Retrospective analysis of acute HBV infections occurred in 1978–79 and 1994–95 in North-East Italy: increasing prevalence of BCP/pre-core mutants in sub-genotype D3

Hepatitis B virus (HBV) infection is a global health problem concerning nearly 257 million chronic carriers at risk of cirrhosis and liver cancer [1]. HBV is an enveloped Hepadnavirus with a circular incomplete double stranded DNA genome of 3.2 Kb [2]. Its genetic variability is greater than that of any other DNA virus, as a result of viral genome replication by a reverse transcriptase lacking the proof-reading function [3].

Sequence analysis has shown that specific nucleotide variations play a role in pathogenesis and/or have impact in public health issues. Among them, mutations in the basal core promoter/pre-core (BCP/pre-core) region (associated with HBeAg negative hepatitis - a severe form of chronic liver disease - and with fulminant acute hepatitis) and those in the S coding region, especially in the main target of antibodies - the “a” determinant (associated with escape from vaccine induced immunity) are particularly well documented [4].

On the basis of nucleotide sequence analysis, HBV has been so far classified in ten genotypes (A to J) and several sub-genotypes, based on divergences of ≥7.5% and ≥ 4%, respectively [5, 6].

HBV genotypes have an ethno-geographic distribution, genotypes D (HBV-D) and A (HBV-A) being the most ubiquitous and responsible for the majority of infections in Northern and Southern Europe [7]. Very high degree of heterogeneity has been described for genotype D, with ten sub-genotypes (D1 to D10) so far reported [8, 9]. The divergences are strictly associated to population-related factors such as spread rate of epidemics and main routes of transmission. During the 1940s–1950s (that included the 2nd World War), HBV-D divergence and epidemic exponential growth took place as a result of unsafe medical practices and unscreened blood and blood derivatives. Another epidemic, mainly linked to HBV-D3, took place in the 1960s–1970s and involved subjects with high risk behaviour, such as intravenous drug users (IDU) [10]. HBV genotype A (particularly sub-genotype A2) largely spread between 1960s and early 1980s as a result of sexual transmission [11].

In Italy, the thirty years between 1940s and 1970s were crucial for the spread of HBV epidemics. At the end of the 1970s, more than 2% of the general population was HBsAg carrier, though with geographical differences; in those years, the circulation of sub-genotypes and viral mutants with clinical relevance were poorly studied because of the absence of appropriate routinely applied laboratory methods. In the following 20–30 years, HBV epidemiology significantly changed as a result of the improvements in socio-economic conditions and the implementation of compulsory vaccination, that started in Italy in 1991 [12].

The study of the HBV strains circulating in a geographical area is usually carried out by studying viruses from chronic patients. However, in most chronic carriers it is unknown when HBV infection was acquired: the transmission event might date back as few as weeks/months or as much as decades. In contrast, the study of acute cases provides an accurate overview of the strains actually transmitted over the enrolment period: being 6 months the upper limit for incubation time, it can be safely assumed that transmission did not occur earlier than 6 months before onset of symptoms.

In the ‘80s and ‘90s, two remarkable events might have impacted on HBV strains transmitted in North-East Italy: (a) a significant increase of acute HBV infections in young males and a parallel increased proportion of drug addicts were observed between 1978 and 1982 [13]; (b) national preventive measures linked to the anti-HIV educational campaign started to be implemented, in 1985 in the area object of the present study, in 1988 at national level. However, whether or not those events produced changes in the HBV strains circulating in this Italian area is unknown. The only available data about HBV strains circulating in North and North-East Italy were from chronic carriers (rather than acute cases) enrolled several years later, in 1993–2005 and 2000–2004 [11, 14]; conversely, studies from acute cases in Italy were either carried out in Southern Italy or, as part of a multicentre study, only included a small survey from North Italy. None of them included acute cases that had occurred earlier than 1999; thus, no data about HBV strains circulating in acute infections in ‘80s and early ‘90s are available [15‐17].

