Occult hepatitis B virus infection in dialysis patients: a multicentre survey


Dr F. Fabrizi, Divisione di Nefrologia e Dialisi, Ospedale Maggiore, IRCCS, Pad. Croff, via Commenda 15, 20122 Milano, Italy.
E-mail: fabrizi@policlinico.mi.it


Background : The epidemiology and clinical significance of occult hepatitis B virus infection (serum hepatitis B surface antigen-negative patients with detectable hepatitis B virus viraemia in serum) remains controversial with only limited information about its prevalence in patients on long-term dialysis.

Aim : To address the epidemiology of occult HBV infection in a large cohort of dialysis patients.

Methods : We screened a large cohort (n = 585) of Italian chronic dialysis patients; from this population, a group of hepatitis B virus surface antigen seronegative patients (n = 213) was tested by Amplicor hepatitis B virus Monitor Test to detect hepatitis B virus viraemia (hepatitis B virus-DNA) in serum.

Results : Occult hepatitis B virus infection was absent (zero of 213 = 0%). Persistent hepatitis B virus surface antigen carriage was less frequent than anti-hepatitis B virus core antibody (anti-hepatitis B core antigen) seropositive status in this study group [1.88% (11 of 585) vs. 36% (216 of 585), P = 0.0001]. No dialysis patients seropositive for anti-hepatitis B core antibody in serum (zero of 123 = 0%) had detectable hepatitis B virus-DNA by polymerase chain reaction technology. No significant association between abnormal biochemical liver tests and serum anti-hepatitis B core antibody was noted in our population. Nominal logistic regression analysis demonstrated an independent and significant relationship between anti-HCV antibody and anti-hepatitis B virus core antibody in serum (Wald chi-square 16.06, P = 0.0001). The rate of seropositive patients for anti-hepatitis B virus core antibody was higher among study patients than controls with normal renal function [36.9% (216 of 585) vs. 21.4% (59 of 275), P = 0.0001]; this difference partially persisted after correction for demographic parameters, and viral markers.

Conclusion : In conclusion, occult hepatitis B virus was absent in our study group. Anti-hepatitis B core antibody was significantly related to presence of anti-HCV antibody supporting shared modes of transmission. Clinical studies based on molecular biology techniques provided with higher sensitivity are planned.


The control of hepatitis B virus (HBV) infection within dialysis centres has been a triumph in the management of end-stage renal disease (ESRD) patients. Recent investigations have shown that the rate of chronic HBV surface antigen (HBsAg) infection among patients undergoing maintenance dialysis in the developed world is now low.1–5 In contrast, the frequency of persistently HBsAg-seropositive dialysis patients is still high in developing countries.6–8

Recent reports have described HBV viraemia, albeit with low level, in patients in whom the conventional marker of HBV infection, HBsAg, is absent. The clinical significance of this virological profile is unclear. Limited data exist on occult HBV among patients on long-term dialysis.9–14 It has been recently suggested that occult HBV infection might be associated with a lower response of hepatitis C virus (HCV) to standard interferon therapy in patients with normal renal function.15–18 Also, occult HBV has been cited as risk factor for hepatocellular carcinoma.18–20 The prevalence of occult HBV infection in renal dialysis patients ranges between 0 and 58% in published reports.9–14

The aim of this survey was to address epidemiology and clinical significance of occult HBV infection in a large cohort of patients receiving regular dialysis in Italy.

Material and methods

Study patients

Patients (n = 585) for the study were recruited from four dialysis units in northern and one in southern Italy, respectively. Blood was drawn for liver biochemistry (glutamic oxaloacetic transaminase [GOT], pyruvic oxaloacetic transaminase [GPT], gamma-glutamyltranspeptidase [GGTP]), HBV serology [HBsAg, antibody against HBsAg (anti-HBs), and antibody to HBc, hepatitis B core antigen (anti-HBc)], and HCV serology (anti-HCV antibody). Additional blood samples were collected from a subgroup (n = 213) of dialysis patients who were randomly selected for nucleic acid testing (NAT) to assess serum HBV viraemia. There were 212 (36.2%) patients in dialysis unit 1, 57 (9.7%) in unit 2, 115 (19.6%) in unit 3, 41 (7.1%) in unit 4 and 160 (27.4%) in unit 5. All patients and controls gave written, informed consent. The study was approved by the institutional and hospital ethical committee. We did not include individuals with a history of alcohol abuse; patients being administered anti-viral therapies and potentially hepatotoxic drugs were excluded. There were no patients with current illicit drug use.

Laboratory measurements

Measurement of serum aspartate aminotransferase (AST; also known as serum glutamic oxaloacetic transaminase), and serum alanine aminotransferase (ALT; also known as serum pyruvic oxaloacetic transaminase) was made with standard automated analyzers. Upper limits for AST and ALT were 45 and 45 IU/L, respectively. Serum γ-glutamyltranspeptidase (GGTP) was measured by a Hitachi analyzer. The normal range for GGTP levels was 11–50 IU/L. All patients were screened by third generation enzyme-linked immunoadsorbent assay anti-HCV 3.0 ELISA (Ortho Diagnostic Systems, Raritan, NY, USA) that detects antibodies (immunoglobulin G, IgG) to proteins derived from three distinct regions of the HCV genome. Australia antigen (HBsAg) was tested by serological assay (chemiluminescent microparticles immunologic assay, CMIA; Abbott Diagnostics Division, North Chicago, IL, USA). All tests were carried out and interpreted strictly in accordance with manufacturers’ protocol.

Anti-HBc testing

Serum antibody against HBc was analysed by Architect anti-HBc assay (Abbott Diagnostics). Architect anti-HBc assay provides a qualitative assessment of antibody against HBc antigen in human serum or plasma by a two-step CMIA. In the first step, microparticles coated by recombinant HBc antigen are diluted with the serum specimen and captured by anti-HBc antibodies. After several washing cycles, pretrigger and trigger solutions are added to the reaction vessels and a chemiluminescent reaction occurs; it is measured by relative light units (RLU). A direct ratio exists between the amount of anti-HBc antibody in serum and RLU measured by Architect anti-HBc assay. Serum specimens giving a chemiluminescent signal over the cut-off are evaluated as positive anti-HBc specimens. Architect anti-HBc assay is used to screen blood and plasma samples to avoid transmission of HBV to patients receiving red blood cell units or blood products. All tests were carried out and interpreted strictly in accordance with manufacturers’ protocol.

HBV viraemia quantitation

The HBV-DNA testing was performed by the Amplicor HBV Monitor Test (Roche Diagnostics, Branchburg, NJ, USA), an in vitro test that utilizes polymerase chain reaction (PCR) nucleic acid amplification and DNA hybridization for the quantitative measurement of hepatitis B viral DNA in human serum or plasma. As described previously,21 the Amplicor HBV Monitor test is based on four major processes: specimen preparation, PCR amplification of target DNA using HBV-specific complimentary primers; hybridization of the amplified products to oligonucleotide probes specific to the target(s) and detection of the amplified products by colorimetric determination. The amplification target for the Amplicor HBV Monitor Test is a portion of the partly single-stranded, circular DNA genome of HBV, defined by the primers HBV-104UB and HBV-104D. The primers amplify a 104 bp sequence in the pre-Core/Core region of the HBV genome. DNA is amplified using one biotinylated and one non-biotinylated oligonucleotide primer. The Amplicor HBV Monitor Test quantitates virus titres from 4 × 102 to 4 × 107 viral particles/mL of serum or plasma. To prevent carryover contamination during PCR, workflow in the laboratory proceeded in a unidirectional manner from the beginning of the procedure to the end; each step of the procedure was performed in a separate room with dedicated equipment.


