Purpose:

Hepatitis B viral (HBV) DNA is frequently integrated into the genomes of hepatocellular carcinoma (HCC) in patients with chronic HBV infection (chronic HBV, hereafter), whereas the frequency of HBV integration in patients after the disappearance of HBV (prior HBV, hereafter) has yet to be determined. This study aimed to detect integration of HBV and adeno-associated virus type 2 (AAV2) into the human genome as a possible oncogenic event.

Experimental Design:

Virome capture sequencing was performed, using HCC and liver samples obtained from 243 patients, including 73 with prior HBV without hepatitis C viral (HCV) infection and 81 with chronic HBV.

Results:

Clonal HBV integration events were identified in 11 (15.0%) cases of prior HBV without HCV and 61 (75.3%) cases of chronic HBV (P < 0.001). Several driver genes were commonly targeted by HBV, leading to transcriptional activation of these genes; TERT [four (5.4%) vs. 15 (18.5%)], KMT2B [two (2.7%) vs. five (6.1%)], CCNE1 [zero vs. one (1.2%)], CCNA2 [zero vs. one (1.2%)]. Conversely, CCNE1 and CCNA2 were, respectively, targeted by AAV2 only in prior HBV. In liver samples, HBV genome recurrently integrated into fibrosis-related genes FN1, HS6ST3, KNG1, and ROCK1 in chronic HBV. There was not history of alcohol abuse and 3 patients with a history of nucleoside analogue treatment for HBV in 8 prior HBV with driver gene integration.

Conclusions:

Despite the seroclearance of hepatitis B surface antigen, HBV or AAV2 integration in prior HBV was not rare; therefore, such patients are at risk of developing HCC.

Translational Relevance

Integration of hepatitis B virus (HBV) into a host genome is known to alter the transcription of driver genes in patients with chronic HBV infection (chronic HBV, hereafter), followed by the development of hepatocellular carcinoma (HCC). However, this has not yet been shown in patients after the disappearance of HBV (prior HBV, hereafter). Here we describe recurrent HBV integration into driver genes such as TERT, KMT2B, CCNA2, and CCNE1, leading to their transcriptional activation in cases of both chronic HBV and prior HBV. HBV integration, followed by the upregulation of driver genes, was not rare in prior HBV. Therefore, even after the seroclearance of HBsAg due to spontaneous regression or hepatitis B treatment, prior HBV cases are still at a high risk of developing HCC.

Integration of hepatitis B virus (HBV) into a host genome is known to alter the transcription of driver genes (1–3). Genetic damage and chromosomal instability with subsequent chromosomal rearrangements and generation of oncogenic chimera can result (4, 5). Thus, in patients with chronic HBV infection (chronic HBV, hereafter), genomic instabilities due to HBV integration may contribute to the development of hepatocellular carcinoma (HCC), even in the absence of cirrhosis (6, 7).

On the basis of PCR in proximity to Alu repeats, the HBV genome was found to insert randomly into human genomes in a small cohort of patients; however, the preferential sites of insertion were not specified (8, 9). On the other hand, development of high-throughput sequencing technologies has enabled genome-wide surveys of HBV integration (1, 2, 10) and the identification of several HBV target genes (9, 11, 12). In addition to HBV integration into known cancer-related genes, such as TERT, KMT2B, and CCNE1 (11–13), new genes that are sites of HBV integration, including CCNA2, FN1, SEND5, ROCK1, and DDX11L1, have been identified and their functional consequences examined (2, 3, 8, 14). Furthermore, transcriptome sequence could detect viral–human chimeric transcript such as HBx-LINE1 (5) and HBV-KMT2B (11). On the other hand, HBV integration in nontumor tissues showed random distribution throughout the host genome, but FN1 was recurrently targeted by HBV in liver samples (15).

Compared with whole-genome sequencing, virome capture sequencing is a cost-effective methodology with higher specificity and sensitivity in detecting viral integration events in the human genome. An experimental and computational method called high-throughput viral integration detection (HIVID) was also reported to be an efficient method to detect HBV integration events (3, 15, 16). With HIVID, HBV sequence fragments were enriched using a set of HBV probes and then sequenced on a high-throughput platform. HIVID detected HBV integration events more effectively, at single base pair resolution, compared with whole-genome sequencing using the same tissue samples (3, 13, 15).

However, there have only been a few studies investigating patients with “occult HBV infection (OBI)” who were negative for serum HB surface antigen (HBsAg), but in whom HBV DNA was detected in their liver tissues (17). Surprisingly, the frequency of detection of integrated HBV using Alu-PCR was similar between patients with OBI (75.5%) and those with chronic HBV (80%). Although this study included only a small number of patients (n = 69), accumulation of data from such patients will elucidate whether prior infection with HBV contributes to HCC development.

Adeno-associated virus type 2 (AAV2) is a nonpathogenic virus. Infection with this virus is reported to induce necrotic death and inhibit proliferation of breast cancer cells (18). In contrast, genes that drive the development of HCC, such as TERT, CCNA2, CCNE1, TNFSF10, and KMT2B are recurrently targeted by AAV2, leading to upregulation of these genes (19). Thus, whether AAV2 exerts tumor-suppressive or oncogenic effect in carcinogenesis remains controversial.

In this study, we investigated the integration of HBV and AAV2 into the genomes of patients with HCC and compared the frequency of viral integration events between patients after HBV seroclearance (prior HBV, hereafter) and those with chronic HBV. Furthermore, we estimated the effect of viral integration on the development of HCC and elucidated the clinical significance of prior HBV infection.

Patients and tissue samples

Patients with HCC who underwent liver resection between 2011 and 2016 in the Department of Digestive Surgery, Nihon University School of Medicine (Tokyo, Japan), were recruited into this study. The study was approved (protocol number: 131) by the institutional review boards of Nihon University, and each participant provided written informed consent. All of the clinical investigations were conducted according to the principles outlined in the Declaration of Helsinki. Surgical specimens were immediately cut into small pieces after resection, snap frozen in liquid nitrogen, and stored at −80°C.

Hepatitis B viral infection

Patients who were negative for HBsAg, but positive for anti-hepatitis B surface antibody (HBsAb) and/or anti-hepatitis B core antibody (HBcAb) were defined as “prior HBV” (20). Patients positive for HBsAb and vaccinated against hepatitis B were not included in the study. Those who were positive for serum HBsAg were considered to be “chronic HBV.” Patients negative for all HBV and hepatitis C virus (HCV)-associated markers were defined as “non-B non-C patients” and included in the study as a negative control. HBcAb levels were measured by Cobas electrochemiluminescence immunoassay (Roche), and patients were determined positive for HBcAb if the index was < 1.0. HBV-DNA was measured by a Taqman PCR assay (Roche) with a detection limit of 2.1 Log copies/mL.

