Chronic hepatitis B (CHB) infection is a large

Transcription

1 AMERICAN ASSOCIATION FOR THE STUDY OFLIVERD I S E ASES HEPATOLOGY, VOL. 64, NO. 2, 2016 Serum Levels of Hepatitis B Surface Antigen and DNA Can Predict Inactive Carriers With Low Risk of Disease Progression Jessica Liu, 1 Hwai-I Yang, 1 Mei-Hsuan Lee, 2 Chin-Lan Jen, 1 Richard Batrla-Utermann, 3 Sheng-Nan Lu, 4 Li-Yu Wang, 5 San-Lin You, 6 and Chien-Jen Chen 1,7 Serum levels of hepatitis B virus (HBV) DNA ( 2000 IU/mL) and hepatitis B surface antigen (HBsAg) (<1000 IU/mL) have been shown to distinguish inactive carriers with high accuracy. The goal of this study was to validate the predictability of onetime measurement of quantitative HBsAg and HBV DNA levels for inactive carrier status and chronic hepatitis B (CHB) progression in a community-based cohort. This study included 1529 participants chronically infected with HBV genotype B or C from the REVEAL-HBV cohort. They were ascertained as inactive or active CHB after 18 months of follow-up. Validity of the one-time measurement was assessed by sensitivity, specificity, and receiver operating characteristic curves, while associations with clinical outcomes were calculated with Cox proportional hazards regressions. The one-time baseline measurement of HBsAg <1000 IU/mL and HBV DNA <2000 IU/mL distinguished inactive carriers from active CHB with a sensitivity, specificity, positive predictive value, negative predictive value, and diagnostic accuracy of 71%, 85%, 83%, 74%, and 78%, respectively. Those identified as inactive carriers using the one-time baseline measurement had multivariate adjusted hazard ratios of 0.36 (95% confidence interval [CI], ) and 0.36 ( ) for hepatocellular carcinoma and liver cirrhosis, respectively, and an adjusted rate ratio of 6.97 (95% CI, ) for HBsAg seroclearance. Areas under the receiver operating characteristic curve of predicting these outcomes using the one-time definition were similar to those obtained when using long-term follow-up defined carrier status for prediction. Conclusion: This study confirms the predictability of a one-time combined HBsAg and HBV DNA measurement for future inactive carriers. This single-point strategy provides new and complementary information useful for management of patients with chronic hepatitis B infection. (HEPATOLOGY 2016;64: ) Chronic hepatitis B (CHB) infection is a large global health problem due to its widespread geographical prevalence and its varying clinical consequences, which range from inflammation and infection to liver cirrhosis and hepatocellular carcinoma (HCC). (1) The characterization of patients with CHB is often divided into hepatitis B e antigen (HBeAg)-positive and HBeAg-negative patients. It is well-known that HBeAg-seropositive patients are at increased risk for HCC, and HBeAg seroconversion is an important milestone for clinical management of these patients. (2,3) On the other hand, HBeAg-negative patients can be defined clinically as inactive carriers or as patients with active CHB. Inactive carriers typically have HBV DNA levels <2,000 IU/mL and normal alanine aminotransferase (ALT) levels, whereas those with active CHB have HBV DNA levels >2,000 IU/mL with fluctuations that reach >20,000 IU/mL in some patients, accompanied by intermittent or persistent ALT elevation. (2,4) Previous studies have shown that compared with individuals with active CHB, inactive carriers have a significantly better prognosis and low rates of cirrhosis and Abbreviations: ALT, alanine aminotransferase; AUROC, area under the receiver operating characteristic; CHB, chronic hepatitis B; CI, confidence interval; HBeAg, hepatitis B e antigen; HBsAg, hepatitis B surface antigen; HBV, hepatitis B virus; HCC, hepatocellular carcinoma. Received September 4, 2015; accepted March 14, This study was supported by the Taiwan Department of Health, Bristol-Myers Squibb, Roche Diagnostics, and Academia Sinica. Copyright VC 2016 by the American Association for the Study of Liver Diseases. View this article online at wileyonlinelibrary.com. DOI /hep Potential conflict of interest: Richard Batrla-Utermann is an employee of Roche Diagnostics. 381

