Anti‐viral therapy is associated with improved survival but is underutilised in patients with hepatitis B virus‐related hepatocellular carcinoma: real‐world east and west experience

Hepatitis B virus (HBV) is the leading cause of hepatocellular carcinoma (HCC) worldwide. It remains incompletely understood in the real world how anti‐viral therapy affects survival after HCC diagnosis.


| INTRODUCTION
Hepatocellular carcinoma (HCC) is the third leading cause of cancer death worldwide with nearly 600 000 deaths in 2008, 1 and hepatitis B virus (HBV) infection is the most common cause of HCC. 2 Up to 20%-40% of patients with HBV infection may develop HCC in the absence of clinically apparent cirrhosis. 3,4 The advent of highly potent anti-viral therapy has offered the possibility of greatly decreasing the incidence of liver-related complications, including cirrhosis and HCC, in patients with chronic hepatitis B (CHB). 5,6 While anti-viral medications are known to reduce the risk of HCC in patients with CHB, 7 it is less well understood how they influence the overall survival of patients with established HCC, with most data only focusing on patients undergoing curative therapy for HCC. 8 A 2013 randomised controlled trial comparing nucleot(s)ide analogues to placebo in patients with HBV-related HCC treated with partial hepatectomy found that anti-viral therapy decreased the risk of HCC recurrence, HCC-related mortality and overall mortality. 9 These findings are supported by an earlier meta-analysis of 9 cohorts and a national database study showing that HCC patients receiving curative therapy for HCC and treated with anti-viral agents had decreased overall mortality and recurrence rate. 10,11 Likewise, anti-viral therapy after radiofrequency ablation is associated with decreased HCC recurrence. 12 The data on anti-viral therapy in patients treated with palliative therapy, including transarterial chemoembolisation (TACE) and sorafenib are more limited, but in the case of TACE, a randomised controlled trial demonstrated that anti-viral therapy increases survival. [13][14][15] Currently, the presence of HCC is not considered an indication for anti-viral medications in international guidelines for the management of CHB. [16][17][18] In addition, lifelong anti-viral medications may not be reimbursed by third-party payers in certain areas such as Taiwan if patients have HCC but no cirrhosis. 19 Thus, the question of how much viral suppression can reduce mortality among HBV-related HCC patients can have important policy implications. Related to this issue is the "cascade of care" for patients with CHB, which describes the reasons for which rates of treatment for CHB are suboptimal. 20 Many patients with CHB have not even been diagnosed, and among those who are diagnosed many have not established care with the medical system. 21 Even among those with access to appropriate medical care, treatment rates of CHB are suboptimal for numerous reasons including patient loss to follow-up, financial difficulties and misconceptions about indications for anti-viral therapy. 22

| Study design and patient population
We performed an international multicentre cohort study of HBVrelated HCC at 5 medical centres. The inclusion criteria were CHB (defined as positive serum hepatitis B surface antigen, detectable HBV DNA, or on anti-viral therapy for a history of CHB) and a new diagnosis of HCC (by pathology or imaging based on 2010 American Association for the Study of Liver Diseases criteria). 23  Selection bias is decreased by the use of consecutive patients.
Study size was not pre-determined and was based on the number of patients diagnosed with HCC between certain time periods.

| Definition of cirrhosis and anti-viral treatment
Laboratory data, imaging findings, and HCC and cancer treatment modalities were obtained from patients' medical records. Patients were designated as having cirrhosis if they were deemed to have cirrhosis based on hepatology notes, or if there was pathological evidence of fibrosis stage 4, clinical evidence of portal hypertension (otherwise-unexplained splenomegaly or platelet count <120 000/lL, ascites, or gastroesophageal varices on imaging), prior hepatic decompensation (hepatic encephalopathy, ascites, variceal gastrointestinal bleeding), or laboratory evidence of decreased synthetic function (total bilirubin >2.0 mg/dL or international normalised ratio >1.2 without alternative explanation). Anti-viral therapy status was determined by chart review and pharmacy records. Criteria for antiviral therapy were based on American Association for the Study of Liver Disease and Asia-Pacific Association for the Study of the Liver guidelines. 16,17 2.3 | Tumour staging and survival outcomes Tumour characteristics were determined by triphasic computed tomography or magnetic resonance imaging. Patients were followed from the date of diagnosis with HCC and either death or last follow-up date in the medical record. For the Stanford and Mayo cohorts, patients not known to be deceased and whose last visit to the medical centre was before January 1 2015, we also performed a National Death Index registry search from 1979-2014. The National Death Index registry is a centralised database of death record information on file in state vital statistics offices with over 90% completion for most states and 99% for the state of California where the Stanford cohort is located. 25 For the Kaohsiung cohort, telephone interview with families was also conducted to obtain additional follow-up data.