Genomic DNA and total RNA extraction

Genomic DNA and total RNA were isolated from tissue specimens using a QIAamp DNA Mini Kit (Qiagen) and TRIzol (Invitrogen), respectively, according to the manufacturers' protocols (21, 22). Genomic DNA and total RNA concentrations were determined by dsDNA BR Assay using a Qubit (Life Technologies) and NanoDrop Spectrophotometer (NanoDrop Technologies), respectively, and RNA integrity was assessed using an Agilent 2100 Bioanalyzer (Agilent Technologies).

Virome capture sequence

We used a SureSelect DNA capture custom kit (Agilent) for the viral capture experiment, in which capture probes were designed with 2 × tiling for the HBV genome (NC_003977.1) and 4 × tiling for the AAV2 genome (NC_001401.2). Five-hundred nanograms of genomic DNA were sheared using Covaris E-210 and SureSelect adapters were ligated using a KAPA HyperPrep kit (KAPA Biosystems) to fragments for amplification. Amplified DNA (750 ng) was hybridized with the virus capture probes and indexed according to the SureSelect protocol. Sequencing libraries were multiplexed, and paired-end 150-bp–long sequencing was performed on a MiSeq (Illumina) platform.

RNA sequence

An RNA sequence library was prepared using a TruSeq Stranded mRNA Library Prep kit (Illumina) and sequenced on a HiSeq 2000 platform. RNA sequences were performed to investigate expression of genes targeted by HBV or AAV2.

Data analysis

Data flow is illustrated in Supplementary Fig. S1. Low-quality reads and adaptor sequences were removed using MiSeq Reporter software (Illumina). Read sequences were mapped to HBV and AAV2 reference genomes and human reference genome (hg19) using BWA-MEM. PCR duplications were removed. Paired-end reads that mapped to the HBV genome at both ends were used for viral typing and a coverage analysis. The paired-end reads that mapped to virus and human genomes were used to detect the breakpoints of virus integration sites. Integration sites that were supported by single-read sequences were considered evidence of “nonclonal integration” events in the genome of a single cell, because not only HCC samples but also corresponding liver tissues of chronic HBV commonly harbor integrated HBV genomes at many sites, resulting in shallow and wide distribution of viral DNA (1). On the other hand, integration sites that were supported by two or more reads were considered evidence of “clonal integration” events in the genomes of cells that proliferated after viral integration.

Statistical analysis

Continuous and categorical variables collected from each group were statistically analyzed using the Wilcoxon rank-sum and Fisher exact test, respectively. All statistical analyses were performed using the JMP 12.0.1 statistical software package (SAS Institute Inc.). In all analyses, P < 0.05 was considered statistically significant.

Sequence data

Sequence data are available on the Japanese Genotype-phenotype Archive (JGA, http://trace.ddbj.nig.ac.jp/jga) under accession number JGAS00000000194.

Patients

A total of 243 patients, comprising 73 prior HBV without HCV infection, 81 chronic HBV, 56 prior HBV with HCV infection, and 33 non-B non-C cases as negative controls, were included in this study. In the 73 prior HBV without HCV, 5 (6.8%), 19 (26.0%), and 49 (67.1%) patients were positive for HBsAb, HBcAb, or both, respectively. HBcAb antibody titers in the 5 patients with HBsAb, but not HBcAb, were relatively lower, perhaps due to spontaneous resolution (23). Serum HBV-DNA was detected in 52 (64.1%) patients in chronic HBV (median, 2.1 log copies/mL; range, 0–7.1), whereas there was one (1.3%) HBV-DNA–positive case in prior HBV without HCV. Five (6.8%) prior HBV without HCV and 42 (51.8%) chronic HBV, respectively, had undergone antiviral treatments with nucleoside analogues before surgery (Supplementary Table S1). The patients with prior HBV without HCV were significantly older than those with chronic HBV (P < 0.001), and males (P = 0.011), alcohol abuse (P = 0.008), and diabetes mellitus (P < 0.001) were significantly more common in this group. On the other hand, the rate of fibrosis grade III or IV of the liver was significantly higher in the chronic HBV group (P = 0.005; Supplementary Table S2).

Viral integration into the human genome

Using virome capture sequencing, we investigated viral integration into the human genome in 243 HCC and 59 (including 29 prior HBV and 30 chronic HBV) liver samples (Table 1). In HCC samples, HBV DNA were detected in 18 (24.6%) prior HBV without HCV, 71 (87.6%) chronic HBV, 3 (5.4%) prior HBV with HCV, and 3 (9.1%) non-B non-C patients. The antibody titers of the 3 patients who were positive for HBV in the non-B non-C group were within normal range but relatively lower than those of other patients in this group, suggesting that these patients might have a history of HBV infection. We identified 57 clonal and 284 nonclonal HBV integration events in the HCC genome of 11 (15.0%) prior HBV without HCV, and 249 clonal and 3,760 nonclonal events in 61 (75.3%) chronic HBV (P < 0.001). Notably, there were no clonal HBV integrations in prior HBV with HCV or in non-B non-C patients, although there were a small number of HBV reads from both groups. On the basis of sequence data, genotype C of HBV was integrated into 16 (88.8%) patients with prior HBV without HCV and 64 (90.1%) patients with chronic HBV. The start and end sites for HBV insertion in the human genome are shown in Supplementary Table S3.

Table 1.