2 LIU ET AL. HEPATOLOGY, August 2016 HCC. Among HBeAg-negative individuals with normal ALT levels, patients with HBV DNA levels <10,000 copies/ml (<2000 IU/mL), either at baseline or throughout follow-up, had significantly decreased rates of both liver cirrhosis and HCC. (5-8) In addition, persistently low viral loads present in inactive carriers also confer much higher rates of hepatitis B surface antigen (HBsAg) seroclearance. (9) Thus, correctly identifying inactive versus active CHB is a crucial step in identifying patients in need of less stringent follow-up, as well as patients who may benefit from additional follow-up and earlier initiation of antiviral therapy. Traditional differentiation between inactive carriers and active CHB is both costly and difficult, because viral loads and ALT levels tend to fluctuate during the natural course of the disease. Therefore, current definitions of an inactive carrier state require at least 1 year of continuous monitoring of viral loads, which require sensitive and costly real-time polymerase chain reaction quantification assays, in addition to continuous monitoring of ALT levels. (2,4) However, several important studies in recent years have shown that quantitative HBsAg levels not only correlate with covalently closed circular DNA (cccdna) levels, (10) they also provide additional predictability of serological and clinical outcomes (11-15) and are good predictors of treatment response. (16,17) In a landmark study among genotype D carriers, Brunetto et al. (18) examined the ability of quantitative HBsAg levels to help distinguish inactive carriers from patients with active CHB. They found that a one-time measurement of HBV DNA levels 2000 IU/mL and HBsAg levels <1000 IU/mL could distinguish inactive carriers with 91.1% sensitivity, 95.4% specificity, 87.9% positive predictive value, and 96.7% negative predictive value, and that using a one-time measurement was comparable with long-term monitoring. (18) These results introduced a novel strategy for effective management of CHB patients. Other studies have also evaluated the ability of a one-time measurement of quantitative HBsAg and HBV DNA to distinguish inactive or reactivated patients. (19) However, previous studies have used small, hospitalbased samples and have not clarified the use of a one-time measurement for prediction of long-term outcomes. Therefore, the goal of the present study was to externally validate the use of a one-time measurement of quantitative HBsAg and HBV DNA to predict future inactive carriers in a community-based cohort of genotype B and C individuals and, using this same one-time measurement, to predict future HCC, cirrhosis, and HBsAg seroclearance. Patients and Methods STUDY COHORT This study included 1529 patients infected with chronic hepatitis B from the REVEAL-HBV cohort, which consists of individuals aged years who were recruited between 1991 and 1992 in Taiwan. Detailed information on enrollment, recruitment, and study procedures have been described previously. (20) All patients were anti-hcv seronegative, HBeAg seronegative, free of liver cirrhosis at study entry, and infected with viral genotype B or C. No patients received antiviral treatment during the study period. In order to be eligible for this study, patients had to have two or three measurements of serum ALT and HBV DNA within 18 months of study entry. All patients provided written informed consent to participate in the study, which was approved by the Institutional Review Board of the College of Public Health, National Taiwan University, Taipei, Taiwan. ASCERTAINMENT OF OUTCOME Patientswereidentifiedasinactivecarriersorashaving active CHB according to their serological profiles after 18 ARTICLE INFORMATION: From the 1 Genomics Research Center, Academia Sinica, Taipei, Taiwan; 2 Institute of Clinical Medicine, National Yang Ming University, Taipei, Taiwan; 3 Roche Diagnostics, Ltd, Basel, Switzerland; 4 Department of Gastroenterology, Chang-Gung Memorial Hospital, Kaohsiung, Taiwan; 5 MacKay Medical College, New Taipei City, Taiwan; 6 Department of Public Health, College of Medicine, Catholic Fu-Jen University, New Taipei City, Taiwan; 7 Graduate Institute of Epidemiology and Preventative Medicine, College of Public Health, National Taiwan University, Taipei, Taiwan. ADDRESS CORRESPONDENCE AND REPRINT REQUESTS TO: Chien-Jen Chen, D.Sc. Genomics Research Center Academia Sinica 128 Academia Road, Section 2 Taipei 115, Taiwan chencj@gate.sinica.edu.tw Tel: Fax:

