Entecavir treatment reduces hepatic events and deaths in chronic hepatitis B patients With liver cirrhosis

Authors

  • Grace Lai-Hung Wong,

    1. Institute of Digestive Disease, the Chinese University of Hong Kong, Hong Kong SAR, China
    2. Department of Medicine and Therapeutics, the Chinese University of Hong Kong, Hong Kong SAR, China
    Search for more papers by this author
  • Henry Lik-Yuen Chan,

    1. Institute of Digestive Disease, the Chinese University of Hong Kong, Hong Kong SAR, China
    2. Department of Medicine and Therapeutics, the Chinese University of Hong Kong, Hong Kong SAR, China
    Search for more papers by this author
  • Christy Wing-Hin Mak,

    1. Department of Medicine and Therapeutics, the Chinese University of Hong Kong, Hong Kong SAR, China
    Search for more papers by this author
  • Stanley King-Yeung Lee,

    1. Department of Medicine and Therapeutics, the Chinese University of Hong Kong, Hong Kong SAR, China
    Search for more papers by this author
  • Zoe Man-Yi Ip,

    1. Department of Medicine and Therapeutics, the Chinese University of Hong Kong, Hong Kong SAR, China
    Search for more papers by this author
  • Andrew Ting-Ho Lam,

    1. Department of Medicine and Therapeutics, the Chinese University of Hong Kong, Hong Kong SAR, China
    Search for more papers by this author
  • Henry Wing-Hang Iu,

    1. Department of Medicine and Therapeutics, the Chinese University of Hong Kong, Hong Kong SAR, China
    Search for more papers by this author
  • Joyce May-Sum Leung,

    1. Department of Medicine and Therapeutics, the Chinese University of Hong Kong, Hong Kong SAR, China
    Search for more papers by this author
  • Jennifer Wing-Yan Lai,

    1. Department of Medicine and Therapeutics, the Chinese University of Hong Kong, Hong Kong SAR, China
    Search for more papers by this author
  • Angeline Oi-Shan Lo,

    1. Institute of Digestive Disease, the Chinese University of Hong Kong, Hong Kong SAR, China
    2. Department of Medicine and Therapeutics, the Chinese University of Hong Kong, Hong Kong SAR, China
    Search for more papers by this author
  • Hoi-Yun Chan,

    1. Institute of Digestive Disease, the Chinese University of Hong Kong, Hong Kong SAR, China
    2. Department of Medicine and Therapeutics, the Chinese University of Hong Kong, Hong Kong SAR, China
    Search for more papers by this author
  • Vincent Wai-Sun Wong

    Corresponding author
    1. Institute of Digestive Disease, the Chinese University of Hong Kong, Hong Kong SAR, China
    2. Department of Medicine and Therapeutics, the Chinese University of Hong Kong, Hong Kong SAR, China
    • Department of Medicine and Therapeutics, 9/F Prince of Wales Hospital, 30-32 Ngan Shing Street, Shatin, Hong Kong . E-mail: wongv@cuhk.edu.hk fax: (852)-2637-3852

    Search for more papers by this author

  • Potential conflict of interest: Grace Lai-Hung Wong has served as an advisory committee member for Otsuka and as a speaker for Echosens. Henry Lik-Yuen Chan is a consultant for Abbott, Bristol-Myer Squibb, Gilead, GlaxoSmithKline, Merck, Novartis, and Roche; is on the speakers' bureau of Echosens, GlaxoSmithKline, and Roche; has received lecture honoraria from Abbott, Bristol-Myer Squibb, Echosens, Gilead, Glaxo-Smith-Kline, Merck, Novartis, and Roche; and has received an unrestricted grant from Roche for hepatitis B research. Vincent Wai-Sun Wong has served as an advisory committee member for Roche, Novartis, Gilead, and Otsuka and has served as a speaker for Bristol-Meyers-Squibb, Roche, Novartis, Abbott Diagnostics, and Echosens.

Abstract

Entecavir is a potent antiviral agent with high genetic barrier to resistance, hence it is currently recommended as first-line antiviral therapy for chronic hepatitis B (CHB). The aim of this study was to investigate the efficacy of entecavir on clinical outcomes and deaths. It was a retrospective-prospective cohort study based on two cohorts of patients. The entecavir cohort included consecutive CHB patients who had received entecavir 0.5 mg/day for at least 12 months. The historical control cohort included untreated patients recruited since 1997 who underwent routine clinical care. The primary outcome was the 5-year cumulative probability of hepatic events, defined as any cirrhotic complications, hepatocellular carcinoma (HCC), and/or liver-related mortality. A total of 1,446 entecavir-treated patients (72% men; age, 51 ± 12 years; follow-up, 36 ± 13 months) and 424 treatment-naïve patients (65% men; age, 41 ± 13 years; follow-up, 114 ± 31 months) were studied. Overall, there was no difference in hepatic events between the entecavir and control cohorts. Among patients with liver cirrhosis (482 entecavir-treated, 69 treatment-naïve), entecavir-treated patients had reduced risks of all clinical outcomes when compared with treatment-naïve patients with cirrhosis after adjusted for model for end-stage liver disease score: hepatic events (hazard ratio [HR], 0.51; 95% confidence interval [CI], 0.34-0.78; P = 0.002), HCC (HR, 0.55; 95% CI, 0.31-0.99; P = 0.049), liver-related mortality (HR, 0.26; 95% CI, 0.13-0.55; P < 0.001), and all-cause mortality (HR, 0.34; 95% CI, 0.18-0.62; P < 0.001). Entecavir-treated patients with cirrhosis who failed to achieve undetectable hepatitis B virus DNA (105/482 [22%]) had comparable risk of hepatic events as the untreated patients. Conclusion: Entecavir therapy reduces the risks of hepatic events, HCC, liver-related and all-cause mortality of CHB patients with liver cirrhosis in 5 years, particularly among those who had maintained viral suppression. (Hepatology 2013;58:1537–1547)

