Journal of Viral Hepatitis, 2013, 20, 890–896
doi:10.1111/jvh.12121
A paucity of liver disease in Canadian Inuit with chronic hepatitis B virus, subgenotype B6 infection G. Y. Minuk,1,2 S. MacRury,1 J. Uhanova,1 S. Caouette,1 N. Coleman,1 K. Cummings,1 B. Larke,3 L. Vardy,4 C. Triet Huyn5 and C. Osiowy5 1Section of Hepatology, Departments of Medicine, University of Manitoba, Winnipeg, MB, Canada; 2Pharmacology and Therapeutics, University of Manitoba, Winnipeg, MB, Canada; 3Public Health Consultant, Edmonton, AB, Canada; 4Centre for Immunization and Respiratory Infectious Diseases, Public Health Agency of Canada, Ottawa, ON, Canada; and 5
National Microbiology Laboratory, Public Health Agency of Canada, Winnipeg, MB,Canada
Received January 2013; accepted for publication April 2013
SUMMARY. Clinical observations suggest that chronic hepatitis B virus (HBV) infections in the Canadian Inuit are less often associated with serious adverse outcomes than has been described in other HBV-infected patient populations. The aim of this study was to document the clinical and biochemical features, liver-related morbidity and all-cause mortality in Canadian Inuit with chronic HBV infections. Administrative databases were reviewed for individuals identified as hepatitis B surface antigen (HBsAg) positive during a 1983–85 seroepidemiological survey of viral hepatitis in Baffin Island, Canada. An equal number of age- and gender-matched HBsAg-negative individuals from the same communities served as controls. Baseline HBV viral loads, genotypes and specific mutations were compared in HBsAgpositive survivors and nonsurvivors. A subset of surviving HBsAg-positive carriers were reassessed 25–30 years following their initial diagnosis for evidence of advanced liver disease and changes to their serological/virological findings. One hundred and forty four HBsAg-positive individuals were identified. All were Canadian Inuit. The mean age at diagnosis was 38 17 years and 69 (61%) were male. Median follow-up was 23 years (range: 2–28 years). Viral quantitation from stored sera could be performed in 70 infected indi-
INTRODUCTION Chronic infection with the hepatitis B virus (HBV) can result in cirrhosis and hepatocellular carcinoma (HCC) in Abbreviation: AFP, alpha-fetoprotein; ALT, alanine aminotransferase; AST, aspartate aminotransferase; Dx, diagnosis; GGT, gammaglutamyl transferase; Hb, haemoglobin; HBsAg, hepatitis B surface antigen; HCC, hepatocellular carcinoma; INR, international normalized ratio for prothrombin times; P.A.N, polyarteritis nodosa; Reg Coef, regression coefficient; WBC, white blood cell count. Correspondence: Dr. Gerald Y. Minuk, University of Manitoba, John Buhler Research Centre, 715 McDermot Ave, Winnipeg, MB R3E 3P4, Canada. E-mail: [email protected]
viduals. The median viral load was 4.3 log 10 IU/ml (range: 2.3–8.8 log 10 IU/ml), and all were genotype B, subgenotype B6. Liver biochemistry, morbidity and all-cause mortality rates were similar in HBsAg-positive carriers and controls. Following multivariate analyses, only age at diagnosis predicted mortality in HBsAg carriers. In a subset of 30 HBsAg-positive survivors who underwent follow-up assessments, clinical, biochemical and radiological examinations of the liver were essentially normal. 23/30 (77%) remained HBsAg positive and 17/19 (90%) HBV-DNA positive. The genotype and prevalence of genomic mutations in this cohort remained largely unchanged, but quantifiable viral loads were significantly lower (P < 0.003). The results of this study suggest that chronic HBV infections in the Canadian Inuit are infrequently associated with serious adverse outcomes. Whether this finding reflects unique features of the host, presence or absence of external factors that influence the course of HBV and/or intrinsic properties of the HBV B6 subgenotype remains to be determined. Keywords: chronic hepatitis B, cirrhosis, hepatitis, hepatocellular carcinoma, Inuit, liver disease, natural history, subgenotype B6.
