DH

M

*‡

BctaawcpctsmtyavHaDcdd.(cvbH

Hpsifrserat0

CLINICAL GASTROENTEROLOGY AND HEPATOLOGY 2009;7:889–893

evelopment of Hepatocellular Carcinoma After Seroclearance ofepatitis B Surface Antigen

YRON JOHN TONG,*,‡ MICHAEL ONG NGUYEN,‡ LORI TERESE TONG,‡ and LAWRENCE MITCHELL BLATT*,§

The Pfleger Liver Institute and the Division of Digestive Diseases, David Geffen School of Medicine, University of California Los Angeles, Los Angeles, California;

The Liver Center, Huntington Medical Research Institutes, Pasadena, California; §Alios Biopharma, South San Francisco, California

awaiaott1H6

rpclcaoH

hi

hptnabiHti

caCan

See related article, Gehring AJ et al, on page682 in Gastroenterology.

ACKGROUND & AIMS: During the natural course ofhronic hepatitis B virus infection, a small proportion of pa-ients experience hepatitis B surface antigen (HBsAg) seroclear-nce. However, the long-term clinical outcomes of this processre not well established. METHODS: Thirty-five patientsith chronic hepatitis B, followed between 1976 and 2008 at a

ommunity liver clinic, experienced HBsAg seroclearance. Tenatients were Caucasian and 25 were Asian. These patientsontinued to undergo surveillance for hepatocellular carcinomahat included test for �-fetoprotein levels and abdominal ultra-ound examinations. The median follow-up time was 185

onths (range, 27– 400 months). RESULTS: During the ini-ial visit to the clinic, the median age of the patients was 41ears (range, 1.5–72 years). Eighteen patients (51.4%) were hep-titis B e antigen (HBeAg) positive 25 (71.4%) were hepatitis Birus DNA positive, and 13 (37.1%) had cirrhosis. At the time ofBsAg loss, the median age was 54 years (range, 13–77 years)

nd all were hepatitis B e antigen- as well as hepatitis B virusNA negative. During the long-term follow-up, 4 patients with

irrhosis developed hepatocellular carcinoma (HCC), which wasiscovered by ultrasound examination. Factors associated withevelopment of HCC were low baseline levels of albumin (P �

04), family histories of HBsAg positivity (P � .01) and HCCP � .04), and age of less than 50 years at the time of HBsAglearance (P � .03). CONCLUSIONS: HCC can still de-elop after HBsAg seroclearance. Thus, surveillance shoulde continued after HBsAg loss in the same manner as forBsAg positive patients.

epatitis B surface antigen (HBsAg) seroclearance is de-fined as the loss of HBsAg detectability in the blood of

atients with chronic hepatitis B. In individuals with longtanding chronic infection, disappearance of circulating HBsAgs an uncommon finding especially in areas of high endemicityor hepatitis B. In many countries in Asia where HBsAg carrierates are up to 15% of the population, the annual HBsAgeroclearance rates were reported to be 0.1% to 1.2%.1–3 In lowndemic areas such as North America and Europe, the annualates of HBsAg loss ranged from 0.1% to 2.1% per year.4 – 8 Therere few studies from the United States. In a report from Alaska,he annual HBsAg clearance rate in Eskimo HBsAg carriers was

.5% per year.9

The outcomes after HBsAg seroclearance are generally favor-ble. Recent studies have indicated that after HBsAg loss, thereas significant improvement in serum alanine aminotransfer-se levels, and liver biopsy findings in these patients showedmprovement in necroinflammation.10 –12 However, serum hep-titis B viral DNA (HBV DNA) was still detectable in 31%–71.4%f patients at the time of HBsAg seroclearance, which decreasedo 21.4% and to 14.3% after 5 and 10 years of follow-up respec-ively.13–15 Intrahepatic HBV DNA also was detected in up to00% of patients after HBsAg loss, and included both integratedBV DNA as well as covalently closed circular DNA in 37%–

3.5% of patients.10,12,16

Clinical complications after HBsAg seroclearance may occur. Ineports from Asia, 1.6%–10.9% of patients with chronic hepatitis Brogressed to cirrhosis, and 13.8% of patients with establishedirrhosis experienced hepatic decompensation.14,17,18 In both of theatter instances, progression of liver disease was associated withoexisting hepatitis C virus or hepatitis delta virus infection.14,19 Inddition, hepatocellular carcinoma (HCC) developed in 2.9%–20%f cirrhosis patients and in 1.9% of noncirrhotic patients who hadBsAg loss.3,8–12,18

In this report, we describe the characteristics of 35 chronicepatitis B patients who experienced HBsAg loss and the clin-

cal sequelae which occurred during long-term follow-up.

MethodsPatientsSince 1976, over 1200 patients with chronic hepatitis B

ave been evaluated in our liver clinic. The HBsAg positiveatients were routinely followed every 3–12 months with liverests including serum albumin, total bilirubin, aspartate ami-otransferase, alanine aminotransferase, alkaline phosphatase,nd platelet counts. Hepatitis B e antigen (HBeAg) and anti-odies to hepatitis B e antigen (anti-HBe) were measured dur-

ng the first clinic visit and every 3–5 years thereafter. TheBsAg was measured every 5–10 years if liver tests were persis-

ently normal, HBeAg was negative, anti-HBe was positive, andf the HBV DNA was persistently negative. Through the years,

Abbreviations used in this paper: Anti-HBc, antibody to hepatitis Bore antigen; Anti-HBe, antibody to hepatitis B e antigen; Anti-HBs,ntibody to hepatitis B surface antigen; Anti-HCV, antibody to hepatitisvirus; HBeAg, hepatitis B e antigen; HBsAg, hepatitis B surface

ntigen; HBV DNA, hepatitis B viral DNA; HCC, hepatocellular carci-oma; HCV RNA, hepatitis C viral RNA.

© 2009 by the AGA Institute1542-3565/09/$36.00

doi:10.1016/j.cgh.2009.04.012

tHcecs

bqtSwtt

tt�itfb

aAadpumao

ss

(H(IHD((rfc

yHHppa1aa1nahrCt

ntfcaa�

T

ASR

F

HAHABAAAP�ACM

a

T

AHAHABAAAP�P

890 TONG ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 7, No. 8

he methodologies and the sensitivities for measurement ofBV DNA were different so that the results could not be

ompared. Therefore, the HBV DNA tests are reported herein asither positive or negative. In patients who presented to thelinic between 1976 –1990, the HBV DNA was measured in seratored at �70°C as previously described.20

A family history of HBsAg positivity and presence of HCC inlood relatives were obtained at the first clinic visit. Whenueried about a history of alcohol consumption, the patients inhis analysis revealed either none or only occasional alcohol use.ome patients who continued to attend our clinic were treatedith antiviral therapy during the course of follow-up. All pa-

ients were observed for development of liver-related complica-ions or progression to HCC.21

Surveillance for HCCAbdominal ultrasound examinations and �-fetoprotein

ests were performed every 6 –12 months in noncirrhotic pa-ients, and every 3– 6 months in patients with cirrhosis. If the-fetoprotein was elevated or a lesion was detected by abdom-

nal ultrasound, then additional tests such as computerizedomography scan or magnetic resonance imaging were per-ormed. Each of the HCC cases described herein had liveriopsy confirmation of malignancy.

