Characteristics of Patients With Chronic Hepatitis B Virus Infection With Genotype E Predominance in Burkina Faso

Study Population and Data Abstraction

This study was approved by the Institutional Review Board of the Burkina Faso National Ethics Committee for Health Research (reference no. 2015-01-006; January 14, 2015). Patients were recruited and tested for HBV viral load from October 1, 2014, through August 30, 2017, at Pietro Annigoni Biomedical Research Center (CERBA) in Ouagadougou, Burkina Faso, West Africa. CERBA provides free HBV viral load testing as well as affordable medical care for low-income patient populations. The protocol for referring patients for HBV viral load testing required that all patients be diagnosed with CHB, confirmed by at least two HBsAg tests at 6-month intervals. A total of 4,163 patients who were HBsAg+ were identified from the CERBA database. Excluded patients included 17 without viral load data and 202 who received treatment before the viral load assay. In total, 3,944 patients with CHB infection (HBsAg+) were included, comprising 3,885 patients without HCC (CHB HCC−) and 59 with HCC (CHB HCC+) (Fig. 1).

Characteristics of patients with CHB, such as age at diagnosis of CHB and HCC, sex, and cirrhosis status, were collected at the time of serologic testing.⁽²⁶⁾ Clinical data on HCC status and HBV treatment were obtained from electronic health records of the Department of Gastroenterology and Hepatology of the Yalgado Ouedraogo National Hospital in Ouagadougou. Cirrhosis and HCC were diagnosed based on radiologic findings. Patients did not routinely have imaging for surveillance because of its limited availability. They were seen by local providers once every 6 to 12 months after the diagnosis of CHB to monitor their clinical symptoms. Patients were referred for ultrasonography or cross-sectional imaging only if cirrhosis was suspected due to jaundice, abnormal liver size, ascites, dilated superficial abdominal veins, or elevated liver enzymes; HCC was suspected due to elevated liver enzymes or a palpable mass; or they had decompensated liver disease.

Specimen Collection and HBV DNA Extraction

Blood samples were collected aseptically into 10 mL ethylenediaminetetraacetic tubes and centrifuged for 15 minutes at 2,000g. HBV viral DNA was extracted from 200 μL of plasma, using a QIAamp DNA Blood Mini Kit (QIAGEN GmbH, Hilden, Germany).

HBV, Hepatitis C Virus, and Human Immunodeficiency Virus Serologic Markers

Laboratory assays, including HBsAg, HBeAg, hepatitis B e antibody (HBeAb), hepatitis B core antibody (HBcAb), aspartate aminotransferase (AST), ALT, human immunodeficiency virus (HIV) antibody, and antibody to hepatitis C virus, were performed. HBV serologic testing for HBsAg, HBeAg, HBeAb, and HBcAb was performed by immunoassay (bioMérieux, Boxtel, the Netherlands).

HBV Viral Load

HBV DNA was quantified in 20-µL reactions containing 5 μL of HBV-infected patient DNA sample using an Oasig 2X quantitative polymerase chain reaction (PCR) Mastermix (Primerdesign Ltd., United Kingdom). An ABI 7500 real-time PCR system (Applied Biosystems, Foster City, CA) was used with the following thermocycling conditions: 95°C for 10 minutes, followed by 50 cycles of 95°C for 10 seconds and 60°C for 1 minute. HBV-specific primers used in this reaction targeted a highly conserved region of the HBV surface gene, which provides accurate detection of genotypes A to H. For these reactions, HBV-plasmid DNA with 10-fold serial dilution was used to generate a standard curve. The data were analyzed using the 7500 Fast Software version 2.1 (Applied Biosystems), and HBV viral load was expressed in IU/mL. The detection limit was set to 20 IU/mL (sensitivity and specificity >95%). The quantitative HBV-specific PCR assays were routinely standardized using the World Health Organization standards for HBV DNA (National Institute for Biological Standards and Control code 97/750).

HBV Genotypes and Subgenotypes

HBV genotyping is not routinely performed for patients in Burkina Faso and was performed on a randomly selected subset in our study. HBV DNA was used in a multiplex PCR to identify HBV genotypes A to F based on the method described by Chen et al.⁽³²⁾ The GeneAmp PCR System 9700 (Applied Biosystems) was used to amplify HBsAg gene sequences with the following PCR conditions: activation (95°C, 15 minutes), 35 cycles of denaturation (94°C, 1 minute), primer annealing (58°C, 1 minute), and extension (72°C, 1 minute), followed by a final extension (72°C, 10 minutes). PCR products were separated by electrophoresis in 2% agarose gels, and DNA was visualized under ultraviolet light. We sequentially first used primers specific for genotypes A, B, and C, followed by a second PCR with primers specific for genotypes D, E, and F. A result of two or more genotypes, for example, C/E, indicates that genotype C was found in the first PCR and E in the second PCR and is reported as having recombinant genotypes E and C. Our group has successfully used this approach and analysis of HBV pre-S1/S2 regions using Sanger sequencing to confirm the predominance of HBV-E in Burkina Faso.^{(22, 23)}