some, but not all individuals [1]. Amongst other variables, age at infection, gender, concurrent liver disorders, viral load and specific mutations to the viral genome have been implicated in the pathogenesis of these sequelae [2]. Recently, data have emerged to suggest that HBV genotypes and subgenotypes might also influence the course and outcome of chronic HBV either independently or in combination with the above variables [3–5]. Chronic HBV infections are common in the Canadian Inuit (Eskimo) population with an estimated prevalence of 5% [6]. Despite this intermediate prevalence, clinical experience and cross-sectional studies of single communities suggest that cirrhosis and HCC are uncommon findings in this population [7]. Thus, the purpose of the present study © 2013 John Wiley & Sons Ltd
Chronic HBV in Canadian Inuit was to formally document the clinical features, liver-related morbidity and all-cause mortality in subjects diagnosed with HBV infection 25–30 years earlier and correlate the outcomes with variables that are thought to influence the course of the disease.
891
alpha-fetoprotein, HBVserology/DNA testing and ultrasounds of the liver were performed. Due to logistical issues with respect to travel, these clinics were held in the three communities with the highest known prevalences of HBV infection (i.e. designed to capture the largest number of HBV carriers).
MATERIALS AND METHODS Statistics Study population The study population was derived from 3267 individuals residing in 12 communities located on Baffin Island, Nunavut, Canada, who participated in a seroepidemiological survey of HBV in the Canadian North in 1983–1985. These individuals represented approximately 30% of the total Island’s population and >95% were Canadian Inuit. As participants in the serosurvey, all subjects had been tested for hepatitis B surface antigen (HBsAg). Those found to be positive were matched for age (5 years) and gender with HBsAg-negative subjects residing in the same communities. Each subject involved in the initial and follow-up study provided informed consent for HBV testing, and the study protocols were approved by the Nunavut Research Institute and University of Manitoba Conjoint Ethics Committee.
Statistical analysis was performed using NCSS statistical software. Categorical variables were expressed as proportions and evaluated using chi-squared analysis (or F-test when warranted). Continuous variables, expressed as means with standard deviations and ranges, were assessed using Student’s t-test or analysis of variance. Univariate and multivariate regression analyses were carried out to determine whether patients’ demographic features, liver biochemistry, viral loads and viral mutations associated with mortality.
RESULTS Study population
The methodologies employed for all serological testing have been described previously [6]. HBsAg-positive sera were processed for HBV-DNA testing, and positive results were quantitated by real-time PCR using a Qiagen artus HBV TM PCR kit (Qiagen, Mississauga, ON, Canada) [8]. For genotype/subgenotype and specific mutation determinations, sequencing of the HBsAg and precore/core coding regions was performed as described previously [9].
Of the 3267 samples obtained in 1983–85, 168 (5.1%) were HBsAg positive. Of these, nursing station, hospital and/or vital statistics records were available in 114 (68%) cases. As shown in Table 1, 69 (61%) were male. The mean age at diagnosis was 38 17 years (range 6–74 years). The median follow-up for this cohort was 23 years (range 2–27 years). As per study design, 69 (61%) of the HBsAg-negative controls were male. The mean age of controls was 34 16 years (range 6–77 years). The median follow-up in the control cohort was also 23 years (range: 12–28 years).
Database reviews
HBV serology
Nursing station, hospital and/or vital statistics records of all HBsAg-positive patients and HBsAg-negative controls were reviewed for biochemical evidence of liver disease, liver-related hospitalizations and all-cause mortality. Groups of diagnoses associated with HBV infection or liver disease were categorized based on International Classification of Diseases (ICD)-8 (1 January 1983–31 December 1993) and ICD-10 (1 January 1994–1 January 2012) codes.
Of the 114 HBsAg-positive carriers 12 (11%) were HBeAg and 90 (79%) anti-HBe positive at diagnosis. The remaining 12 samples were negative for both HBeAg and antiHBe or had insufficient volume for testing. Forty-six (40%) of the 114 HBsAg-positive carriers were anti-HBs-positive controls compared with 24/114 (21%) HBsAg-negative controls (P < 0.005). Of note, these samples were obtained prior to the availability of HBV vaccines in the Canadian North.
Clinical assessments
HBV-DNA testing
Hepatitis B surface antigen (HBsAg)-positive patients still residing in three communities were invited to attend a nursing station clinic wherein a complete history, physical examination, laboratory testing for liver biochemistry,
Seventy-eight of the 114 (68%) HBsAg-positive samples were HBV-DNA positive. Of these, viral quantitation could be performed in 70 (90%). The median viral load was 4.3 log 10 IU/ml (range: 2.3–8.8 log 10 IU/ml).
