The Long-Term Course of Chronic Hepatitis B

Sana	29.11.2020
Hajmi	101.6 Kb.
	#154875

Bog'liq
hep.510300109

ITO

ARCO

RESTE

ACONO

ALOGERO

AMMA

` ,

LESSANDRA

ACCARO

ARCO

IUNTA

IUSEPPE

ARTORANA

ATRIZIA

USCHI

IERO

L. A

LMASIO

AND

NTONIO

RAXI

1

The aim of this study was to assess the long-term

outcome in hepatitis B virus (HBV)-infected patients accord-

ing to HBV, hepatitis C virus (HCV), and hepatitis D virus

(HDV) replication, focusing on survival, liver failure, and

hepatocellular carcinoma (HCC). A cohort of 302 hepatitis

B surface antigen (HBsAg)-positive subjects (mean age,

34

ⴞ 15.3 years; male/female 214/88; 39 subjects under 14

years) with biopsy-proven chronic hepatitis (86 with cirrho-

sis) was prospectively assessed, with a median follow-up of

94

ⴞ 37.6 months. One hundred nine patients received

interferon alfa (IFN). At baseline, 86 subjects (28.5%) were

hepatitis B e antigen (HBeAg)-positive (wild-type HBV), 80

(26.5%) were HBeAg-negative, HBV-DNA–positive, 76

(25.2%) had HDV infection, 43 (14.2%) had dual HBV/HCV

infection, and 17 (5.6%) were negative for HBV-DNA,

anti-HCV, and anti-HDV. During follow-up, decompensation

of disease occurred in 46 subjects: 8 developed HCC, 36

developed ascites, and 2 developed jaundice. Five patients

underwent transplantation. Thirty-five subjects died: 33 of

hepatic and 2 of nonhepatic causes. Overall mortality was

5.2-fold that of the general population (95% CI: 3.6-7.3; 35

deaths observed, 6.7 expected; P F .0001). By Cox regres-

sion analysis, survival was independently predicted by

young age, absence of cirrhosis at baseline, and sustained

normalization of aminotransferases during follow-up. Sur-

vival without decompensation was independently predicted

by the same factors and by IFN treatment. Chronic hepatitis

B infection increases mortality in comparison with the

general population in our area regardless of specific virologi-

cal profiles at presentation. Presence of cirrhosis and

persistent necroinflammation markedly increase the risk of

death. (H

EPATOLOGY

1999;30:257-264.)

Chronic infection with hepatitis B virus (HBV) can progress

to cirrhosis, liver failure, and hepatocellular carcinoma (HCC),

and is a major cause of mortality.

1-3

Estimates on the rate of

progression are contradictory, making the proposed Mark-

ovian model

of the long-term course of chronic hepatitis B

unreliable. A relevant confounding effect stems from the

inclusion in some earlier studies

of subjects with human

immunodeficiency virus (HIV) infection, a known modifier

of immune response to HBV and a cause of reduced life

expectancy. Because interferon alfa (IFN) is now in wide-

spread use as the first-line therapy for chronic hepatitis B,

6,7

no additional prospective long-term studies on the course of

untreated disease are feasible. On the other hand, long-term

prospective cohort studies to evaluate the benefits of IFN

therapy on true end-points, i.e., prevention of cirrhosis, liver

failure, HCC, and death, are difficult to perform because of

the prolonged and slow course of the disease. Two long-term

cohort studies from Western countries

8,9

have shown that

clearance of HBV-replicative markers in patients undergoing

IFN therapy is frequent and long-lasting. These results are

less frequent in Chinese patients.

To the present, a long-

term benefit of IFN treatment in patients with chronic

hepatitis who are hepatitis B e antigen (HBeAg)-positive has

been supported by Wong et al.,

by the use of a mathematical

simulation model, and subsequently confirmed by Niederau

in a cohort of 103 treated patients. Data on the long-term

benefit of IFN in subjects with anti-HBe–positive, HBV-DNA–

positive chronic hepatitis B (the so-called HBeAg minus

mutant) or in HBV carriers coinfected with other viruses

(hepatitis D virus [HDV], hepatitis C virus [HCV]) are still

unavailable.

Since the early 1980s, we have followed a large cohort of

patients with chronic HBV infection to assess the natural

course of liver disease and their ultimate outcome. This

analysis aims to estimate the link between ‘‘surrogate’’ mark-

ers of disease (e.g., indexes of viral replication, of necroinflam-

matory activity, and of disease stage) to ‘‘hard’’ end-points

(clinically relevant outcomes such as decompensation, HCC,

or death). The wide availability, since 1987, of IFN as an

antiviral treatment for HBV gave us a chance to estimate,

outside the somewhat-limited context of randomized, con-

trolled trials, whether antiviral therapy would alter in the

long-term the course of liver disease.

