The majority of all cases of hepatocellular carcinoma (HCC) worldwide are found in the Asia-Pacific region, making this a leading public health problem. This high incidence of HCC in the Asia-Pacific region is largely due to the high prevalence of chronic viral hepatitis.^1,2

In fact, approximately 80% of HCC cases worldwide can be attributed to chronic infection with either hepatitis B virus (HBV) or hepatitis C virus (HCV). These infections, particularly when left untreated, lead to cirrhosis, which is present in 80% to 90% of patients with HCC. Furthermore, because most HCC cases occur in countries with fewer healthcare resources (and, therefore, fewer surveillance options), HCC is often discovered at an advanced stage and accompanied by liver dysfunction, making this a highly lethal cancer.

Approximately 5% of the global population is infected with HBV, and 75% of patients with HBV are Asian. Within HBV endemic areas of the Asia-Pacific region, HBV infection is frequently acquired through mother-to-child transmission, which contributes to earlier onset of HCC than in nonendemic areas.^3,4

As previously mentioned, Asian patients with HBV have often been infected since birth. Yang and colleagues⁵ found that the proportion of US patients with a diagnosis of HCC at younger than 40 years of age was highest in Asian-Pacific Islanders vs other groups. The table in this slide showing demographic factors associated with early-onset HCC demonstrates that, compared with White patients, Asian patients have an adjusted odds ratio (OR) of 1.92 (95% CI: 1.55-2.37; P <.01)) for the development of HCC before 40 years of age.

In the landmark real-world REVEAL study of 3653 untreated hepatitis B surface antigen (HBsAg)–positive adults in Taiwan, 2 key risk factors for HCC were cirrhosis and a high HBV DNA level. The threshold for significant increase in HCC risk occurs at an HBV DNA of 10,000 copies/mL, or approximately 2000 IU/mL. Above that threshold, we can see that HCC risk increases dramatically.⁶

Family history of HCC is an important predictor of HCC risk, as well. The population-based cohort study of 22,472 individuals in Taiwan by Loomba and colleagues⁷ showed that the cumulative incidence of HCC is highest in patients with a positive family history who are also HBsAg seropositive. Tong and colleagues⁸ analyzed data from HBsAg-positive patients (N = 413), 173 with HCC and 240 without HCC, for risk factors associated with HCC occurrence. The results underscore the association between age of HCC onset and family history of HCC. As shown in this age distribution chart, HCC onset is decades earlier in sons and daughters who have parents with HCC compared with the age of onset in the first generation.

Given that both HBV and HCV infections increase the risk of HCC, what is the HCC risk in patients who are coinfected with HBV and HCV? The results are mixed. This meta-analysis by Cho and colleagues⁹ showed that in case-control trials, coinfection was associated with a supra-additive risk of HCC (OR: 62.2; 95% CI: 38-103; P_{heterogeneity} = .002) vs HCV (OR: 27.0; 95% CI: 19-39) or HBV (OR: 25.4; 95% CI: 18-36) monoinfection. When stratified by region, nonendemic regions observed a subadditive effect of coinfection, whereas HBV endemic and HCV endemic regions showed an additive and supra-additive effect of coinfection, respectively. A subadditive is possible because infection by one virus can inhibit infection by the other virus.

Other risk cofactors of potential importance in the development of HCC include alcohol misuse and metabolic problems like diabetes mellitus, metabolic syndrome, and fatty liver. Certain HBV genotypes may increase HCC risk as well. Hosaka and colleagues¹⁰ demonstrated that of the 2 most common genotypes in Asia, genotype C is associated with a higher risk of HCC than genotype B. In addition, suboptimal antiviral therapy and reduced viral suppression is associated with a higher risk of HCC compared with more effective therapy. In this Kaplan-Meier curve of cumulative HCC incidence by type of nucleos(t)ide analogue therapy, risk of developing HCC was lower in those receiving entecavir (ETV) vs the older agent lamivudine, the latter being associated with greater drug resistance.¹⁰

Of course, the best way to reduce risk of HBV-related HCC is to prevent HBV infection. Therefore, the Asian Pacific Association for the Study of the Liver (APASL) recommendation for primary prophylaxis is universal HBV vaccination of infants in all countries, particularly in HBV endemic areas. Most Asia-Pacific countries have approved this concept, and Taiwan, for example, has practiced universal HBV vaccination for close to 40 years, with concomitant decreases in HCC incidence.

Secondary prophylaxis remains critically important at this time, despite HBV vaccination, because we still have a large population of adults with chronic HBV infection. In these patients, we need to reduce the risk of HCC by providing very good, potent antiviral therapy as a long-term treatment to suppress their virus.²

There are 2 nucleos(t)ide analogues currently recommended for first-line treatment of HBV, tenofovir and ETV. Tenofovir is formulated as tenofovir disoproxil fumarate (TDF) or tenofovir alafenamide, and I will discuss these 2 formulations together as they contain the same active component. These are the drugs commonly used today, but the question remains whether one is better than the other. There have been multiple studies investigating this question in the recent 2-3 years.¹¹

In 2020, we conducted a retrospective analysis involving more than 29,000 patients in Hong Kong with HBV who received treatment with TDF or ETV from January 2008 to June 2018. Patients with other viral coinfections, patients previously treated with nucleos(t)ide analogues or peginterferon, and patients with HCC, other cancers, or liver transplant at baseline or within 6 months thereafter were excluded. The median follow-up period of this study was 3.6 years. After propensity score weighting to make the 2 cohorts with chronic HBV infection comparable, we found that TDF treatment reduced the risk of HCC compared with ETV treatment (subdistribution HR: 0.36; 95% CI: 0.16-0.80; P = .013). This suggests that treatment with TDF can reduce the HCC risk 60% further than treatment with ETV.¹¹

This question was studied not just in Hong Kong, but in many countries around the world—mostly in Asia, but also some from the West—and these studies show quite similar findings. In this meta-analysis by Choi and colleagues,¹² treatment with TDF vs ETV was associated with a lower risk of HCC (HR: 0.80; 95% CI: 0.69-0.93; P <.003). Overall, when many propensity, score-matched cohorts are analyzed collectively, we see an approximately 20% reduction in the risk of developing HCC when using TDF vs ETV for treatment of HBV.

To summarize, HCC risk assessment in patients who are HBsAg positive, the risk factor most strongly associated with HCC is advanced fibrosis/cirrhosis. Therefore, all patients should have the severity of fibrosis assessed. Currently, we typically use noninvasive assessments, such as transient elastography or serum biomarkers, for assessing fibrosis. Other groups of patients with an increased risk of HCC include those with a family history of first-degree relatives with HCC; Asian or Black men older than 40 years of age or Asian women older than 50 years of age; and persons with HBV coinfected with other viruses like HCV or hepatitis delta virus. (We advise screening for other viruses in patients with HBV.) Emerging HCC risk factors include alcohol misuse, coexistence of nonalcoholic fatty liver disease, and certain HBV genotypes (eg, genotype C in Asia). Finally, potent antiviral therapy with sustained HBV DNA suppression reduces the risk of HCC, so we prefer to use potent first-line nucleos(t)ide analogues to treat HBV.

From previous studies, we know that it is cost-effective to conduct surveillance for HCC when the incidence is ≥1.5% per year in patients with cirrhosis, regardless of etiology. In patients with chronic HBV infection with cirrhosis, the annual HCC risk is 3% to 5% per year and, therefore, warrants surveillance. In noncirrhotic patients, the cost-effective threshold for surveillance has been suggested to be at an annual HCC incidence >0.2%. As shown here, Asian women older than 50 years of age and Asian men older than 40 years of age have annual HCC incidences >0.2%. The APASL recommends HCC surveillance in all patients with cirrhosis and in patients with HBV without cirrhosis if they are Asian women older than 50 years of age, Asian men older than 40 years of age, Africans older than 20 years of age, or persons with a positive family history of HCC.²

Similar to other major regions of the world, we in the Asia-Pacific region continue to use ultrasonography as our first-line screening tool, in part because it is noninvasive and widely available. Ultrasonography is available in most clinics or hospitals in the region and the cost is relatively low. We also understand, however, the disadvantages of ultrasonography, including that the results are very operator dependent and the sensitivity in patients with obesity and with certain disease states is low. For example, if the lymph nodes are very fatty or the patient has thick subcutaneous fat, it is quite difficult to adequately visualize the whole liver.

In the Asia-Pacific region, we are now increasingly using CT, as it is becoming more widely available and has moderate to high resolution and high sensitivity. But because of the relatively high dose of radiation with CT, we prefer not to use this repeatedly, especially in young patients. The costs are also higher compared with ultrasonography.

MRI is also used for surveillance in some parts of the world. Because MRI is expensive and requires a large space to install, the availability is limited in lower-resource countries. The advantages of MRI include its high resolution and high sensitivity. Some healthcare providers will use noncontrast MRIs first; if using gadolinium as a contrast agent, there is a risk of gadolinium accumulation with repeated use.¹³

The sensitivity of these surveillance imaging modalities varies by nodule size: the sensitivity of each modality decreases with nodule size. As demonstrated in these data from Yu and colleagues,¹⁴ for nodules <2 cm, the sensitivity of ultrasonography is approximately 21%. For CT and MRI, the sensitivity increases, ranging from 40% to 47% for nodules <2 cm. For large nodules 2- to 4-cm in size, CT and MRI were shown to have very satisfactory sensitivities of 80% and 86%, respectively. And for nodules >4 cm, all modalities work quite well.

Apart from imaging, we incorporate serum biomarkers in surveillance recommendations because of high utility and convenience. We have been using α-fetoprotein (AFP) for decades because it is commonly available, it can be ordered with a routine liver panel, and the costs are relatively low. We also understand its disadvantages: AFP may be elevated in the presence of liver injury, such as in those with liver inflammation due to active hepatitis, which means that AFP levels may increase even in the absence of cancer. This lack of specificity is one major limitation. Similarly, approximately one third of HCCs do not secrete AFP, meaning that the sensitivity may also be low. On the other hand, a subfraction of AFP, known as AFP-L3, has been shown to be elevated in and highly specific for HCC. APASL guidelines recommend surveillance with AFP and ultrasonography every 6 months; healthcare professionals in some regions are also using AFP-L3.¹⁵

Another serum biomarker for HCC surveillance is des-γ-carboxy-prothrombin (DCP), also known as proteins induced by vitamin K absence-II (PIVKA-II). Because this biomarker is involved in the prothrombin pathway, it is quite specific for HCC, particularly in those with higher neovascular invasion. DCP/PIVKA-II has been widely used in some Asian countries, such as Japan and Korea. In fact, Japanese guidelines recommend DCP in addition to AFP for routine HCC surveillance.^15,16

A newer approach to biomarkers is to combine them with some other important risk factors to develop an HCC risk score. The GALAD model was developed using data from 670 patients in the United Kingdom: 331 patients with HCC and 339 patients with chronic liver disease without HCC. In the GALAD acronym, G stands for gender (sex), A stands for age, L stands for AFP-L3, A stands for AFP, and D stands for DCP. The theory is that in combining patient risk factors and serum biomarkers in this formula, one can generate a probability of developing HCC.¹⁷

The GALAD score has been validated in Asian and European cohorts (N = 6834) and found to satisfactorily discriminate between patients with chronic liver disease without HCC (n = 4404) and with HCC (n = 2430). In fact, the GALAD model outperformed each of the 3 component biomarkers alone (AFP, DCP, or AFP-L3) for predicting HCC, with an AUROC of >0.90 for all cohorts. This score has been advocated as a type of surveillance tool in some parts of the world.¹⁸

To summarize the APASL guidance on HCC screening and surveillance, we would advocate the use of ultrasonography plus AFP every 6 months in all patients with cirrhosis, regardless of etiology. For those patients who are HBsAg positive without cirrhosis, we would also consider other risk factors as previously mentioned, and these patients would also be offered ultrasonography plus AFP every 6 months.

In fact, we see consensus about the use of ultrasonography for surveillance around the world. Not only the APASL guidelines, but also the American Association for the Study of the Liver (AASLD) and European Association for the Study of the Liver (EASL) guidelines recommend routine surveillance with ultrasonography because of its wider availability—we can offer this scan to essentially all of our patients. There are still some discrepancies among the major liver societies about the use of AFP in conjunction with ultrasonography: APASL recommends AFP, EASL does not, and AASLD considers AFP optional. The major societies agree that AFP should not be used alone (except in cases where ultrasonography is not available). Ideally, AFP is always used in conjunction with an imaging modality.^2,19

We have discussed that AFP is not highly specific in patients with active HBV infection with hepatic necroinflammation. But if patients are receiving potent antiviral treatment, necroinflammation will be under control. In a prospective analysis of more than 1500 patients with chronic HBV infection receiving ETV treatment, our group showed that AFP levels decreased in the first 12 months after treatment initiation and then stabilized. That is why I still consider AFP a very useful tool in patients who have been receiving antiviral treatment. We showed that when using an AFP cutoff of 20 µg/L, the specificity of AFP for HCC is very high at approximately 99%. But we need to bear in mind that the sensitivity will also be reduced at this cutoff. Using a cutoff of 6 µg/L, we showed that specificity decreased to 80% but sensitivity increased to 81%, which is quite satisfactory.²⁰