The control group included 275 healthy out-patients consecutively admitted to the Institute of Hygiene and Preventive Medicine, Maggiore Hospital, IRCCS, Milano, Italy, over the years 2000–2004. All were anti-HIV-negative, and HBsAg-negative. About 133 (48.3%) were males, and the mean age was 44.76 ±17.06 years. Individuals with a history of alcohol abuse and/or illicit drug use were not included on the study.

Statistical analysis

Serum aminotransferase, and GGTP levels were transformed into decimal logarithm to obtain normal distribution and then subjected to statistical tests. Student's t-test and chi-square analysis were used, when appropriate. We conducted multivariate analysis (nominal logistic regression analysis) in the study group: serum anti-HBV core antibody status was assumed as dependent variable, and independent variables were as follows: (1) age, (2) gender, (3) race, (4) underlying nephropathy, (5) time on dialysis, (6) dialysis type (haemodialysis, HD vs. peritoneal dialysis, PD), (7) anti-HCV antibody status, (8) GOT, (9) GPT, (10) dialysis location and (10) HBsAg status. Univariate analysis was conducted using microsoft excel 97 (Microsoft Corp, Redmond, WA, USA); multivariate analysis was conducted using the statistical package jmp in (1996; SAS Institute, Cary, NC, USA). A double-sided P-value of <0.05 was considered significant.


Study group

In the study group there were 585 patients on maintenance dialysis (Table 1). The mean age was 63.14 ± 14.43 years, the mean time on dialysis was 75.22 ± 83.9 months. Fifty-four (9.2%) patients underwent PD, and 531 (90.8%) received maintenance HD. All patients undergoing PD were on continuous ambulatory peritoneal dialysis (CAPD). There were 570 (97.4%) Caucasian patients, and 15 (2.6%) non-Caucasian individuals. About 251 (42.9%) were females, and 334 (57.1%) were males.

Table 1.  Anti-HBV core antibody-positive vs. -negative patients: demographic, biochemical and virological parameters
 Anti-HBV core antibody-positive patients (n = 216)Anti-HBV core antibody-negative patients (n = 369) P-value
  1. HBV, hepatitis B virus; RLU, relative light unit; HBsAg, HBV surface antigen; HCV, hepatitis C virus; N.S., not significant.

Gender, F (%)81 (37.5%)170 (46%)0.0001
Race, Caucasian (%)210 (97.2%)360 (97.5%)RLU
Age (years)63.14 ± 14.4363.96 ± 14.25N.S.
Time on dialysis (months)90.98 ± 103.6166.26 ± 69.390.0023
HBsAg-positive patients (n)11 (5.1%)00.0001
Anti-HCV-positive patients (n)63 (29.2%)47 (12.7%)0.0001
Peritoneal dialysis patients (n)26 (12.0%)28 (7.5%)N.S.
GOT (IU/L)16.28 ± 10.614.93 ± 7.6N.S.
GPT (IU/L)16.18 ± 10.315.14 ± 9.04N.S.
GGPT (IU/L)39.8 ± 51.031.1 ± 43.87N.S.

Anti-HBV core antibody rate was 36.9% (216 of 585). The distribution of anti-HBV core-positive patients among the dialysis units was as follows: 27.8% (59 of 212) in unit 1, 47.3% (27 of 57) in unit 2, 52.2% (60 of 115) in unit 3, 26.8% (11 of 41) in unit 4 and 36.9% (59 of 160) in unit 5; there was a significant difference in the rate of anti-HBV core antibody-positive patients among the five dialysis units (P = 0.001).

Frequency of chronic HBsAg carriage was 1.88% (11 of 585); anti-HBV core antibody-positive status was much more frequent than persistent HBsAg carriage, 36.9% (216 of 585) vs. 1.88% (11 of 585), P = 0.0001.

A group of patients (211 of 585 = 36%) was tested for anti-HBV core IgM antibody test, 59 (59 of 211 =27.9%) of whom had serum anti-HBc antibody, all were anti-HBV core IgM antibody-negative.