3 HEPATOLOGY, Vol. 64, No. 2, 2016 LIU ET AL. FIG. 1. Flow chart of study participants. months of follow-up (Fig. 1). First, patients were categorized according to their baseline HBV DNA levels. Among those who had HBV DNA levels <2000 IU/mL at baseline, patients whose HBV DNA levels remained persistently <2000 IU/mL with normal ALT throughout 18 months of follow-up were classified as inactive carriers. However, patients who had periodic HBV DNA elevations that were below or above 20,000 IU/mL were classified as having active CHB and high-viremia active CHB, respectively. Moreover, among the patients who had HBV DNA levels 2000 IU/mL at baseline, patients who had periodic HBV DNA elevations below or above 20,000 IU/mL were also classified as active CHB or high-viremia active CHB, respectively (Fig. 1). In total, after 18 months of follow-up, there were 777 inactive carriers, 497 of whom had active CHB and 255 of whom had high-viremia active CHB. HCC was detected by way of follow-up examinations and computerized linkages with the National Cancer Registry and National Death Certification databases until December 31, Identified cases were confirmed through medical chart reviews by gastroenterologists according to the following criteria: histopathologic confirmation; positive lesions detected by at least two different imaging techniques (e.g., abdominal ultrasonography, angiogram, computed tomography); or positive lesions detected by one imaging technique combined with a serum a-fetoprotein level >400 ng/ml. Cirrhosis was diagnosed through high-resolution abdominal ultrasonography, which was confirmed by gastroenterologists and followed up until June 30, 2004, as described elsewhere. (8) HBsAg seroclearance was defined as the first instance at which a patient tested negative for HBsAg and remained seronegative thereafter, and was followed up until June 30, LABORATORY METHODS Blood collection and health examinations were performed at study entry and every 6-12 months at followup. Serostatus of HBsAg, HBeAg, and anti-hcv and serum levels of HBV DNA and ALT were tested using commercial kits: HBsAg and HBeAg by radioimmunoassay (Abbott Laboratories, North Chicago, IL), anti- 383

4 LIU ET AL. HEPATOLOGY, August 2016 Characteristic TABLE 1. Baseline Characteristics of the Study Cohort (N ) Inactive Carriers (n 5 777) Active CHB (n 5 497) High-Viremia Active CHB (n 5 255) P Sex Women 299 (38.5) 168 (33.8) 69 (27.1) Men 478 (61.5) 329 (66.2) 186 (72.9) Age, years, mean (SD) 46.6 (9.8) 45.7 (9.5) 47.4 (9.4) 0.05 ALT, U/L Mean (SD) 13.3 (7.9) 19.4 (26.5) 24.1 (24.9) <0.001 < (64.4) 288 (57.9) 114 (44.7) < (35.6) 170 (34.3) 104 (40.8) >45 0 (0) 39 (7.9) 37 (14.5) HBV DNA, IU/mL Log copies, mean (SD) 2.1 (0.4) 3.6 (0.9) 4.4 (1.2) <0.001 Undetectable 410 (52.8) 0 (0) 0 (0) <0.001 Undetectable-2, (47.2) 155 (31.2) 48 (18.8) 2,000-20,000 0 (0) 258 (51.9) 73 (28.6) 20, ,000 0 (0) 60 (12.1) 83 (32.6) >200,000 0 (0) 24 (4.8) 51 (20.0) HBsAg, IU/mL Log IU, mean (SD) 1.8 (1.7) 2.8 (0.9) 3.1 (0.6) <0.001 < (44.7) 78 (15.7) 13 (5.1) < (26.4) 200 (40.2) 86 (33.7) (28.9) 219 (44.1) 156 (61.2) HBV genotype* B 253 (60.0) 341 (76.3) 181 (74.5) <0.001 B1C/C 169 (40.0) 106 (23.7) 62 (25.5) *Data only available for those with detectable (>57 IU/mL) HBV DNA levels. Data are presented as n (%) unless noted otherwise. Abbreviation: SD, standard deviation. HCV by way of enzyme immunoassay using secondgeneration test kits (Abbott Laboratories), ALT by way of a serum chemistry autoanalyzer (model 736; Hitachi Co., Tokyo, Japan) using commercial reagents, and serum HBV DNA levels by way of polymerase chain reaction (COBAS Amplicor, Roche Diagnostics, Indianapolis, IN) for baseline samples and real-time polymerase chain reaction (COBAS TaqMan, Roche Diagnostics, Indianapolis, IN) for follow-up samples. Serum HBsAg levels at study entry were quantified retrospectively using the Elecsys HBsAg II Quant assay (Roche Diagnostics GmbH, Mannheim, Germany), which has a lower limit of detection of 0.05 IU/mL. STATISTICAL ANALYSIS Differences between follow-up groups were analyzed using the Kruskal-Wallis test or the chi-squared test where appropriate. The cutoffs of HBV DNA <2000 IU/mL and HBsAg <1000 IU/mL were previously reported to be optimal cutoffs. In our cohort, these same cutoffs were determined to be optimal using