Laboratory testing
© 2013 John Wiley & Sons Ltd
892
G. Y. Minuk et al.
Male Gender Age at Dx Age at F-up Median F-up
HBsAg Positive (N = 114)
Controls (N = 114)
P-value
69 (61%) 38 years (6–74) 58 years (20–89) 23 years (2–27)
69 (61%) 34 years (6–77) 56 years (25–88) 23 years (12–28)
0.10 0.25 0.11
Table 1 Demographic features and length of follow-up in HBsAg-positive and HBsAg-negative (controls) Canadian Inuit
Dx, diagnosis; F-up, follow-up; HBsAg, hepatitis B surface antigen. Based on HBsAg sequencing results, all samples were genotype B, subgenotype B6. Based on precore/core region sequencing, the precore (nt 1896) mutation was present in 57/78 (73%) samples, wild type in 17/78 (22%) and mixed in 1/78 (1.3%). Three samples were PCR negative. Basal core promoters (T1762/A1764) mutations were absent. One of the 114 HBsAg-negative controls was HBV-DNA positive, but at a viral load too low to quantitate or sequence. Due to the unexpected high prevalence of concurrent anti-HBs (40%) in HBsAg-positive samples, sequences from HBsAg-positive/anti-HBs-positive and HBsAg-positive/antiHBs-negative samples were aligned and translated, focusing on the antigenic determinant region from amino acids 100–160 of the HBsAg protein. Sixteen of 73 (22%) HBsAg/anti-HBs-positive sequences had mutations in this region compared with 21/116 (18%) HBsAg-positive/antiHBs-negative samples. One of the HBsAg-positive/anti-HBsnegative samples had the G145R ‘S escape’ mutation.
Liver enzyme and function testing The results of liver enzyme and function testing from chart reviews of HBsAg-positive and HBsAg-negative controls are provided in Table 2. At last follow-up visit, 12/66 (18%) HBsAg-positive individuals had elevated alanine aminotransferase (ALT) levels vs 7/52 (14%) HBsAg-negative controls. Low albumin and elevated bilirubin levels were
detected in 4/32 (13%) and 3/55 (4.5%) HBsAg-positive subjects compared with 0/14 (0%) and 0/50 (0%) controls, respectively (P = 0.17 and 0.13). Prolonged prothrombin times or elevated INR values were present in 3/11 (27%) and 5/22 (23%) HBsAg-positive carriers vs 0/4 (0%) and 2/14 (14%) controls, respectively (P = 0.53).
Hospitalizations Hospitalizations for hepatic disorders (Table 3) were similar in HBsAg-positive (20/114, 14%) and HBsAg-negative (11/114, 9.7%) controls (P = 0.13). In both populations,
Table 3 Liver-related hospitalizations in 144 HBsAgpositive and HBsAg-negative (controls) Canadian Inuit
Cirrhosis Hepatitis Hepatomegaly Glomerulonephritis P.A.N. Cholecystectomy HCC Total
HBsAg Positive
Controls
P-value
1 1 1 1 0 16 0 20
0 0 0 0 1 8 2 11
0.22 0.47 0.13
(1%) (1%) (1%) (1%) (11%) (14%)
(1%) (5.3%) (1.4%) (9.7%)
HBsAg, hepatitis B surface antigen; HCC, hepatocellular carcinoma; P.A.N., polyarteritis nodosa.
Test (normal values)
HBsAg Positive
Controls
P-value
ALT (0–55 U/L) Albumin (35–50 g/L) Total Bilirubin (0–20.5 lmol/L) INR (1.0–1.2) Elevated ALT Low Albumin Elevated Total Bilirubin Prolonged INR
45 28 37 8.0 8.3 5.0
45 48 40 4.0 8.4 4.0
0.96 0.18 0.86
1.2 (0.7) 12/66 (18%) 4/32 (13%) 3/55 (4.5%) 8/35 (23%)
1.1 0.2 7/52 (14%) 0/14 (0%) 0/50 (0%) 2/14 (14%)
0.55 0.49 0.17 0.13 0.53
Table 2 Liver biochemistry in HBsAgpositive and HBsAg-negative (controls) Canadian Inuit
ALT, alanine aminotransferase; HBsAg, hepatitis B surface antigen; INR, international normalized ratio for prothrombin times. © 2013 John Wiley & Sons Ltd
Chronic HBV in Canadian Inuit Table 4 Hepatitis B virus features in HBsAg-positive Canadian Inuit survivors and non survivors at baseline testing
Mean log viral load Genotype B Subtype B6 Precore (nt. 1896) BCP (nt. 1762/64)
hospitalizations for cholecystectomies were most common (N = 16 (11%) and 8 (5.3%), respectively).