| Anti-viral therapy indications
Four different standards of anti-viral therapy were used: American Association for the Study of Liver Disease guidelines, 16 Asia-Pacific Association for the Study of the Liver, 26 Ministry of Health and Welfare for the Republic of Korea 27 and National Health Insurance Administration for Taiwan. 28 Local guidelines were defined as American Association for the Study of Liver Disease guidelines for United States centres, Ministry of Health and Welfare reimbursement criteria for Korean centres and National Health Insurance Administration for the Taiwan centre (Table S1).

| Statistical analysis
Descriptive statistics were reported as proportion (%) for categorical variables, and mean AE standard deviation (SD) or median (and range) for continuous variables. Normally distributed continuous variables were compared by Student's t tests. Nonparametric statistics were applied when continuous variables were not normally distributed.
Chi-squared tests were used to compare categorical variables. In this study, the primary outcome was overall survival of HCC patients.
Person-years of follow-up were calculated for each patient as the time from dates of HCC diagnosis to the date of death or to the last date when patients were last known to be alive. Mortality rates by various disease status were calculated and expressed per 100 person-years. Kaplan-Meier methods were utilised to depict the overall survival of patients with or without anti-viral therapy; patients lost to follow-up were censored. Statistical differences in overall survival by various subgroups were compared and examined by log-rank tests. Cox proportional hazards models were used to estimate the hazard ratios (HRs) and 95% (CIs) relating anti-viral therapy and other risk factors with overall survival. Lead time analysis was performed as previously described. 29,30 Statistical significance was defined as a two-tailed P value <0.05. Missing data were excluded from analysis. All statistical analyses were performed using STATA 11.0 (Stata Corporation, College Station, TX, USA).

| Baseline clinical/tumour characteristics
The overall cohort consisted of 2518 patients, of whom 73% had cirrhosis, 81% were male, and 98% were Asian. Baseline clinical and laboratory characteristics of the patients in this cohort by anti-viral therapy status are shown in Table 1. Among the patients with cirrhosis, 30% had had a history of hepatic decompensation with variceal bleeding, hepatic encephalopathy, or ascites, and 38% had Child-Pugh-Turcotte (CPT) class B or C disease.
In total, 49% of patients had ever received anti-viral therapy at any time ( Figure 1). However, only 17% were on treatment at time of HCC diagnosis ( Figure 1). Proportion of patients receiving anti-viral therapy before their diagnosis of HCC did not differ between those with or without cirrhosis (17.9% vs 15.3%; P = 0.12; Figure S1A).   Of the patients who were not on therapy at time of HCC diagnosis, 37% were later started on anti-viral therapy ( Figure 1).
Patients with cirrhosis were more likely to receive anti-viral therapy after HCC diagnosis and more likely to receive anti-viral therapy at any time than those without cirrhosis (49.9% vs 43.1%; P < 0.001; Figure S1A). Of note, anti-viral therapy is reimbursed in Korea for patients with detectable HBV DNA and HCC, 27 but HBV DNA was detected in 68% of patients who never received anti-viral therapy after HCC diagnosis.
Though rates of anti-viral therapy were well below 50% at time of HCC diagnosis at all centres, it was higher at the US centres than the Taiwan or Korea centres (40% vs 12% vs 10%; P < 0.001 for three-way comparison; Figure S1B). After HCC diagnosis, 68% of the US and 61% of the Korea centre patients were on anti-viral therapy, compared to only 31% in the Taiwan centre ( Figure S1B; P < 0.001 for three-way comparison  When comparing cancer-directed treatment modalities based on anti-viral therapy status (Table S3), patients receiving anti-viral therapy were more likely to receive any treatment and most individual treatments including resection, LT, and liver-directed therapy. As with cirrhosis, the difference in liver-directed therapy was driven primarily by palliative transarterial chemoembolisation and external radiation therapy. person-years in the study population. Overall, patients with cirrhosis had increased mortality rate compared to those without cirrhosis (P < 0.001). As expected, mortality was higher with higher CPT class, BCLC stage and use of curative therapies (P < 0.05 for all). Figure 4 shows overall survival based on anti-viral therapy status.