Abbreviations
ALT

alanine aminotransferase

anti-HBe

antibody against hepatitis B e antigen

CHB

chronic hepatitis B

CI

confidence interval

HBeAg

hepatitis B e antigen

HBsAg

hepatitis B surface antigen

HBV

hepatitis B virus

HCC

hepatocellular carcinoma

HR

hazard ratio

INR

international normalized ratio

MELD

model for end-stage liver disease

SBP

spontaneous bacterial peritonitis

Chronic hepatitis B (CHB) is the leading cause of liver cirrhosis and hepatic events, including various cirrhotic complications and hepatocellular carcinoma (HCC) in Asia.[1] A high level of hepatitis B virus (HBV) DNA has been consistently shown to be an independent risk factor for the development of cirrhosis and HCC.[2, 3] Antiviral therapy is effective in suppressing HBV DNA and reducing the risk of hepatic events and HCC.[4, 5] Lamivudine, the first generation oral nucleoside analogue was shown to reduce the incidence of hepatic events in patients with advanced fibrosis or compensated cirrhosis by 50%.[6] However, the clinical benefit may be negated by lamivudine resistance and virologic relapse after cessation of treatment.[6, 7] Therefore, lamivudine is no longer the first-line antiviral therapy for CHB. Instead, nucleos(t)ide analogues with high genetic barrier to resistance—namely, entecavir and tenofovir disoproxil fumarate—are recommended by international guidelines.[8]

Entecavir is a more potent antiviral agent than lamivudine and adefovir.[11] Most entecavir-treated patients achieve maintained viral suppression and low risk of drug resistance in long-term studies.[14, 15] Histological improvement was observed in 66 of the 69 patients (96%) who had received entecavir for 3 to 7 years.[16] In a randomized trial of 191 patients with decompensated cirrhosis, entecavir-treated patients tended to have lower incidence of HCC than those treated with adefovir at 48 weeks.[17] However, due to the short duration of follow-up, the difference between the two groups was uncertain, and some patients might have already harbored undiagnosed HCC before antiviral therapy. Moreover, the findings cannot be extrapolated to patients with milder disease. The efficacy of entecavir to reduce hepatic events remains unclear.

Nowadays, it is unethical and impossible to perform a placebo-controlled trial to evaluate the efficacy of entecavir on reducing hepatic events in CHB patients. In this long-term cohort study, we included a real-life cohort of entecavir-treated CHB patients and a historical cohort of untreated hepatitis B patients. We aimed to investigate the efficacy of entecavir therapy on hepatic events and deaths.

Patients and Methods

Study Population

This was a retrospective-prospective cohort study based on two cohorts of patients. Cohort 1 (an entecavir cohort) included consecutive CHB patients who had received entecavir 0.5 mg/day with indications according to Asian Pacific Association for the Study of the Liver guidelines[9, 18] for at least 12 months in the Hepatitis Clinic, Prince of Wales Hospital, Hong Kong, from December 2005 to June 2012. Entecavir-treated patients were retrieved from a list of all patients who underwent HBV DNA testing at our hospital. Since October 2009, all patients newly started on entecavir were recruited into a longitudinal observational study in a prospective manner. In cohort 2 (a historical control cohort), treatment-naïve patients were recruited from December 1997 to July 2000 from the Hepatitis Clinic, Princess of Wales Hospital, and underwent routine clinical care until the mid- to late 2000s, at the time that antiviral treatments not been readily available or reimbursable.[19] Both cohorts included patients with positive hepatitis B surface antigen (HBsAg) for at least 6 months and a life expectancy of more than 1 year (i.e., lack of any significant acute and/or chronic comorbid conditions that might lead to an estimated life expectancy less than 1 year). Patients with other chronic liver diseases, HCC at baseline, or Child-Pugh class C cirrhosis, hepatic decompensation, autoimmune hepatitis, coinfection with hepatitis C virus, other serious concurrent illness (e.g., alcoholism, uncontrolled diabetes, or cancer) were excluded. This study was approved by the Clinical Research Ethics Committee of the Chinese University of Hong Kong. All patients in the prospective cohorts provided informed written consent.

Clinical Evaluation

The baseline visit was defined as the time of entecavir being started in the entecavir cohort, and the time of recruitment in the control cohort. At baseline, patients from both cohorts underwent an evaluation including full medical history, physical examination, complete blood count, prothrombin time and international normalized ratio (INR), liver and renal biochemistries, HBsAg, hepatitis B e antigen (HBeAg) and antibodies (anti-HBe), HBV DNA, and transabdominal ultrasonography. In our laboratory, HBV DNA was measured by Taqman real-time polymerase chain reaction assay validated against the EUROHEP standard with a linear range of detection from 20 to 2 × 108 IU/mL.[23]

The patients were followed up once every 3 to 6 months. During each visit, patients' symptoms, clinical events, drug history and adherence were recorded. Liver biochemistry and α-fetoprotein were checked every visit. HBsAg, HBeAg, and anti-HBe were checked at least yearly. In the entecavir group, HBV DNA was checked every 6 months. Maintained virologic response was defined as undetectable serum HBV DNA until the last visit.[15] Ultrasonography of the abdomen was performed every 1 to 2 years for surveillance of hepatocellular carcinoma, or more frequently if the α-fetoprotein level rose above 20 μg/L.