Survivors (N = 45)
Nonsurvivors (N = 33)
5.1 2.6 (2.3–8.8) 100% 100% 62% 0%
3.1 0.2 (2.5–4.1) 100% 100% 88% 0%
893
P-value 0.04
0.02
However, only age at the time of sampling remained significant after including all three variables in the multivariate model.
Deaths A total of 44/114 (39%) HBsAg-positive individuals died during the follow-up period compared with 45/114 (40%) HBsAg-negative controls (P = 0.89). The mean ages at death were 68 12 years (range: 43–89 years) vs 66 16 years (range: 27–68 years), respectively (P = 0.88). Liver-related deaths were 0/44 (0%) and 2/45 (1.4%), respectively (P = 0.47).
HBsAg survivors vs non survivors Table 4 provides the viral loads, genotypes and mutations in surviving vs deceased HBsAg-positive carriers. Viral loads were significantly higher in survivors vs nonsurvivors (5.1 2.6 vs 3.1 0.2 log 10 IU/L, P = 0.04). All subjects were infected with genotype B, subgenotype B6. More nonsurvivors had precore (nt 1896) mutations than survivors (88% vs 62%, P = 0.02). No individual tested positive for the basal core promoter (nt 1762/64) mutations. As shown in Table 5, following univariate analysis, three variables had statistically significant associations with death as an outcome: age at the time of initial sampling, viral load and the presence of precore mutations.
Clinical assessments A total of 52 HBsAg-positive individuals were still residing in the three communities selected for follow-up evaluations. Thirty (58%) appeared at the nursing station and consented to clinical assessments, ultrasound examinations of the abdomen, liver biochemistry, hepatocellular carcinoma tumour markers and further HBV testing. Of these individuals, none provided a history or had physical stigmata in keeping with decompensated liver disease (jaundice, fluid retention, encephalopathy, cachexia, etc.). Ultrasound examinations of the liver were positive for fatty infiltration in 6/30 (20%) subjects, and 1/30 (3.3%) had an irregular liver edge suggestive of cirrhosis. No patient had ascites, splenomegaly or space occupying lesions within the liver. The demographic features and results of complete blood counts, liver biochemistry and alpha-fetoprotein testing in this cohort are provided in Table 6. Follow-up serological testing revealed that 23/30 (77%) previously HBsAg-positive subjects remained positive. In the seven individuals who had become HBsAg negative, 5/7 (71%) seroconverted to anti-HBs positive. All 30 subjects were anti-HCV negative.
Table 5 Predictors of mortality in HBsAg-positive Canadian Inuit Univariate
HBsAg pos Male gender Age at Dx ALT (0–55 U/L) Albumin (35–50 g/L) Total bilirubin (0–20.5 lmol/L) INR (1.0–1.2) Viral load Precore mutation
OR
95% CI
1.27 0.70
0.8–2.1 0.4–1.3
Multivariate Reg Coef
0.11 0.02 0.07 0.07 2.18 0.32 4.4
1.3–14.7
P 0.36 0.23 0.0000 0.54 0.39 0.68 0.73 0.01 0.02
OR
0.93
95% CI
0.1–7.1
Reg Coef
P
0.15
0.00002
0.08
0.74 0.94
ALT, alanine aminotransferase; Dx, diagnosis; HBsAg, hepatitis B surface antigen; INR, international normalized ratio for prothrombin times; Reg Coef, regression coefficient. © 2013 John Wiley & Sons Ltd
894
G. Y. Minuk et al. Mean SD
Age (years) Male Hb (134–166 g/L) WBC (4.0–11.0 9 109/L) Platelets (156–416 9 109/L) ALT (0–55 U/L) AST (5–34 U/L) Alkaline Phosphatase (40–150 U/L) GGT (12–64 U/L) Total Bilirubin (0–20.5 lmol/L) AFP (
doi:10.1111/jvh.12121