Virus reads coverage into HCC genome

Prior HBV HCV(−)Chronic HBVPrior HBV HCV(+)non-B non-C
HCC (n = 73)Liver (n = 29)HCC (n = 81)Liver (n = 30)HCC (n = 56)HCC (n = 33)
Hepatitis B virus       
 Positive, casea, n (%) 18 (24.6) 21 (72.4) 71 (87.6) 30 (100) 3 (5.3) 3 (9.0) 
 Read number 5,224 ± 8,228 1,277 ± 2,452 7,331 ± 19,487 36,985 ± 57,475 26.0 ± 22.7 120.0 ± 150.6 
 Nonclonal integration site 249 73 3,760 3,873 
 Clonal integration site 57 11 281 372 
 Integrated gene 5.2 ± 4.4 2.0 ± 1.4 4.6 ± 5.1 8.9 ± 6.5 
 Integrated case, n (%) 11 (15.0) 4 (13.7) 61 (75.3) 24 (80) 
Genotype       
 B, n (%) 1 (5.5)  6 (8.4)  
 C, n (%) 16 (88.8)  64 (90.1)  1 (33.3) 2 (66.6) 
 Othersb, n (%) 1 (5.5)  1 (1.4)  1 (33.3) 
AAV type 2       
 Positive, casea, n (%) 3 (4.1) 2 (6.8) 1 (1.2) 2 (6.6) 2 (6.0) 
 Read number 6,772 ± 11,223 1,476/20 120 202/18 2,290/130 
 Integrated case, n (%) 2 (2.7) 1 (3.0) 
 Driver gene integration, case, n (%) 8 (10.9) 21 (25.9) 
Prior HBV HCV(−)Chronic HBVPrior HBV HCV(+)non-B non-C
HCC (n = 73)Liver (n = 29)HCC (n = 81)Liver (n = 30)HCC (n = 56)HCC (n = 33)
Hepatitis B virus       
 Positive, casea, n (%) 18 (24.6) 21 (72.4) 71 (87.6) 30 (100) 3 (5.3) 3 (9.0) 
 Read number 5,224 ± 8,228 1,277 ± 2,452 7,331 ± 19,487 36,985 ± 57,475 26.0 ± 22.7 120.0 ± 150.6 
 Nonclonal integration site 249 73 3,760 3,873 
 Clonal integration site 57 11 281 372 
 Integrated gene 5.2 ± 4.4 2.0 ± 1.4 4.6 ± 5.1 8.9 ± 6.5 
 Integrated case, n (%) 11 (15.0) 4 (13.7) 61 (75.3) 24 (80) 
Genotype       
 B, n (%) 1 (5.5)  6 (8.4)  
 C, n (%) 16 (88.8)  64 (90.1)  1 (33.3) 2 (66.6) 
 Othersb, n (%) 1 (5.5)  1 (1.4)  1 (33.3) 
AAV type 2       
 Positive, casea, n (%) 3 (4.1) 2 (6.8) 1 (1.2) 2 (6.6) 2 (6.0) 
 Read number 6,772 ± 11,223 1,476/20 120 202/18 2,290/130 
 Integrated case, n (%) 2 (2.7) 1 (3.0) 
 Driver gene integration, case, n (%) 8 (10.9) 21 (25.9) 

aHaving > 9 HBV or AAV2 reads.

bGenotype A, D, and H in prior HBV without HCV, prior HBV with HCV, and chronic HBV, respectively. Genotype could not be determined due to small number of virus read (< 20 per sample) in 4 patients. Mean values are presented with standard deviation. HCC, hepatocellular carcinoma; HBV, hepatitis B virus; HCV, hepatitis C virus.

On the other hand, the HBV genome was detected in the livers of 21 (72.4%) prior HBV and 30 (100%) chronic HBV. The read number per patient was much lower in prior HBV than chronic HBV.

Viral integration into driver genes

Although most insertion sites were distributed randomly throughout the genome, in both prior and chronic HBV, known driver genes were targeted for HBV integration in 6 (8.2%) prior HBV without HCV and 21 (25.3%) chronic HBV (P = 0.005): TERT in 4 (5.4%) and 15 (18.5%) patients (P = 0.015), KMT2B in 2 (2.7%) and 5 (6.1%) patients (P = 0.446), CCNE1 in 0 and 1 (1.2%) patient (P = 1.000), and CCNA2 in 0 and 1 (1.2%) patient (P = 1.000), respectively. One patient harbored integrated HBV in both TERT and CCNA2 (Fig. 1). In addition to the genes described above, cancer-related genes, including HGF, CSMD3, AR, and NF2 were targeted by HBV in both prior and chronic HBV. There were also many noncancer-related genes such as those in the LINC00486, CWH43, DEBT, and DDX11L1 gene families that showed evidence of recurrent HBV genome integrations, which were assumed to be passenger events (Supplementary Table S4).

Figure 1.

Landscape of genes targeted for viral integration. Top, patient characteristics according to the scale below. Bottom, the distribution of genes showing virus integration. The black box indicates integrated genes. NUCs, nucleoside analogues.

Figure 1.

Landscape of genes targeted for viral integration. Top, patient characteristics according to the scale below. Bottom, the distribution of genes showing virus integration. The black box indicates integrated genes. NUCs, nucleoside analogues.

Close modal

AAV2 was found to integrate into 3 genes in two prior HBV and 1 non-B non-C patients. CCNE1 and CCNA2 were targeted by AAV2 only in prior HBV, while SLC6A5 was integrated in a non-B non-C patient.

The HBV genome was recurrently integrated into fibrosis-related genes in liver samples; FN1 was found in three (10.0%), HS6ST3 in two (6.7%), KNG1in two (6.7%), and ROCK1 in two (6.7%) chronic HBV, while no gene was significantly targeted by HBV in prior HBV.

Clinical impact of viral integration on HCC development

There were no heavy drinkers in the 8 cases of prior HBV with HBV/AAV2 integration into driver genes. On the other hand, 4 cases of the 8 patients had diabetes mellitus, but they were histologically diagnosed having no or mild steatosis. Three (37.5%) patients with HBV/AAV2 integration had advanced liver fibrosis, including 2 patients with grade IV and 1 with grade III fibrosis, and the livers of 5 (62.5%) patients were diagnosed as having low grade 0−II fibrosis (Table 2; Supplementary Table S5). Among 6 prior HBV with driver gene integration, all were positive for both HBsAb and HBcAb, and 3 (50%) patients had a history of treatment for HBV using nucleoside analogues. One prior HBV patient who was positive for HBV-DNA had no virus integration. On the other hand, 14 (33.3%) patients out of 21 chronic HBV with driver gene integration were treated for HBV prior to surgery (P = 0.134; Supplementary Table S1), and the frequency of HBeAg was not significantly different between the patients with and without driver gene integration [six (28.5%) vs. 11 (18.3%), P = 0.357].

Table 2.

