Recent High Incidence of Fulminant Hepatitis in Samara, Russia: Molecular Analysis of Prevailing Hepatitis B and D Virus Strains

Abstract

Until 1991, the Russian city of Samara was largely isolated from other parts of Russia and the rest of the world. Very recently, Samara has seen an alarming increase in the incidence of hepatitis. The proportion of fulminant cases is unusually high. We wanted to assess the roles of hepatitis B virus (HBV) and hepatitis D virus (HDV) in acute viral hepatitis in this region by analyzing the prevailing strains of both and by determining their genotypes and possible origin. Serum samples were screened for different serological markers and by PCR followed by direct sequencing. Of the 94 HBV-positive samples (80% of which were acute infections), 37 (39%) were also HDV positive. Sixty-seven percent of the patients had anti-HCV antibodies. Twenty-five percent of all patients in the study had fulminant hepatitis. Statistically significant sex differences were found among fulminant cases. For HBV, the core promoter sequences of 62 strains were determined and all but one were found to be of genotype D. None of these had any deletions. Only one strain, from a patient with fulminant fatal hepatitis, showed multiple mutations. The pre-S2 region sequences of 31 HBV strains were also compared. Phylogenetically, these fell into two distinct groups within genotype D, suggesting different origins. For HDV, part of the region encoding the δ-antigen was sequenced from four strains. All proved to be of genotype I and were similar to Far Eastern and Eastern European strains. The contribution of intravenous drug use to the sharp increase in viral hepatitis in this unique setting is discussed.

Hepatitis B virus (HBV) belongs to the family Hepadnaviridae and has several unique properties. It has a very compact circular genome, with overlapping reading frames, and, unlike any other animal DNA virus, replicates through an intermediate reverse transcription step (3). The mutation rate of HBV is not known, but the enhanced rate arising from reverse transcription without proofreading is held in check by the compactness of the genome, which limits the number of viable mutations possible. Genetic analysis of HBV has shown there to be six different genotypes, A to F, based on an intergroup divergence in nucleotide sequence of 8% or more (19, 20). These genotypes vary in geographical prevalence and, in part, in clinical and serological outcome.

Approximately half of all acute HBV infections are subclinical. At the opposite range of the scale, the fulminant cases constitute less than 1% of acute hepatitis B infections. Following reports of outbreaks of fulminant hepatitis B, it has been suggested that specific mutations in the HBV genome often detected in these cases are associated with the development of fulminant disease (10, 14).

An important cause of fulminant viral hepatitis is co- or superinfection of HBV with hepatitis D virus (HDV) (4, 17). During both acute and chronic infection, HDV infection often leads to a more severe disease (26). This subviral human pathogen, having an RNA genome of only 1.7 kb, produces only one protein, which appears in two forms, and is dependent on HBV for packaging (21).

Three phylogenetically distinct genotypes of HDV have been reported. HDV strains are more heterologous than HBV strains, and strains differing in more than 20% of their nucleotide sequences constitute different genotypes (29). Genotypes II and III have only been isolated in eastern Asia and in northern South America, respectively, whereas genotype I is more widespread geographically (1, 7). Clinically, the disease pattern resulting from infection with HDV genotype I is very variable, ranging from mild to severe disease. Genotype II usually gives rise to a milder hepatitis, while genotype III appears to lead more often to fulminant hepatitis (1).

The city of Samara, located in the southeastern part of European Russia and the fourth largest city of the country, has had an alarming increase in the incidence of viral hepatitis over the last 2 years. Prior to 1991, this city was relatively isolated, both from other countries and from other parts of Russia. As there appears to be an unusually large proportion of fulminant hepatitis cases, we set out to assess the roles of principally HBV and HDV in acute viral hepatitis in this region. Having found HDV to be prevalent among HBV-infected individuals, we attempted to determine the genotype and possible origin of the HDV and HBV strains circulating in Samara between 1997 and 1999.

