, versus ICI-VEGF, exhibited improved median OS (22.6 months versus 15.9 months) and higher ORR (47.3% versus 29.7%) [222].

An important caveat is that these phase 3 clinical trials were performed in select populations with preserved liver function (e.g., Child-Pugh class A liver disease) and good performance (ECOG PS 0 or 1) status, limiting generalizability of the efficacy and safety of treatment in patients with more comorbid conditions [223]. ICI treatment should be considered and further studied in patients with advanced liver dysfunction given its tolerability; a meta-analysis found that Child-Pugh class B HCC patients had a comparable safety profile from ICI treatment and response rate compared to Child-Pugh class A patients but poorer prognosis in overall survival due to higher competing risk of death from liver failure [224]. Furthermore, although various combination therapies are being studied, the identification of biomarkers to predict the response to systemic therapies such as ICI or TKI is limited with few relevant studies to date [225]. Lastly, the financial burden of systemic treatment is substantial and should be better understood and taken into consideration in selecting a treatment strategy for HCC [226].