Characteristics of prior HBV

Driver gene integration (+) (n = 8)Driver gene integration (−) (n = 65)P
Age, years 70.5 (63–76) 70 (44–85) 0.875 
Sex, male, n (%) 7 (87.5) 61 (93.9) 0.539 
Alcohol, n (%) 29 (44.6) 0.003 
Diabetes mellitus, n (%) 4 (50.0) 31 (47.7) 0.902 
Treatment historya, n (%) 3 (37.5) 2 (3.0) 0.008 
Child-Pugh, A, n (%) 8 (100) 60 (92.3) 0.272 
ICGR15, % 10.8 (5.6–27.8) 13.6 (3.1–57.7) 0.127 
AFP, ng/mL 4 (1–405) 6 (1–449,211) 0.274 
DCP, mAU/mL 113 (12–10,429) 119 (6–75,000) 0.453 
Tumor size, cm 3.5 (2.1–6.5) 4.3 (0.7–17) 0.771 
Vascular invasion, n (%) 2 (25.0) 14 (21.5) 0.596 
Differentiation grade, well/mod/poor 1/7/0 10/48/7 1.000 
Fibrosis grade, III/IV, n (%) 3 (37.5) 25 (38.5) 0.958 
Driver gene integration (+) (n = 8)Driver gene integration (−) (n = 65)P
Age, years 70.5 (63–76) 70 (44–85) 0.875 
Sex, male, n (%) 7 (87.5) 61 (93.9) 0.539 
Alcohol, n (%) 29 (44.6) 0.003 
Diabetes mellitus, n (%) 4 (50.0) 31 (47.7) 0.902 
Treatment historya, n (%) 3 (37.5) 2 (3.0) 0.008 
Child-Pugh, A, n (%) 8 (100) 60 (92.3) 0.272 
ICGR15, % 10.8 (5.6–27.8) 13.6 (3.1–57.7) 0.127 
AFP, ng/mL 4 (1–405) 6 (1–449,211) 0.274 
DCP, mAU/mL 113 (12–10,429) 119 (6–75,000) 0.453 
Tumor size, cm 3.5 (2.1–6.5) 4.3 (0.7–17) 0.771 
Vascular invasion, n (%) 2 (25.0) 14 (21.5) 0.596 
Differentiation grade, well/mod/poor 1/7/0 10/48/7 1.000 
Fibrosis grade, III/IV, n (%) 3 (37.5) 25 (38.5) 0.958 

NOTE: Data are presented as median with range, if not specified. Driver gene integration (+) group contains 6 patients with HBV and 2 patients with AAV type 2 integration into the known driver genes.

Abbreviations: AFP, alpha-fetoprotein; DCP, des-gamma carboxyprothrombin; HBV, hepatitis B virus; HBsAb, anti-hepatitis B surface antibody; HBcAb, anti-hepatitis B core antibody; ICGR15, indocyanine green clearance rate at 15 minutes; mod, moderately; poor, poorly.

aPatients who were treated by nucleoside analogue before surgery.

Breakpoints of the integrated genes

The distribution of HBV integration breakpoints in TERT and KMT2B was similar between prior HBV without HCV and chronic HBV (Fig. 2A and B). Among the 19 patients with HBV integration near the TERT gene, HBV was integrated into the promoter region in 16 patients and into the second intron in 3 patients. In the KMT2B gene, HBV was integrated into the third exon and intron in 2 and 3 patients, respectively. Notably, the HBV breakpoints of CCNE1 and CCNA2 were in the fourth and second introns, respectively (Fig. 2C and D); breakpoints that have been reported as a recurrent integration region for AAV2. AAV2 also integrated into the promoter region and first intron of CCNE1 [AAV2 integration site, nucleotides 4,272−4,627 (356 bp)] and CCNA2 [4,057−4,600 (544 bp)] in prior HBV.

Figure 2.

Distribution of HBV integration site. A, Location of HBV integration sites in TERT. B, Location of HBV integration sites in KMT2B. C, Location of HBV and AAV2 integration sites in CCNE1. An AAV2 insertion with nucleotide numbers and fragment length in patient 888 is shown below (white box). D, Location of HBV and AAV2 integration sites in CCNA2. An AAV2 insertion with nucleotide numbers and fragment length in patient 615 is shown below (white box). The black and gray bars with patient numbers indicate integration sites in each gene. Patient 473 harbored an integrated HBV genome in TERT and CCNA2 simultaneously. UTR, untranslated region.

Figure 2.

Distribution of HBV integration site. A, Location of HBV integration sites in TERT. B, Location of HBV integration sites in KMT2B. C, Location of HBV and AAV2 integration sites in CCNE1. An AAV2 insertion with nucleotide numbers and fragment length in patient 888 is shown below (white box). D, Location of HBV and AAV2 integration sites in CCNA2. An AAV2 insertion with nucleotide numbers and fragment length in patient 615 is shown below (white box). The black and gray bars with patient numbers indicate integration sites in each gene. Patient 473 harbored an integrated HBV genome in TERT and CCNA2 simultaneously. UTR, untranslated region.

Close modal

HBV genome breakpoints

We investigated the distribution of HBV integration breakpoints, which might be associated with the development of HCC (Fig. 3). Sixty-four breakpoints were found in cases with prior HBV without HCV infection and 265 breakpoints were found in chronic HBV. Among these HBV integration breakpoints, 22 (34.3%) and 126 (47.5%) were clonal breakpoints, at nucleotides between 1,400 and 1,900 bp in the HBV genome where the HBx gene is located, in prior HBV and chronic HBV, respectively (P = 0.068).

Figure 3.

Distribution of HBV breakpoints. A, Frequency of HBV breakpoints in patients with prior HBV without HCV infection. B, Frequency of HBV breakpoints in patients with chronic HBV. S, S region; C, Pre-core/core region; P, Polymerase region; X, HBx region.

Figure 3.

Distribution of HBV breakpoints. A, Frequency of HBV breakpoints in patients with prior HBV without HCV infection. B, Frequency of HBV breakpoints in patients with chronic HBV. S, S region; C, Pre-core/core region; P, Polymerase region; X, HBx region.

Close modal

In addition to the HBx gene, we found a high rate of integration events between 2,750 and 2,800 bp. These fragments were identified in the TTGGGG repeat of the DDX gene family, including DDX11L1, DDX1L2, and DDX11L16 in the same patient. The TTGGGG repeat is also present in other sites of the human genome, suggesting that fragments between nucleotides 2,750 and 2,800 of the HBV genome were falsely mapped due to repetitive sequences. Therefore, we assumed that the HBV genome breakpoints were similar between prior HBV and chronic HBV.