MATERIALS AND METHODS

Patients.

Consecutive patients admitted with acute hepatitis to the Department of Infectious Diseases, University Clinics of Samara, Samara, Russia, between October 1997 and May 1999 were included in the study. The patients were bled as part of the routine clinical procedure, and an aliquot of serum was separated and frozen pending further testing for hepatitis markers and nucleic acid analysis in Sweden. After excluding 15 cases lacking sufficient patient data and 7 cases with no biochemical and/or serological signs of viral hepatitis, 105 patients remained in the study. There were 84 men and 21 women, and the mean age was 24 years (range, 15 to 76). The vast majority of the patients were intravenous drug users (IVDUs). All patients but one had elevated alanine aminotransferase levels. Twenty-seven patients had fulminant hepatitis, as defined by encephalopathy and coagulopathy accompanied by deteriorated liver function tests, including elevated bilirubin levels. There were limited volumes of serum available from some patients. In these cases, priority was given to testing for HBV DNA by PCR, followed by anti-HCV antibody, HBsAg, and HDVAg and/or anti-HDV antibody.

Serological tests.

HBsAg, anti-HBc immunoglobulin M (IgM), HBeAg, anti-HBe antibody, anti-HCV antibody, and anti-HAV IgM were tested by AXsym (Abbott Laboratories, North Chicago, Ill.). HDVAg and anti-HDV antibody were tested by an in-house radioimmunoprecipitation assay (5).

Extraction and amplification of HBV DNA.

HBV DNA was extracted from serum by the phenol-chloroform method as previously described (15). Two sets of primers (KL28-KL6 and KL12-KL33; Table 1), amplifying the core promoter-precore region and the pre-S–S region, respectively, were used in the PCR. The amplification followed a protocol described earlier, with slight modifications of cycling temperatures (94°C for 1 min, 45°C for 1 min, and 72°C for 2 min) (15).

TABLE 1.

Primers used for amplification and sequencing of HBV and HDV^a

Primer	Sequence	Position
A	5′-GAA GGA AGG CCC TCG AGA ACA AGA-3′	1290–1267b
EF3	5′-TGC CAT GCC GAC CCG AAG AGG AAA-3′	885–908b
KL6	5′-GGA AAG AAG TCA GAA GGG A-3′	1974–1956c
KL12	5′-GGG TCA CCA TAT TCT TGG-3′	2814–2832c
KL15	5′-CTC AGG CTC AGG GCA TA-3′	3083–3099c
KL28	5′-GAG ACC ACC GTG AAC GCC-3′	1611–1628c
KL33	5′-ACC ACT GAA CAA ATG GCA CTA-3′	701–680c

^aAll primers were used for amplification and sequencing of HBV and HDV except for KL15, which was used just for sequencing. Primers A (1) and EF3 were used for HDV and the KL primers were used for HBV. 

^bNumbering is according to Makino et al. (18). 

^cNumbering is according to Okamoto et al. (20). 

During the course of the study, standard precautions were taken to prevent carryover between samples, both at the time of sampling and during all pre- and postamplification work (12), and there was never any evidence of contamination.

Extraction and amplification of HDV RNA.

HDV RNA was isolated using the RNAWIZ RNA Isolation Reagent (Ambion, Austin, Tex.). A total of 140 μl of serum was incubated in 280 μl of RNAWIZ for 5 min. After the addition of 84 μl of pure chloroform, the samples were incubated for another 10 min and then centrifuged for 15 min in a microcentrifuge. The upper aqueous phase was transferred to a new microcentrifuge tube, to which 210 μl of sterile double-distilled water (ddH₂O) was also added. An addition of 420 μl of isopropanol was followed by a 10-min incubation and then centrifugation for 15 min. The RNA pellet was washed in 420 μl of cold 75% ethanol, centrifuged for 5 min, and subsequently air dried for 10 to 15 min. The pellet was resuspended in 80 μl of sterile ddH₂O and, if necessary, the suspension was stored at −20°C. All steps were performed at room temperature, and mixing was achieved by inverting the tubes a few times. To amplify the purified HDV RNA, the Access RT-PCR System Kit (Promega, Madison, Wis.) was used according to the manufacturer's instructions, with 20 μl of RNA suspension serving as template for the reverse transcription. Primers EF3 and A (1) (Table 1) were used in the ensuing PCR, amplifying a 400-nucleotide (nt)-long region of the HDV genome proposed for the classification of HDV genotypes.