Statistical AnalysisCategorical data were summarized by using frequencies

nd were analyzed by �2 methods for assessment of differences.ll continuous data were descriptively summarized with meansnd standard deviations, and further analyzed for assessment ofifferences by analysis of variance (ANOVA) with post hocair-wise Student t tests. Due to the small sample size, contin-ous data were analyzed by both the parametric and nonpara-etric tests. The data reported herein are from parametric tests

nd these results were confirmed by nonparametric methodol-

able 1. Baseline Characteristics of 35 HBsAg PositivePatients Prior to HBsAg Seroclearance

ge, y a 41 (2–72)ex (male:female) 24:11ace, n (%)Asian 25 (71.4)Caucasian 10 (28.6)

amily history, n (%)HBsAg positivity 18 (51.4)Hepatocellular carcinoma 17 (48.6)

BeAg positive, n (%) 17 (48.6)nti-HBe positive, n (%) 18 (51.4)BV DNA positive, n (%) 25 (71.1)lbumin (g/dL)a 4.4 (2.7–5.1)ilirubin (mg/dL)a 0.8 (0.2–2.9)lkaline phosphatase (U/L)a 75 (26–245)spartate aminotransferase (U/L)a 43 (15–430)lanine aminotransferase (U/L)a 67 (5–565)latelet (�109/L)a 208 (70–366)-fetoprotein (ng/mL)a 3 (1.5–309)nti-HCV positive, n (%) 2 (5.7)irrhosis, n (%) 13 (37.1)onths follow-upa 185 (27–400)

Median (range).

gy in each instance. Analyses were conducted by using SAS a

oftware version 9.1 (SAS Institute Inc, Cary, NC). Statisticalignificance was defined as 2-sided P values � .05.

ResultsDuring a median follow-up period of 185 months

range, 27– 400 months; mean 205.6 months � 97 SD), 35BsAg positive patients had HBsAg seroclearance. Twenty-four

69%) were males and 25 (71%) were Asian (5 born in the USA).n the Asian patients, 17 (68%) had family members who wereBsAg positive and 10 (40%) had a family history of HCC.uring the first visit to our clinic, the median age was 41 years

range, 1.5–72 yrs), 18 (51.4%) were HBeAg positive, and 2571.4%) were HBV DNA positive. The baseline laboratory testesults are shown in Table 1. Two (5.7%) patients were coin-ected with hepatitis C virus. Thirteen (37.1%) patients hadirrhosis.

At the time of HBsAg clearance, the median age was 54ears (range, 13–77 years). Eighteen patients who wereBeAg positive during the first visit had all seroconverted toBeAg negative. Thus, at the time of HBsAg loss, the 35atients were anti-HBe positive, HBV DNA negative, and hadersistently normal levels of serum alanine aminotransferasend �-fetoprotein (Table 2). During the course of follow-up,4 patients received antiviral therapy (7 with interferonlone, 5 with interferon plus lamivudine (4) or adefovir (1),nd 1 each with lamivudine or adefovir. Prior to treatment,1 of the 14 patients were HBeAg positive, and 3 were HBeAgegative with positive HBV DNA. Two of the latter patientslso were positive for antibodies to hepatitis C virus andepatitis C viral RNA, and both were treated with interferon,ibavirin, and lamivudine resulting in clearance of hepatitis

and HBsAg loss. Six of the patients who received antiviralherapy had cirrhosis.

After HBsAg seroclearance, all 35 patients had persistentlyormal liver tests and none had seroreversion of HBsAg. Twen-y-six patients subsequently developed anti-HBs. During furtherollow-up, 31 patients remained clinically stable. However, withontinued routine surveillance for HCC, 4 patients had appear-nce of liver lesions on abdominal ultrasound examinationsnd each had liver biopsy confirmation of HCC (Table 3). The-fetoprotein was normal in 3 of the 4 patients who developed

able 2. Characteristics of 35 Patients at the Time ofHBsAg Seroclearance

ge, y a 54 (13–77)BeAg positive, n (%) 0 (0)nti-HBe positive, n (%) 35 (100)BV DNA positive, n (%) 0 (0)lbumin (g/dL)a 4.6 (2.8–4.8)ilirubin (mg/dL)a 0.7 (0.4–1.4)lkaline phosphatase (U/L)a 64 (30–235)spartate aminotransferase (U/L)a 22 (13–34)lanine aminotransferase (U/L)a 22 (9–59)latelet (�109/L)a 211 (55–317)-fetoprotein (ng/mL)a 2.3 (1.2–5)rior antiviral treatment, n (%) 14 (40)Interferon 7Interferon � lamivudine (4) or adefovir (1) 5Lamivudine or adefovir 2

Median (range).

