Seventy-three HCC patients with hepatitis B surface antigen (HBsAg) and 19 controls with HBsAg (without HCC) were seen at the Department of Internal Medicine, Kangnam St. Mary’s Hospital, Catholic University Medical College, Seoul, Korea, between 1990 and 1993 (Table 1). All were negative for antibody to the hepatitis C virus (anti-HCV). Of the 73 HCC patients, the diagnosis of HCC was made histologically in 61 and was based on elevated levels of alpha-fetoprotein (AFP) (>400ng/ml) and CT scan in 12. Cirrhosis was visualized at laparoscopy in 46/61 (75%) HCC patients. Serum samples were collected at the time of the initial hospital admission and were stored at −20°C until tested in the present study. The 19 controls had histologic evidence of chronic active hepatitis and included 9 with HBeAg and 10 with antibody to HBeAg (anti-HBe) (HBeAg-negative).

For patients and controls, sera were tested for HBsAg, antibody to HBsAg, HBeAg, anti-HBe and antibody to the hepatitis B core antigen using commercially available radioimmunoassays (Abbott Laboratories, North Chicago, IL, USA), and antibody to the hepatitis C virus (anti-HCV) using commercially available enzyme immunoassays (Abbott Laboratories).

One hundred microliters of each serum sample were mixed with 200ng of proteinase K solution (20mg/ml; Boehringer Mannheim, Mannheim, Germany) in 100 μl of 100mM NaCl, 10 mM Tris-HCl (pH 8.0), 25mM ethylenediaminetetraacetic acid (pH 8.0), and 0.5% sodium dodecyl sulfate, and incubated at 56°C for 2 hours. This was sequentially extracted with an equal volume of phenol, with an equal volume of a solution of phenol: chloroform: isoamyl alcohol (25:24:1), and with an equal volume of chloroform. The extracted sample was mixed with 3 μl glycogen and one-tenth volume of 3M sodium acetate (pH 5.2) and precipitated overnight with three volumes of cold 100% ethanol. The precipitated DNA was washed with 70% ethanol, air dried, and dissolved in 100 μl TE buffer (10mmol/L Tris-HCl and 0.1mmol/L EDTA, pH 8.0).

The reaction mixture (100 μl) contained 50 μl of prepared DNA template, 10mmol/L Tris (pH 8.3), 50mmol/L KCl, 1.5mmol/L MgCl₂, 1% (w/v) gelatin, 0.5 units of Taq polymerase (Perkin-Elmer; Norwalk, CT, USA), 200mmol/L each of the four deoxynucleoside triphosphates and 1 mmol/L each of sense primer (nucleotides 1689–1708; 5′-ACCTTGAGGCATACTTCAAA) and antisense primer (nucleotides 2058–2078; 5′-CAGAATAGCTTGCCTGAGTGC) for a DNA fragment of 390 bp, which includes the precore and proximal core regions of the HBV genome. The reaction mixture was overlaid with a drop of mineral oil, heated at 95°C for 5 minutes, and amplified with a programmable thermal cycler (Perkin-Elmer). Thirty-five amplification cycles were performed, each consisting of denaturation of template DNA at 95°C for 1 min, primer annealing at 55°C for 1 minute, and polymerization at 72°C for 2 minutes. After the last cycle, the temperature was maintained at 72°C for 10 minutes for complete polymerization. Ten microliters of each polymerase chain reaction (PCR) product were separated by electrophoresis in a 2% agarose gel and stained with ethidium bromide to visualize the amplified DNA fragment.

For sera with no detectable HBV DNA by the PCR process described above, DNA was re-extracted from 250 μl of serum and two PCR amplifications were performed. In the first step, amplification conditions were the same as above; in the second amplification step, 5 μl of the PCR product were amplified for 30 cycles using the same primers with primer annealing conditions at 60°C for 1 minute.

Care was taken to avoid cross-contamination between different steps in the procedure; preparation of the DNA template and preparation of the PCR mixture were conducted in a separate room from the amplification, gel electrophoresis, and purification of PCR products. All samples and solutions were handled with aerosol-resistant pipette tips, and fresh aliquots of reagent solutions were used for each experiment. Positive and negative controls were run with each amplification.

Sequencing of DNA amplified by PCR was performed according to the Sanger dideoxy chain termination method using a commercially available sequencing kit (“fmol”; Promega, Madison, WI, USA). The PCR product was purified using a microconcentrator (microcon-100; Amicon, Danvers, MA, USA). Ten pmol of a sequencing primer (nucleotide 1776–1797; 5′-GGCTGTAGGCATAAATTGGTC) for the sense strand of the precore and proximal core regions were end-labeled with 10 pmol of [γ-³²P]-adenosine triphosphate (Amersham, Arlington Heights, IL, USA). The sequencing product was electrophoresed with 6% polyacrylamide/7 M urea gel, dried at 80°C for 1 hour with a gel drier, and autoradiographed at −70°C for 2–24 hours.

Data were analyzed using an unpaired t-test, chi-square test and Mann-Whitney U test.

Among 73 HCC patients, HBV-DNA was detected in the serum of 64 (88%), including all 29 HBeAg-positive (100%) and 35/44 (80%) HBeAg-negative cases. All 19 controls with chronic hepatitis B had HBV-DNA in their serum (100%), including 10 HBeAg-negative centrals (Table 1, 2).

Among those cases whose sequencing results were sufficient for analysis, 50/58 (86%) of the HCC patients and 12/16 (69%) of the controls with chronic hepatitis B had an HBV with a mutation at the 83rd nucleotide of the precore region (TGG →TAG) that changes codon 28 from tryptophan to a stop codon (Table 3).

Among HCC patients, the precore codon 28 stop mutant was detected in 23/28 (82%) HBeAg-positive patients and 27/30 (90%) HBeAg-negative patients. Among controls with chronic hepatitis B, this mutation was detected in 5/9 (56%) HBeAg-positive controls and 7/7 (100%) HBeAg-negative controls (Table 4).

Of the 28 HBeAg-positive HCC patients, 5 (18%) had wild-type HBV alone, 13 (46%) had a mixed infection with both the wild-type and mutant HBV in their serum, since two bands could be detected at the second nucleotide of the precore codon 28 (Fig. 1), and 10 (36%) were infected with the mutant-type alone. Of the 9 HBeAg-positive controls, 4 (44%) had the wild-type alone and 5 (56%) had a mixture of both strains (Table 4).

Of the 30 HBeAg-negative HCC patients, 3 (10%) had wild-type HBV alone, 8 (27%) had a mixed infection with both strains and 19 (63%) were infected with the mutant-type alone. Of the 7 HBeAg-negative controls, 4 (57%) had a mixed infection of both strains and 3 (43%) were infected with the mutant-type alone (Table 4).

Sequencing results from HCC patients were available for 50 of the 61 patients on whom laparoscopic examination was performed. Among those, cirrhosis was observed in 34/43 (79%) patients with the precore codon 28 stop mutant, including 22/24 (92%) with the mutant-type alone, compared to 12/19 (63%) HCC patients with a mixture of wild- and mutant-type HBV (p<0.03) (Table 5).

The 19 anti-HBe-positive HCC patients with mutant HBV alone were older (median age 53 years; range 37–64 years), compared to the 17 HBeAg-positive HCC patients with co-infection of both wild-type and mutant HBV (median age 43 years; range 23–66 years) (p<0.05). They also had lower serum aspartate aminotransferase levels (82±3IU/L vs. 192±75IU/L) (p<0.03) and lower serum gamma-glutamyl transpeptidase levels (107±2IU/L vs. 214±35IU/L) (p<0.03).