Despite HBV being DNA virus, it harbors mutations at the same level as an RNA virus (24). We investigated HBV S gene sequence from HCC samples with integrated HBV genomes. Among the 13 samples from prior HBV in which HBV was detected, 6 HCC samples harbored the full-length S gene (Supplementary Fig. S2), and there were only two mutations, S17A and S6T, in 2 patients, that were not observed in chronic HBV. Therefore, we assumed that there was no significant difference in the presence of HBV S gene sequences between prior HBV and chronic HBV.

Transcriptional effects of viral integration

To investigate the effects of viral integration on TERT, KMT2B, CCNE1, and CCNA2 transcription, RNA sequencing was performed on 29 HCC samples with and 40 HCC samples without viral integration into driver genes. Nineteen HCC samples with HBV integration near TERT exhibited high levels of TERT mRNA compared with 50 HCC samples without HBV integration near this gene [average reads per kilo-base of exon per million mapped sequence reads (RPKM), 5.6 ± 4.7 vs. 0.5 ± 1.7, P < 0.001; Fig. 4A]. Expression of KMT2B was also elevated in 7 HCC samples with HBV integration (RPKM, 9.3 ± 4.2) compared with 62 HCC samples without HBV integration near KMT2B (RPKM, 5.6 ± 3.0, P = 0.008; Fig. 4B). Notably, CCNE1 and CCNA2 were remarkably upregulated by AAV2 as well as HBV integration (P < 0.001; Fig. 4C and D). Hierarchical clustering using 29 HCC samples with viral integration showed that HCC samples were mainly divided into two classes by integrated driver genes, not by HBV status (Supplementary Fig. S3), suggesting that viral integration seriously affected the expression profiling of liver cancer.

Figure 4.

Transcriptional effect of HBV integration into the human genome. RNA sequencing was performed on 15 patients with prior HBV without HCV, 40 with chronic HBV, and 12 prior HBV with HCV, and in 2 non-B non-C patients. A, Expression of TERT in 19 and 50 HCC samples, with and without HBV integration into the target gene, respectively. B, Expression of KMT2B in seven and 62 HCC samples, with and without HBV integration to the target gene, respectively. C, Expression of HBV and AAV2-integrated CCNE1. D, Expression of HBV and AAV2-integrated CCNA2. The broken lines represent the average RPKM of HCC samples, with and without viral integration into each driver gene, that were subjected to an RNA sequence analysis. RPKM, average reads per kilo-base of exon per million mapped sequence reads.

Figure 4.

Transcriptional effect of HBV integration into the human genome. RNA sequencing was performed on 15 patients with prior HBV without HCV, 40 with chronic HBV, and 12 prior HBV with HCV, and in 2 non-B non-C patients. A, Expression of TERT in 19 and 50 HCC samples, with and without HBV integration into the target gene, respectively. B, Expression of KMT2B in seven and 62 HCC samples, with and without HBV integration to the target gene, respectively. C, Expression of HBV and AAV2-integrated CCNE1. D, Expression of HBV and AAV2-integrated CCNA2. The broken lines represent the average RPKM of HCC samples, with and without viral integration into each driver gene, that were subjected to an RNA sequence analysis. RPKM, average reads per kilo-base of exon per million mapped sequence reads.

Close modal

Analysis of the RNA sequence data revealed several types of gene-fusion events between HBV/AAV2 and genes into which they had integrated, with an HBV-TERT fusion in 9 (47.3%) cases and HBV-KMT2B fusion in 6 (85.7%) cases. In addition, expression levels of CCNE1 and CCNA2 were high and chimeric transcripts including HBV-CCNE1, HBV-CCNA2, AAV2-CCNE1, and AAV2-CCNA2 were detected in all cases (Supplementary Table S3).

Overexpression of TERT, KMT2B, CCNE1, and CCNA2 were observed in tumors of prior HBV as well as of chronic HBV, suggesting that integration of HBV or AAV2 into these driver genes leads to upregulation of these genes and plays a pivotal role in HCC development even after the disappearance of HBsAg.

Despite the disappearance of serum HBsAg after spontaneous regression or hepatitis B treatment with nucleoside analogues, for example, some patients still develop liver cancer (17, 25). This study showed that the HBV genome (from both chronic and prior HBV) commonly integrated into HCC driver genes. Therefore, a part of HCC development in prior HBV was due to HBV integration into the human genome before the disappearance of HBsAg, and prior HBV cases should be considered at higher risk of HCC.

Consistent with previous reports, we found that the HBV genome randomly integrated into the entire human genome and recurrently integrated into known driver genes, including TERT, KMT2B, CCNE1, and CCNA2, in HCC samples of chronic HBV (2, 3, 8, 9), but most of the HBV integration events were passenger events. Overall integration of the HBV genome into the genomes of prior HBV was not as frequent as in chronic HBV, but TERT and KMT2B were recurrently targeted by the HBV genome, and the frequency of driver gene targeting in prior HBV was higher than that in chronic HBV based on clonal HBV integration events. This could be because hepatocytes with nononcogenic virus integrations disappeared and only those with integrations in driver genes could survive after HBsAg clearance. Furthermore, driver genes integrated by AAV2 were the same as those integrated by HBV. These driver genes were significantly upregulated as a result of viral genome integration, suggesting that hepatocytes in which driver genes into which HBV and AAV2 viral genomes had integrated during a prior HBV infection were oncogenically transformed, leading to the development of liver cancer.

In addition to the known driver genes described above, we found several cancer-related genes such as HGF (26), CSMD3 (22), and AR (27), and a suppressor gene NF2 (28) that showed HBV genome integration. SLC6A5, which can be hypermethylated in cervical cancer cases, was targeted by AAV2 (29). Although viral integration into these genes was not common in this study, a study with a larger cohort of participants could clarify whether these HBV integration target genes are associated with HCC development.

Several genes, including those in the LINC00486, CWH43, DEBT, and DDX11L1 families, showed recurrent HBV DNA integration events in some HCC samples in this study. These genes harbor repetitive sequences; poly (G) site in LINC00486, TTCCA repeat in CWH43, GGGTTA repeat in DEBT, and TTGGGG repeat in DDX11L family members. These repetitive sequences exist in other sites of the human genome with the possibility that they will be counted as virus-integration genes. Thus, it is difficult to confirm by HIVID analysis whether such genes were really targeted for HBV DNA integration. Because LINC00486, CWH43, DEBT, and DDX11L have not been reported to be associated with cancer, we excluded these genes as driver genes with regard to the development of HCC.