HaTycHpdwspoaH.ho

d

2aohrDHeawt

pLpaPr

aobtChfoa

svtitttimc

T

NF

F

A

B

T

ASRFHHAAIHHRC

Aa d ex

August 2009 HEPATOCELLULAR CARCINOMA AFTER HBsAg CLEARANCE 891

CC. The time interval between HBsAg seroclearance to HCCppearance in these 4 patients was 15, 16, 20, and 24 months.he median age at HCC detection was 69 years (range, 55–73ears). The 4 patients who developed HCC were male, all hadirrhosis, and 3 had family members with HCC. At the time ofCC discovery, the HBeAg was negative and anti-HBe wasositive in all 4 patients. In 1 patient, the HBV DNA becameetectable when the HCC appeared. In 3 patients, the tumorsere treated with transcatheter arterial chemoembolization. A

econd HCC was subsequently detected in 1 of these latteratients 3 years later, which also was similarly treated. Onether patient had hepatic resection of a left lobe HCC. Factorsssociated with HCC development by univariate analysis afterBsAg seroclearance were low baseline serum albumin (P �

04), a family history of HBsAg positivity (P � .01), a familyistory of HCC (P � .04), and age less than 50 years at the timef HBsAg seroclearance (P � .03) (Table 4).

The data from 4 patients with HBsAg seroclearance whoeveloped HCC are illustrated in Figure 1.

DiscussionHBsAg seroclearance is an unusual event. In a report of

45 HBsAg positive asymptomatic carriers in Taiwan, the meannnual rate of HBsAg loss was only 1.2% per year.3 In this study,lder age at presentation and sustained clinical remission ofepatitis were associated with HBsAg seroclearance. In anothereport from Korea, HBeAg negativity and undetectable HBVNA at enrollment were associated with subsequent loss ofBsAg.22 However, in a study from Europe, HBeAg positivity at

ntry was a significant predictive factor for HBsAg seroclear-nce.8 Another report from Hong Kong indicated that patientsith genotype B more often had HBsAg seroclearance than

hose with genotype C.10

Another important factor which may influence HBsAg loss wasrevious treatment with interferon or oral antiviral agents.8,23

oss of HBsAg after interferon therapy in chronic hepatitis Batients was reported to be 3%–7.8%.22,24 HBsAg seroclearancefter treatment with lamivudine or adefovir was less than 1%.rior to HBsAg clearance, 14 of 35 (40%) of our patients had

able 3. Characteristics of 4 Patients Who Developed Hepat

A

ge at HCC diagnosis, y 55ex Maleace Asianamily history HCC BrotherBeAg/anti-HBe at HCC diagnosis �/�BV DNA at HCC diagnosis Negativenti-HBs �FP at HCC diagnosis (ng/mL) 2.8

nterval between ultrasound examinationa 4 monthsCC size (cm) 2.5CC treatment Resectionecurrent HCC Nolinical status Alive

FP, alphafetoprotein; TACE, transcatheter arterial chemoembolizatioTime from the last negative ultrasound to the first positive ultrasoun

eceived a course of antiviral therapy consisting of interferon

lone, interferon plus lamivudine or adefovir, and lamivudiner adefovir alone. Eleven of the patients were HBeAg positive ataseline and seroconverted to HBeAg negative after antiviralherapy. Also, both patients who were coinfected with hepatitis

virus had a sustained virologic response with clearance ofepatitis C as well as HBsAg loss after treatment with inter-

eron, ribavirin, and lamivudine. The remaining 21 of 35 (60%)f our patients who had HBsAg seroclearance did not receiventiviral therapy and therefore had spontaneous HBsAg loss.