PATIENTS AND METHODS

Patients.

All newly diagnosed consecutive hepatitis B surface

antigen (HBsAg)-positive subjects, either self-referred or referred by

attending physicians, observed between January 1982 and Decem-

ber 1991 in our Liver Clinic, which serves a tertiary referral

function, were considered for the study. They were eligible if they

were HBsAg-positive, had a persistent elevation of alanine transami-

nase (ALT) levels (at least two times the upper normal limits during

the previous 6 months), and had chronic hepatitis with or without

cirrhosis on liver biopsy. Patients with antibodies against human

immunodeficiency virus (HIV); a history of renal dialysis, organ

Abbreviations: HBV, hepatitis B virus; HCC, hepatocellular carcinoma; HIV, human

immunodeficiency virus; IFN, interferon alfa; HBeAg, hepatitis B e antigen; HDV,

hepatitis D virus; HCV, hepatitis C virus; HBsAg, hepatitis B surface antigen; ALT,

alanine transaminase; PCR, polymerase chain reaction; SMR, standardized mortality

ratio.

From the

Cattedra di Medicina Interna, Istituto di Clinica Medica I, Universita` di

Palermo, Italy; and

ISMEDA, Consiglio Nazionale delle Ricerche, Palermo, Italy.

Received August 28, 1998; accepted April 1, 1999.

Supported in part by grants from the Ministero della Universita` e della Ricerca

Scientifica e Tecnologica (60%).

Address reprint requests to: Prof. Antonio Craxı`, Cattedra di Medicina Interna,

Istituto di Clinica Medica I, Piazza delle Cliniche 2, 90127, Palermo, Italy. E-mail:

craxanto@unipa.it; fax 39-091-6552156.

1999 by the American Association for the Study of Liver Diseases.

0270-9139/99/3001-0034$3.00/0

257

transplantation, and those with advanced cirrhosis (Child-Pugh

class B or C), HCC; or a history of portal hypertensive bleeding were

excluded. None of the patients had received steroids, or immunosup-

pressive or antiviral drugs before entry into the study. Alcohol abuse

was defined as an habitual intake of more than 60 g/d for at least 5

years.

Three hundred two patients (59%) of 512 HBsAg-positive sub-

jects observed during the accrual period met the eligibility criteria

and were enrolled. Most patients not fulfilling inclusion criteria had

advanced liver disease (25%) or HIV infection (4%), while a

minority (12%) were chronic HBsAg carriers without clinical

evidence of liver disease. Before entry into the study, all patients

underwent a full clinical evaluation, with biochemical and virologi-

cal tests, ultrasonography, and a percutaneous liver biopsy.

Virological Assessment.

Sera taken at the time of biopsy and at each

follow-up visit were stored at

⫺30°C. Markers of HBV infection

(HBsAg, HBeAg, and anti-HBe antibodies), HDV infection (IgG and

IgM anti-HDV antibodies), and HIV infection (anti-HIV1/2 antibod-

ies) were tested by commercial enzyme immunoassays (Abbott

Laboratories, North Chicago, IL). Serum HBV DNA was detected by

dot-blot hybridization with standard methods,

with a sensitivity

level above 0.5 pg HBV DNA/mL. HCV antibodies were tested on

stored sera by second-generation enzyme immunoassay and con-

firmed by immunoblotting (ORTHO EIA2 and RIBA2, Ortho Diag-

nostics, Raritan, NJ). HCV RNA was detected in serum by polymer-

ase chain reaction (PCR), using nested primers from the 5

noncoding region of HCV.

The detection limit was about 10

to 10

genome equivalents/mL. Intrahepatic expression of HDAg was

assessed by immunofluorescence on frozen liver section using

human polyclonal fluorescein isothiocyanate–conjugated anti-HD.

Histological Evaluation.

Liver biopsies were taken before entry into

the study by Menghini needle (Hepafix 16 mm, Braun Melsungen,

Germany). Only samples at least 25 mm long after fixation were

considered suitable for diagnosis. All specimens were read for

diagnostic purposes by a pathologist according to standard criteria.

14

Outcomes.

Follow-up time was defined as the number of months

from liver biopsy to clinical events, death, or the last contact with

the patient. We included only newly diagnosed cases and calculated

the survival rate after the histological diagnosis for all subjects to get

a standardized level of uniformity of the starting point for the

follow-up of all patients. Follow-up data were obtained through

periodic controls at our Clinic. Clinical examination and routine

laboratory investigations were performed monthly during IFN

treatment and every 6 months in all other cases. Ultrasonography

and

␣-fetoprotein was repeated every 6 months in cirrhotic patients

and yearly in the remaining patients.