Statistical Analysis

Baseline characteristics were compared between patients with and without HCC and reported as percentages for categorical variables and median (interquartile range [IQR]) for continuous variables. Continuous data, including age, AST, ALT, HBV DNA viral load, and log₁₀ viral load, were compared using the Mann-Whitney U test. Categorical data, including sex, HBV serology, cirrhosis status, and coinfection status, were compared using the χ² or Fisher’s exact test. The associations of baseline characteristics with cirrhosis, HCC, and HBV viral load were analyzed by univariate and multivariable logistic regression. All analyses were performed using JMP version 10 software (SAS Institute Inc., Cary, NC). P < 0.05 was considered statistically significant.

Characteristics of Burkina Faso Patients Who Were CHB HCC+ or CHB HCC−

There were 59 patients who were CHB HCC+ and 3,885 patients who were CHB HCC−. Patients who were CHB HCC+ were significantly older (median, 43; IQR, 34-50; range, 25-73 years) than patients who were CHB HCC− (median, 32; IQR, 26-39; range, 1-85 years; P < 0.001) (Table 1). The percentage of male patients was higher in the CHB HCC+ cohort (n = 39; 66.1%) than in the CHB HCC− cohort (n = 1,717; 44.2%; P < 0.001). In the CHB HCC− cohort, 10 patients had multiple infections, including 5 with HBV/HCV (0.13%) and 5 with HBV/HIV (0.13%). No coinfections were found in the CHB HCC+ cohort. Fifty-four of the 3,885 (1.4%) CHB HCC- patients had cirrhosis, while 17 of 33 (51.5%) CHB HCC+ patients with available cirrhosis status had cirrhosis (P < 0.001).

The median serum aminotransferase levels (AST and ALT) reflective of active hepatitis were significantly higher in the CHB HCC+ group in than the CHB HCC− group (median AST, 90.5; IQR, 39.2-197.5 vs. median, 25.0; IQR, 19.0-34.0; P < 0.001, respectively, and median ALT, 59.5; IQR, 36.0-102.8 vs. median, 21.0; IQR, 14.0-31.0; P < 0.001, respectively). The median AST/ALT ratio was 1.4 (IQR, 1.1-2.7) in the CHB HCC+ group and 1.2 (IQR, 0.9-1.6) in the CHB HCC− group (P = 0.002).

Of the patients who were CHB HCC−, 2.5% (96/3,807) were HBeAg+ and 95.8% (3,593/3,752) were HBeAb+. The distribution of HBeAg+ by age was 4% (9/225), <20 years; 4.1% (54/1,317), 20-29 years; 1.9% (26/1,398), 30-39 years; 1.2% (7/569), 40-49 years; and 0% (0/298), ≥50 years. All 58 patients who were CHB HCC+ with available data were HBeAg− and HBeAb+. There was no significant difference in HBeAg or HBeAb serologic status between patients who were CHB HCC− and CHB HCC+ (P = 0.40 and 0.53, respectively).

Factors Associated With Cirrhosis in Patients Who Were CHB HCC−

Of the CHB HCC- patients, the 54 patients with cirrhosis were significantly older and had significantly higher HBV viral loads than the 3,831 patients without cirrhosis, with an odds ratio (OR) of 1.08 (95% confidence interval [CI], 1.06-1.11; P < 0.001) per 1-year increase in age and 1.26 (95% CI, 1.12-1.43; P < 0.001) per 1 log₁₀ IU/mL increase in viral load (Table 2). Male sex was not significantly associated with cirrhosis status (OR, 1.26; 95% CI, 0.71-2.28; P = 0.44).

Factors Associated With HCC in Patients With CHB

Comparing CHB HCC+ and CHB HCC- patients, older age, male sex, and higher HBV viral load were significantly associated with HCC (OR, 1.08; 95% CI, 1.06-1.10; P < 0.001 per 1-year increase in age; OR, 2.03; 95% CI, 1.11-3.69; P = 0.02 for male sex; and OR, 1.48; 95% CI, 1.31-1.67; P < 0.001 per 1 log₁₀ IU/mL increase in viral load) (Table 2).

HBV Viral Load of Patients Who Were CHB HCC+ or CHB HCC−

The median HBV viral load of patients who were CHB HCC+ (median, 83,593; IQR, 1,405-830,902 IU/mL) was significantly higher than that of patients who were CHB HCC− (median, 502; IQR, 0-6,475 IU/mL; P < 0.001) (Fig. 2).