In the majority of instances, the outcomes after HBsAgeroclearance are favorable. The serum alanine aminotransferasealues remained normal and antibody to hepatitis B core antigeniters gradually decreased.11,15 Improvement in necroinflammationn liver biopsies after HBsAg loss was also noted.10,11,12 However,he serum HBV DNA was detectable in 31%–98% of patients athe time of HBsAg seroclearance,10,11,14,16 and remained so in upo 14% of patients even 10 years after HBsAg loss.15 Also,ntrahepatic HBV DNA was detected in liver tissue in the

ajority of patients and may be in the form of covalently closedircular DNA.10,12,17

able 4. Analysis of Factors Associated With HepatocellularCarcinoma in Patients With HBsAg Seroclearance

Factor HCC No HCC P value

umber of patients 4 31amily history .01HBsAg �

Yes 3/4 13/31No 1/4 18/31

amily history .04HCC

Yes 3/4 7/31No 1/4 24/31

ge at HBsAgseroconversion

.03

�50 years 14 0�50 years 4 17

aseline albumin, 3.75 � 0.88 4.31 � 0.45 .04

ular Carcinoma After HBsAg Seroclearance

Patient

B C D

67 69 73Male Male MaleAsian Caucasian AsianBrother None Brother; father

�/� �/� �/�Positive Negative Negative

� � �4 342 3.5

7 months 6 months 3 months8 1.7 5.5

TACE TACE TACENo No YesDead Alive Alive

amination.

ocell

n.

gm/dL � SD

rhdcpwdh

oaImdphhdsgtHohtodwpHhssd

n

lmsmpcfOsmvfmpmmctn

ltwpiflp

sHetcHv

892 TONG ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 7, No. 8

Clinical complications following HBsAg seroclearance areare. In patients with HBsAg loss, progression from chronicepatitis to cirrhosis was reported to be 1.6%–5.6% and hepaticecompensation occurred in 2.5% of patients with pre-existingirrhosis.3,14,19 However, many of these latter patients had su-erimposed hepatitis C virus or hepatitis delta virus infectionhich may have contributed to the progression of their liverisease even after HBsAg seroclearance.14,20 None of our patientsad liver-related complications after HBsAg seroclearance.

The most serious clinical sequelae after HBsAg loss is devel-pment of HCC. The time interval between HBsAg seroclear-nce and HCC detection ranged from 9 months to 9 years.8 –14,18

n the latter reports, up to one third of patients developed HCCore than 5 years after HBsAg loss. The annual rates of HCC

evelopment in patients without and with cirrhosis were re-orted to be 0.21% and 0.45% to 6.2% respectively, and wereigher in patients who were coinfected with hepatitis C orepatitis delta virus.8,10,12,14,18 Significant factors associated withevelopment of HCC included a history of perinatal transmis-ion, greater than 50 years duration of HBsAg positivity, age ofreater than 50 years at the time of HBsAg seroclearance, andhe presence of cirrhosis.10,12 The mean age of our patients atCC development was 66 years and all had cirrhosis at the time

f the initial visit. Also, as recently reported, our patients whoad HBsAg seroclearance after the age of 50 years were more apto develop HCC than those who had HBsAg loss under the agef 50 years.25 Reported herein, 3 of the Asian patients whoeveloped HCC had a family history of HCC and this findingas an independent risk factor for HCC development in ouratients after HBsAg loss. Thus, the 7 additional patients withBsAg seroclearance who had family members with HCC are atigh risk for developing HCC themselves and will require closeurveillance for HCC. Further studies are needed to identifypecific genetic factors which may predispose these patients forevelopment of HCC.