the Youden s index. To examine the predictive accuracy, sensitivities, specificities, and receiver operating characteristic curves were calculated. For the prediction of HCC, cirrhosis, and HBsAg seroclearance, person-time was calculated as the time from study entry until either outcome occurred, death, or last follow-up, whichever came first. The associations with HCC, cirrhosis, and HBsAg seroclearance were calculated with Cox proportional hazards regression analyses. All analyses were performed with SAS software (version 9.3; SAS Institute, Cary, NC), and statistical significance was determined using two-tailed tests (P < 0.05). Results BASELINE CHARACTERISTICS In total, among 1529 HBeAg seronegative individuals in this study, there were 777 inactive carriers, 497 of whom had active CHB and 255 of whom had highviremia active CHB (Fig. 1, Table 1). Men were more likely to have high-viremia active CHB and less likely to be inactive carriers than women. There was a higher proportion of men among patients with high-viremia CHB versus inactive carriers (72.9% and 61.5% men versus 27.1% and 38.5% women, respectively; P ). ALT and HBV DNA levels were significantly different by outcome groups. An increasing trend of higher baseline HBsAg levels was also seen 384

5 HEPATOLOGY, Vol. 64, No. 2, 2016 LIU ET AL. Predictive Criteria TABLE 2. Prediction of Inactive Carriers Based on Serum HBV DNA and HBsAg Levels All Participants (N *) Excluded 497 Active CHB (n ) HBV DNA <2000 IU/mL HBsAg <1000 IU/mL HBV DNA <2000 IU/mL and HBsAg <1000 IU/mL HBV DNA <2000 IU/mL HBsAg <1000 IU/mL HBV DNA <2000 IU/mL and HBsAg < 1000 IU/mL Sensitivity Specificity Positive predictive value Negative predictive value Diagnostic accuracy *Inactive carriers versus all active CHB. Inactive carriers versus high-viremia active CHB. with high-viremia active CHB and active CHB compared with inactive carriers (61.2% and 44.1% with HBsAg levels 1000 IU/mL versus 28.9%; P < 0.001) (Table 1). Median (interquartile range) baseline HBsAg levels (log10 IU/mL) were significantly different between inactive carriers, active CHB, and highviremia active CHB at 2.21 ( ), 2.87 ( ), and 3.15 ( ), respectively. The correlation between baseline HBsAg and HBV DNA levels was moderately low, with a correlation coefficient R (P < 0.001). PREDICTION OF INACTIVE CARRIERS USING HBSAG LEVELS We examined the ability of baseline HBsAg levels to predict future inactive carriers identified through 18 months of follow-up. Baseline HBsAg levels were able to differentiate inactive carriers from patients with active CHB with an area under the receiver operating characteristic (AUROC) of When limited to inactive carriers and those with high-viremia active CHB, the AUROC for baseline HBsAg levels increased to The diagnostic performances of single-point measurements of HBV DNA and HBsAg and their combination in distinguishing inactive carriers from all active CHB or high-viremia active CHB is shown in Table 2. Using 1000 IU/mL of HBsAg alone as a cutoff, HBsAg was able to distinguish inactive from all active CHB (n ) with a sensitivity of 71%, specificity of 50%, positive predictive value of 59%, negative predictive value of 63%, and diagnostic accuracy of 61%. When limited to inactive carriers and high-viremia active CHB (n51032), specificity, positive predictive value, and diagnostic accuracy improved. When HBsAg <1000 IU/mL was combined with HBV DNA <2,000IU/mL,theone-timebaselinemeasurement could distinguish inactive carriers from all active CHB with a sensitivity, specificity, positive predictive value, negative predictive value, and diagnostic accuracy of 71%, 85%, 83%, 74%, and 78%, respectively. When limited to inactive carriers and high-viremia active CHB, the resulting sensitivity, specificity, positive predictive value, TABLE 3. Incidence Rates and Hazard (Rate) Ratios of Cirrhosis, HCC, and HBsAg Seroclearance After Baseline Identification of Inactive Carriers (N ) Defined Inactive Carrier at Baseline* n (%) Controls, n (%) Cases, n (%) Person- Years Incidence Rate (per 10,000 Person-Years) Crude Hazard Ratio (95% CI) Multivariate Adjusted Hazard Ratio (95% CI) Liver cirrhosis No 865 (57) 772 (55) 93 (78) Yes 664 (43) 637 (45) 27 (22) ( ) 0.36 ( ) HCC No 865 (57) 811 (56) 54 (78) Yes 664 (43) 649 (44) 15 (22) ( ) 0.36 ( ) HBsAg seroclearance No 865 (57) 808 (66) 57 (19) Yes 664 (43) 419 (34) 245 (81) ( ) 6.97 ( ) *Defined using baseline HBV DNA <2000 IU/mL and HBsAg <1000 IU/mL. Adjusted for sex, age, and ALT levels. AUROC 385