Analysis and sequencing of viral DNA.

The products of the HBV and HDV PCRs were first analyzed on a 1.8% agarose gel and then directly sequenced as previously described (11), using the primers in Table 1.

Phylogenetic analyses.

Nucleotide sequences of both HBV and HDV were compared to previously reported sequences and aligned by the Clustal method in the program MEGALIGN of the Lasergene99 software package (DNASTAR, Madison, Wis.). CLUSTALX version 1.8 (28) was then used to generate the data files to be used in a combination of the programs DNADIST and FITCH. This produced phylogenetic trees using the distance matrix method, which estimates the pairwise genetic similarity (or distance) between strains. The method starts by grouping the two most homologous strains, then adding strains of increasing variability one by one. The resulting trees were finally visualized with DRAWTREE. In a radial tree based on a distance matrix, the genetic distance between any two strains is displayed by the combined length of the branches connecting them. The programs DNADIST, FITCH, RETREE (for layout modifications), and DRAWTREE are all part of the Phylogeny Inference Package (PHYLIP) version 3.5c (2) (http://evolution.genetics.washington.edu/phylip.html).

Statistical methods.

The chi-square test with Yates' correction was used.

Nucleotide sequence accession numbers.

The 31 Samara pre-S2 sequences can be found in GenBank under accession no. AF247933 to AF247963. The four new HDV sequences have been deposited in GenBank under accession no. AF247964 to AF247967.

RESULTS

Serological testing.

Serum samples from 105 patients from Samara with signs of acute viral hepatitis were analyzed. Due to limited volumes of serum, all markers could not be analyzed in every sample. Of the 105 patients, 94 had an HBV infection as defined by positive HBV DNA PCR and/or HBsAg and/or anti-HBcIgM (Table 2). Eighty samples were HBV DNA PCR positive. Of the 84 samples tested for anti-HBc IgM, the majority (80%) were positive, indicating acute HBV infection.

TABLE 2.

Percentage positive, by gender, of patients tested for different viral hepatitis agents

Virus	No. of patients tested	% Positive
		Males	Females	Total
HAV	54	17	25	19
HBV	105	89	90	90
HCV	100	70	53	67
HDV	103	40	19	36

Antibodies to HCV were detected in 67 of the 100 samples tested. There was a considerable, although not significant, difference between the number of females and males infected with HCV: 10 of 19 and 57 of 81, respectively. Fifty-four samples were tested for the presence of IgM antibodies to hepatitis A virus (anti-HAV IgM). Ten samples were positive, indicating acute HAV infection.

Infection with HDV, detected by anti-HDV antibody and/or HDVAg, was found in 37 of the 94 HBV-positive samples (39%). HDV infection also appeared more commonly, but not significantly so, in males than in females. Coinfection with HBV and HDV, indicated by a positive anti-HBc IgM result, was found in 23 patients. HDV superinfection of a chronic HBV carrier occurred in nine cases, and in five HDV-positive cases, the absence of an anti-HBc IgM result precluded any further characterization.

Molecular analysis of HBV.