| Mortality rates and overall survival
Survival was significantly higher in patients receiving anti-viral therapy ( Figure 4A), and notably both among those with cirrhosis (42% vs 25%; P < 0.001; Figure 4B) and those without cirrhosis (58% vs 36%; P < 0.001; Figure 4C). Subgroup analysis of 5-year survival based on anti-viral treatment status is shown in per 100 person-years, P = 0.0008, Figure S2A and Table S5). This We also compared patients receiving anti-viral therapy only after HCC diagnosis with those not receiving anti-viral therapy at all.
Here, there was no significant difference in maximum tumour size, proportion of multifocal tumours, or vascular invasion (Table S6) 27 and National Health Insurance Administration for Taiwan (green) 28 (P < 0.001) but were no different in patients receiving anti-viral therapy after HCC diagnosis (P = 0.58). Table 4 shows predictors of mortality among HBV-related HCC patients. On unadjusted analysis, prognostic factors associated with increased mortality included younger age, male sex, cirrhosis, decompensated cirrhosis (CPT stage B and C), higher MELD score, more advanced BCLC stage and the Taiwan centre (P < 0.05 for all). Conversely, factors associated with decreased morality included treatment with surgery (resection or liver transplant) or with either sorafenib or liver-directed therapy, anti-viral therapy at any time, duration of anti-viral therapy both before and after HCC diagnosis and anti-viral therapy with newer agents (entecavir or tenofovir) (P < 0.05 for all). We included relevant predictors associated with mortality in the multiple regression models to estimate the adjusted HR and 95% CI of each predictor: age, sex, cirrhosis status, MELD, treatment type, BCLC stage and country. We also included anti-viral treatment status, stratified as no therapy, therapy only after HCC diagnosis and therapy before HCC diagnosis. In this model, anti-viral therapy either before or only after HCC diagnosis was independently associated with decreased mortality (adjusted HR 0.62 and 0.79, respectively; P < 0.001; Table 4). In this model, the Taiwan centre was no longer independently associated with increased mortality. On subanalysis of the patients for whom screening information was available, both screening and anti-viral therapy were associated with increased survival in a multivariate analysis model (Table S7).

| DISCUSSION
In this study, we characterised a cohort of patients with HBV-related HCC stratified by anti-viral therapy utilisation and cirrhosis status.
We found that the use of anti-viral medications at any time in HBVrelated HCC patients was associated with a 20%-40% reduction in overall mortality of these patients, a sizable effect especially when compared to the modest survival benefits seen with many standard therapies for HCC such as palliative liver-directed therapy and sorafenib. 31,32 The benefit of anti-viral therapy holds across a range of different cancer stages including BCLC stage C/D and treatment types and even in patients receiving supportive care only. In addition, while there were significant differences in the rates of anti-viral utilisations and overall mortality among US vs Taiwan vs Korea centres, there was no difference in overall survival based on country of study sites in this multicentre international study after adjustment was made for anti-viral therapy use.
There is extensive evidence that anti-viral therapy in patients with CHB decreases risk of liver-related complications including liver decompensation and HCC development. 6 without cirrhosis to those with cirrhosis and advanced liver disease, from those with early to advanced cancer stage, and from those receiving curative therapy to those receiving only palliative therapy or even supportive care only. Anti-viral therapy could increase survival following HCC diagnosis in either the long-or short term through different mechanisms. In the long run, anti-viral therapy could decrease HCC recurrence and/or HCC progression. Previous studies showed anti-viral therapy was associated with decreased HCC recurrence and increased survival among patients with HBVrelated HCC undergoing surgery with curative intent. 9,10 The longterm beneficial effects would be more significant in patients with early-stage HCC and compensated liver disease. In the short term, anti-viral therapy may counter the destabilising effect by HCC on liver function, which may be more important in patients with more advanced HCC and/or more impaired liver function. This study found that the increase in survival with anti-viral therapy was seen in a range of severity of liver disease and HCC stage, and, if anything, may have been more pronounced in patients with more advanced disease. In addition, choice of anti-viral therapy used may be important: use of newer anti-viral agents, ie, tenofovir or entecavir, was associated with improved survival compared to use of lamivudine or adefovir (Table 4).
Disappointingly, in this multinational cohort, there was a strikingly low rate of anti-viral therapy. In particular, there was a much lower rate of anti-viral use in the Asian sites compared to the US    Figure S1).
It should be noted that anti-viral treatment rate before HCC diagnosis was suboptimal in all of our study centres. This included US patients from 2 major university referral centres, which suggests that financial coverage is unlikely to be the only major barrier to In summary, we report here the largest cohort of diverse HBVrelated HCC patients from several medical centres from 3 countries.
We found that anti-viral therapy at any time was significantly associated with 20%-40% lower mortality and this beneficial effect was independent of age, cirrhosis status, severity of cirrhosis, cancer stage and cancer treatment. Unfortunately, this study also found an alarmingly low rate of anti-viral therapy utilisation in centres in the US as well as Asia with the majority of patients not receiving any anti-viral therapy before their HCC diagnosis, even though a large proportion of them met both Asian and US treatment guideline criteria for therapy. Our data support more widespread use of anti-viral therapy in patients with HBV-related HCC, while highlighting the needs for improved linkage-to-care and earlier treatment with antiviral therapy in high-risk patients. In addition, the discrepancy between guidelines of management of patients with compensated cirrhosis has significant real-world implications on which patients are eligible for anti-viral therapy. Additional prospective studies are needed to understand and overcome the barriers to appropriate management of patients with HBV infection.