6 LIU ET AL. HEPATOLOGY, August 2016 negative predictive value, and diagnostic accuracy was 71%, 91%, 96%, 51%, and 76%, respectively (Table 2). PREDICTION OF CIRRHOSIS, HCC, AND HBSAG SEROCLEARANCE We also examined the ability of the baseline identification of inactive carriers using a one-time baseline measurement of HBsAg <1000 IU/mL and HBV DNA <2000 IU/mL to predict the future incidence of liver cirrhosis, HCC, and HBsAg seroclearance (Table 3). Baseline identified inactive carrier status was significantly associated with a low risk of liver cirrhosis and HCC, and were highly predictive of HBsAg seroclearance in multivariate analyses. Compared with those defined as active CHB at baseline, those defined as inactive carriers at baseline had multivariate adjusted hazard ratios (95% confidence interval [CI]) of 0.36 ( ) (P < 0.001) and 0.36 ( ) (P < 0.001) for liver cirrhosis and HCC, as well as an adjusted rate ratio (95% CI) of 6.97 ( ) (P < 0.001) for HBsAg seroclearance. Moreover, a model including baseline-identified inactive carrier status, sex, age, and ALT levels predicted liver cirrhosis, HCC, and HBsAg seroclearance with AUROCs of 0.72, 0.79, and 0.78, respectively (Fig. 2). Given the predictive accuracy reported above, we also investigated the frequency of misclassification as inactive carriers. In carriers with worse clinical outcomes, 6 (5%) patients with cirrhosis and 2 (3%) patients with HCC were misclassified as inactive carriers, whereas among carriers with favorable outcomes (HBsAg seroclearance), 21 (7%) of patients with seroclearance were misclassified. There was no significant difference in the frequency of misclassification between carriers with worse clinical outcomes and carriers with favorable outcomes. We also investigated the predictability of inactive carrier status as defined through the traditional longterm follow-up of 18 months for the future development of HCC and cirrhosis, as well as the seroclearance of HBsAg. Only outcomes occurring after the 18-month follow-up were included for prediction (Table 4). Follow-up revealed that inactive carriers still had significantly lower hazard ratios of liver cirrhosis and HCC, with adjusted hazard ratios (95% CI) of 0.43 ( ) (P < 0.001) and 0.31 ( ) (P < 0.001), respectively. Inactive carriers also had higher rates of HBsAg seroclearance, with an adjusted rate ratio (95% CI) of 2.64 ( ) (P < 0.001). A model including baseline-identified inactive carrier FIG. 2. AUROCs. Baseline versus follow-up defined inactive carriers for prediction of cirrhosis (A), HCC (B), and HBsAg seroclearance (C). 386

7 HEPATOLOGY, Vol. 64, No. 2, 2016 LIU ET AL. TABLE 4. Incidence Rates and Hazard (rate) Ratios of Cirrhosis, HCC, and HBsAg Seroclearance After Follow-up Identification of Inactive Carriers Inactive Carrier* n (%) Controls, n (%) Cases, n (%) Person- Years Incidence Rate (per 10,000 Person-Years) Crude Hazard Ratio (95% CI) Multivariate Adjusted Hazard Ratio (95% CI) Liver Cirrhosis No 740 (50) 666 (47) 74 (70) Yes 775 (50) 743 (53) 32 (30) ( ) 0.43 ( ) HCC No 752 (49) 699 (48) 53 (78) Yes 776 (51) 761 (52) 15 (22) ( ) 0.31 ( ) HBsAg seroclearance No 750 (52) 684 (56) 66 (30) Yes 694 (48) 540 (44) 154 (70) ( ) 2.64 ( ) *Cases occurring within 18 months of study entry were not included. Adjusted for sex, age, and ALT levels. AUROC status, sex, age, and ALT levels predicted liver cirrhosis, HCC, and HBsAg seroclearance with AUROCs of 0.70, 0.79, and 0.76, respectively (Fig. 2). Discussion To our knowledge, this is the first study to externally validate the usage of a one-time measurement of serum HBsAg <1000 IU/mL and HBV DNA <2000 IU/mL to differentiate inactive carriers from active CHB among a large community-based cohort of genotype B and C carriers. In contrast to the study by Brunetto et al. (18) that used a hospital-based cohort focused on genotype D and with more limited sample size, our study used a community-based cohort, examined genotypes B and C, did not include patients with cirrhosis at baseline, and was able to examine long-term clinical outcomes such as liver cirrhosis, HCC, and HBsAg seroclearance. Our study had two main findings. First, we found that a single-point measurement of HBsAg and HBV DNA levels <1000 IU/mL and <2000 IU/mL was able to distinguish future inactive carriers with moderate accuracy. The accuracy of distinguishing future inactive carriers in our cohort was adequate, but not as accurate as the prediction reported previously, with a sensitivity and specificity of 71% and 85%, respectively, versus the 91% and 95% reported by Brunetto et al. (18) Additional analyses using combinations of different cutoffs ranging from undetectable to 20,000 IU/mL of HBV DNA and 100 IU/mL to 1000 IU/mL of HBsAg were also analyzed to determine whether better cutoffs could be identified. However, the singlepoint HBV DNA <2000 IU/mL and HBsAg <1000 IU/mL cutoff still provided the optimal sensitivity and specificity in this study. There may be several reasons for the inconsistency in predictive accuracy between the two studies. First, the inconsistency may be due to differences in viral genotype. Previous studies have shown differing baseline HBsAg levels and kinetics during antiviral therapy. (16,21,22) Moreover, the correlation coefficient for HBsAg levels and HBV DNA levels among genotype D carriers in the study by Brunetto et al. (18) was as high as 0.638, and the log HBV DNA/HBsAg ratio was only 0.25 among inactive carriers, showing a much higher correlation between HBsAg and HBV DNA levels among genotype D carriers than the low correlation between serum HBV DNA and HBsAg levels seen among HBeAg-seronegative patients in our cohort (R ). Within the REVEAL cohort, which comprised only genotype B and C patients, many individuals who were defined as inactive carriers through 18 months of follow-up did not have HBsAg levels <1000 IU/mL at baseline, even though they were later defined to be inactive carriers through repeated measurements of HBV DNA and ALT levels. Thus, these individuals were misclassified as having active CHB at baseline, resulting in a lower sensitivity. In the same way, individuals with baseline HBsAg levels <1000 IU/mL did not necessarily remain inactive carriers in the future, contributing to the lower positive predictive value as well. Thus, low levels of HBsAg did not necessarily represent low levels of HBV DNA within patients. This discrepancy may be due to the large differences in correlations, also explaining the differences in accuracy of distinguishing inactive carriers from those with active CHB between the two studies. Indeed, a previous study (10) found that lower levels of pregenomic RNA led to significantly impaired viral productivity among HBeAg-seronegative patients, resulting in a very low correlation between serum HBV 387

8 LIU ET AL. HEPATOLOGY, August 2016 DNA and intrahepatic cccdna. On the other hand, production of HBsAg was not impaired and correlated well with cccdna. This explains the lack of correlation seen between HBV DNA and HBsAg levels in our HBeAg-seronegative cohort, as well as other HBeAg-seronegative cohorts, and emphasizes the importance of HBsAg as a marker of liver infection and immune control. (10,23) Similarly, other studies have also found weak correlations between HBV DNA and HBsAg among genotype B and C carriers (24-26) while finding higher correlations among genotype A and D carriers. (27) However, these differences in correlations may be explained by differing extracellular expression of viral DNA and HBsAg between different hepatitis B virus genotypes and may also help to explain the lower accuracy of this validation study among genotype B and C carriers when compared with previously reported higher accuracy in genotype A and D patients from Europe. (28) The second major finding of this study was that inactive carriers defined by a one-time baseline measurement of HBsAg and HBV DNA <1000 IU/mL and <2000 IU/mL could also predict future incidence of liver cirrhosis, HCC, and HBsAg seroclearance. Because rate ratios of cirrhosis and HCC were similar when using both baseline, and because follow-up defined inactive carriers, this shows that the one-time carrier definition is comparable to inactive carriers defined through continuous follow-up. Furthermore, AUROCs for predicting cirrhosis, HCC, and HBsAg seroclearance using a model including baseline defined inactive carriers, sex, age, and ALT levels, were similar, and even slightly higher, than AUROCs for predicting cirrhosis, HCC, and HBsAg seroclearance using a model that included follow-up defined inactive carriers. In addition to the current study, several recent studies have already shown the ability of quantitative HBsAg levels to accurately predict HBV DNA undetectability, HBsAg seroclearance, and occurrence of liver cirrhosis and HCC. (11-15,24) Moreover, when taken with previous studies that have shown good correlation between cccdna and serum HBsAg, differences in HBsAg levels across disease phases, and a strong decline in HBsAg levels in response to immunomodulatory therapy, (10,17,29,30) the results of this study once again strongly support the role of quantitative HBsAg levels in reflecting the ability of the host s immune system to adequately control HBV infection. The results of this study also show that quantitative HBsAg provides added and complementary information in differentiating patients in different phases of disease. Although this one-time measurement cannot replace the importance of long-term monitoring, it provides important information that may affect future follow-up strategies and treatment assessment of HBeAg-seronegative patients. There are some limitations to be noted. This study used patients with two or three measurements within 18 months to define inactive carriers. Although it would be best to use the most visits possible, we could not do so, because most patients third visit occurred an average of 5.4 years after baseline. This gap in time between the second and third visits could have created room for misclassification, because we did not have detailed ALT or DNA information during this time. Therefore, we limited the analysis to patients with at least two visits within 18 months. Future studies should use more frequent follow-ups. In addition, this study consisted only of genotype B and C carriers; further studies in larger cohorts of patients with other genotypes would help to further validate the results of this study. Finally, this study did not examine long-term trends of seromarkers; future studies should investigate the ability of a onetime measurement to predict long-term trajectories of HBsAg and HBV DNA. In conclusion, this study confirms the ability of a one-time combined HBsAg and HBV DNA measurement to predict future inactive carriers, although the prediction accuracy may not be as high as it is in genotype D carriers. This single-point measurement is also able to predict future liver cirrhosis, HCC, and HBsAg seroclearance, with accuracy equal to prediction using inactive carrier status defined through traditional repeated follow-ups. This single-point strategy may provide new and complementary information useful for simplifying or tailoring management of patients with chronic hepatitis B infection. REFERENCES 1) Dienstag JL. Hepatitis B virus infection. N Engl J Med 2008; 359: ) EASL clinical practice guidelines: management of chronic hepatitis B virus infection. J Hepatol 2012;57: ) Yang HI, Lu SN, Liaw YF, You SL, Sun CA, Wang LY, et al. Hepatitis B e antigen and the risk of hepatocellular carcinoma. N Engl J Med 2002;347: ) Lok AS, McMahon BJ. Chronic hepatitis B: update HEPATOLOGY 2009;50: ) Tseng TC, Liu CJ, Chen CL, Wang CC, Su TH, Kuo SF, et al. Serum hepatitis B virus-dna levels correlate with longterm adverse outcomes in spontaneous hepatitis B e antigen seroconverters. J Infect Dis 2012;205:

9 HEPATOLOGY, Vol. 64, No. 2, 2016 LIU ET AL. 6) Chen CF, Lee WC, Yang HI, Chang HC, Jen CL, Iloeje UH, et al. Changes in serum levels of HBV DNA and alanine aminotransferase determine risk for hepatocellular carcinoma. Gastroenterology 2011;141: ) Chen CJ, Yang HI, Su J, Jen CL, You SL, Lu SN, et al. Risk of hepatocellular carcinoma across a biological gradient of serum hepatitis B virus DNA level. JAMA 2006;295: ) Iloeje UH, Yang HI, Su J, Jen CL, You SL, Chen CJ. Predicting cirrhosis risk based on the level of circulating hepatitis B viral load. Gastroenterology 2006;130: ) Liu J, Yang HI, Lee MH, Lu SN, Jen CL, Wang LY, et al. Incidence and determinants of spontaneous hepatitis B surface antigen seroclearance: a community-based follow-up study. Gastroenterology 2010;139: ) Volz T, Lutgehetmann M, Wachtler P, Jacob A, Quaas A, Murray JM, et al. Impaired intrahepatic hepatitis B virus productivity contributes to low viremia in most HBeAg-negative patients. Gastroenterology 2007;133: ) Liu J, Lee MH, Batrla-Utermann R, Jen CL, Iloeje UH, Lu SN, et al. A predictive scoring system for the seroclearance of HBsAg in HBeAg-seronegative chronic hepatitis B patients with genotype B or C infection. J Hepatol 2013;58: ) Liu J, Yang HI, Lee MH, Batrla-Utermann R, Jen CL, Lu SN, et al. Distinct seromarkers predict different milestones of chronic hepatitis B progression. HEPATOLOGY 2014;60: ) Liu J, Tseng TC, Yang HI, Lee MH, Batrla-Utermann R, Jen CL, et al. Predicting HBsAg seroclearance in HBeAg-negative chronic hepatitis B patients: external validation of a scoring system. J Infect Dis doi: /infdis/jiu ) Lee MH, Yang HI, Liu J, Batrla-Utermann R, Jen CL, Iloeje UH, et al. Prediction models of long-term cirrhosis and hepatocellular carcinoma risk in chronic hepatitis B patients: risk scores integrating host and virus profiles. HEPATOLOGY 2013;58: ) Tseng TC, Liu CJ, Yang HC, Su TH, Wang CC, Chen CL, et al. High levels of hepatitis B surface antigen increase risk of hepatocellular carcinoma in patients with low HBV load. Gastroenterology 2012;142: e3. 16) Brunetto MR, Moriconi F, Bonino F, Lau GK, Farci P, Yurdaydin C, et al. Hepatitis B virus surface antigen levels: a guide to sustained response to peginterferon alfa-2a in HBeAg-negative chronic hepatitis B. HEPATOLOGY 2009;49: ) Moucari R, Mackiewicz V, Lada O, Ripault MP, Castelnau C, Martinot-Peignoux M, et al. Early serum HBsAg drop: a strong predictor of sustained virological response to pegylated interferon alfa-2a in HBeAg-negative patients. HEPATOLOGY 2009;49: ) Brunetto MR, Oliveri F, Colombatto P, Moriconi F, Ciccorossi P, Coco B, et al. Hepatitis B surface antigen serum levels help to distinguish active from inactive hepatitis B virus genotype D carriers. Gastroenterology 2010;139: ) Martinot-Peignoux M, Lapalus M, Laouenan C, Lada O, Netto-Cardoso AC, Boyer N, et al. Prediction of disease reactivation in asymptomatic hepatitis B e antigen-negative chronic hepatitis B patients using baseline serum measurements of HBsAg and HBV-DNA. J Clin Virol 2013;58: ) Chen CJ, Iloeje UH, Yang HI. Long-term outcomes in hepatitis B: the REVEAL-HBV study. Clin Liver Dis 2007;11: ) Moucari R, Marcellin P. Quantification of hepatitis B surface antigen: a new concept for the management of chronic hepatitis B. Liver Int 2011;31(Suppl 1): ) Moucari R, Martinot-Peignoux M, Mackiewicz V, Boyer N, Ripault MP, Castelnau C, et al. Influence of genotype on hepatitis B surface antigen kinetics in hepatitis B e antigen-negative patients treated with pegylated interferon-alpha2a. Antivir Ther 2009;14: ) Thompson AJ, Nguyen T, Iser D, Ayres A, Jackson K, Littlejohn M, et al. Serum hepatitis B surface antigen and hepatitis B e antigen titers: disease phase influences correlation with viral load and intrahepatic hepatitis B virus markers. HEPATOLOGY 2010;51: ) Tseng TC, Liu CJ, Su TH, Wang CC, Chen CL, Chen PJ, et al. Serum hepatitis B surface antigen levels predict surface antigen loss in hepatitis B e antigen seroconverters. Gastroenterology 2011;141: ) Nguyen T, Thompson AJ, Bowden S, Croagh C, Bell S, Desmond PV, et al. Hepatitis B surface antigen levels during the natural history of chronic hepatitis B: a perspective on Asia. J Hepatol 2010;52: ) Su TH, Hsu CS, Chen CL, Liu CH, Huang YW, Tseng TC, et al. Serum hepatitis B surface antigen concentration correlates with HBV DNA level in patients with chronic hepatitis B. Antivir Ther 2010;15: ) Jaroszewicz J, Calle Serrano B, Wursthorn K, Deterding K, Schlue J, Raupach R, et al. Hepatitis B surface antigen (HBsAg) levels in the natural history of hepatitis B virus (HBV)-infection: a European perspective. J Hepatol 2010;52: ) Sugiyama M, Tanaka Y, Kato T, Orito E, Ito K, Acharya SK, et al. Influence of hepatitis B virus genotypes on the intra- and extracellular expression of viral DNA and antigens. HEPATOLOGY 2006;44: ) Werle-Lapostolle B, Bowden S, Locarnini S, Wursthorn K, Petersen J, Lau G, et al. Persistence of cccdna during the natural history of chronic hepatitis B and decline during adefovir dipivoxil therapy. Gastroenterology 2004;126: ) Nguyen T, Desmond P, Locarnini S. The role of quantitative hepatitis B serology in the natural history and management of chronic hepatitis B. Hepatol Int 2009;3: