• Open Access

Coinfection with hepatitis B and D: Epidemiology, prevalence and disease in patients in Northern California


  • Conflict of Interest: None.
  • Financial support: Unrestricted grant from California Pacific Medical Center Philanthropic Foundation for the initial study. Grant support for statistical analysis: This project was partially supported by the National Institutes of Health, Grant UL1TR000100 during year 3 and beyond of CTSA funding. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH.


Dr Robert G Gish, 200 West Arbor Drive, MC8413, San Diego, CA 92103-8413, USA. Email: rgish@ucsd.edu


Background and Aims

With no report on the overall prevalence and ramifications of hepatitis Delta virus (HDV) infection in the United States for more than two decades, the characteristics of chronic hepatitis B virus (CHB) patients coinfected with HDV, including clinical presentation, rate of hepatitis C virus tri-infection, and HDV viral load, were assessed.


At California Pacific Medical Center, a retrospective chart review was conducted on all CHB patients.


Of 1191 CHB patients, 499 had been tested for HDV, with 42 (8%) determined to be coinfected; half of these were also hepatitis C virus-infected. Cirrhosis was present in 73% of the coinfected, 80% of the tri-infected, but only 22% of the monoinfected. Twenty-nine patients (69%) were Caucasian non-Hispanic; 10 (24%) were Asians and Pacific Islanders. Of 39 patients for whom HBV-DNA quantification at time of HDV presentation was available, 22 (56%) had undetectable levels; four (10%) had levels > 100 000 IU/mL.


HDV affects individuals of all ages and various ethnic groups. Although HBV viral loads are lower, rates of cirrhosis are higher in coinfected patients and higher still in the tri-infected. Our data support revising screening guidelines to advocate for all patients with HBV to be screened for HDV in order to both give the individual patient important information related to the possible need for treatment and to support the public health goal of reducing transmission by educating HDV-negative patients about the need for protection against superinfection and HDV-infected patients about the need to protect against transmission to others.


Approximately 15 million people worldwide are infected with hepatitis D (hepatitis Delta) virus (HDV), with estimates as high as 80 000 in the US. HDV is a circular, single-stranded RNA virus similar to plant viroids that is coated in hepatitis B surface antigen (HBsAg) and thus requires the presence of hepatitis B virus (HBV) for its propagation and presumptively for its pathogenicity.[1] The nucleocapsid consists of a 1.7 kB circular RNA with two hepatitis D antigen (HDAg): small HDAg (S-HDAg) and large HDAg (L-HDAg). The outer coat contains the three HBV envelope proteins termed large, medium, and small HBsAg. The virus is uncoated after entering the hepatocyte, and a signal in HDAg translocates the nucleocapsid to the nucleus. The antigen then uses host RNA polymerases that treat the genome as a double-stranded DNA leading to the production of three forms of RNA. The messenger RNA (mRNA) contains the open reading frame of the HDAg. Replication of the circular HDV-RNA template occurs via a rolling mechanism. HDV-RNA is first synthesized as a linear molecule that is self-cleaved by an HDV ribozyme into monomers. These monomers are then ligated to form circular RNA. HDAg is the only protein encoded by the HDV genome. The open reading frame of the mRNA generates both isoforms of HDAg, S-HDAg and L-HDAg. The S-HDAg supports viral replication, while the L-HDAg is essential for virion assembly. Post-translational modification of L-HDAg involves several processes, including prenylation, phosphorylation, sumoylation, and methylation. In the nucleus, molecules of L-HDAg form complexes with S-HDAg and new constructs of genomic RNA; these complexes are exported to the Golgi membranes. Once there, these complexes associate with HBV envelope proteins to create an infectious virion. There are two modes of infection by HDV with HBV: coinfection and superinfection. In the former, an individual is coinfected by HBV and HDV at the same time; in the latter, an individual previously infected with HBV is superinfected with HDV at a later time.