The aim of the present study was to evaluate if the epidemics observed in 1978–1982 in North East Italy as well as the anti-HIV educational campaigns (started locally in 1985 and nationally in 1988) had any impact on HBV genotypes/sub-genotypes and clinically relevant mutants circulating in acute cases in this Italian area. The aim was addressed by comparing HBV sequences from acute cases enrolled in 1978–79 and 1994–95, i.e. before and after those events. Two viral genome regions of HBV (a 374 nt BCP/pre-core/core region and a 1295 nt region spanning the entire S open reading frame) were sequenced and subjected to phylogenetic analysis as well as search for BCP/pre-core and S mutations.

The demographic, clinical and virological characteristics at hospital admission of the 80 acute hepatitis B patients included in the study are reported in Table 1. All patients were HBsAg and anti-HBc IgM positive. They were enrolled in two periods 16 years apart: 50 cases in 1978–79 and 30 cases in 1994–95. The two groups were similar for demography, distribution of HBV markers and clinical features.

Co-infection with HDV and HCV was detected in 4 and 24% cases, respectively, in the 1978–79 group vs. 10 and 25.8% in the 1994–95 group (p = 0.36 for HDV co-infection and p = 0.79 for HCV co-infection).

In one half cases, none of the investigated risk factors for HBV infection was reported. The IDU risk factor was significantly more frequent in the 1994–95 group (8/30, 26.7%) than in the 1978–79 group (4/50, 8%) (p = 0.048). The frequency of other risk factors (sexual or household contacts and professional exposure) did not differ between the two groups (p = 0.16).

In North-East Italy, remarkable changes of the epidemiology of acute hepatitis B were observed between 1978 and 1982; they were mainly related to the spread of parenteral drug abuse; the epidemics peaked in 1982–1983 [13]. However, possible changes of the circulating HBV strains remained uninvestigated, mainly because the routine methods for molecular characterisation became largely applied only several years later.

The first descriptions of the HBV strains circulating in Northern Italy were reported in 2006 and 2007, when two studies described the strains detected in two surveys of chronic carriers enrolled, respectively, between 2000 and 2004 and between 1993 and 2005 [11, 14]. In the present retrospective study, HBV strains from the same area (North-East Italy) dating back 1978–79 and 1994–95 were analysed, but exclusively from acute cases: this provides an accurate overview of the strains actually transmitted in those years because, in contrast with chronic carriers, the time point of infection can be safely defined.

The studied acute cases had preceded (1978–1979 cases) or had occurred in the early years of (1994–95 cases) public health interventions that will then prove crucial for the modification of HBV epidemiology in Italy: the wide application of the preventive measures linked to the anti-HIV educational campaign (started in 1988) and the implementation of HBV mass vaccination (started in 1991).

The results show that the same genotypes and sub-genotypes were stably transmitted in this Italian area through 16 years. In particular, HBV-D was the most prevalent in both the 1978–79 and the 1994–95 groups, while HBV-A accounted for the remaining cases, with the unique exception of an HBV-E strain detected in one single patient in 1978. Even at the sub-genotype level, the D3, D2 and D1 in decreasing abundance and the A2 were detected in both groups at similar frequency. Thus, the epidemiological and public health events occurred between the two investigated periods did not impact on the distribution of transmitted HBV genotypes/subgenotypes.

Search for BCP/pre-core mutations showed no such mutations in genotype A cases. At least for position 1896, the finding is in agreement with previous reports: the G1896A pre-core mutation has been rarely observed in genotype A, as it would disrupt an essential stem-loop structure in the ε signal essential for pre-genomic RNA packaging [23]. In contrast, the analysis has highlighted a significantly increased prevalence of BCP/pre-core mutants in acute HBV-D infections, particularly in HBV-D3, from 1978 to 79 to 1994–95.