In liver samples, FN1 (15), HS6ST3 (30), KNG1 (31), and ROCK1 (32), genes associated with liver, pulmonary or cardiac fibrosis were found to have integrated HBV genomes only in chronic HBV. In contrast, no significant gene was found to be associated with integration in prior HBV, which could be because read numbers per sample were much less in liver samples of prior HBV compared with chronic HBV. Given that FN1 has been found recurrently integrated by the HBV genome in the liver, these genes are believed to be associated with liver fibrosis in HBV infection. Different from HCC samples, cancer-related genes were not targeted in liver samples. Although HBV integration into TERT was found in two liver samples, the HBV genome was detected in the same position as that of the HCC genome in both cases; therefore, we assumed that this was evidence of contamination in sampling of the resected specimens.

Given the disappearance of HBsAg, it is acceptable that the patients were older and male, alcohol abuse, and diabetes mellitus were more frequent in prior HBV compared with the chronic HBV. It was intriguing that there was no patient with a history of alcohol abuse among the 8 cases of prior HBV with HBV or AAV2 integration into driver genes. Although there were four patients with diabetes mellitus, they did not histologically diagnosed having steatosis, which is one of the major causes of HCC (33). These clinical characteristics suggested that viral integration was the primary event for hepatocarcinogenesis in these patients. In addition, the livers of 5 of these 8 patients were normal or showed early-stage fibrosis (grade 0 or I). Cirrhotic patients are known to be at risk for HCC after HBsAg clearance (25). However, our data showed that HCC can frequently develop in patients with the livers showing grade 0/I fibrosis, owing to virus integration into driver genes in the human genome. Therefore not only advanced fibrosis, but also viral integration into driver genes might contribute, perhaps synergistically, to HCC development.

In chronic HBV, the frequency of HBV integration into the human genome was not different between patients who had and had not undergone treatment with nucleoside analogues before surgery. On the other hand, the number of patients who were treated by antiviral therapy was larger in prior HBV and evidence of HBV integration, suggesting that HBV integration into the host genome is an early event during chronic infection. Given that HCC risk depends on the duration of the infection, prior HBV without antiviral therapy might have other causes of hepatocarcinogenesis in addition to HBV integration into driver genes. However, the number of such patients was too small to conclude that nucleoside analogues affected the integration of HBV. Therefore, whether treatment by nucleoside analogues is associated with hepatocarcinogenesis should be investigated in a larger cohort. In addition, we advocate universal HBV vaccination to prevent the need for antiviral treatments.

In patients with HCV infection, OBI positively drives the development of HCC through its direct proto-oncogenic effects, as well as indirectly by causing persistent hepatic inflammation and fibrosis (34–36). In this study, in contrast to patients with prior HBV without HCV infection, there was no HBV integration into the host genomes in 56 prior HBV with HCV infection. Given that advanced liver fibrosis was more frequent in prior HBV with HCV than in those without HCV infection, prior HBV and HCV infection, together, appeared to induce hepatocarcinogenesis in these patients, perhaps as the result of chronic inflammation and fibrosis rather than to the direct oncogenic result of HBV infection, such as integration into the host genome (37).

Consistent with previous data, HBV integration sites in the human genome were concentrated in the promoter region in TERT and intron 3 of KMT2B in both prior HBV and chronic HBV (2, 3). HBV integration into intron 4 of CCNE1 has been reported elsewhere (2), but these integration sites were not identified as fragile sites in the human genome (38). It is noteworthy that these recurrent integration sites of HBV were similar to those of AAV2 in HCC (19); therefore, preferential integration regions must exist in driver genes but not in passenger genes (39), regardless of the type of virus that integrates (3).

Detection of chimeric transcripts by RNA sequencing is dependent on the expression levels of integrated genes. Consequently, gene fusions between HBV and TERT or KMT2B could not be detected in some cases because of their lower expression levels compared with those of other genes upregulated in HCC. On the other hand, expression levels of CCNE1 and CCNA2 were higher, and chimeric HBV/AAV2 and CCNE1/CCNA2 transcripts were detected in all cases.

High expression of a transactivator gene, HBx, can contribute to the development of HCC, both in transgenic mice and humans (40, 41). Consistently, more HBx integrants were detected, compared with the integrants of the other three HBV genes (HBs, HB precore/core, and HB polymerase), dominant in both prior HBV and chronic HBV, as described previously (42). Notably for driver genes, most HBV integrants were in the HBx gene, suggesting that HBV-integrated genes, except for known driver genes, would not initiate hepatocarcinogenesis. On the other hand, the AAV2 integration sites in CCNE1 and CCNA2 identified in this study were the same as those detected previously (19), suggesting that the region between nucleosides 4,200 and 4,600 of the AAV2 genome may be associated with cancer development.

HBV integration in patients with OBI, determined in previous study using the Alu-PCR technique, was remarkably frequent compared with that found in the current study (approximately 75% vs. 15%; ref. 17). In addition, viral genomes integrated into known driver genes such as PI4K2A and SCARB1 identified by Alu-PCR were not detected in chronic HBV or prior HBV in the current study. Similarly, commonly integrated HCC driver genes, detected in other studies with large cohorts of patients with chronic HBV (2, 3) and in this study using high-throughput sequence technologies were not identified as target genes for HBV integration by Alu-PCR. These discrepancies could be attributed to differences in the experimental methods used to detect HBV integration (17). Therefore, it is necessary to compare integrated genes from the same sample using different experimental approaches, and to estimate the accuracy of different methods in detecting viral integration events.

Taken together, although overall integration of HBV into the human HCC genome in chronic HBV was much more frequent than that in prior HBV, HBV integration, and subsequent upregulation of driver gene was not rare even in prior HBV. Therefore, despite the disappearance of serum HBsAg, prior infection with HBV is one of the significant risk factors for the development of HCC.

No potential conflicts of interest were disclosed.

The content is solely the responsibility of the authors.