Traditionally, HDV testing was conducted only in HBV patients from high-risk groups and in those with advanced liver disease. Members of high-risk groups were generally defined as including individuals who paid for sex, intravenous (IV) drug users, patients on hemodialysis, and those receiving blood products (including patients who had received transfusions and hemophiliacs). In a European study, 29.6% of HBV patients from such high-risk groups had anti-HDV antibodies.[2] In comparison, the prevalence of anti-HDV antibodies in HBV-infected patients who were not members of such groups was only 1.5%. In the United States, studies from the 1970s to the mid-1980s reported a prevalence of HDV infection of 3.8% in HBsAg-positive blood donors from 49 of 57 regions of the American Red Cross,[3] 15% in HIV-coinfected CHB patients in Los Angeles,[4] 30% of CHB patients in Illinois state facilities for the developmentally disabled,[5] 42% of drug-using CHB patients at several US Veterans Administration facilities,[6] and 67% of injection drug users in New York City.[7] A more recent study reporting data collected in 2005–2006 found that 50% of injection drug users with CHB were HDV-infected.[8] There has been no report on the overall prevalence of HDV infection in the US for more than two decades.

The most dire complications of chronic HBV infection are the development of cirrhosis, hepatic decompensation, hepatocellular carcinoma, and death because of liver failure.[9] Rates of these complications are higher in those coinfected with HDV.[10, 11] The prevalence of chronic HBV infection in the United States is lower than in highly endemic regions.[12] However, in developed countries, there is a higher prevalence of HBV among those who have immigrated from countries with high or intermediate prevalence,[9] making patients' countries of origin of interest. The highest rates of chronic HBV infection are thought to be in Africa, Southeast Asia, the Middle East, the interior Amazon River basin, and parts of the Caribbean and Pacific islands.[13]

In order to determine the prevalence of HDV infection and the characteristics of those infected, patients at a tertiary care center in northern California who were dually infected with HBV and HDV or tri-infected with HBV, HDV, and hepatitis C virus (HCV) were evaluated. This investigation of the epidemiology and disease state of HBV/HDV coinfection and HBV/HCV/HDV tri-infection can potentially contribute to an increased awareness of HDV infection, the ramifications of coinfection or tri-infection in terms of disease progression, and the need for prevention and treatment. It may also point to the need for updates in HBV screening guidelines.

Patients and methods

Chronic HBV carriers were extracted from California Pacific Medical Center's complete database of patients from northern California. Chronic HBV carrier status was established by a positive HBsAg laboratory value (HBsAg+ for > 6 months). Using electronic patient records collected from 2002 to 2007 as well as transplant patient records collected between 1989 and 2002, we found that of 1191 patients defined as chronic HBV carriers, 499 had been tested for HDV using HDV antigen or anti-HDV tests (Quest assays) or both. The Quest HDV assays (total antibody, immunoglobulin M [IgM], and antigen) have an interassay coefficient of variation for the positive control of between 10% and 20%. The results are reported qualitatively based on an index value, calculated by dividing the specimen optical density (OD) value by the OD value of a calibrator serum included in every run; an index value < 1.0 is considered negative, while an index value ≥ 1.0 is considered positive. The analytical sensitivity and specificity (i.e. the results of the Quest assays compared with other assays for the same analytes) are > 90%. Quest does not provide data for clinical sensitivity (i.e. what proportion of patients with independently confirmed HDV infection are positive in a given assay). HIV testing was not routine in this population. Age at time of diagnosis and the patient's country of origin and race were identified. The countries of origin were divided into the following geographical categories: North America (United States), Southeast Asia (Cambodia, Indonesia, Laos, and Vietnam), East Asia (China and Mongolia), Middle East (Lebanon, Saudi Arabia, and Syria), Eastern Europe (Moldova and Romania), and Africa (Ivory Coast). When available, social history of IV drug use was noted.

Each individual was defined as having cirrhosis or no evidence of cirrhosis from analysis of liver biopsy records. Patients without liver biopsy records were analyzed by a second set of criteria. Individuals with a platelet count of fewer than 150 000 were defined as having advanced fibrosis or cirrhosis, and were scored as cirrhosis. Where available, HBV-DNA levels at the time of diagnosis were included in our analysis. The tests had been performed according to the manufacturer's instructions using the COBAS Amplicor HBV Monitor test (Roche Diagnostics, Indianapolis, IN, USA), which measures HBV-DNA quantitatively. The highly conserved HBV precore/core region and the internal standard DNA were amplified in the same well but hybridized in separate wells. The amount of HBV-DNA was calculated from the ratio of the HBV well to the internal standard well, and the copy number per milliliter was calculated from a standard curve. The dynamic range of the Amplicor test was 3 × 102 to 2 × 105 copies/mL. Patients were also separated based on HCV antibody positivity; those who were HCV antibody-positive were labeled as HBV/HCV/HDV “tri-infected.” Tri-infected patients were further categorized as cirrhosis positive or negative, with the exception of one patient whose cirrhosis status was unknown.

Although there are no clinically available tests for HDV-RNA that are approved in the United States, in the subset of patients for whom stored serum was available (12% of the patients diagnosed as HDV-infected), HDV-RNA quantification was provided by Jeff Glenn's research laboratory at Stanford School of Medicine using a sensitive and reproducible real-time reverse transcription–polymerase chain reaction assay to quantify the amount of HDV-RNA in patient serum. Delta RNA was extracted from 50 uL of patient serum using Trizol, and the isolated RNA was reverse-transcribed using MuLV reverse transcriptase and random primers. A standard curve composed of serial dilutions of a previously quantified, in vitro-transcribed HDV-RNA preparation was reverse-transcribed in parallel. The resulting complementary DNAs were then subjected to a Taqman-based real-time PCR, and viral load values of the serum samples were calculated from the standard curve. The limit of sensitivity is approximately 100 RNA copies. HDV-RNA quantities were matched with the quantitative lab values of HBV-DNA at the closest available time point.


Of 499 CHB patients tested for HDV, 42 (8%) were confirmed to be HDV-infected: 29 patients by a positive hepatitis Delta antibody (HDab+), 7 patients by both a HDab+ and an HDAg, and 6 patients by a positive delta antigen only (Table 1). Of the 36 patients who tested positive for anti-HD, 34 were confirmed by total antibody only and two had laboratory tests that were positive for both total antibody and IgM. HDV-RNA quantification was available for 5 of the 42 HDV patients. All of these patients (100%) were shown to be positive for HDV-RNA, thus confirming that positive delta serology is a strong correlate of active disease. Of the HBV/HDV-coinfected patients, one-half (21 patients) were proposed to be also HCV-infected (tri-infected) as determined by antibody testing to HCV; HCV-RNA data were not available. The HCV status of one patient was unknown. Of the 42 HBV/HDV-coinfected patients, 26 had been tested for HIV; only one was HIV+.

Table 1. Test results of 42 patients categorized as HDV-infected at time of diagnosis
HDV antibody+ onlyHDV antigen+ onlyBoth HDV Ab+ and HDV Ag+Anti-HDV IgM+HBeAg-HBeAg+
  1. †HBeAg status was unavailable for 14 patients.
  2. HBeAg, hepatitis B e antigen; HDV, hepatitis Delta virus; HDV ab, HDV antibody; HDV Ag, HDV antigen; IgM, immunoglobulin M.

Of the 41 HDV-infected patients with known cirrhosis status, 73% (30 patients) were diagnosed with cirrhosis. By comparison, 17% (78) of the 457 HDV-tested CHB patients who were only HBV-infected and 22% (262) of the total CHB population of 1191 patients had cirrhosis. Pearson's chi-squared test was used to test the independence of HDV co- or tri-infection and cirrhosis diagnosis (Table 2). Based on the test, the finding of cirrhosis in 30 of 41 HDV-infected patients (including co- and tri-infected patients) compared with 78 of 457 patients with HBV only is significant evidence to reject independence (P value <.0001).

Table 2. Cirrhosis diagnosis for patients with HBV infection only versus patients with either HBV/HDV coinfection or HBV/HDV/HCV tri-infection
 Type of infectionTotal
HBV infection onlyCo- or tri-infection
  1. Pearson's chi-squared test with Yates' continuity correction was used to test the independence of co- or tri-infection, and cirrhosis diagnosis; the test statistic (66.46, P value ≤.0001) indicates that there is significant evidence to reject independence.
  2. HBV, hepatitis B virus; HCV, hepatitis C virus; HDV, hepatitis Delta virus.
No cirrhosis37911390

In addition, 80% of the proposed tri-infected (16 of 20 patients with known cirrhosis status; status of one tri-infected patient was not known) were found to have cirrhosis. To uncouple the effect of HCV tri-infection from HDV/HBV only, Fisher's exact test was used (as an expected cell count was below five) to test the independence of only HDV/HBV-coinfection and cirrhosis diagnosis (Table 3). Based on the test, the finding of 14 cases of cirrhosis out of 21 patients with HDV/HBV coinfection only compared with 78 cases of cirrhosis out of 457 patients with HBV only is significant evidence to reject independence (P value <0.0001).

Table 3. Cirrhosis diagnosis for patients with HBV infection only or HBV/HDV coinfection only
 Type of infectionTotal
HBV infection onlyHBV/HDV coinfection only
  1. Fisher's exact test was used to test the independence of coinfection and cirrhosis diagnosis instead of Pearson's chi-squared, as there is an expected cell count below 5; the P value (≤ .0001) indicates that there is significant evidence to reject independence.
  2. HBV, hepatitis B virus; HDV, hepatitis Delta virus.
No cirrhosis3797386

Thirty-one of the HDV-infected patients (74%) were male. The geographical origin was available for 41 of the 42 HDV patients. Twenty-six patients (63%) were born in North America, six (15%) in South East Asia, three (7%) in each of East Asia and the Middle East, two (5%) in Eastern Europe, and one (2%) in Africa. As shown in Table 4, most patients (29; 69%) were Caucasian non-Hispanic; 10 (24%) were Asian and Pacific Islanders, 2 (5%) were Hispanic Caucasian, and 1 (2%) was native African. The mean age at time of presentation was 48. A wide age range of patients presented with HDV (median 47.0 years, range 30–79). Eleven patients (26%) reported past IV drug use. Information on the probable mode of transmission was available for 28 patients: injection drug use (10 patients), sexual transmission (4 patients), medical transmission (2 patients), and vertical or childhood transmission (12 patients); for 14 patients, there was no information available related to the probable mode of transmission. The treatment history for these patients was not available in their medical charts.

Table 4. Demographics of study population
HDV demographicsn
  1. API, Asian and Pacific Islander; HDV, hepatitis Delta virus.
Age< 50/> 5024/18
RaceCaucasian non-Hispanic/API/other29/10/3

Quantification of HBV-DNA at the approximate time of HDV presentation was available in 39 of the 42 patients. Twenty-two of 39 patients (56%) had undetectable levels of HBV-DNA at time of HDV presentation; four patients (10%) were found to have HBV-DNA levels > 100 000 IU/mL (Fig. 1). Of the five patients with HDV-RNA quantification, four had HDV-RNA values greater than 107 copies/mL and associated HBV viral loads that were undetectable (Table 5). The remaining patient had an HDV-RNA quantity of 2.46 × 105 copies/mL and an associated HBV viral load of 4.17 × 105 IU/mL.

Figure 1.

Hepatitis B virus (HBV) DNA quantification at time of hepatitis Delta virus (HDV) presentation in 39 of the 42 HDV-infected patients. *For the six patients shown as “positive,” only qualitative HBV-DNA testing was available; the quantitative test was either not done or not available.

Table 5. HDV-RNA quantities matched with the quantitative lab values of HBV-DNA at the closest available time point
Patient #HDV-RNA quantity in copies/mLAssociated HBV quantity in IU/mL
  1. HBV, hepatitis B virus; HDV, hepatitis Delta virus.
12.46 × 1054.17 × 105
23.5 × 108Undetectable
34.30 × 109Undetectable
41.2 × 107Undetectable
52.8 × 109Undetectable


This study demonstrates that HDV affects individuals of all adult ages and a wide variety of ethnicities. Our results show that males were infected at a higher rate than females; however, this was most likely due to our small sample size. Although the majority of participants in this study were born in the United States, the country of origin did not appear to be a barrier to infection. Numerous studies have shown the prevalence of HDV infection in specific countries; however, few have examined HDV in a worldwide context.

In accordance with prior research, our data show that patients coinfected with HBV/HDV have a substantially increased rate of cirrhosis, with a 51% higher risk when compared with the entire population of HBV patients. Not unexpectedly, individuals with HBV/HCV/HDV tri-infection had rates of cirrhosis (80%) even higher than individuals with either chronic HBV monoinfection (22%) or HBV/HDV coinfection (73%).

Prior studies have investigated the interplay between HDV and HBV in coinfected patients. Lee demonstrated that 8 of 55 coinfected patients had undetectable HBV-DNA viral loads and 31 of 55 (56%) had low HBV viral loads (under 100 000 IU/mL), supporting that HDV may inhibit HBV replication.[14] Similarly, a large percentage of patients in our study had low HBV viral loads. Using 6.5 million copies/mL (1.2 million IU/mL) as a reference for high HBV-DNA load,[15] we found that only 2 (6%) of the 29 patients with available HBV-DNA quantification at the time of HDV presentation had high HBV viral loads. Additionally, the four patients with HDV-RNA quantities above 107 had undetectable HBV-DNA, while the patient with the lowest HDV-RNA value had a detectable HBV viral load greater than 105.

This study was limited by its retrospective nature and by the fact that HDV testing was not available for the entire population of CHB patients. Our analysis would have been stronger if HDV-RNA could have been tested in a higher percentage of the HDV seropositive patients. Unfortunately, stored serum samples were only available for HDV-RNA testing for 12% of these patients. Additionally, because we used the time at which a patient tested positive for either the HDV antibody or antigen as the time of HDV presentation (which was most likely well after they were originally infected), we were unable to accurately determine the time of HDV infection in our patient population.

Despite these limitations we believe that this investigation expands our understanding of the ramifications of HDV infection in terms of disease progression, providing data that clearly point to the need for prevention and treatment, and supports the idea of revising HBV screening guidelines. In particular, we believe that the substantially increased risk of cirrhosis we observed in the HDV-infected combined with the relatively low cost of screening for the anti-HDV antibody support a recommendation to guidelines committees to consider advocating for all patients with HBV to be screened for HDV coinfection and as an absolute minimum, all those who have advanced liver disease or are members of high-risk groups. Screening all patients who are HBsAg-positive for HDV could result in earlier diagnosis and possible treatment intervention. Multiple studies have indicated that there is at least some benefit from interferon treatment,[16-19] with the most recent publication showing that treatment with peginterferon alfa-2a for 48 weeks resulted in sustained HDV-RNA clearance in about one quarter of HDV-infected patients.[20] As new treatments for HDV such as prenylation inhibitors emerge,[21] higher rates of HDV viral clearance on treatment may become evident. In addition, universal screening of CHB patients for HDV infection could support the important public health goal of reducing transmission. With universal screening, HDV-negative patients could be educated about the need for protection against superinfection and HDV-infected patients educated about the need to protect against transmission to others.