The primary outcome was survival. For the calculation of survival

times, liver transplantation was combined with death. The second-

ary outcome was event-free survival, i.e., survival without disease

decompensation. The latter included any of the following: death,

liver transplantation, HCC, ascites, jaundice, encephalopathy, and

portal hypertensive bleeding.

Statistical Evaluation.

Continuous variables were expressed as

mean

⫾ SD. The ␹

and Student’s t test were performed as

appropriate, all P values being two-tailed. A mortality analysis was

performed using the person-years method. For the purpose of this

analysis, a tabulation of patients at risk was constructed in terms of

age at diagnosis and interval from diagnosis. Each year that a patient

survives contributes 1 patient-year at a specific age in the tabulation.

The number of expected deaths was obtained by multiplying each

cell of the table of the person-years at risk by the reported age- and

sex-specific mortality rate for Sicily for the year 1992 (ISTAT data).

The standardized mortality ratio (SMR) was calculated by dividing

the observed number of deaths by the expected number of deaths.

The exact 95% CIs for the SMR were estimated by Poisson’s

distribution. Changes in mortality of the study cohort were regarded

as significant if the 95% CIs of the SMR did not include unity.

The Kaplan-Meier method was used to estimate the length of

survival and of survival without clinical events.

The following

variables were considered for univariate analysis: age, sex, alcohol

abuse, baseline ALT, cirrhosis, IFN treatment, baseline virological

class, and sustained ALT normalization during follow-up. Variables

with a P value

⬍ .05 at univariate were included in the final

multivariate model.

The Cox proportional hazard model

was used to assess survival

and event-free survival in a multiple regression analysis. All analyses

were conducted with the Statistical Analysis System, version 6.08.

For the Cox regression, the PHREG procedure was used. All P values

were two-tailed, and all the CIs were 95%. We verified that the most

important issue of Cox’s model, i.e., the assumption of proportional

hazard, was not violated. The proportional assumption was also

checked with analytical and graphical methods. The log minus log

plot (i.e., the logarithm of the cumulative hazard function) was used

to check the proportionality assumption of the model. The vertical

equidistance between the curves for all the variables shows that the

assumption does not seem to be violated. Moreover, we performed a

test of trend in the hazard ratio by adding into the model a new

variable, which represents the interaction effect between the prognos-

tic variable and the follow-up time. Because this variable was not

statistically significant, we can conclude that there is no trend

(increase or decrease) over time in the hazard ratio. To estimate

expected survival time for a hypothetical patient with a combination

of prognostic factors, the estimate survival function was com-

puted.

18

RESULTS

Features of Patients at Baseline.

The 302 patients (88 women

and 214 men) had a mean age of 34

⫾ 15.3 years (range, 4-72

years). Thirty-nine subjects were under 14 years of age.

Twenty-eight (9.3%) had a history of heavy drinking. Patients

were assigned to one of six virological classes according to

their baseline markers of HBV, HDV, and HCV replication

(Table 1). Wild-type HBV infection (HBeAg-positive, HBV-

DNA–positive) was detected in 86 patients (28.5%), whereas

80 patients (26.5%) were presumably infected by a predomi-

nance of the HBeAg minus mutant (HBeAg-negative, HBV-

DNA–positive). Patients under 14 years of age had, by

comparison with adults, a higher rate of HBeAg positivity

⬍ .001). Seventy-six patients (25.2%) had an HDV

infection, as shown by IgG and IgM anti-HDV positivity and

intrahepatic expression of HDVAg. In all anti-HDV–positive

patients, no HBV replication was detectable by HBV-DNA

dot-blot testing. HCV infection was found in 43 subjects

(14%), all anti-HCV– and HCV-RNA–positive. Twenty-three

of them were HCV-RNA–positive, HBV-DNA–negative (prob-

able HCV superinfection of HBV carriers), while 20 were both

HCV-RNA– and HBV-DNA–positive (HBV/HCV active coin-

fection). Seventeen subjects (5.6%) were negative for HBV-

DNA, anti-HCV, and anti-HDV antibodies, thus showing no

evidence of viral replication. Four of 17 had a history of

alcohol abuse. Histological evidence of cirrhosis was found at

baseline in 86 patients (28.5%). Cirrhosis was significantly

more common (Table 1) among subjects with HDV infection

⬍ .05) and with HBV/HCV active coinfection (P ⬍ .05).

One hundred nine patients received IFN treatment. They

were fully comparable with untreated subjects for all the

baseline clinical features, except ALT levels (Table 2), and for

having presented at the Liver Clinic during later years of

recruitment.