These findings can help to explain a previously observed major change in epidemiology of HBV in Italy over 30 years. By the end of the 1970s, in Italy most chronic carriers were HBeAg(+): the carrier cohort showed low proportion of HBeAg(−) subjects; however, at the beginning of the year 2000 the scenario was inverted, with high proportion of HBeAg(−) carriers [12, 24, 25]. Two factors had been invoked to explain this change: (a) reduction of the overall HBV incidence (as result of prevention campaigns and vaccination) that reduced replenishment of the carrier cohort with new HBeAg(+) carriers and (b) aging of the HBV carrier cohort, leading to selection of BCP/pre-core mutants and, consequently, change from HBeAg(+) to HBeAg(−) phenotype [12, 24, 25]. The present study suggests a third cooperating factor: the increased transmission of BCP/pre-core mutants. The increased detection of BCP/pre-core mutants in acute cases implied replenishment of the carrier cohort with an increased proportion of HBeAg(−) cases; in turn, the widespread parenteral drug abuse likely resulted in increased likelihood to transmit BCP/pre-core mutant viruses, due to their preferential circulation in IDUs, giving a contribution to their spread. The epidemic of acute HBV infections observed in North East Italy between 1978 and 1982 was mainly caused by parenteral drug abuse: the proportion of IDUs reached 42% of total acute cases [13]. In the next years, between 1983 and 1988, the HBV incidence dropped and the proportion of IDUs in acute cases also decreased markedly; however, the number of new subjects starting drug abuse did not decline and the proportion of IDUs, after a minimum in 1987, increased again in 1988 [13]. In the present study, the proportion of IDUs in the 1994–95 group was 26.7%, so 3-fold more frequent than in 1978–79 group and also higher than in 1988; significantly, the IDU risk factor was mostly reported by patients infected by HBV-D3, a sub-genotype highly circulating among IDUs [8]. The tendency of D3 variants found in IDUs to group into a phylogenetic clade (Fig. 3 and Additional file 1: Figure S1, “IDU enriched” clade) suggests that their circulation was limited, to a certain degree, to IDUs; however, spread into the general population had also occurred, because strains from cases with different or unknown risk factors were also interspersed in the same clade. Thus, increased circulation of BCP/pre-core mutants in acute cases likely cooperated with the two other previously reported factors (see above) to change the proportion of HBeAg(−) cases in the Italian carrier cohort, from the low proportion at the end of 1970s to the high proportion observed at the beginning of the 2000s [12, 24, 25].

Regarding analysis of the S region, an above mentioned study of HBV in the same geographical area in chronic carriers between 1993 and 2005 reported frequent mutation of the “a” determinant, including the G145A/R mutation involved in vaccine escape [11].

In the present work no such mutations were observed. Two main differences between the two studies are likely responsible for this result: (a) different infection status (acute vs chronic cases) and (b) different sampling period (1978–79 and 1994–95 vs 1993–2005). It is not surprising that no escape mutations were observed in the present study. Acute cases had been enrolled either several years before (1978–79 group) or just a few years after (1994–95 group) the introduction of mandatory vaccination, which started in Italy in 1991 in two population cohorts: 3 months of age newborns and 12 years old children. Thus, in 1978–79 any strain with a vaccine escape polymorphism would have not been subjected to any vaccine selection, while in 1994–95 the vaccinated cohorts were 0–4 years and 12–16 years, so the selection pressure for escape mutants was likely still negligible. In contrast, the survey analysed in De Maddalena et al. included chronic cases, sampled about a decade later (between 1993 and 2005): the reported escape mutants could have been either the result of natural selection during chronic infection because of their ability to evade the host immune response or had been directly acquired when escape mutant strains had possibly started spreading (cases were in fact sampled up to 14 years after introduction of mandatory vaccination) [11].

In conclusion, the present study shows that the proportion of genotypes and sub-genotypes in acute cases in North-East Italy remained unchanged over 16 years, from 1978 to 79 to 1994–95. In contrast, the proportion of BCP/pre-core mutants increased more than three-fold, mostly in HBV-D3, a sub-genotype highly circulating in IDUs; accordingly, the IDU risk factor was also more frequently reported, suggesting that IDUs contributed to the spread of the BCP/pre-core mutants.

These findings contribute to explain a previously described major change in HBV epidemiology in Italy: the proportion of HBeAg(−) cases in the carrier cohort changed from a low proportion in late 1970s to a high proportion at the beginning of the 2000s. In addition to other recognized factors, the increased circulation of BCP/pre-core mutants likely represents a further factor that contributed to this change.