Conception and design: K. Tatsuno, Y. Midorikawa, M. Moriyama, K. Moriya

Development of methodology: K. Tatsuno

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): Y. Midorikawa

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): S. Yamamoto

Writing, review, and/or revision of the manuscript: K. Tatsuno, Y. Midorikawa

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): G. Nagae, H. Nakagawa

Study supervision: T. Takayama, K. Koike, H. Aburatani

We thank Kaori Shiina and Hiroko Meguro for valuable technical assistance. This research was supported by AMED under Grant Number JP18hk0102049 (to Y. Midorikawa), JP18ck0106265 (to H. Aburatani), and JP18fk0210040 (to K. Koike), and a Grant-in-Aid for Scientific Research (C) JP17K07195 (to K. Tatsuno) from the Ministry of Education, Culture, Sports, Science, and Technology (MEXT), Japan.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1.
Jiang
Z
,
Jhunjhunwala
S
,
Liu
J
,
Haverty
PM
,
Kennemer
MI
,
Guan
Y
, et al
The effects of hepatitis B virus integration into the genomes of hepatocellular carcinoma patients
.
Genome Res
2012
;
22
:
593
601
.
2.
Sung
WK
,
Zheng
H
,
Li
S
,
Chen
R
,
Liu
X
,
Li
Y
, et al
Genome-wide survey of recurrent HBV integration in hepatocellular carcinoma
.
Nat Genet
2012
;
44
:
765
9
.
3.
Zhao
LH
,
Liu
X
,
Yan
HX
,
Li
WY
,
Zeng
X
,
Yang
Y
, et al
Genomic and oncogenic preference of HBV integration in hepatocellular carcinoma
.
Nat Commun
2016
;
7
:
12992
.
4.
Chiu
YT
,
Wong
JK
,
Choi
SW
,
Sze
KM
,
Ho
DW
,
Chan
LK
, et al
Novel pre-mRNA splicing of intronically integrated HBV generates oncogenic chimera in hepatocellular carcinoma
.
J Hepatol
2016
;
64
:
1256
64
.
5.
Lau
CC
,
Sun
T
,
Ching
AK
,
He
M
,
Li
JW
,
Wong
AM
, et al
Viral-human chimeric transcript predisposes risk to liver cancer development and progression
.
Cancer Cell
2014
;
25
:
335
49
.
6.
Nagaya
T
,
Nakamura
T
,
Tokino
T
,
Tsurimoto
T
,
Imai
M
,
Mayumi
T
, et al
The mode of hepatitis B virus DNA integration in chromosomes of human hepatocellular carcinoma
.
Genes Dev
1987
;
1
:
773
82
.
7.
Shafritz
DA
,
Shouval
D
,
Sherman
HI
,
Hadziyannis
SJ
,
Kew
MC
. 
Integration of hepatitis B virus DNA into the genome of liver cells in chronic liver disease and hepatocellular carcinoma. Studies in percutaneous liver biopsies and post-mortem tissue specimens
.
N Engl J Med
1981
;
305
:
1067
73
.
8.
Jiang
S
,
Yang
Z
,
Li
W
,
Li
X
,
Wang
Y
,
Zhang
J
, et al
Re-evaluation of the carcinogenic significance of hepatitis B virus integration in hepatocarcinogenesis
.
PLoS One
2012
;
7
:
e40363
.
9.
Murakami
Y
,
Saigo
K
,
Takashima
H
,
Minami
M
,
Okanoue
T
,
Brechot
C
, et al
Large scaled analysis of hepatitis B virus (HBV) DNA integration in HBV related hepatocellular carcinomas
.
Gut
2005
;
54
:
1162
8
.
10.
Fujimoto
A
,
Totoki
Y
,
Abe
T
,
Boroevich
KA
,
Hosoda
F
,
Nguyen
HH
, et al
Whole-genome sequencing of liver cancers identifies etiological influences on mutation patterns and recurrent mutations in chromatin regulators
.
Nat Genet
2012
;
44
:
760
4
.
11.
Dong
H
,
Zhang
L
,
Qian
Z
,
Zhu
X
,
Zhu
G
,
Chen
Y
, et al
Identification of HBV-MLL4 integration and its molecular basis in Chinese hepatocellular carcinoma
.
PLoS One
2015
;
10
:
e0123175
.
12.
Li
X
,
Zhang
J
,
Yang
Z
,
Kang
J
,
Jiang
S
,
Zhang
T
, et al
The function of targeted host genes determines the oncogenicity of HBV integration in hepatocellular carcinoma
.
J Hepatol
2014
;
60
:
975
84
.
13.
Totoki
Y
,
Tatsuno
K
,
Covington
KR
,
Ueda
H
,
Creighton
CJ
,
Kato
M
, et al
Trans-ancestry mutational landscape of hepatocellular carcinoma genomes
.
Nat Genet
2014
;
46
:
1267
73
.
14.
Ding
D
,
Lou
X
,
Hua
D
,
Yu
W
,
Li
L
,
Wang
J
, et al
Recurrent targeted genes of hepatitis B virus in the liver cancer genomes identified by a next-generation sequencing-based approach
.
PLoS Genet
. 
2012
;
8
:
e1003065
.
15.
Furuta
M
,
Tanaka
H
,
Shiraishi
Y
,
Unida
T
,
Imamura
M
,
Fujimoto
A
, et al
Characterization of HBV integration patterns and timing in liver cancer and HBV-infected livers
.
Oncotarget
2018
;
9
:
25075
88
.
16.
Li
W
,
Zeng
X
,
Lee
NP
,
Liu
X
,
Chen
S
,
Guo
B
, et al
HIVID: an efficient method to detect HBV integration using low coverage sequencing
.
Genomics
2013
;
102
:
338
44
.
17.
Saitta
C
,
Tripodi
G
,
Barbera
A
,
Bertuccio
A
,
Smedile
A
,
Ciancio
A
, et al
Hepatitis B virus (HBV) DNA integration in patients with occult HBV infection and hepatocellular carcinoma
.
Liver Int
2015
;
35
:
2311
7
.
18.
Alam
S
,
Bowser
BS
,
Conway
MJ
,
Israr
M
,
Tandon
A
,
Meyers
C
. 
Adeno-associated virus type 2 infection activates caspase dependent and independent apoptosis in multiple breast cancer lines but not in normal mammary epithelial cells
.
Mol Cancer
2011
;
10
:
97
.
19.
Nault
JC
,
Datta
S
,
Imbeaud
S
,
Franconi
A
,
Mallet
M
,
Couchy
G