258

DI MARCO ET AL.

EPATOLOGY

July 1999

Follow-up.

The mean length of follow-up after the initial

evaluation was 94

⫾ 37.6 months (range, 6-180 months) for

the entire group. Twenty-nine subjects were lost before the

first follow-up visit, and are counted as dropouts. Their main

baseline features did not differ significantly from the whole

series of patients. During follow-up, 2 of the 109 treated

patients (1.8%) and 11 of the 193 untreated patients (5.6%)

cleared HBsAg from blood (P

⫽ not significant). None of

these 13 patients died or developed clinical complications.

ALT levels became sustainedly normal in 88 patients (29.1%)

without a significant difference between IFN-treated (26.6%)

and untreated (30.6%) cases. Sixty-four (74.4%) of the 86

subjects (35 IFN-treated) infected with wild-type HBV cleared

HBeAg during follow-up. Among the 64 subjects clearing

HBeAg, only 45 (52.3%) had a sustained ALT normalization.

HBV DNA was eventually cleared from serum in 74 of 186

patients (40%) who were HBV-DNA–positive at baseline (30

of 71 IFN-treated and 44 of 115 untreated).

Disease decompensation occurred in 46 patients: 8 devel-

oped HCC, 36 developed ascites, and 2 developed jaundice.

Five patients who decompensated underwent liver transplan-

tation (1 for HCC and 4 for chronic liver failure). Thirty-five

subjects died, 33 of hepatic causes and 2 of nonhepatic causes

(1 from trauma at 25 months, and another from lung cancer

at 28 months). Liver disease did not decompensate in patients

who did not have cirrhosis at baseline.

A fivefold excess mortality (SMR: 5.25; 95% CI: 3.66-7.30)

in the total cohort of patients with chronic hepatitis B as

compared with a matched general population was observed

(Table 3). Table 4 shows the SMRs of patients with chronic

hepatitis B according to baseline liver histology, virological

classes, and IFN treatment. Chronic HBV infection signifi-

cantly increased mortality both in cirrhotic patients (SMR:

18.6; 95% CI: 12-27) and in patients without cirrhosis (SMR:

6.20; 95% CI: 2.98-11.4). SMRs showed that the worse risk of

death occurred in patients with dual HBV/HCV-replicative

infection, followed by HDV infection, wild-type HBV, and

HBeAg minus mutant HBV.

The cumulative survival (until liver transplantation or

death) and the cumulative survival without disease decompen-

sation are shown in Figs. 1 and 2. Survival (Fig. 1) and

event-free survival (Fig. 2) were significantly longer in

patients with sustained normalization ALT during follow-up

versus all others (P

⬍ .01 and ⬍ .001, respectively ), and in

patients without cirrhosis at baseline in comparison with

cirrhotic patients (P

⬍ .0001 and P ⬍ .0001, respectively).

The risk of death and of disease decompensation were lower

T

ABLE

1. Baseline and Follow-up Features of Patients According to Etiology of Hepatitis

Wild-Type

HBV (HBeAg-positive,

HBV-DNA–positive)

HBeAg Minus

Mutant HBV

(HBeAg-negative,

HBV-DNA–positive)

HDV Infection

(anti-HDV–positive,

HBV-DNA–negative)

HCV Infection

(HCV-RNA–positive,

HBV-DNA–negative)

Dual HBV/HCV-

Replicative Infection

(HCV-RNA–positive,

HBV-DNA–positive)

All

Replicative

Markers

Negative

Patients (n)

80

76

Mean age (yr)

24.1

⫾ 16.8

40.3

⫾ 11.9

33.1

⫾ 12.6

⫾ 10.7

38.6

⫾ 13.6

42.2

⫾ 10.7

Sex (M/F)

57/29

62/18

53/23

13/10

13/7

16/1

Alcohol abuse (n)

9

9

ALT (UI/L)

87.7

⫾ 64.4

100.9

⫾ 81

108.7

⫾ 89

107.3

⫾ 67.4

77.4

⫾ 40.9

⫾ 48.3

Platelets (

⫻1,000/mm

)

232.7

⫾ 71.4

190.8

⫾ 52.6

175

⫾ 69.9

170.9

⫾ 61.9

140.7

⫾ 40.4

203.7

⫾ 56.1

Liver biopsy (n)

Chronic hepatitis

Mild

Moderate/severe

Cirrhosis

23

28

IFN-treated patients (n)

Mean total IFN dose,

560

678

853

407

414

—

range (MU)

72-1440

216-1944

108-2880

112-900

144-1035

Mean follow-up (mo)

⫾ 38.4

86.7

⫾ 34.1

93.3

⫾ 41.8

106.1

⫾ 33.3

82.7

⫾ 33.2

112.5

⫾ 35

Events (n)

Death (n)

6

14

Clearance of HBsAg (n)

Clearance of HBeAg (n)

—

—

—

Clearance of HBV DNA (n)

—

Long-term ALT normalization (n)

22

9

NOTE. Plus-minus values are means

⫾ SD. Normal values: ALT ⬍28 UI/L.