It is important to remember that AFP is a screening tool, and it is not recommended as a confirmatory test in small HCC because approximately one third of such cases would be AFP negative. The APASL guidelines advocate using a higher cutoff when using AFP for routine surveillance in conjunction with ultrasonography, specifically 200 ng/mL (200 µg/L). Although this cutoff avoids some of the false-positive AFP elevations associated with necroinflammation and hepatocyte regeneration, I think setting the cutoff this high may miss a few HCC. In my practice, we may use 20 µg/L or even 6 µg/L as the cutoff in patients who have been receiving antiviral treatment with good viral suppression and stable liver enzymes, as demonstrated in our previous study.^2,20

The algorithm in this slide reflects APASL guidance on the subsequent workup if routine surveillance detects nodules. If nodule(s) are detected by ultrasonography, we recommend a dynamic imaging test to confirm the pattern—is it hypervascular, and if so, with or without washout? If there is washout, it is quite diagnostic for HCC. But if there is no washout, we rescan to specifically look at the hepatobiliary phase. In some borderline cases, we may advise the use of biopsy. If there is a nonhypervascular tumor, the workup may include contrast-enhanced ultrasonography. Ultimately, if there is any uncertainty, we may either repeat a dynamic imaging scan in 3-6 months or consider biopsy. In any case, if we confirm that there is no suspicious nodule, we can return to routine surveillance by ultrasonography every 6 months.²

The previous algorithm can be too complicated for some practice settings, so an alternative approach is to just use dynamic CT or MRI. The main question here is regarding the hypervascular pattern: Is it with or without washout? If there is no washout (or if it is nonhypervascularized), then the scan is repeated (for nodules <1 cm) or the nodule is biopsied (if ≥1 cm). This approach has been adopted in many parts of the world, including the Asia-Pacific regions.²

There are many factors to consider in HCC treatment, making a multidisciplinary approach extremely helpful. The first 2 factors to consider are the presence of extrahepatic metastases and liver function as assessed by the Child-Pugh stage. Next, do the liver tumor size(s) and location(s) indicate that it is resectable? The degree of macrovascular invasion, tumor number, and tumor size affect treatment options. The best case scenario would be no metastases, well-compensated liver function, and resectable tumor(s). Resection is one of the best options to prevent recurrence, but in patients with advanced age or other high surgical risks, local ablation for nodules <3 cm or with nodule numbers fewer than 3 is also a very good approach. Because now we have very good tools like radiofrequency ablation, relatively small nodules can be nonsurgically treated with reasonably good tumor control.

In cases with >3 nodules, transarterial chemoembolization (TACE) or systemic therapy has been adopted quite widely. And in the presence of decompensated liver disease (if the tumor is still within the transplant criteria) liver transplantation would treat both the cancer and the decompensated cirrhosis simultaneously. But of course, this depends on the availability of the liver graft. Systemic therapy, particularly immunotherapy, has been used in the last few years with very good disease control, even in patients with macrovascular invasion or extrahepatic metastasis. Other systemic therapies have also been proven to prolong survival.

Surgeons, oncologists, and radiologists are important members of the multidisciplinary team required to manage HCC treatment, but the hepatologist is the referral source and is the central healthcare professional directing patient care throughout the treatment.²

There are some Asia-Pacific considerations regarding the multidisciplinary management of HCC treatment. First, we need to recognize the inherent difficulty in defining strict criteria for resectability that are generalizable to various medical environments. Liver resection is a first-line curative treatment for HCC among patients with Child-Pugh class A, but it is up to the local multidisciplinary team to confirm resectability in terms of tumor burden and liver functional reserve.

Second, Asia-Pacific guidance recognizes radiofrequency ablation for patients with resectable tumors and TACE for patients with macrovascular invasions without extrahepatic metastases as treatment options widely used in this region, even though we know that there are insufficient data to recommend them as standard approaches. Here again, the multidisciplinary team of surgeons, oncologists, and radiologists provides input on treatment choices based on locally available resources.

Third, if TACE is ineffective after a few sessions (with evidence of disease progression), APASL recommends conversion from TACE to systemic therapy, based on evidence from retrospective studies from Europe and Japan.²

In conclusion, how should we manage patients with HBV at risk for HCC? Hepatologists need to see patients with HBV regularly to monitor their viral disease and assess response to treatment. Currently, we are lucky to have very good antiviral agents, so this part is usually quite easily managed. Effective surveillance will identify HCC at an early stage, so we can offer curative treatment. Subspecialty care and engagement with the multidisciplinary team or tumor board can help determine the best treatment options, but hepatologists are the referring specialist and must guide the care of patients with HBV and HCC to ensure all options are considered.