The model for end-stage liver disease (MELD) score was calculated using serum creatinine, bilirubin, and INR of prothrombin time[11] according to the following formula: 9.57 × Ln (creatinine mg/dL) + 3.78 × Ln (bilirubin mg/dL) + 11.20 × Ln INR + 6.43.[24] Radiologic cirrhosis was defined as coarse liver echotexture with nodularity and small liver size or the presence of features of portal hypertension (e.g., ascites, splenomegaly, and varices) noted on liver imaging.[22]

Clinical Outcomes

The primary outcome of this study was the cumulative probability of a composite endpoint of hepatic events occurring during the study period. The secondary outcomes were the cumulative probability of HCC, all-cause mortality, and liver-related mortality. Hepatic events were defined as any cirrhosis complications, HCC, and/or liver-related mortality. Cirrhosis complications (excluding HCC) included ascites, spontaneous bacterial peritonitis (SBP), variceal bleeding, hepatic encephalopathy, hepatorenal syndrome, and/or hepatic decompensation. Ascites was defined as free peritoneal fluid identified by ultrasound or computed tomography scans, or as clinically evident ascites confirmed by paracentesis. SBP was defined as an ascitic fluid polymorph count of 250/mm3 or above with or without positive bacterial culture. Variceal bleeding was defined as hematemesis and/or melena, with esophageal or gastric varices identified during upper gastrointestinal endoscopy, and no other lesion that might explain the bleeding episode. The diagnosis of HCC was established based on liver tumor feature detected by imaging techniques, tumor histology and/or high α-fetoprotein level (above 400 ng/mL), or other diagnostic criteria generally accepted for HCC.[25] Liver-related mortality was defined as death related to cirrhosis complications and/or HCC.

Statistical Analyses

Statistical analysis was performed by SPSS version 20.0 (SPSS, Chicago, IL). Continuous variables were expressed in mean ± standard deviation or median (range) as appropriate. Qualitative and quantitative differences between subgroups were analyzed using chi-square test or Fisher's exact test for categorical parameters, and Student t test or Mann-Whitney test for continuous parameters as appropriate. In order to minimize the inequality of follow-up duration in two cohorts, all the clinical outcomes were estimated by survival analysis to allow censoring the patients who had shorter follow-up durations, and presented as 3-year or 5-year cumulative probabilities. The rates of hepatic events and deaths at 5 years of entecavir-treated patients were compared with the control group by Cox proportional hazard model, and the risks are expressed as hazard ratios (HRs) and 95% confidence intervals (CIs). The effect of entecavir therapy on clinical outcomes was adjusted for baseline MELD score if indicated. Sensitivity analyses on clinical parameters and 3-year cumulative probability of hepatic events were also performed. The Kaplan-Meier method was used to estimate the cumulative probability of hepatic events and deaths. The log-rank test was used to compare time-to-event curves between the patient entecavir and control groups. All statistical tests were two-sided. Statistical significance was taken as P < 0.05.

Results

Patient Characteristics

During the study period, 1,505 CHB patients took entecavir for at least 12 months at the hepatitis clinics of Prince of Wales Hospital, Hong Kong. After excluding 28 patients who had evidence of preexisting HCC at baseline, 9 patients with Child-Pugh C cirrhosis, and two patients with coinfection with hepatitis C, the final analysis included 1,446 entecavir-treated patients who were followed for 36 ± 13 months. Four hundred forty-five (30%) patients were exposed to other antiviral therapies before entecavir treatment; 28 (2%) previously received pegylated interferon of a mean duration of 10 ± 3 months, and 417 (28%) had been put on other oral nucleos(t)ide analogue(s) of a mean duration of 16 ± 10 months before the entecavir therapy.

The control cohort included 424 treatment-naïve CHB patients followed for 114 ± 31 months. The entecavir cohort included more male patients (72% versus 65%; P = 0.02) and patients with older age (51 ± 12 versus 41 ± 13 years; P < 0.001), lower platelet counts (171 ± 61 × 109/L versus 181 ± 68 × 109/L; P = 0.003), higher MELD score (5.3 ± 4.6 versus 4.2 ± 3.2; P < 0.001), and radiologic cirrhosis (33% versus 16%; P < 0.001) than the control cohort (Table 1).