T

ABLE

2. Baseline Features of Patients According to IFN Treatment

Treated

Untreated

P

Patients (n)

109

193

Mean age (yr)

32.5

⫾ 15

34.7

⫾ 15

n.s.

Sex (M/F)

72/37

51/142

n.s.

Alcohol abuse (n)

n.s.

ALT (UI/L)

114.8

⫾ 78

⫾ 75

.01

Platelets (

⫻1,000/mm

)

204.8

⫾ 73

192

⫾ 67

n.s.

Wild-type HBV (n)

51

n.s.

HBeAg minus mutant HBV (n)

n.s.

HDV infection (n)

47

n.s.

HCV infection (n)

n.s.

Dual HBV/HCV-replicative infection (n)

13

n.s.

All replicative markers negative (n)

n.s.

Chronic hepatitis (n)

Mild

17

25

n.s.

Moderate/Severe

108

Cirrhosis (n)

n.s.

NOTE. Plus-minus values are means

⫾ SD. Normal values: ALT ⬍28 UI/L.

Abbreviation: n.s., not significant.

EPATOLOGY

Vol. 30, No. 1, 1999

DI MARCO ET AL.

259

in IFN-treated than in untreated patients (P

⬍ .05 and P ⬍

.01, respectively), while HBV-DNA clearance did not signifi-

cantly affect these outcomes. The rate of mortality and of

decompensation did not differ significantly between patients

with wild-type HBV and HBeAg minus mutant, between

patients with or without HCV infection, and between patients

with or without HDV infection. Subjects with dual HBV/HCV-

replicative infection had a significantly reduced survival and

event-free survival as compared with all others (P

⬍ .01).

Multivariate analysis showed that older age, cirrhosis at

baseline, and abnormal ALT during follow-up were indepen-

dent predictors of decompensation and death (Table 5). HBV-

replicative status, as expressed by serum HBV DNA or

HBeAg, was not significantly predictive upon univariate

analysis. Regression analysis of significant variables showed

that disease decompensation occurred more often among

patients not treated with IFN. Antiviral therapy, however, did

not significantly affect the death rate.

The estimated probabilities of liver-related death for hypo-

thetical patients, according to age at presentation, stage of

disease, and ALT behavior, are shown in Figs. 3 and 4. At the

extremes of the range, the 5-year probability of death for a

60-year-old cirrhotic patient with persistently abnormal ALT

is 32% (Fig. 3), and for a 30-year-old noncirrhotic subject

who normalizes ALT, it is less than 2% (Fig. 4).

DISCUSSION

This study prospectively evaluates the long-term clinical

outcome in a large cohort of treated and untreated HBV-

infected patients, assessed according to the HBV-, HDV-, and

HCV-replicative patterns in a clinical setting outside random-

ized clinical trials. All patients came from Southern Italy, a

geographical area where HBV infection, up to 10 years ago,

took place mostly during the first years of life, and where

coinfection with HDV is not uncommon.

Our cohort is thus

representative of a specific sociodemographic reality, and

results may not be generalizable to populations with different

modes of HBV transmission and different immunogenetics,

such as Asia, Central Europe, or North America.

The effectiveness of IFN treatment in chronic hepatitis B,

in normalizing biochemical, virological, and histological

surrogate markers of response, has been widely proven by

many RCTs and by two meta-analyses.

6,7

Although these

markers of cure are clinically and biologically plausible, their

link to the true end-points of chronic liver disease (i.e.,

decompensation and mortality) remains undetermined. These

relevant health outcomes cannot be directly addressed by

RCTs because of the prolonged course of the disease. More-

over, clinical trials have been conducted largely among

patients with a typical (HBeAg-positive) virological profile.

Therefore, the transferability of results to the whole spectrum

of HBV-infected subjects is questionable. To our knowledge,

only two studies compare the long-term clinical course of

treated and untreated patients with HBeAg-positive chronic

hepatitis,

4,11

while data on the long-term outcome of anti-HBe–

positive chronic hepatitis after IFN therapy are still unavail-

able. The benefit of IFN in prolonging survival in HBeAg-

positive patients was first suggested by Wong

using a

mathematical simulation model, and subsequently confirmed

by Niederau

in a cohort of 103 treated subjects.