All samples were tested for HBV DNA by PCR using two separate sets of primers. Eighty samples were positive with primer pair KL28 and KL6. A 159-nt long region (positions 1742 to 1900 according to the numbering of Okamoto et al. [20]), spanning the core promoter region and precore gene, was directly sequenced. There were no deletions in the core promoter region and only one strain (no. 13) belonged to HBV genotype A. The remaining strains were all of genotype D. Commonly seen mutations in the core promoter (nucleotide positions 1762 and 1764, AGG to AGA or TGA) and in the precore region (G-to-A mutation at position 1896 leading to a translational stop) were found in some strains. HBeAg and anti-HBe status was not tested in all HBV-positive samples, but in 13 of 17 samples with anti-HBe antibody and lacking HBeAg, both the core promoter and precore region had the wild-type sequences. One strain (no. 98) had multiple mutations in both the core promoter and the precore gene, including G-to-A mutations at precore positions 1896 and 1899. The patient infected with this strain had fulminant fatal hepatitis.

The PCR using primer pair KL12 and KL33 amplified the whole pre-S region and most of the S gene, and 78 samples were positive in this PCR. No major deletions were found in these samples, as judged by the size of the bands in the agarose gels. Thirty-one of these samples were randomly chosen for further analysis by direct sequencing. The whole of the pre-S1 and pre-S2 genes and a region of the S gene spanning codon 118 to 170, encompassing the a-determinant, were sequenced. No insertions or deletions were found, and there were no mutations in the a-determinant. The sequences of the S gene confirmed that all strains characterized but one belonged to genotype D.

The pre-S2 region, being the most variable region sequenced, was chosen for phylogenetic analysis. The sequences from the 30 Samara genotype D samples were compared to those from 12 other, geographically different strains of the same genotype (8, 9; three unpublished strains). The resulting radial tree clearly shows that the prevailing strains from Samara fall into two distinct clusters which are separate from strains with other geographic origins (Fig. 1).

FIG. 1.

Radial tree showing inferred relationships between pre-S2 sequences of 30 Samara HBV strains and 12 other HBV strains (9 previously reported and 3 unpublished). The abbreviation Sam indicates new strains from Samara. If sequences were identical between two or more Samara strains, only one strain is represented in the tree, which is the case for Sam7 (12 identical strains) and Sam35 (8 identical strains). Other strains are from India (In), Iran (Ir), Kosovo (Ko), Lebanon (Le), New Zealand (NZ), Romania (Ro), Samoa (Sa), Somalia (So), Sweden (Sw), and Turkey (Tu).

Fulminant hepatitis.

One quarter of the 105 consecutive hepatitis patients included in this study had fulminant hepatitis (25.7%). Table 3 shows the viral etiology of the fulminant cases. Of the 27 fulminant cases, 13 were infected with HDV. A significant sex difference between HDV-positive and HDV-negative patients with fulminant hepatitis was seen; males were HDV positive more often (P < 0.05). This skewed sex distribution was also seen in the total proportion of males versus females acquiring fulminant hepatitis, as females were seen to have a statistically significant higher risk of fulminant disease (P < 0.005). When the 14 HDV-negative patients with fulminant disease were studied more closely, 7 of the 14 were also anti-HCV positive. With the exception of sample no. 98, commonly occurring mutations in the core promoter and precore region were not more prevalent in the HBV strains infecting these patients (Table 4).

TABLE 3.

Viral etiology and sex distribution of 27 patients with acute fulminant hepatitis in Samara, Russia^a

Infective agent(s)	Male	Female	Total
HBV only	2	5	7
HBV + HCV	3	4	7
HBV + HDV	3	0	3
HBV + HCV + HDV	8	2	10
Total	16	11	27

^aTotal number of patients with hepatitis in the study: 105 (84 male, 21 female). 

TABLE 4.

Characteristics of 14 fulminant HDV-negative hepatitis patients

Sample no.	Sex	Age	HCV infection	HBV PCR	Sequence analysis (core promoter/precore)
2	M	17	Yes	−	NDa
14	M	37	Yes	+	Wild/wild
15	F	30	Yes	+	Wild/wild
16	F	76	No	+	Wild/stop28b
25	F	22	No	+	AGAc/wild
48	M	20	No	(+)d	ND
58	F	28	No	+	Wild/wild
59	F	19	Yes	+	Wild/wild
65	F	16	Yes	+	Wild/wild
74	F	18	No	+	Wild/wild
83	M	18	Yes	+	Wild/wild
97	F	24	No	(+)	ND
98e	M	39	No	+	TGA/stop28f
117	F	18	Yes	−	ND

^aND, not done. 