Table 1. Baseline Clinical Characteristics of Patients
 All PatientsPatients Without Radiologic CirrhosisPatients with Radiologic Cirrhosis
Entecavir Cohort (n = 1,466)Control Cohort (n = 424)PEntecavir Cohort (n = 984)Control Cohort (N = 355)PEntecavir Cohort (n = 482)Control Cohort (n = 69)P
  1. Data are expressed as the mean ± SD or no. (%).
  2. Abbreviation: NA, not applicable.
Follow-up, months36 ± 13114 ± 31<0.00139 ± 13118 ± 26<0.00135 ± 1396 ± 44<0.001
Male sex1049 (72)276 (65)0.02693 (70)221 (62)0.007356 (74)55 (80)0.30
Age, years51 ± 1241 ± 13<0.00149 ± 1239 ± 13<0.00155 ± 1151 ± 11<0.001
Platelet count, ×109/L171 ± 61181 ± 680.003195 ± 51195 ± 580.95130 ± 55112 ± 740.001
Prothrombin time, seconds13 ± 311 ± 2<0.00112 ± 311 ± 2<0.00114 ± 412 ± 20.01
Creatinine, mmol/L84 ± 7081 ± 180.3984 ± 8179 ± 170.5684 ± 4589 ± 220.04
Albumin, g/L44 ± 1339 ± 4<0.00145 ± 1239 ± 3<0.00143 ± 1336 ± 5<0.001
Total bilirubin, μmol/L23 ± 6211 ± 9<0.00123 ± 719 ± 6<0.00124 ± 4118 ±150.003
ALT, IU/L145 ± 31984 ± 1130.31165 ± 34685 ± 1150.09116 ± 26080 ± 970.15
α-Fetoprotein, μg/L34 ± 326 ± 15<0.00125 ± 1685 ± 12<0.00152 ± 4814 ± 260.02
HBeAg  0.48  0.61  0.81
Positive443 (30)155 (37) 307 (31)136 (38) 136 (28)19 (28) 
Negative1023 (70)269 (63) 677 (69)219 (62) 346 (72)50 (73) 
HBV DNA, log IU/mL5.0 ± 2.15.0 ± 2.10.215.0 ± 2.35.1 ± 2.20.735.1 ± 1.94.7 ± 1.70.03
Child-Pugh score  0.89  0.91  0.003
51,247 (85)357 (84) 877 (89)316 (89) 370 (77)41 (59) 
6114 (8)51 (12) 56 (6)35 (10) 58 (12)16 (23) 
≥7105 (7)16 (4) 51 (5)4 (1) 54 (11)12 (18) 
MELD score5.3 ± 4.64.2 ± 3.2<0.0014.1 ± 4.73.9 ± 3.40.236.6 ± 4.97.2 ± 4.10.11
Radiologic cirrhosis482 (33)69 (16)<0.001NANA NANA 
Previous antiviral therapy445 (30)0 (0) 335 (34)0 (0) 110 (23)0 (0) 
Pegylated interferon28 (2)NA 16 (2)NA 12 (2)NA 
Nucleos(t)ide analogues417 (28)NA 319 (32)NA 98 (20)NA 

In the radiologic cirrhosis subgroup, entecavir-treated patients were older (55 ± 11 versus 51 ± 11 years; P < 0.001), and had higher platelet counts (130 ± 55 × 109/L versus 112 ± 74 × 109/L; P = 0.001) and serum HBV DNA levels (5.1 ± 1.9 versus 4.7 ± 1.7 logs IU/mL; P < 0.001) than controls.

Cumulative Probability of Hepatic Events and Deaths

At a mean follow-up duration of 36 ± 13 months, 130 patients in the entecavir cohort developed hepatic events. The most commonly encountered hepatic events were variceal bleeding (n = 76), followed by HCC (n = 54), ascites (n = 28), liver-related mortality (n = 22), hepatic encephalopathy (n = 16), hepatorenal syndrome (n = 8), and SBP (n = 7). All-cause mortality developed in 38 patients in this cohort. Apart from the 22 patients who died of liver-related causes (15 patients died of HCC, 7 patients died of liver failure), 7 patients died of other malignancies, 4 patients died of sepsis, 2 patients died of renal failure, 1 patient died of heart failure, 1 patient died of respiratory failure, and 1 patient died of exfoliative skin disorder.

At a mean follow-up duration of 114 ± 31 months, 89 patients in the control cohort developed hepatic events. The most commonly encountered hepatic events were HCC (n = 53), followed by variceal bleeding (n = 37), liver-related mortality (n = 33), ascites (n = 31), SBP (n = 15), hepatic encephalopathy (n = 11), and hepatorenal syndrome (n = 3). All-cause mortality developed in 60 patients in the control cohort. Apart from the 33 patients who died of liver-related causes (22 patients died of HCC, 11 patients died of liver failure), 12 patients died of other malignancies, 7 patients died of sepsis, 3 patients died of renal failure, two patients died of heart failure, 2 patients died of intra-cranial hemorrhage, and one patient died of sudden cardiac arrest.

There was no statistical difference in the cumulative probabilities of hepatic events (P = 0.23), HCC (P = 0.82), liver-related mortality (P = 0.20) and all-cause mortality (P = 0.20) between the entecavir and the control cohorts. The 3-year cumulative probabilities among patients in the entecavir and control cohort were as follows: hepatic events, 9.5% (95% CI, 8.7%-10.3%) versus 8.0% (95% CI, 6.7%-9.3%); HCC, 3.9% (95% CI, 3.3%-4.5%) versus 3.7% (95% CI, 2.8%-4.6%); liver-related mortality, 1.9% (95% CI, 1.5%-2.3%) versus 1.7% (95% CI, 1.1%-2.3%); and all-cause mortality, 3.0% (95% CI, 2.5%-3.5%) versus 2.4% (95% CI, 1.7%-3.1%). The 5-year cumulative probabilities were: hepatic events, 12.4% (95% CI, 10.9%-13.9%) versus 10.8% (95% CI, 9.3%-12.3%); HCC, 6.6% (95% CI, 5.1%-8.1%) versus 6.5% (95% CI, 5.3%-7.7%); liver-related mortality, 2.2% (95% CI, 1.7%-2.7%) versus 4.6% (95% CI, 3.6%-5.6%); and all-cause mortality, 3.8% (95% CI, 3.1%-4.5%) versus 6.3% (95% CI, 5.1%-7.5%).