Our data show that patients with chronic hepatitis B have a

fivefold excess in mortality when compared with the general

population. Subjects without cirrhosis at presentation remain

asymptomatic for many years after diagnosis, even if their

ultimate life expectancy is reduced (SMR: 6.2). When cirrho-

sis is already present, the prognosis is even worse (SMR: 18),

and a careful follow-up program is required. In particular,

because cirrhotic patients between 40 and 50 years of age

with persistently abnormal aminotransferases have a 5-year

mortality beyond 20%, liver transplantation should be consid-

ered for some of these patients. Some methodological issues

should, however, be considered. In the first place, studies of

prognosis of a chronic disease should ideally involve all

patients from a large and unselected population. Even if we

did try to enroll all potential candidates, an intrinsic referral

bias as a result of the status of our Unit (academic, tertiary

care) might skew the disease spectrum toward the observa-

tion of more severe cases. Moreover, comparisons of SMRs

T

ABLE

3. SMRs in Patients With Chronic Hepatitis B According to Age and Sex Groups as Compared With the General Population of Sicily in 1992

Age

Group

(yr)

Men

Women

Total

Number

of Patients

Observed

Deaths

Expected

Deaths

SMR

Number

of Patients

Observed

Deaths

Expected

Deaths

SMR

Number

of Patients

Observed

Deaths

Expected

Deaths

SMR

95% CI

0-29

0.46

10.80

0.15

—

106

0.61

8.20

2.66-19.11

30-39

72

6

0.75

8.02

0.09

21.41

0.84

9.52

4.10-18.75

40-49

0.91

7.71

0.11

9.40

1.02

7.84

3.38-15.45

50-59

23

6

1.26

4.76

0.16

6.16

1.42

4.93

1.98-10.16

⫹

1.88

1.06

0.90

5.57

2.78

2.52

1.01-5.19

Total

214

5.26

4.94

1.41

6.39

302

6.67

5.25

3.66-7.30

T

ABLE

4. SMR in Patients With Chronic Hepatitis B According to Baseline

Liver Histology, Virological Classes, and IFN Treatment as Compared

With the General Population of Sicily in 1992

Group

Number

of

Patients

Observed

Deaths

Expected

Deaths

SMR

95% CI

Cirrhosis

25

1.34

18.60 12.03-27.53

No Cirrhosis

216

10

1.60

6.20

2.98-11.41

Wild-type HBV

0.47

17.0

7.33-33.49

HBeAg minus mutant

HBV

80

6

1.15

5.20

1.91-11.34

HDV infection

14

0.51

27.0

14.74-45.36

HCV infection

0.37

—

Dual HBV/HCV-

replicative infection

7

0.14

49.0

19.65-100.94

All replicative markers

negative

0.32

6.20

0.75-22.38

IFN-treated

109

0.80

8.60

3.45-17.72

Untreated

193

1.47

18.90 12.57-27.41

260

DI MARCO ET AL.

EPATOLOGY

July 1999

among different subgroups from the same cohort must be

interpreted cautiously, because different SMRs may have

different standards. The SMRs are thus not comparable

between them and should not be compared to evaluate the

relative effect of subgroups with different mean ages.

A point of practical relevance can be drawn from this study:

only three easily definable clinical variables (i.e., age, cirrho-

sis, and persistently normal ALT) were found by multivariate

analysis to independently predict survival. HBV DNA was

not a significant predictor. In fact, we did not find a strict

correlation between serum HBV-DNA clearance and overall

or event-free survival. Discordance with other reports

might be the result of the use in our study of an insensitive

assay. Dot-blot hybridization for HBV DNA is obviously less

sensitive than PCR. It was, however, the only available

technique at the time of enrolling our cohort, and has since

. 1.

Cumulative survival in

all patients (solid line) and according

to presence/absence of preexisting

cirrhosis and to normal/abnormal

ALT levels.

IG

. 2.

Cumulative survival

without complications in all patients

(solid line) and according to pres-

ence/absence of preexisting cirrhosis

and to normal/abnormal ALT levels.

EPATOLOGY

Vol. 30, No. 1, 1999

DI MARCO ET AL.

261

remained the current clinical standard on which HBV-

replicative status is assessed in clinical practice. In fact, when

tested by PCR, most HBsAg-positive individuals without

detectable liver damage have a positive HBV-DNA test by

PCR, while the latter is a valuable tool for quantifying HBV

viral load.

21-23

We have found that patients with dual HBV/HCV-

replicative infection were more likely to have cirrhosis at

presentation. These findings are consistent with Fong’s data.