^bTGG-to-TAG mutation in precore codon 28, creating a translational stop. 

^cAGG-to-AGA mutation in core promoter, nucleotide position 1764, often associated with an anti-HBe positive phenotype. 

^d(+), Sample was too weak in HBV DNA PCR to be sequenced. 

^eFatal fulminant hepatitis B. 

^fThis strain had multiple changes in the core promoter and precore regions. 

Molecular analysis of HDV.

In order to determine the genotype of the HDV strains prevailing in Samara, the genomes of four randomly chosen strains were isolated, reverse transcribed, amplified by PCR, and then sequenced using primers EF3 and A. The 257-nt sequence from nt 934 to 1190 (numbering according to Makino et al. [18]) was then used for BLAST searches in GenBank, revealing that Samara HDV belongs to genotype I. The Samara strains also clustered with other genotype I sequences in the subsequent phylogenetic analysis and not with either of the two other genotypes (data not shown).

The Samara HDV sequences were compared to 47 previously reported HDV genotype I sequences representing different geographical regions. The resulting radial phylogenetic tree is shown in Fig. 2. The tree shows two major clusters, with the Samara HDV strains located between them. Some homology to Far Eastern and Eastern European strains is seen. Interestingly, there is also a close match with another Russian strain. Unfortunately, there is no further epidemiological information about this strain (24).

FIG. 2.

Radial tree indicating relationship between 51 strains of HDV genotype 1. The abbreviation Sam indicates new strains from Samara. Other strains originated from Albania (Al), Archangelos, Rhodes, Greece (Ar), Canada (Ca), France (Fr), mainland Greece (Gr), Italy (It), Japan (Ja), Nauru, South Pacific (Na), Romania (Ro), Russia (Ru), Taiwan (Ta), Turkey (Tu), and the United States (US), and all were obtained from GenBank. The strain denominations are according to the respective authors, as found in GenBank.

DISCUSSION

In Samara, a city in the southeastern part of European Russia, there has been a sudden and sharp rise in the incidence of viral hepatitis over the last 2 years. Compared to the figures for the whole country, which have been stable over the last 4 years, the incidence of HBV infection has doubled and that of HCV has tripled in Samara between 1998 and 1999 (23; unpublished observations). When an epidemiological phenomenon like this is observed, it is often caused by a change either in the population or in the infecting agent. Until the beginning of the 1990's, Samara and its surrounding region were isolated and foreign travel into the area was not permitted. After this ban was lifted, increased travel brought an increase in illicit drug commerce and the number of IVDUs (mainly young) increased. It is possible, therefore, that the rising incidence of viral hepatitis in Samara is the result of the increasing illicit intravenous drug use.

The situation is similar to that seen in southern Sweden several decades ago, when HDV infection was introduced in 1973 and spread among the IVDUs, with a yearly rise in incidence (5). An unusually large proportion of HDV infections among the IVDUs were fulminant (16), an observation also made in Ireland (25). Fulminant HDV hepatitis has also been described from South America (4, 17). Later studies investigating fulminant strains from the South American studies have shown that HDV genotype III appears to be associated with more aggressive disease (1).

In the present study, as many as 39% of the HBV-positive patients were also infected with HDV, and one-third of these HDV-positive patients had fulminant hepatitis. In order to assess the prevailing HDV genotype in Samara, four strains isolated from patients with different age and sex patterns were genotyped, and they were all of genotype I. The sequences used for HDV genotyping in this study were slightly shorter than those used by some other groups (1, 29), and after analyzing the sequences, we found that alignment with the shorter ones still gives a good representation of the comparison of the entire genome (data not shown).

The possibility of mutations in HBV DNA giving rise to more aggressive disease and being another cause of the high proportion of fulminant cases seen in this study was then investigated. The specific mutations that have been pinpointed most often as being associated with fulminant disease are located in the core promoter at nucleotide positions 1762 and 1764 and in the precore gene at nucleotide position 1896 (6, 22, 27). Not all HBV strains from fulminant non-HDV cases could be sequenced in this study, but in 7 of 10 strains, both the core promoter and the precore regions were of the wild type. This supports the study by Laskus et al. (13), in which fulminant hepatitis B (of mainly genotype A or D) in the United States could not be correlated with core promoter or precore mutations.

Only one strain (no. 98) showed multiple changes in both the core promoter and precore region. No larger deletions could be seen in the pre-S and S genes. The patient fatally infected by this strain was HCV and HDV negative, and no other explanation for the fulminant course of illness could be found.

The most notable finding in the group of patients with fulminant hepatitis was the significant difference in sex distribution. That a larger proportion of males were infected with HDV and acquired fulminant HDV infection can be explained by intravenous drug use being by far more common among young males in Samara (unpublished observation). The disproportionately high number of females with non-HDV fulminant viral hepatitis, however, is not explained by differences in social behavior. To our knowledge, such a difference in sex distribution has not been described before.

Taking into account the HBV sequences from both the core promoter-precore region and the pre-S–S region in this study, there were remarkably few changes when comparing them to a consensus sequence from strains of the same genotype. No deletions, and only a limited number of point mutations, were found. This is in line with the view that some of these commonly occurring mutations arise from selection through immunological pressure in the chronic carrier (13). Eighty percent of the patients in this study who were tested for anti-HBc IgM were positive, and thereby had, by definition, acute HBV infections. This, taken together with the observed sharply rising incidence, implies that HBV (and HDV) have recently gained access to this community.

A phylogenetic analysis of prevailing strains in a community together with strains isolated from other regions can serve two main purposes. Firstly, it will demonstrate whether many different strains circulate in the community. Secondly, it may indicate the geographic origin of the prevailing community strains. The advantage of using the distance matrix method and displaying the results as a radial tree, as we chose to do in this study, is that it demonstrates very clearly how far apart different strains are from each other phylogenetically.

Phylogenetic analysis of the HBV pre-S2 gene demonstrated two clear clusters of genotype D strains present in Samara that were clearly separate from strains of the same genotype isolated from other regions of the world. It thus appears that two main HBV strains circulate in the Samara region and that they are distinct from strains from the other regions studied.

In contrast, the phylogenetic tree emerging from comparisons of four Samara HDV strains with 47 HDV strains from other regions, showed that the Samara strains were similar to each other, but did not cluster close together. Such clusters could be seen among strains isolated in the United States. The similarity between HDV strains in the United States and also between strains in some regions of Greece have been suggested to represent the recent introduction of HDV into these populations (24). The Samara HDV strains appeared to be phylogenetically closest to Far Eastern and Eastern European strains.

In summary, this study has shown that the abruptly increasing number of cases of viral hepatitis in the previously isolated Russian city of Samara is caused by HBV, HCV, and HDV, with the majority of patients having an acute HBV infection. Although many of the patients studied were also infected with HCV, acute HCV infection is often subclinical. Acute HBV infection, on the other hand, more frequently leads to the kind of severe disease seen in many of the patients in Samara. An unusually large proportion of these patients have fulminant cases, and a disproportionately large number of non-HDV fulminant cases in females has also been demonstrated. Two main HBV strains of genotype D appear to circulate in the community, whereas the HDV strains are phylogenetically more distinct from each other. There is a potentially serious situation in a community with such a rising incidence of HBV, HCV, and HDV if other blood-borne viruses, such as human immunodeficiency virus, are introduced.