A subset (213 of 585 = 36.4%) of HBsAg-negative patients was tested by PCR to detect HBV-DNA, no sample had detectable HBV-DNA including 123 (123 of 213 = 57.7%) specimens from patients with detectable anti-HBc antibody in serum.

Data on anti-HBs antibody were available from a subset of dialysis patients [54% (316 of 585)]; the rate of anti-HBc seropositivity was 37.9% (120 of 316). There were 25 HBsAg-negative/anti-HBc-positive/anti-HBs-negative patients [20.8% (25 of 120)] on maintenance dialysis.

No relationship was observed between anti-HBV core antibody status and liver biochemical tests (Table 1). In the subgroup of anti-HBV core antibody-positive patients, there was a significant association between anti-HCV antibody-positive status and raised aminotransferase activity (Table 2). In the whole study group, multivariate analysis demonstrated an independent relationship between anti-HCV-positive antibody status and anti-HBV core antibody-positive status (Table 3).

Table 2.  Serum aminotransferase activity in HBsAg-negative, anti-HBV core antibody-positive patients (n = 205)
 Anti-HBV core antibody-positive patientsP-value
Anti-HCV-positive/ anti-HCV-negative (n = 58) Anti-HCV-negative (n = 147)
  1. HBV, hepatitis B virus; HBsAg, HBV surface antigen; HCV, hepatitis C virus.

GOT (IU/L)16.27 ± 10.6414.36 ± 6.630.03
GPT (IU/L)16.20 ± 10.3114.43 ± 8.350.05
Table 3.  Nominal logistic regression analysis: parameter estimates and effect test (n = 585)
 Number of parametersd.f.Wald chi-squareP-value
  1. HBsAg, hepatitis B virus surface antigen; HCV, hepatitis C virus; ESRD, end-stage renal disease.

ESRD, cause557.5792070.1609
Dialysis, time111.5983080.1573
Dialysis, type117.5792070.0707
Dialysis, location447.7246860.1022

Control group

There were 59 (21.4%) anti-HBV core antibody-positive individuals and five (1.8%) anti-HCV-positive patients. In the control group, anti-HBV core antibody-positive status was affected by age; the rate of anti-HBV core antibody positivity was higher in older controls [33.3% (10 of 30; >60 years) vs. 9.4% (six of 64; <40 years), P = 0.010]. Anti-HBV core antibody-positive controls were significantly older than anti-HBV core-negative individuals, 54.35 ± 16.41 vs. 42.10 ± 16.33 years (P = 0.0001).

Study patients were significantly older than controls, 63.14 ± 14.43 vs. 44.76 ± 17.06, P = 0.0001. A significant difference occurred in the distribution of female gender between study patients and controls, 42.9% (251 of 585) vs. 51.6% (142 of 275), P = 0.021. The rate of seropositivity for anti-HBV core antibody was higher among study patients than controls, 36.9% (216 of 585) vs. 21.4% (59 of 275), P = 0.0001. The frequency of anti-HCV antibody seropositivity was higher in study than control individuals, 18.8% (110 of 585) vs. 1.8% (five of 275), P = 0.0001.

As listed in Table 4, the rate of anti-HBV core antibody was higher in dialysis patients than controls, after correction for race, gender, age and viral markers.

Table 4.  Frequency of serum anti-HBV core antibody in age-, race-, gender (female)-matched study and healthy groups
Age (years)Dialysis patients (%)Healthy controls (%)P-value
  1. HBV, hepatitis B virus; N.S., not significant.

>6030.3 (53/175)33.3 (10/30)N.S.
50–5939.4 (15/38)36.6 (11/30)N.S.
40–4938.8 (7/18)33.3 (6/18)N.S.
<4034.2 (13/38)14.6 (12/82)0.021


Occult HBV infection was absent in this multicentre survey that is HBV viremia was not detected in HBsAg-negative patients. This result is in contrast with other studies where the rate of detectable HBV-DNA in serum has ranged between 58 and 0%.9–14 A number of possible explanations exist including sensitivity of molecular biology techniques, size and virological features of the patient groups. Of note, in the study of Cabrerizo et al.12, 13 the rate of occult HBV was 57.6% (19 of 33) but the frequency of serum HBV markers was 51.5% (17 of 33). In the series of Dueymes et al.10 the HBV-DNA rate was 13.9% (six of 43) with an anti-HBV core antibody seropositivity of 64.7% (33 of 51). Also, the ‘background’ frequency of HBV infection could play a role; geographical differences in the prevalence of occult HBV usually occur and are mostly related to endemicity of HBV infection.22, 23

In our study, the rate of anti-HBV core antibody was higher in dialysis patients than controls after correction for several demographic parameters, and viral markers; we found a significant relationship between anti-HBV core antibody and HCV infection, as detected by anti-HCV serum antibody. These findings should support the notion that transmission of HBV among patients undergoing regular dialysis remains a concern in developed countries. Time on dialysis was significantly greater in anti-HBV core antibody-positive than -negative patients on dialysis; however, this link was lost by multivariate analysis. Serological markers of HBV and HCV infection in the same patient reflect shared modes of transmission for these viruses, not necessarily implying infection in the dialysis unit but also possible routes such as illicit drug use.

The clinical significance of isolated anti-HBV core antibody in serum (i.e. anti-HBc-positive status in HBsAg-negative patients) is still unclear in dialysis patients as well as in non-uraemic populations. Some reports suggested that anti-HBV core antibody-positive status implies a potentially infectious state. Several studies of HBV transmission in recipients of blood,24, 25 liver26–28 and kidney29, 30 grafts from anti-HBV core antibody-positive donors have been reported. Depending on the anti-HBc test used and the prevalence of true positive results, a certain proportion of anti-HBV core antibody-positive patients may be false-positive.31, 32 The diagnostic specificity and sensitivity of the Architect anti-HBc assay are claimed by the test kit manufacturers to be >99%.33 Anti-HBV core antibody can reflect late HBV immunity with anti-HBs concentrations low or undetectable. Nevertheless, no clinical history of acute icteric hepatitis was elicited in our patients with anti-HBV core antibody; acute HBV, however, not infrequently can be subclinical. Alternatively, anti-HBV core antibody-positive patients may have an unresolved HBV infection respectively a chronic infection in a late or low grade replicative state.22 Persistent infection was not confirmed in our patients with anti-HBV core antibody as detectable HBV-DNA was not found. However, HBV-DNA has been found detectable only temporarily, when patients are followed for years in some reports;34 virological follow-up are ongoing in our patients. Finally, the frequent coincidence of anti-HBc and HCV markers has been seen in non-uraemic populations35, 36 and is supported by in vitro experiments.37 HCV/HDV coinfection may suppress HBV replication leading to isolated anti-HBV core antibody status.38–40 Different mutations down-regulating the replication of HBV or the secretion of HBsAg have also been shown to play a role.41–43 Whether HBV mutations are associated with anti-HBV core antibody seropositivity remain to be evaluated among dialysis patients.

We did not find any relationship between increased biochemical liver tests and anti-HBV core-seropositive status in our dialysis population. Serum transaminase values are lower in patients with chronic renal failure (CRF) regardless they are dialysis-dependent44, 45 or not46– this hampers the recognition of liver damage on the basis of liver biochemical tests. A significant association between increased serum aminotransferase levels and persistent HBsAg carriage in dialysis population has been already noted.47 The association between raised serum transaminase activity and positive anti-HCV antibody status was confirmed in this study. Thus, what impact (if any) of anti-HBV core antibody-seropositive status in our dialysis population remains unclear.

In conclusion, we did not find occult HBV infection in our population of dialysis patients. Anti-HBV core antibody was more common among dialysis than control groups and significantly linked with anti-HCV antibody suggesting shared modes of transmission. Additional studies provided with more sensitive NAT assays are under way.


This study was supported in part by the grant ‘Project Glomerulonephritis’ in memory of Pippo Neglia.