In our study, patients with double infection showed a worse

clinical course. However, firm conclusions on this matter are

hampered by the small number of patients with HBV and

HCV infection enrolled in both studies. Further prospective

studies on the prognosis of patients with dual HBV and HCV

viral infection are needed.

In contrast to other studies,

8,9

we have observed a very low

rate (3%) of clearance of HBsAg over a mean follow-up of 10

years without a significant difference between treated and

untreated subjects. This may reflect differences between the

studies in the features of the patients, in the inclusion criteria,

and perhaps in the schedule of IFN administered. Possible

explanations may lie in the high prevalence of perinatal

infection in Southern Europe, with a long duration of

infection in many of our subjects, and/or in the high

prevalence of infection with the HBeAg minus mutant HBV.

None of the patients who cleared HBsAg died or developed

clinical complications. As already reported by Niederau et

al.,

our patients who cleared HBeAg either spontaneously or

by IFN had a significant better clinical course than those in

whom HBeAg persisted.

Univariate analysis of each patient’s subgroup from this

long-term prospective study including a wide spectrum of

patients with different virological profiles showed that the

prognosis was better for treated than for untreated patients.

IG

. 3.

Estimated probability of

death for patients with chronic hepa-

titis B with cirrhosis in eight differ-

ent clinical settings.

T

ABLE

5. Adjusted Relative Risk for Decompensation and Death in 302 Patients With Chronic Hepatitis B

Variable

Code

␤

SE

P

Relative Risk (95% CI)

Decompensation

Age

Continuous

0.034

0.012

.005

1.034 (1.010-1.059)

Cirrhosis

⫽ Absent

1.071

0.348

.002

2.917 (1.475-5.768)

⫽ Present

IFN treatment

⫽ No

⫺1.169

0.451

.01

0.311 (0.128-0.752)

⫽ Yes

ALT normalization

⫽ No

⫺1.405

0.451

.002

0.245 (0.101-0.594)

⫽ Yes

Death

Age

Continuous

0.042

0.015

.004

1.043 (1.014-1.073)

Cirrhosis

⫽ Absent

1.114

0.410

.007

3.047 (1.365-6.801)

⫽ Present

ALT normalization

⫽ No

⫺1.424

0.544

.009

0.241 (0.083-0.699)

⫽ Yes

NOTE. For Decompensation, Model

␹

⫽ 50.296 with 5 df, P ⬍ .0001; adjusted for dual HBV/HCV-replicative infection. For Death, Model ␹

⫽ 40.31 with

5 df, P

⬍ .0001; adjusted for dual HBV/HCV-replicative infection and IFN treatment.

262

DI MARCO ET AL.

EPATOLOGY

July 1999

Nonetheless, the multivariate analysis pointed out that IFN

therapy was positively associated only to event-free survival,

but not to overall survival. Although the baseline features

were not significantly different between treated and untreated

patients, the lack of randomization does not allow us to

exclude that the benefit observed in the IFN group is the

result of a selection bias. Therefore, the better prognosis of

the treated group must be interpreted cautiously. Although

the reduction in the number of events in treated patients

suggests a long-term benefit of IFN therapy, available evi-

dence is still insufficient to confirm that IFN prolongs life in

HBV-infected patients. More data are needed, particularly in

patients with an atypical form of hepatitis B (i.e., HBeAg

minus mutant HBV and HCV/HBV dual infection) and in

those with HDV infection.

Acknowledgment:

The authors thank Dr. Tommaso Strof-

folini for providing public health data on mortality rates and

his helpful discussion.

REFERENCES

1. Liaw YF, Tai DI, Chu CM, Chen TJ. The development of cirrhosis in

patients with chronic type B hepatitis: a prospective study. H

EPATOLOGY

1988;8:493-496.

2. De Jongh FE, Janssen HLA, De Man RA, Hop WCJ, Schalm SW, Van

Blankenstein M. Survival and prognostic indicators in hepatitis B surface

antigen–positive cirrhosis of the liver. Gastroenterology 1992;103:1630-

1635.

3. Fattovich G, Giustina G, Schalm SW, Hadziyannis S, Sanchez-Tapias J,

Almasio P, Christensen E, et al. Occurrence of hepatocellular carcinoma

and decompensation in Western European patients with cirrhosis type B.

EPATOLOGY

1995;21:77-82.

4. Wong JB, Koff RS, Tine´ F, Pauker SG. Cost-effectiveness of interferon-

alpha 2b treatment for hepatitis B e antigen–positive chronic hepatitis B.

Ann Intern Med 1995;122:664-675.

5. Perrillo RP, Aach RD. The clinical course and chronic sequelae of

hepatitis B virus infection. Semin Liver Dis 1981;1:15-25.