Comparing patients who were CHB HCC− and those who were CHB HCC+, we found a significantly higher percentage of patients with HBV viral load ≥2,000 IU/mL (73.3% vs. 34.4%) and ≥20,000 IU/mL (60.0% vs. 18.7%) in the CHB HCC+ group than the CHB HCC− group (P < 0.001 and P < 0.001, respectively).

Among the CHB HCC− cohort, there were 1,337 patients (34.4%) with an HBV viral load ≥2,000 IU/mL, 728 patients (18.7%) with an HBV viral load ≥20,000 IU/mL, and 946 (24.4%) patients with undetectable viral DNA.

Among the CHB HCC+ cohort, there were 44 patients (73.3%) with an HBV viral load ≥2,000 IU/mL, 36 patients (60%) with an HBV viral load ≥20,000 IU/mL, and 4 patients (6.8%) with undetectable viral DNA.

Among the age groups, HBV viral loads of patients with HCC were significantly higher than those in patients without HCC, except in patients from 20 to 29 years of age. When patients in different age groups were classified by viral load, we found that individuals <20 years of age had the highest percentage of viral load ≥20,000 IU/mL (n = 80, 35.2%). Patients from 20 to 29 years of age had the second highest percentage of viral load ≥20,000 IU/mL (n = 272, 20.5%) compared with <15% of patients in the other age groups. Most patients had viral loads of 60 IU/mL or less (30.4%-35.3% across all age groups). There were significant differences in viral load among age groups (P = 0.004). We also observed similar trends of differential viral loads among different age groups between male and female patients (Fig. 3). However, with regression analysis, HBV viral load was not correlated with age (r = −0.028; 95% CI, −0.059-0.003).

Characteristics of CHB in Patients With Different HBV Genotypes

We performed HBV genotyping in 364 patients who were CHB HCC− and 41 patients who were CHB HCC+. Characteristics of patients with genotype E, other genotypes excluding genotype E, and nongenotyped patients are shown in Table 3. Of the 364 patients who were CHB HCC−, 256 (70.3%) had genotype E, 47 (12.9%) had genotype C/E, 21 (5.8%) had genotype C, and 14 (3.8%) had genotype B/C/E. Among 41 patients with HCC with genotyping, genotype E was found in 32 patients (78.0%). The other less common genotypes are shown in Table 1. There was no difference in numbers of patients with genotype E, C, or C/E among patients who were CHB HCC‒ and CHB HCC+ (P = 0.83, 0.36, and 0.76, respectively).

Among the patients who were CHB HCC−, we compared the 256 patients with genotype E to the 47 patients with genotype C/E and the 21 patients with genotype C (Table 3). There were fewer patients who were HbeAg+ in the genotype E group (6.3%) than the genotype C/E (14.9%) or genotype C (9.5%) groups (P = 0.06 and P = 0.58, respectively). More patients with genotype E (14.5%) than patients with genotypes C/E (10.6%) or C (4.8%) had cirrhosis (P = 0.47 and P = 0.16, respectively). The HBV viral load was lower in the genotype E group (median, 13,424; IQR, 1,463-310,013 IU/mL) than the C/E group (median, 19,496,410; IQR, 1,248,321-80,494,145 IU/mL;P < 0.001) or the C group (median, 939,129; IQR, 9,800-2,795,475 IU/mL; P = 0.003).

Characteristics of CHB With Genotype E With and Without HCC

Of the 256 patients who were CHB HCC− with exclusively genotype E, there were 16 (6.3%, 1 missing) patients who were HBeAg+ and 232 (92.1%, 4 missing) who were HBeAb+. None of the 32 patients with HCC were HBeAg+ (Supporting Table S1). Patients with HCC had significantly elevated levels of AST and ALT and a higher AST/ALT ratio compared with the patients without HCC (P < 0.001, P < 0.001, and P = 0.004, respectively; Supporting Table S1). HBV viral load was not significantly different between patients with and without HCC (P = 0.107).

Patient Classification by Phase of CHB Infection

Patients were classified into four phases of chronic HBV infection according to the 2016 American Association for the Study of Liver Diseases (AASLD) guidelines: (1) immunotolerance, (2) HBeAg+ immune active, (3) inactive CHB viral infection, and (4) HBeAg− immune reactivation.⁽³¹⁾ When strictly classified according to AASLD criteria, only 1,317 (33.9%) patients could be categorized. Patients who did not meet the inclusion requirements of any specific group were categorized on the basis of presence of HBeAg, elevated ALT level, and viral load, respectively. The number of patients in each phase by age group is shown in Fig. 4, with most patients categorized in the inactive CHB and HBeAg− immune reactivation phases.