, et al
Recurrent AAV2-related insertional mutagenesis in human hepatocellular carcinomas
.
Nat Genet
2015
;
47
:
1187
93
.
20.
Yotsuyanagi
H
,
Shintani
Y
,
Moriya
K
,
Fujie
H
,
Tsutsumi
T
,
Kato
T
, et al
Virologic analysis of non-B, non-C hepatocellular carcinoma in Japan: frequent involvement of hepatitis B virus
.
J Infect Dis
2000
;
181
:
1920
8
.
21.
Midorikawa
Y
,
Tsutsumi
S
,
Nishimura
K
,
Kamimura
N
,
Kano
M
,
Sakamoto
H
, et al
Distinct chromosomal bias of gene expression signatures in the progression of hepatocellular carcinoma
.
Cancer Res
2004
;
64
:
7263
70
.
22.
Midorikawa
Y
,
Yamamoto
S
,
Tsuji
S
,
Kamimura
N
,
Ishikawa
S
,
Igarashi
H
, et al
Allelic imbalances and homozygous deletion on 8p23.2 for stepwise progression of hepatocarcinogenesis
.
Hepatology
2009
;
49
:
513
22
.
23.
Lok
AS
,
McMahon
BJ
. 
Chronic hepatitis B.
Hepatology
2007
;
45
:
507
39
.
24.
Orito
E
,
Mizokami
M
,
Ina
Y
,
Moriyama
EN
,
Kameshima
N
,
Yamamoto
M
, et al
Host-independent evolution and a genetic classification of the hepadnavirus family based on nucleotide sequences
.
Proc Natl Acad Sci U S A
1989
;
86
:
7059
62
.
25.
Papatheodoridis
GV
,
Lampertico
P
,
Manolakopoulos
S
,
Lok
A
. 
Incidence of hepatocellular carcinoma in chronic hepatitis B patients receiving nucleos(t)ide therapy: a systematic review
.
J Hepatol
2010
;
53
:
348
56
.
26.
Bottaro
DP
,
Rubin
JS
,
Faletto
DL
,
Chan
AM
,
Kmiecik
TE
,
Vande Woude
GF
, et al
Identification of the hepatocyte growth factor receptor as the c-met proto-oncogene product
.
Science
1991
;
251
:
802
4
.
27.
Ohnishi
S
,
Murakami
T
,
Moriyama
T
,
Mitamura
K
,
Imawari
M
. 
Androgen and estrogen receptors in hepatocellular carcinoma and in the surrounding noncancerous liver tissue
.
Hepatology
1986
;
6
:
440
3
.
28.
Benhamouche
S
,
Curto
M
,
Saotome
I
,
Gladden
AB
,
Liu
CH
,
Giovannini
M
, et al
Nf2/Merlin controls progenitor homeostasis and tumorigenesis in the liver
.
Genes Dev
2010
;
24
:
1718
30
.
29.
Wang
R
,
van Leeuwen
RW
,
Boers
A
,
Klip
HG
,
de Meyer
T
,
Steenbergen
RD
, et al
Genome-wide methylome analysis using MethylCap-seq uncovers 4 hypermethylated markers with high sensitivity for both adeno- and squamous-cell cervical carcinoma
.
Oncotarget
2016
;
7
:
80735
50
.
30.
Lu
J
,
Auduong
L
,
White
ES
,
Yue
X
. 
Up-regulation of heparan sulfate 6-O-sulfation in idiopathic pulmonary fibrosis
.
Am J Respir Cell Mol Biol
2014
;
50
:
106
14
.
31.
Niu
R
,
Liu
Y
,
Zhang
Y
,
Zhang
Y
,
Wang
H
,
Wang
Y
, et al
iTRAQ-based proteomics reveals novel biomarkers for idiopathic pulmonary fibrosis
.
PLoS One
2017
;
12
:
e0170741
.
32.
Haudek
SB
,
Gupta
D
,
Dewald
O
,
Schwartz
RJ
,
Wei
L
,
Trial
J
, et al
Rho kinase-1 mediates cardiac fibrosis by regulating fibroblast precursor cell differentiation
.
Cardiovasc Res
2009
;
83
:
511
8
.
33.
Zakhari
S
. 
Bermuda Triangle for the liver: alcohol, obesity, and viral hepatitis
.
J Gastroenterol Hepatol
. 
2013
;
28
Suppl 1
:
18
25
.
34.
Adachi
S
,
Shibuya
A
,
Miura
Y
,
Takeuchi
A
,
Nakazawa
T
,
Saigenji
K
. 
Impact of occult hepatitis B virus infection and prior hepatitis B virus infection on development of hepatocellular carcinoma in patients with liver cirrhosis due to hepatitis C virus
.
Scand J Gastroenterol
2008
;
43
:
849
56
.
35.
Ikeda
K
,
Marusawa
H
,
Osaki
Y
,
Nakamura
T
,
Kitajima
N
,
Yamashita
Y
, et al
Antibody to hepatitis B core antigen and risk for hepatitis C-related hepatocellular carcinoma: a prospective study
.
Ann Intern Med
2007
;
146
:
649
56
.
36.
Squadrito
G
,
Cacciola
I
,
Alibrandi
A
,
Pollicino
T
,
Raimondo
G
. 
Impact of occult hepatitis B virus infection on the outcome of chronic hepatitis C
.
J Hepatol
2013
;
59
:
696
700
.
37.
Raimondo
G
,
Pollicino
T
,
Squadrito
G
. 
What is the clinical impact of occult hepatitis B virus infection?
Lancet
2005
;
365
:
638
40
.
38.
Fungtammasan
A
,
Walsh
E
,
Chiaromonte
F
,
Eckert
KA
,
Makova
KD
. 
A genome-wide analysis of common fragile sites: what features determine chromosomal instability in the human genome?
Genome Res
2012
;
22
:
993
1005
.
39.
Budzinska
MA
,
Shackel
NA
,
Urban
S
,
Tu
T
. 
Sequence analysis of integrated hepatitis B virus DNA during HBeAg-seroconversion
.
Emerg Microbes Infect
2018
;
7
:
142
.
40.
Diamantis
ID
,
McGandy
CE
,
Chen
TJ
,
Liaw
YF
,
Gudat
F
,
Bianchi
L
. 
Hepatitis B X-gene expression in hepatocellular carcinoma
.
J Hepatol
1992
;
15
:
400
3
.
41.
Kim
CM
,
Koike
K
,
Saito
I
,
Miyamura
T
,
Jay
G
. 
HBx gene of hepatitis B virus induces liver cancer in transgenic mice
.
Nature
1991
;
351
:
317
20
.
42.
Toyoda
H
,
Kumada
T
,
Kaneoka
Y
,
Murakami
Y
. 
Impact of hepatitis B virus (HBV) X gene integration in liver tissue on hepatocellular carcinoma development in serologically HBV-negative chronic hepatitis C patients
.
J Hepatol
2008
;
48
:
43
50
.

Supplementary data