Hepatic Events and Deaths in Patients With or Without Cirrhosis

In the subgroup of patients without radiologic cirrhosis, the cumulative probabilities of all clinical outcomes were similar in the entecavir and control cohort (P values ranged from 0.19 to 0.64). At 3 years (the median follow-up of the entecavir cohort), the cumulative probability of hepatic events among patients without radiologic cirrhosis in the entecavir and control cohort was 3.6% (95% CI, 2.9%-4.3%) and 2.3% (95% CI, 1.5%-3.1%), respectively. The 3-year cumulative probabilities among patients without radiologic cirrhosis in the entecavir and control cohort were as follows: HCC, 1.4% (95% CI, 1.0%-1.8%) versus 1.7% (95% CI, 1.0%-2.4%); liver-related mortality, 0.8% (95% CI, 0.4%-1.2%) versus 0.3% (95% CI, 0%-0.6%); and all-cause mortality, 1.3% (95% CI, 0.9%-1.7%) versus 1.2% (95% CI, 0.6%-1.8%). The 5-year cumulative probabilities were: hepatic events, 6.1% (95% CI, 4.3%-8.9%) versus 5.2% (95% CI, 4.1%-6.3%); HCC, 3.3% (95% CI, 1.3%-5.3%) versus 3.0% (95% CI, 2.1%-3.9%); liver-related mortality, 0.9% (95% CI, 0.5%-1.3%) versus 1.2% (95% CI, 0.6%-1.8%); and all-cause mortality, 1.7% (95% CI, 1.2%-2.2%) versus 2.0% (95% CI, 1.2%-2.8%).

By Kaplan-Meier analysis, the 3-year and 5-year cumulative probabilities of hepatic events among patients with radiologic cirrhosis in the entecavir cohort was significantly lower than that in the control cohort (P = 0.001) (Fig. 1A). The 3-year cumulative probabilities of hepatic events among patients with radiologic cirrhosis in the entecavir cohort and control cohort were 21.6% (95% CI, 19.5%-23.7%) and 33.9% (95% CI, 28.1%-39.7%) (P = 0.001), respectively. Entecavir-treated patients with cirrhosis also had a lower cumulative probability of HCC (P = 0.036) (Fig. 1B), liver-related mortality (P < 0.001) (Fig. 1C), and all-cause mortality (P < 0.001) (Fig. 1D) compared with the untreated patients with cirrhosis. The 3-year cumulative probabilities among patients with radiologic cirrhosis in the entecavir cohort and control cohort were as follows: HCC, 9.1% (95% CI, 7.6%-10.6%) versus 14.5% (95% CI, 11.0%-18.0%) (P = 0.04); liver-related mortality, 3.8% (95% CI, 2.8%-4.8%) versus 7.5% (95% CI, 4.3%-10.7%) (P < 0.001); and all-cause mortality, 6.3% (95% CI, 5.0%-7.6%) versus 10.2% (95% CI, 7.8%-12.6%) (P < 0.001). The 5-year cumulative probabilities were: hepatic events, 25.5% (95% CI, 22.7%-28.3%) versus 45.8% (39.7%-51.9%) (P = 0.001); HCC, 13.8% (95% CI, 11.3%-16.3%) versus 26.4% (95% CI, 20.7%-32.1%) (P = 0.036); liver-related mortality, 5.0% (95% CI, 3.6%-6.4%) versus 23.3% (95% CI, 18.0%-28.6%) (P < 0.001); and all-cause mortality, 8.3% (95% CI, 6.6%-10.0%) versus28.2% (95% CI, 22.7%-33.7%) (P < 0.001).

Figure 1.

Kaplan-Meier analysis of cumulative probability of (A) hepatic events, (B) HCC, (C) liver-related mortality, and (D) all-cause mortality among patients with cirrhosis in the entecavir and control cohorts.

Virologic Response and Clinical Events in Patients With Cirrhosis

In the entecavir cohort, maintained viral suppression was achieved in 377 (78%) of the 482 patients with cirrhosis and in 759 (77.1%) of the 984 patients without cirrhosis. The overall cumulative probabilities of undetectable HBV DNA at years 1, 2, and 3 were 63.6%, 72.2%, and 76.5%, respectively; those among patients without cirrhosis at years 1, 2, and 3 were 65.8%, 74.6%, and 77.9%, respectively. The rate and time to undetectability in control patients was not available, as HBV DNA was not regularly monitored in the control cohort. Instead, we only collected the baseline level in the control cohort.

Maintained viral suppression was achieved in 360 of 372 (97%) of patients with cirrhosis who were treatment-naïve before entecavir therapy. Among the 105 patients who failed to achieve maintained viral suppression, 93 (89%) received previous antiviral therapy: 56 patients received lamivudine, 11 patients received adefovir, 7 patients received telbivudine, 4 patients received pegylated interferon, and 15 patients received combinations of nucleos(t)ide analogues and/or pegylated interferon (also see Supporting Information).