Because HBsAg seroclearance is an uncommon event, a large

umber of chronic hepatitis B patients must be observed for a H

ong period of time. Thus, the majority of such reports areainly from Asian countries where hepatitis B is endemic. A

tudy from Europe of 32 patients with HBsAg loss required aulticenter study which involved at least 7 countries.8 The

atients reported herein are from a community-based liverlinic in Pasadena, California, which has been a referral sourceor over 1200 HBsAg positive patients for close to 3 decades.ur cohort of 35 patients who had HBsAg seroclearance con-

isted of both Asian and Caucasian patients. The Asian patientsost likely contracted hepatitis B by perinatal transmission or

ia intrafamilial spread of hepatitis B virus since 68% of theiramily members also were HBsAg positive and 40% had family

embers with HCC. This finding is not only observed in Asianatients with HBsAg seroclearance but also in those who re-ain HBsAg positive.26 In contrast, our Caucasian patientsost likely contracted hepatitis B as adults either by sexual

ontact, blood transfusion, or by needle stick exposure. None ofhe Caucasian patients had family histories of HBsAg positivityor had family members with HCC.

In a recent report, the serum albumin levels were consistentlyower throughout the course of follow-up in our HBsAg posi-ive patients who later died from liver-related complications orho developed HCC.20 A low baseline albumin level in ouratients who progressed to HCC after HBsAg loss was an

ndependent risk factor for HCC development, and was a re-ection of the established liver damage which was alreadyresent when these patients first presented to our clinic.

In summary, chronic hepatitis B patients who have HBsAgeroclearance still may develop HCC. Thus, surveillance forCC must be continued indefinitely in patients who have

stablished cirrhosis at the time of HBsAg loss, and especially inhose with a family history of HCC, and those who had HBsAglearance after the age of 50 years. Early detection of a smallCC may result in successful treatments which improve sur-

ival as demonstrated in our patients who developed HCC after

Figure 1. (A) Patient A: a 55-year-oldAsian male. (B) Patient B: a 67-year-old Asian male. (C) Patient C: a 69-year-old Caucasian male. (D) PatientD: a 73-year-old Asian male. AFP, al-phafetoprotein; ALT, alanine amino-transferase.

BsAg seroclearance.

1

1

1

1

1

1

1

1

1

1

2

2

2

2

2

2

2

R

Sm

A

t

C

August 2009 HEPATOCELLULAR CARCINOMA AFTER HBsAg CLEARANCE 893

References

1. Wu TT, Hsu HC, Chen DS, et al. Clearance of hepatitis B surfaceantigen (HBsAg) after surgical resection of hepatocellular carci-noma. J Hepatol 1987;4:45–51.

2. Liaw YF, Sheen IS, Chen TJ, et al. Incidence, determinants andsignificance of delayed clearance of serum HBsAg in chronichepatitis B virus infection: a prospective study. Hepatology 1991;13:627–631.

3. Chu CM, Liaw YF. HBsAg seroclearance in asymptomatic carriersof high endemic areas: appreciably high rates during a long-termfollow-up. Hepatology 2007;45:1187–1192.

4. Sampliner RF, Hamilton FA, Iseri OA, et al. The liver histology andfrequency of clearance of the hepatitis B surface antigen (HBsAg)in chronic carriers. Am J Med Sci 1979;277:17–22.

5. Feinman SV, Berris B, Cooter N, et al. Results of long-termprospective study of the hepatitis B surface antigen (HBsAg)carrier state. Hepatogastroenterology 1982;29:58–61.

6. de Franchis R, Meucci G, Vecchi M, et al. The natural history ofasymptomatic hepatitis B surface antigen carriers. Ann InternMed 1993;118:191–194.

7. Da Silva LC, Madruga CL, Carrilho FJ, et al. Spontaneous hepatitis Bsurface antigen clearance in a long-term follow-up study of patientswith chronic type B hepatitis. Lack of correlation with hepatitis C andD virus superinfection. J Gastroentrol 1996;31:696–701.

8. Fattovich G, Giustina G, Sanchez-Tapias J, et al. Delayed clear-ance of serum HBsAg in compensated cirrhosis B: relation tointerferon alpha therapy and disease prognosis. Am J Gastroen-terol 1998;93:896–900.

9. McMahon BJ, Holck P, Bulkow L, et al. Serologic and clinicaloutcomes of 1536 Alaska native chronically infected with hepa-titis B virus. Ann Intern Med 2001;135:759–768.

0. Yuen MF, Wong DK, Sablon E, et al. HBsAg seroclearance inchronic hepatitis B in the Chinese: virology, histological andclinical aspects. Hepatology 2004;39:1694–1701.

1. Arase Y, Ikeda K, Suzuki Y, et al. Long-term outcome afterhepatitis B surface antigen seroclearance in patients withchronic hepatitis B. Am J Med 2006;119:71–16.

2. Ahn SH, Park YN, Park JY, et al. Long-term clinical and histolog-ical outcomes in patients with spontaneous hepatitis B surfaceantigen seroclearance. J Hepatol 2005;42:188–194.

3. Adachi H, Kaneko S, Matsushita E, et al. Clearance of HBsAg inseven patients with chronic hepatitis B. Hepatology 1992;16:1334–1337.

4. Huo TI, Wu JC, Lee PC, et al. Seroclearance of hepatitis B surfaceantigen in chronic carriers does not necessarily imply a goodprognosis. Hepatology 1999;28:231–236.

5. Arase Y, Suzuki F, Suzuki Y, et al. Long-term presence of HBV inthe sera of chronic hepatitis B patients with HBsAg seroclear-

ance. Intervirology 2007;50:161–165.

6. Loriot M, Marcellin P, Walker F, et al. Persistence of hepatitis Bvirus DNA in serum and liver from patients with chronic hepatitisB after loss of HBsAg. J Hepatol 1997;27:251–258.

7. Pollicino T, Squadrito G, Cerenzia G, et al. Hepatitis B virusmaintains its pro-oncogenic properties in the case of occult HBVinfection. Gastroenterology 2004;126:102–110.

8. Chen CM, Karayiannis P, Flower MJ, et al. Natural history of chronichepatitis B virus infection in Taiwan: studies of hepatitis B virus DNAin serum. Hepatology 1985;5:431–434.

9. Liaw YF, Chen YC, Sheen IS, et al. Impact of acute hepatitis Cvirus superinfection in patients with chronic hepatitis B virusinfection. Gastroenterology 2004;126:1024–1029.

0. Tong MJ, Blatt LM, Kao J-H, et al. Precore/basal core promotermutants and HBV DNA levels as predictors for liver deathsand hepatocellular carcinoma. World J Gastroenterol 2006;12:6620–6626.

1. Tong MJ, Hsien C, Hsu L, et al. Treatment recommendation forchronic hepatitis B: an evaluation of current guidelines based ona natural history in the United States. Hepatology 2008;48:1070–1078.

2. Kim JH, Lee JH, Park SJ, et al. Factors associated with naturalseroclearance of hepatitis B surface antigen and prognosis afterseroclearance: a prospective follow-up study. Hepatogastroenter-ology 2008;55:578–581.

3. Ruiz-Moreno M, Otero M, Millan A. Clinical and histological out-come after hepatitis B e antigen to antibody seroconversion inchildren with chronic hepatitis B. Hepatology 1999;29:572–575.

4. Lok ASF, McMahon BJ. Chronic hepatitis B. Hepatology 2007;45:507–539.

5. Yuen MF, Wong DKH, Fund J, et al. HBsAg seroclearance in chronichepatitis B in Asian patients: replicative level and risk of hepatocel-lular carcinoma. Gastroenterology 2008;135:1192–1199.

6. Tong MJ, Weiner JM, Ashcavai MW, et al. A comparative study ofhepatitis B viral markers in the family members of Asian and non-Asian patients with hepatitis B surface antigen-positive hepatocel-lular carcinoma and with chronic hepatitis B infection. J Infect Dis1979;140:506–512.

eprint requestsAddress requests for reprints to: Dr Myron J. Tong, Liver Center, 660

outh Fair Oaks Avenue, Pasadena, California 91105. e-mail:yrontong@hmri.org; fax: (626) 397-5827.

cknowledgmentsThe authors thank Carlos Hsien and Hai-En Sun for technical assis-

ance during the preparation of this manuscript.

onflicts of interest

The authors disclose no conflicts.