A paucity of liver disease in Canadian Inuit with chronic hepatitis B virus, subgenotype B6 infection G. Y. Minuk,1,2 S. MacRury,1 J. Uhanova,1 S. Caouette,1 N. Coleman,1 K. Cummings,1 B. Larke,3 L. Vardy,4 C. Triet Huyn5 and C. Osiowy5 1Section of Hepatology, Departments of Medicine, University of Manitoba, Winnipeg, MB, Canada; 2Pharmacology and Therapeutics, University of Manitoba, Winnipeg, MB, Canada; 3Public Health Consultant, Edmonton, AB, Canada; 4Centre for Immunization and Respiratory Infectious Diseases, Public Health Agency of Canada, Ottawa, ON, Canada; and 5
National Microbiology Laboratory, Public Health Agency of Canada, Winnipeg, MB,Canada
Received January 2013; accepted for publication April 2013
SUMMARY. Clinical observations suggest that chronic hepatitis B virus (HBV) infections in the Canadian Inuit are less often associated with serious adverse outcomes than has been described in other HBV-infected patient populations. The aim of this study was to document the clinical and biochemical features, liver-related morbidity and all-cause mortality in Canadian Inuit with chronic HBV infections. Administrative databases were reviewed for individuals identified as hepatitis B surface antigen (HBsAg) positive during a 1983–85 seroepidemiological survey of viral hepatitis in Baffin Island, Canada. An equal number of age- and gender-matched HBsAg-negative individuals from the same communities served as controls. Baseline HBV viral loads, genotypes and specific mutations were compared in HBsAgpositive survivors and nonsurvivors. A subset of surviving HBsAg-positive carriers were reassessed 25–30 years following their initial diagnosis for evidence of advanced liver disease and changes to their serological/virological findings. One hundred and forty four HBsAg-positive individuals were identified. All were Canadian Inuit. The mean age at diagnosis was 38 17 years and 69 (61%) were male. Median follow-up was 23 years (range: 2–28 years). Viral quantitation from stored sera could be performed in 70 infected indi-
INTRODUCTION Chronic infection with the hepatitis B virus (HBV) can result in cirrhosis and hepatocellular carcinoma (HCC) in Abbreviation: AFP, alpha-fetoprotein; ALT, alanine aminotransferase; AST, aspartate aminotransferase; Dx, diagnosis; GGT, gammaglutamyl transferase; Hb, haemoglobin; HBsAg, hepatitis B surface antigen; HCC, hepatocellular carcinoma; INR, international normalized ratio for prothrombin times; P.A.N, polyarteritis nodosa; Reg Coef, regression coefficient; WBC, white blood cell count. Correspondence: Dr. Gerald Y. Minuk, University of Manitoba, John Buhler Research Centre, 715 McDermot Ave, Winnipeg, MB R3E 3P4, Canada. E-mail: [email protected]
viduals. The median viral load was 4.3 log 10 IU/ml (range: 2.3–8.8 log 10 IU/ml), and all were genotype B, subgenotype B6. Liver biochemistry, morbidity and all-cause mortality rates were similar in HBsAg-positive carriers and controls. Following multivariate analyses, only age at diagnosis predicted mortality in HBsAg carriers. In a subset of 30 HBsAg-positive survivors who underwent follow-up assessments, clinical, biochemical and radiological examinations of the liver were essentially normal. 23/30 (77%) remained HBsAg positive and 17/19 (90%) HBV-DNA positive. The genotype and prevalence of genomic mutations in this cohort remained largely unchanged, but quantifiable viral loads were significantly lower (P < 0.003). The results of this study suggest that chronic HBV infections in the Canadian Inuit are infrequently associated with serious adverse outcomes. Whether this finding reflects unique features of the host, presence or absence of external factors that influence the course of HBV and/or intrinsic properties of the HBV B6 subgenotype remains to be determined. Keywords: chronic hepatitis B, cirrhosis, hepatitis, hepatocellular carcinoma, Inuit, liver disease, natural history, subgenotype B6.
some, but not all individuals [1]. Amongst other variables, age at infection, gender, concurrent liver disorders, viral load and specific mutations to the viral genome have been implicated in the pathogenesis of these sequelae [2]. Recently, data have emerged to suggest that HBV genotypes and subgenotypes might also influence the course and outcome of chronic HBV either independently or in combination with the above variables [3–5]. Chronic HBV infections are common in the Canadian Inuit (Eskimo) population with an estimated prevalence of 5% [6]. Despite this intermediate prevalence, clinical experience and cross-sectional studies of single communities suggest that cirrhosis and HCC are uncommon findings in this population [7]. Thus, the purpose of the present study © 2013 John Wiley & Sons Ltd
Chronic HBV in Canadian Inuit was to formally document the clinical features, liver-related morbidity and all-cause mortality in subjects diagnosed with HBV infection 25–30 years earlier and correlate the outcomes with variables that are thought to influence the course of the disease.
891
alpha-fetoprotein, HBVserology/DNA testing and ultrasounds of the liver were performed. Due to logistical issues with respect to travel, these clinics were held in the three communities with the highest known prevalences of HBV infection (i.e. designed to capture the largest number of HBV carriers).
MATERIALS AND METHODS Statistics Study population The study population was derived from 3267 individuals residing in 12 communities located on Baffin Island, Nunavut, Canada, who participated in a seroepidemiological survey of HBV in the Canadian North in 1983–1985. These individuals represented approximately 30% of the total Island’s population and >95% were Canadian Inuit. As participants in the serosurvey, all subjects had been tested for hepatitis B surface antigen (HBsAg). Those found to be positive were matched for age (5 years) and gender with HBsAg-negative subjects residing in the same communities. Each subject involved in the initial and follow-up study provided informed consent for HBV testing, and the study protocols were approved by the Nunavut Research Institute and University of Manitoba Conjoint Ethics Committee.
Statistical analysis was performed using NCSS statistical software. Categorical variables were expressed as proportions and evaluated using chi-squared analysis (or F-test when warranted). Continuous variables, expressed as means with standard deviations and ranges, were assessed using Student’s t-test or analysis of variance. Univariate and multivariate regression analyses were carried out to determine whether patients’ demographic features, liver biochemistry, viral loads and viral mutations associated with mortality.
RESULTS Study population
The methodologies employed for all serological testing have been described previously [6]. HBsAg-positive sera were processed for HBV-DNA testing, and positive results were quantitated by real-time PCR using a Qiagen artus HBV TM PCR kit (Qiagen, Mississauga, ON, Canada) [8]. For genotype/subgenotype and specific mutation determinations, sequencing of the HBsAg and precore/core coding regions was performed as described previously [9].
Of the 3267 samples obtained in 1983–85, 168 (5.1%) were HBsAg positive. Of these, nursing station, hospital and/or vital statistics records were available in 114 (68%) cases. As shown in Table 1, 69 (61%) were male. The mean age at diagnosis was 38 17 years (range 6–74 years). The median follow-up for this cohort was 23 years (range 2–27 years). As per study design, 69 (61%) of the HBsAg-negative controls were male. The mean age of controls was 34 16 years (range 6–77 years). The median follow-up in the control cohort was also 23 years (range: 12–28 years).
Database reviews
HBV serology
Nursing station, hospital and/or vital statistics records of all HBsAg-positive patients and HBsAg-negative controls were reviewed for biochemical evidence of liver disease, liver-related hospitalizations and all-cause mortality. Groups of diagnoses associated with HBV infection or liver disease were categorized based on International Classification of Diseases (ICD)-8 (1 January 1983–31 December 1993) and ICD-10 (1 January 1994–1 January 2012) codes.
Of the 114 HBsAg-positive carriers 12 (11%) were HBeAg and 90 (79%) anti-HBe positive at diagnosis. The remaining 12 samples were negative for both HBeAg and antiHBe or had insufficient volume for testing. Forty-six (40%) of the 114 HBsAg-positive carriers were anti-HBs-positive controls compared with 24/114 (21%) HBsAg-negative controls (P < 0.005). Of note, these samples were obtained prior to the availability of HBV vaccines in the Canadian North.
Clinical assessments
HBV-DNA testing
Hepatitis B surface antigen (HBsAg)-positive patients still residing in three communities were invited to attend a nursing station clinic wherein a complete history, physical examination, laboratory testing for liver biochemistry,
Seventy-eight of the 114 (68%) HBsAg-positive samples were HBV-DNA positive. Of these, viral quantitation could be performed in 70 (90%). The median viral load was 4.3 log 10 IU/ml (range: 2.3–8.8 log 10 IU/ml).
Laboratory testing
© 2013 John Wiley & Sons Ltd
892
G. Y. Minuk et al.
Male Gender Age at Dx Age at F-up Median F-up
HBsAg Positive (N = 114)
Controls (N = 114)
P-value
69 (61%) 38 years (6–74) 58 years (20–89) 23 years (2–27)
69 (61%) 34 years (6–77) 56 years (25–88) 23 years (12–28)
0.10 0.25 0.11
Table 1 Demographic features and length of follow-up in HBsAg-positive and HBsAg-negative (controls) Canadian Inuit
Dx, diagnosis; F-up, follow-up; HBsAg, hepatitis B surface antigen. Based on HBsAg sequencing results, all samples were genotype B, subgenotype B6. Based on precore/core region sequencing, the precore (nt 1896) mutation was present in 57/78 (73%) samples, wild type in 17/78 (22%) and mixed in 1/78 (1.3%). Three samples were PCR negative. Basal core promoters (T1762/A1764) mutations were absent. One of the 114 HBsAg-negative controls was HBV-DNA positive, but at a viral load too low to quantitate or sequence. Due to the unexpected high prevalence of concurrent anti-HBs (40%) in HBsAg-positive samples, sequences from HBsAg-positive/anti-HBs-positive and HBsAg-positive/antiHBs-negative samples were aligned and translated, focusing on the antigenic determinant region from amino acids 100–160 of the HBsAg protein. Sixteen of 73 (22%) HBsAg/anti-HBs-positive sequences had mutations in this region compared with 21/116 (18%) HBsAg-positive/antiHBs-negative samples. One of the HBsAg-positive/anti-HBsnegative samples had the G145R ‘S escape’ mutation.
Liver enzyme and function testing The results of liver enzyme and function testing from chart reviews of HBsAg-positive and HBsAg-negative controls are provided in Table 2. At last follow-up visit, 12/66 (18%) HBsAg-positive individuals had elevated alanine aminotransferase (ALT) levels vs 7/52 (14%) HBsAg-negative controls. Low albumin and elevated bilirubin levels were
detected in 4/32 (13%) and 3/55 (4.5%) HBsAg-positive subjects compared with 0/14 (0%) and 0/50 (0%) controls, respectively (P = 0.17 and 0.13). Prolonged prothrombin times or elevated INR values were present in 3/11 (27%) and 5/22 (23%) HBsAg-positive carriers vs 0/4 (0%) and 2/14 (14%) controls, respectively (P = 0.53).
Hospitalizations Hospitalizations for hepatic disorders (Table 3) were similar in HBsAg-positive (20/114, 14%) and HBsAg-negative (11/114, 9.7%) controls (P = 0.13). In both populations,
Table 3 Liver-related hospitalizations in 144 HBsAgpositive and HBsAg-negative (controls) Canadian Inuit
Cirrhosis Hepatitis Hepatomegaly Glomerulonephritis P.A.N. Cholecystectomy HCC Total
HBsAg Positive
Controls
P-value
1 1 1 1 0 16 0 20
0 0 0 0 1 8 2 11
0.22 0.47 0.13
(1%) (1%) (1%) (1%) (11%) (14%)
(1%) (5.3%) (1.4%) (9.7%)
HBsAg, hepatitis B surface antigen; HCC, hepatocellular carcinoma; P.A.N., polyarteritis nodosa.
Test (normal values)
HBsAg Positive
Controls
P-value
ALT (0–55 U/L) Albumin (35–50 g/L) Total Bilirubin (0–20.5 lmol/L) INR (1.0–1.2) Elevated ALT Low Albumin Elevated Total Bilirubin Prolonged INR
45 28 37 8.0 8.3 5.0
45 48 40 4.0 8.4 4.0
0.96 0.18 0.86
1.2 (0.7) 12/66 (18%) 4/32 (13%) 3/55 (4.5%) 8/35 (23%)
1.1 0.2 7/52 (14%) 0/14 (0%) 0/50 (0%) 2/14 (14%)
0.55 0.49 0.17 0.13 0.53
Table 2 Liver biochemistry in HBsAgpositive and HBsAg-negative (controls) Canadian Inuit
ALT, alanine aminotransferase; HBsAg, hepatitis B surface antigen; INR, international normalized ratio for prothrombin times. © 2013 John Wiley & Sons Ltd
Chronic HBV in Canadian Inuit Table 4 Hepatitis B virus features in HBsAg-positive Canadian Inuit survivors and non survivors at baseline testing
Mean log viral load Genotype B Subtype B6 Precore (nt. 1896) BCP (nt. 1762/64)
hospitalizations for cholecystectomies were most common (N = 16 (11%) and 8 (5.3%), respectively).
Survivors (N = 45)
Nonsurvivors (N = 33)
5.1 2.6 (2.3–8.8) 100% 100% 62% 0%
3.1 0.2 (2.5–4.1) 100% 100% 88% 0%
893
P-value 0.04
0.02
However, only age at the time of sampling remained significant after including all three variables in the multivariate model.
Deaths A total of 44/114 (39%) HBsAg-positive individuals died during the follow-up period compared with 45/114 (40%) HBsAg-negative controls (P = 0.89). The mean ages at death were 68 12 years (range: 43–89 years) vs 66 16 years (range: 27–68 years), respectively (P = 0.88). Liver-related deaths were 0/44 (0%) and 2/45 (1.4%), respectively (P = 0.47).
HBsAg survivors vs non survivors Table 4 provides the viral loads, genotypes and mutations in surviving vs deceased HBsAg-positive carriers. Viral loads were significantly higher in survivors vs nonsurvivors (5.1 2.6 vs 3.1 0.2 log 10 IU/L, P = 0.04). All subjects were infected with genotype B, subgenotype B6. More nonsurvivors had precore (nt 1896) mutations than survivors (88% vs 62%, P = 0.02). No individual tested positive for the basal core promoter (nt 1762/64) mutations. As shown in Table 5, following univariate analysis, three variables had statistically significant associations with death as an outcome: age at the time of initial sampling, viral load and the presence of precore mutations.
Clinical assessments A total of 52 HBsAg-positive individuals were still residing in the three communities selected for follow-up evaluations. Thirty (58%) appeared at the nursing station and consented to clinical assessments, ultrasound examinations of the abdomen, liver biochemistry, hepatocellular carcinoma tumour markers and further HBV testing. Of these individuals, none provided a history or had physical stigmata in keeping with decompensated liver disease (jaundice, fluid retention, encephalopathy, cachexia, etc.). Ultrasound examinations of the liver were positive for fatty infiltration in 6/30 (20%) subjects, and 1/30 (3.3%) had an irregular liver edge suggestive of cirrhosis. No patient had ascites, splenomegaly or space occupying lesions within the liver. The demographic features and results of complete blood counts, liver biochemistry and alpha-fetoprotein testing in this cohort are provided in Table 6. Follow-up serological testing revealed that 23/30 (77%) previously HBsAg-positive subjects remained positive. In the seven individuals who had become HBsAg negative, 5/7 (71%) seroconverted to anti-HBs positive. All 30 subjects were anti-HCV negative.
Table 5 Predictors of mortality in HBsAg-positive Canadian Inuit Univariate
HBsAg pos Male gender Age at Dx ALT (0–55 U/L) Albumin (35–50 g/L) Total bilirubin (0–20.5 lmol/L) INR (1.0–1.2) Viral load Precore mutation
OR
95% CI
1.27 0.70
0.8–2.1 0.4–1.3
Multivariate Reg Coef
0.11 0.02 0.07 0.07 2.18 0.32 4.4
1.3–14.7
P 0.36 0.23 0.0000 0.54 0.39 0.68 0.73 0.01 0.02
OR
0.93
95% CI
0.1–7.1
Reg Coef
P
0.15
0.00002
0.08
0.74 0.94
ALT, alanine aminotransferase; Dx, diagnosis; HBsAg, hepatitis B surface antigen; INR, international normalized ratio for prothrombin times; Reg Coef, regression coefficient. © 2013 John Wiley & Sons Ltd
894
G. Y. Minuk et al. Mean SD
Age (years) Male Hb (134–166 g/L) WBC (4.0–11.0 9 109/L) Platelets (156–416 9 109/L) ALT (0–55 U/L) AST (5–34 U/L) Alkaline Phosphatase (40–150 U/L) GGT (12–64 U/L) Total Bilirubin (0–20.5 lmol/L) AFP (