The cumulative number (and probability) of patients who developed hepatic events at 3, 6, and 12 months after entecavir treatment was 11 (2.3%), 16 (3.3%), and 42 (8.7%), respectively. That of those patients who achieved maintained viral suppression at 3, 6, and 12 months was 8 (2.1%), 13 (3.4%), and 29 (7.7%), respectively, and that of those patients who did not achieve maintained viral suppression was 3 (2.8%), 3 (2.8%), and 13 (12.4%), respectively.

Patients who achieved maintained viral suppression had a similar cumulative probability of hepatic events (P = 0.11) (Fig. 2A) and HCC (P = 0.33) (Fig. 2B), but significantly lower cumulative probability of liver-related mortality (P < 0.001) (Fig. 2C) and all-cause mortality (P = 0.001) (Fig. 2D) when compared with those who did not achieved maintained viral suppression. The 3-year cumulative probabilities among patients who could or could not achieve maintained viral suppression were as follows: hepatic events, 15.1% (95% CI, 10.4%-19.8%) versus 20.6% (95% CI, 18.3%-22.9%); HCC, 8.7% (95% CI, 7.1%-10.3%) versus 10.7% (95% CI, 6.9%-14.5%); liver-related mortality, 1.2% (95% CI, 0.3%-2.1%) versus 10.6% (95% CI, 7.0%-14.2%); and all-cause mortality, 4.3% (95% CI, 3.0%-5.6%) versus 13.9% (95% CI, 10.0%-17.8%).

Figure 2.

Kaplan-Meier analysis of cumulative probability of (A) hepatic events, (B) HCC, (C) liver-related mortality, and (D) all-cause mortality with respect to the maintained viral suppression among patients with cirrhosis in the entecavir cohort.

Compared with the untreated patients with cirrhosis, entecavir-treated patients with cirrhosis who achieved maintained viral suppression had a significantly lower cumulative probability of hepatic events (P = 0.001), HCC (P = 0.02), liver-related mortality (P < 0.001), and all-cause mortality (P < 0.001) (Fig. 3). On the other hand, entecavir-treated patients with cirrhosis who did not achieve maintained viral suppression did not have significantly different cumulative probabilities of hepatic events (P = 0.23), HCC (P = 0.63), liver-related mortality (P = 0.57), or all-cause mortality (P = 0.51) when compared with the untreated patients with cirrhosis (Fig. 3).

Figure 3.

Kaplan-Meier analysis of cumulative probability of (A) hepatic events, (B) HCC, (C) liver-related mortality, and (D) all-cause mortality among patients with cirrhosis in the entecavir cohort who could or could not achieve maintained viral suppression. The comparisons were performed against patients with cirrhosis in the control cohort.

Adjustment of Time-to-Event According to Virologic Response

Because the effect of entecavir was caused by HBV suppression, the time-to-event was adjusted for the interval before patients achieving viral suppression by subtracting the time to HBV DNA negativity from time to hepatic events if patients developed hepatic events after viral suppression. On the other hand, those patients who developed hepatic events before viral suppression were regarded as those who could not achieve maintained viral suppression.

Sixty-two patients with cirrhosis developed hepatic events after achieving maintained viral suppression; the median time to HBV DNA negativity was 6 months (interquartile range, 5-8 months), and the median time to hepatic event was 18 months (interquartile range, 16-27 months). Twenty-two patients with cirrhosis developed hepatic events before achieving maintained viral suppression; hence they were regarded as patients who could not achieve maintained viral suppression. After these adjustments, patients who achieved maintained viral suppression had lower cumulative probabilities of all clinical outcomes compared with patients who could not achieve maintained viral suppression (Supporting Fig. 1).

Sensitivity Analyses on Clinical Parameters and Hepatic Events

Because the severity of liver cirrhosis would be a major confounder for all liver-related complications, we adjusted all clinical outcomes for MELD score among patients with liver cirrhosis using the Cox proportional hazard model. Patients with liver cirrhosis in the entecavir cohort had reduced risks of all clinical outcomes when compared with those in the control cohort, after adjustment for MELD score: hepatic events (adjusted HR, 0.51; 95% CI, 0.34-0.78; P = 0.002); HCC (adjusted HR, 0.55; 95% CI, 0.31-0.99; P = 0.049); liver-related mortality (adjusted HR, 0.26; 95% CI, 0.13-0.55; P < 0.001); and all-cause mortality (adjusted HR, 0.34; 95% CI, 0.18-0.62; P < 0.001) (Table 2).

Table 2. Efficacy of Entecavir Therapy on Clinical Outcomes Adjusted for MELD Score and Maintained Viral Suppression Among Patients With Liver Cirrhosis
Clinical OutcomesHR (95% CI)P
Hepatic events0.51 (0.34-0.78)0.002
HCC0.55 (0.31-0.99)0.049
Liver-related mortality0.26 (0.13-0.55)<0.001
All-cause mortality0.34 (0.18-0.62)<0.001

To further assess the impact of different clinical parameters on clinical outcome, the 3-year cumulative probabilities of hepatic events in different subgroups of patients with cirrhosis were compared between the entecavir and control cohorts (Table 3). The 3-year cumulative probability of hepatic events in the entecavir cohort were significantly lower among patients of male sex (32.1% versus 53.0%; P < 0.001), platelet count ≤100 × 109/L (25.5% versus 50.3%; P < 0.001), serum albumin ≤35 g/L (47.6% versus 68.2%; P = 0.04), serum total bilirubin >18 μmol/L (28.2% versus 56.3%; P = 0.007), serum alanine aminotransferase (ALT) ≤2× upper limit of normal (35.6% versus 45.6%; P = 0.03) or >2× upper limit of normal (23.4% versus 71.4%; P = 0.003), baseline HBV DNA >200,000 IU/mL (27.6% versus 59.1%; P < 0.001), Child-Pugh score ≥7 (54.8% versus 79.5%; P = 0.02), and MELD score <10 (23.4% versus 40.6%; P < 0.001) or ≥10 (48.6% versus 75.2%; P < 0.001), when compared with the control cohort.

Table 3. Sensitivity Analyses on Clinical Parameters and 3-Year Cumulative Probability of Hepatic Events Among Patients With Liver Cirrhosis
 −Year Cumulative Probability (95% CI)HR (95% CI)PaAdjusted HR (95% CI)Pb
Entecavir Cohort (n = 482)Control Cohort (n = 69)
  1. Abbreviation: ULN, upper limit of normal.
  2. aBy Cox proportional hazard model.
  3. bHR was adjusted for MELD score and maintained viral suppression.
Sex      
Male32.1% (28.6%-35.7%)53.0% (44.8%-61.2%)2.1 (1.3-3.4)<0.0011.9 (1.2-2.8)0.001
Female20.7% (16.1%-25.3%)36.4% (21.9%-50.9%)1.6 (0.6-4.2)0.161.3 (0.5-3.6)0.27
Platelet count, ×109/L      
≤100 ×109/L25.5% (22.3%-28.7%)50.3% (42.5%-58.1%)2.5 (1.6-4.0)<0.0012.3 (1.5-3.7)<0.001
>100 ×109/L18.6% (14.2%-23.0%)26.8% (22.3%-31.3%)1.0 (0.3-2.9)0.160.9 (0.2-1.7)0.33
Albumin      
≤35 g/L47.6% (40.8%-54.4%)68.2% (58.5%-77.9%)1.7 (1.0-3.0)0.041.6 (0.9-2.2)0.07
>35 g/L29.3% (25.4%-33.2%)26.1% (16.9%-35.3%)0.8 (0.4-1.8)0.670.9 (0.5-1.7)0.73
Total bilirubin      
≤18 μmol/L22.1% (18.1%-26.1%)31.4% (22.6%-40.8%)1.4 (0.7-2.5)0.281.3 (0.6-2.2)0.35
>18 μmol/L28.2% (23.4%-33.0%)56.3% (44.4%-68.2%)2.1 (1.1-3.9)0.0071.8 (1.0-2.7)0.02
ALT, IU/L      
≤2× ULN35.6% (32.8%-38.4%)45.6% (38.9%-52.3%)1.5 (1.0-2.4)0.031.4 (1.0-2.1)0.04
>2× ULN23.4% (17.1%-29.7%)71.4% (54.3%-88.5%)4.4 (1.5-12.9)0.0033.7 (1.4-8.9)0.008
HBeAg      
Positive27.9% (22.2%-33.6%)40.2% (26.2%-54.2%)1.8 (0.8-4.1)0.171.6 (0.9-3.4)0.41
Negative32.0% (28.2%-35.8%)52.8% (44.5%-61.1%)1.9 (1.1-3.1)0.0011.8 (1.1-2.7)0.006
Baseline HBV DNA      
≤2,000 IU/mL34.0% (26.6%-41.4%)36.4% (21.9%-50.9%)0.9 (0.3-2.9)0.840.8 (0.2-2.1)0.91
>2,000-200,000 IU/mL29.5% (23.9%-35.1%)47.4% (35.9%-58.9%)1.6 (0.9-3.1)0.071.4 (0.7-2.7)0.11
>200,000 IU/mL27.6% (23.5%-31.7%)59.1% (47.6%-70.6%)3.4 (1.9-6.1)<0.0013.0 (1.8-5.2)<0.001
Child-Pugh score      
521.3% (18.1%-24.5%)30.0% (19.8%-40.2%)1.4 (0.9-3.3)0.091.4 (0.9-3.3)0.09
645.6% (36.0%-55.2%)50.0% (37.5%-62.5%)1.3 (0.6-2.8)0.381.4 (0.7-2.9)0.28
≥754.8% (46.8%-63.8%)79.5% (67.5%-91.5%)2.0 (1.0-4.1)0.021.7 (1.0-3.5)0.04
MELD score      
<1023.4% (20.3%-26.5%)40.6% (31.9%-49.3%)2.3 (1.3-3.9)0.0082.3 (1.3-3.9)0.008
≥1048.6% (41.9%-55.3%)75.2% (65.0%-85.4%)2.8 (1.6-4.7)<0.0012.8 (1.6-4.7)<0.001

Discussion

This was the one of the first large-scale real-life cohort studies demonstrating the efficacy of entecavir therapy in reducing various complications of chronic hepatitis B. The beneficial effects were observed clearly in patients with cirrhosis but not in patients without cirrhosis. Hepatic events and HCC were reduced by approximately 50%, whereas liver-related and all-cause mortality was reduced by 74% and 66%, respectively, among patients with cirrhosis treated by entecavir after adjusting for MELD score. Maintained viral suppression, which was achieved in most entecavir-treated patients, was important for patients with cirrhosis in order for them to benefit from the reduced risk of clinical outcomes.

Despite the clear benefits illustrated in patients with cirrhosis, we could not identify any difference in the rates of clinical outcomes between the subgroups of patients without cirrhosis in two cohorts. In fact, there is no evidence concerning the beneficial effect on cirrhosis complications and mortality by antiviral therapy among CHB patients without cirrhosis. The efficacy of lamivudine was also only demonstrated among patients with advanced fibrosis or cirrhosis.[6] Noticeably, the rates of clinical outcomes were very low in untreated patients without cirrhosis, hence demonstrating that the even lower rates of the clinical outcomes in entecavir-treated patients without cirrhosis became challenging. We believe that a significantly larger sample size and longer follow-up duration is needed to demonstrate any beneficial effect of entecavir on patients without cirrhosis.

Although maintained virologic response is an important surrogate endpoint, the effect of entecavir on clinical outcomes was unclear. In previous studies, HCC could still develop in patients with complete viral suppression or even HBsAg seroclearance.[7, 26] A recent multicenter European study involved 372 entecavir-treated patients (26% of them had cirrhosis) followed for a median of 20 months demonstrated that a virologic response to entecavir (defined as serum HBV DNA levels < 80 IU/mL) reduced the probability of clinical events in patients with cirrhosis compared with those without a virologic response.[27] In this European cohort, 87% and 93% achieved virologic response at week 96 and 144 of entecavir treatment, respectively, hence the subgroup of patients without virologic response was small (n = 26). Furthermore, there was no untreated control group for comparison. In the current study, we have shown patients who could not achieve a maintained virological response did not have significantly lower risk of all clinical outcomes in 5 years compared with untreated patients. Most of the patients who failed to achieve virological response had been exposed to other antiviral therapies, particularly lamivudine and/or telbivudine. Some of them might have developed entecavir resistance, which might lead to reactivation of HBV and subsequently disease progression. On the other hand, because maintained viral suppression could be achieved in 97% of patients who were previously treatment-naïve, entecavir will be most beneficial if it is used as a first-line therapy.

In our sensitivity analysis, we found that the risk of hepatic events is much reduced by entecavir among patients with advanced liver disease and/or portal hypertension, as reflected by low platelet counts, low serum albumin, high serum total bilirubin, high serum HBV DNA levels, and high Child-Pugh score and/or MELD score. HBeAg status and ALT levels had little effect on the benefit of entecavir. This observation supports the current clinical guidelines that entecavir should be considered in patients with advanced liver cirrhosis even when HBV DNA is low and ALT level is normal.[8] However, because HCC cannot be completely prevented by entecavir, regular HCC surveillance is still needed even if/when undetectable HBV DNA is achieved.[28, 29]

Our study has the strength of large sample size and long follow-up, which increased the statistical power and reliability of the results. Inclusion of a historical cohort as the control group rendered meaningful comparisons to establish the beneficial effects of entecavir therapy. Nonetheless, our study also has a few limitations. The retrospective nature of certain clinical information in the entecavir cohort might be biased by incomplete data collection. Patient inclusion and identification through the database of HBV DNA testing might miss those not adherent to follow-up. Fortunately, the compliance of HBV DNA monitoring every 6 months was high in entecavir-treated patients (>98%), as the test was free of charge to the patients. Moreover, patients who started taking entecavir after October 2009 had all the clinical information prospectively collected, which also minimized the chance of incomplete data collection. Diagnosis of liver cirrhosis radiologically might miss advanced fibrosis and early cirrhosis. To avoid false labeling, radiologic cirrhosis was diagnosed with features of portal hypertension, thus indicating patients with more established cirrhosis. This definition can be more adopted by physicians in real-life clinical practice. The difference in baseline clinical characteristics in two cohorts—namely, that the untreated control cohort might have milder disease activity, or that the entecavir cohort were older, had lower platelet count, and had higher MELD scores—might affect the analysis on the difference of clinical outcomes. However, all these factors were not in favor of the entecavir cohort, such that it might only lessen the difference between the cohorts. Therefore, the beneficial effect of entecavir might be even more than we demonstrated. Furthermore, the sensitivity analyses on different clinical parameters, the reduced risks of hepatic events, and other outcomes were consistently demonstrated. Another limitation was the possibility of a cohort effect due to the recruitment of the control cohort at an earlier time than the entecavir cohort, which was impossible to adjust using statistical methods. The relatively small number of patients in the cirrhosis subgroup of the control cohort and shorter follow-up duration (3 years) of the patients in the entecavir cohort were other limitations. The different follow-up period might introduce bias on the clinical outcomes. This bias was mostly compensated by the survival analysis to allow censoring the patients who had shorter follow-up durations. Further analysis of the data of the noncirrhosis subgroup after a longer follow-up duration is warranted to evaluate the efficacy of entecavir on patients without cirrhosis. Finally, due to the study design, subjects in the entecavir group could not have experienced death in the first year. In fact, the information of all patients who had received entecavir was captured, and the morality benefit was still present when all entecavir-treated patients with cirrhosis were analyzed regardless of treatment duration.

In conclusion, entecavir therapy reduces the risks of hepatic events, HCC, liver-related, and all-cause mortality of CHB patients with liver cirrhosis. Patients with more advanced liver disease can achieve maintained viral suppression benefit most from this antiviral therapy. Our results provide important evidence to support the current recommendation of entecavir as a first-line antiviral therapy for CHB patients.

Ancillary

Advertisement