6. Tine´ F, Liberati A, Craxı` A, Almasio P, Pagliaro L. Interferon treatment in

patients with chronic hepatitis B: a meta-analysis of the published

literature. J Hepatol 1993;18:154-162.

7. Wong DK, Cheung AM, O’Rourke K, Naylor CD, Detsky AS, Heathcote J.

Effect of alpha-interferon treatment in patients with hepatitis B e

antigen–positive chronic hepatitis B. A meta-analysis. Ann Intern Med

1993;119:312-323.

8. Korenman J, Baker B, Waggoner J, Everhart JE, Di Bisceglie AM,

Hoofnagle JH. Long-term remission of chronic hepatitis B after alpha-

interferon therapy. Ann Intern Med 1991;114:629-634.

9. Carren˜o V, Castillo I, Molina J, Porres JC, Bartolome J. Long-term

follow-up of hepatitis B chronic carriers who responded to interferon

therapy. J Hepatol 1992;15:102-106.

10. Lok AS, Chung HT, Liu VW, Ma OC. Long-term follow-up of chronic

hepatitis B patients treated with interferon alfa. Gastroenterology

1993;105:1833-1838.

11. Niederau C, Heintges T, Lange S, Goldmann G, Niederau CM, Mohr L,

Haussinger D. Long-term follow-up of HBeAg-positive patients treated

with interferon alfa for chronic hepatitis B. N Engl J Med 1996;334:1422-

1427.

12. Realdi G, Alberti A, Rugge M, Bortolotti F, Gigoli AM, Tremolada F, et al.

Seroconversion from hepatitis B e antigen to anti-HBe in chronic

hepatitis B virus infection. Gastroenterology 1980;79:195-199.

13. Yoshioka K, Kakumu S, Wakita T, Ishikawa T, Itoh Y, Takayanagi M,

Higashi Y, et al. Detection of hepatitis C virus by polymerase chain

reaction and response to interferon-a therapy: relationship to genotypes

of hepatitis C virus. H

EPATOLOGY

1992;16:293-299.

14. Scheuer, PJ. Classification of chronic hepatitis: a need for reassessment. J

Hepatol 1991;13:372-374.

15. Peto R, Pike MC, Armitage P, Breslow NE, Cox DR, Howard SV. Design

and analysis of randomized clinical trials requiring prolonged observa-

tion. II. Analysis and examples. Br J Cancer 1977;35:1-39.

16. Cox DR. Regression models and life tables (with discussion). J Roy

Statist Soc B 1972;34:187-220.

17. SAS Institute Inc. SAS guide for personal computers. Version 6.08. Cary,

NC: SAS Institute Inc., 1988.

18. Schlichting P, Christensen E, Andersen PK, Fauerholdt L, Juhl E,

Hemming P, Tygstrup N, et al. Prognostic factors in cirrhosis identified

by Cox’s regression model. H

EPATOLOGY

1983;3:889-895.

19. Craxı` A, Tine` F, Vinci M, Almasio P, Camma` C, Garofalo G, Pagliaro L.

Transmission of hepatitis B and hepatitis delta viruses in the households

of chronic hepatitis B surface antigen carriers: a regression analysis of

indicators of risk. Am J Epidemiol 1991;134:641-650.

20. Rothman KJ. Modern Epidemiology. Boston: Little, Brown & Co., 1986.

21. Jardi R, Buti M, Rodriguez-Frias F, Cortina M, Esteban R, Guardia J,

Pascual C. The value of quantitative detection of HBV-DNA amplified by

PCR is the study of hepatitis B infection. J Hepatol 1996;24:680-685.

. 4.

Estimated probability of

death for patients with chronic hepa-

titis B without cirrhosis in eight

different clinical settings.

EPATOLOGY

Vol. 30, No. 1, 1999

DI MARCO ET AL.

263

22. Simpson PR, Yu XH, Redza ZM, Anson JG, Chan SH, Lin Y. Quantifica-

tion of hepatitis B virus DNA using competitive PCR and a scintillation

proximity assay. J Virol Methods 1997;69:197-208.

23. Gerken G, Gomes J, Lampertico P, Colombo M, Rothaar T, Trippler M,

Colucci G. Clinical evaluation and applications of the Amplicor HBV

Monitor test, a quantitative HBV DNA PCR assay. J Virol Methods

1998;74:155-165.

24. Fong TL, Di Bisceglie AM, Waggoner JG, Banks SM, Hoofnagle JH. The

significance of antibody to hepatitis C virus in patients with chronic

hepatitis B. H

EPATOLOGY

1991;14:64-67.

264

DI MARCO ET AL.

EPATOLOGY

July 1999

Download 101.6 Kb.

Do'stlaringiz bilan baham: