Table 1.
Design of studies.
| Country | Design | ||
|---|---|---|---|
| Study period | • Inclusion criteria | ||
| Reference | Study population | • Exclusion criteria | Limitations |
| China | Ethnic Chinese CHB patients with liver biopsy or transient elastography between 2005 and 2012 | Cohort | Kaplan–Meier and Cox regression analysis based on age at HBeAg seroclearance as the end point instead of time from baseline to HBeAg seroclearance; effect of metabolic risks modification during the follow-up period on immune virological response and eventually clinical outcomes not determined |
| 2005–2013 | • HBsAg positive for > 6 months | ||
| Hsiang et al. 201413 | • HCV or HIV co-infection; other causes of hepatitis, including autoimmune liver disease, Wilson's disease and haemochromatosis; > 20 g alcohol consumption/day; previous interferon therapy; follow-up period < 12 months | ||
| Taiwan | Adults > 20 y of age with CHB from randomly sampled enrolees from the NHIRD claims data | Cohort | Possibility of misclassification of both the risk and the outcome of interest (diabetes and cirrhosis); unknown validity of the ICD-9 codes to define cirrhosis, esophageal varices, or decompensated cirrhosis; possibility of undetected cirrhosis leading to the development of diabetes; potentially important variables related to diabetes (severity and treatment of diabetes, triglyceride level, and BMI) not included. |
| 1997–2009 | • NR | ||
| Huang et al. 201314 | • Diagnosis of cirrhosis or esophageal varices before the inception point for follow-up; alcoholic cirrhosis or biliary cirrhosis; alcoholic liver disease or hepatitis C infection | ||
| Taiwan | Patients with a chart-documented history of CHB infection | Cohort | NR |
| NRHuo et al. 200027 | • Seropositive for HBsAg ≥ 6 months; absence of liver cirrhosis; available data for serum biochemical studies and viral serology; at least 2 clinical visits/year; a follow-up for at least one year; non-alcoholism; no previous therapy with interferon or other anti-viral agent | ||
| • NR | |||
| Greece | HBeAg-negative CHB patients admitted to undergo liver biopsy | Cohort | Causative effect of DM on the histological progression of chronic viral hepatitis not directly supported by data |
| 1998–2003 | • NR | ||
| Papatheodoridis et al, 200628 | • Malignancy; antiviral or immunosuppressive therapy within the last 6 months; inadequate biopsy; HBV and HCV co-infection; detectable antibodies against hepatitis delta virus or HIV | ||
| China | Patients with CHB identified in primary care clinics and hospitals in different regions of Hong Kong | Cohort | Transient elastography instead of liver biopsy was used as the diagnostic tool to define liver fibrosis. Given that the natural history of liver fibrosis progression occurs over decades, a follow-up period of close to 4 y may still be too short. |
| 2006–2008 | • NR | ||
| Wong et al. 201429 | • evidence of hepatitis C virus; men who consumed >30 g of alcohol/w and women who consumed >20 grams of alcohol/w; secondary causes of hepatic steatosis; decompensated liver disease; complications of liver surgery/liver transplantation | ||
| Taiwan | > 40 y of age participants from the cancer screening program conducted by the Tainan | Cohort | Use of abdominal girth instead of BMI; patients with fasting blood sugar >126 mg/dL defined as DM cases; pathology reports unavailable; no active surveillance |
| 2004–2007 | County Health Bureau | • NR | |
| Chen et al. 201331 | • Participants with incomplete data or HCC at baseline were excluded | ||
| Taiwan | Residents ≥ 35 y old | Cohort | Abdominal ultrasonography screening not performed for seronegative participants; |
| 1997–2004 | • NR | Possibility of overestimating the risk of developing HCC in viral hepatitis–positive participants; effect of antiviral therapies on development of HCC not investigated; small sample size for incident HCC cases | |
| Wang et al. 200934 | • Diagnosed with HCC before the screening | ||
| Taiwan | Participants of cancer screening program held between 1991 and 1992 | Cohort | Self-report of diabetes status; no information on high-density lipoprotein cholesterol levels, and blood pressures; small numbers of HCC cases with extreme obesity; risk factors based on BMI measurement at enrolment; definition of extreme obesity (BMI 30 kg/m2) may not apply globally |
| Chen et al. 200812 | • NR | ||
| • Subjects with liver cancer before/at enrollment | |||
| China | Cases: Hospitalized patients with HBV related HCC at Jinan Infectious | Case-control | Overweight and obese subjects not included; all HCC and cirrhosis subjects not diagnosed histologically. |
| 2004–2008 | Disease Hospital | • Hospitalized for HCC or CHB; ≥ 30 y of age; HBsAg positive; anti-HCV negative; no history of cancer other than HCC or hepatitis other than hepatitis B; no cancer treatment; no treatment with nucleotide/nucleosides or interferon; residence of Shandong Province | |
| Li et al. 201215 | Controls: CHB patients without HCC hospitalized at Jinan Infectious | • NR | |
| Disease Hospital (hospital cross-sectional CHB controls) | |||
| Taiwan | Cases: All hospitalized inpatients first diagnosed with HCC in the computerized database | Case-control | Biased control group in terms of risk factors in association with HCC; difficult to investigate the relationship between DM and HCC; no information on fasting insulin or postprandial glucose level; effect of antiviral treatment of both HBV and HCV infection not considered; obesity regarded as an independent factor in association with HCC. |
| 2004–2005 | Controls: Subjects randomly selected from those participating in the health check-up program at the hospital (controls) | • Cases: ICD-9-CM code 155.0. Controls: NR | |
| Ko et al. 201233 | • Cases: Other primary liver cancer indicated in ICD-9-CM code 155.0 were excluded after reviewing the medical chart; colorectal cancer (ICD-9-CM code 153.0–153.9, 154.0, 154.1 and 154.8), gastric cancer (ICD-9-CM code 151.0–151.9), pancreatic cancer (ICD-9-CM code 157.0–157.8) breast cancer (ICD-9-CM code 174.0–174.9) or lung cancer (ICD-9-CM code 162.0–162.9) | ||
| Taiwan | Enrollees from the NHIRD between 1997–2009 | Cohort | Possibility of misclassification of DM and HCC cases; DM diagnosis was not systematically screened at regular intervals; possible under-recognized or under-reported cases; unknown validity of the 9th ICD codes; all variables related to DM or CHB (severity of diabetes, glucose level, HbA1c level, triglyceride level, body mass index, HBV DNA level, HBeAg and family history of HCC) not included; synergism between obesity and alcohol in relation to incident diabetes unknown |
| 1997–2009 | • NR | ||
| Fu et al. 201532 | • Patients with the HCC diagnosis before the inception point; patients with the diagnosis of cirrhosis, either alcoholic/biliary origin; patients with alcoholic liver disease/hepatitis C infection | ||
| China | Patients diagnosed with LC from June 2003 to July 2013 | Cohort | Retrospective design. |
| 2003–2013 | • NR | ||
| Xiong et al. 201535 | • NR | ||
| Korea | Multicenter, retrospective study of CHB patients who underwent LB and TE before starting antiviral therapy | Cohort | Retrospective design; the number of patients who developed HCC was small (8.9%), which was related to the characteristics of the study population who were receiving antiviral therapy |
| 2005–2015 | • NR | ||
| Seo et al. 201636 | • Failure to obtain reliable LS values, An invalid LS value; Delay between LB and TE>1 month; Starting antiviral therapy >1 month after LB; Presence/history of HCC at enrolment, HCC development within 6 month after enrolment; History of previous antiviral therapy; History of decompensated cirrhosis, Child Pugh class B/C cirrhosis at enrolment; Unsuitable quality of LB specimen for appropriate interpretation; Coinfection with hepatitis C/D or HIV; Right-sided heart failure; pregnancy | ||
| USA | Non-institutionalized civilian US population with CHB | Cohort | Possibility of liver disease subjects being classified as controls; possibility of underlying NAFLD or ALD in patients with normal liver enzymes; miss-classification of NAFLD cases as ALD due to relatively low threshold for definition of excessive alcohol consumption; histological data absent |
| 1988–2006 | • ≥ 17 y of age; complete demographic, social history, and clinical data available | ||
| Stepanova et al, 201040 | • NR | ||
| New Zealand | HBV patients with cirrhosis, who sought care at the hospital | Cohort | Retrospective design; possible observational biases and record errors; small size of the cohort; protective effect of metformin on hepatocarcinogenesis not assessed; anthropometric measurement not consistently performed in all patients; assessment for metabolic syndrome un-available |
| 2000–2010 | • NR | ||
| Hsiang et al. 201537 | • Co-infection with hepatotrophic viruses or HIV; cirrhosis complication in the year before DM diagnosis, type 1 DM patients | ||
| USA | Individuals who were KPNC health plan members in March 1996 and who were diagnosed with hepatitis B either before or after that date | Cohort | NR |
| 1996–2005 | • KNPC member between March 1995 and February 1996 | ||
| Szpakowski et al. 201341 | • Co-infection with HIV or HCV | ||
| Taiwan | Patients who underwent primary liver resection for HCC identified from the Cancer Registry Database of the hospital | Cohort | Small patient number in the dual viral infection |
| 1996–1999 | • Complete medical history, including serological markers for hepatitis B and C.; ≥ 6 months follow-up period | ||
| Huo et al. 200338 | • NR | ||
| Spain | Cohort: HBsAg positive subjects identified among all blood donors between 1972 and 1985 | Nested case-control | Liver morbidity and risk factor exposures not determined in all subjects as 55% of alive HBsAg positive subjects did not participate in the study; HCV and HIV not detected in subjects who died before 1985 |
| 1972–2000 | Nested case-control study: All HBsAg positive men of the cohort who died because of liver disease (cases) and HBsAg positive men without liver disease (controls). | • NR | |
| Ribes et al. 200639 | • HBV seromarkers negative subjects | ||
| China | Cases: DM patients ≥ 18 y of age who underwent cadaveric related liver transplantation for HBV-related liver disease | Case-control | NR |
| 2003–2007 | Control: Liver transplant recipients without DM matched for age, gender, primary liver disease and model for end-stage liver | • NR | |
| Ling et al. 201142 | • NR | ||
| USA | HBsAg positive patients | Cross-sectional | Retrospective design, lack of complete histologic data and HBV genotyping for CHB patients. |
| 2000–2006 | • ≥ 18 y of age; negative for anti-HCV and anti-HIV antibodies | ||
| Bondini et al. 200743 | • Alcohol intake > 20 g/day; drug use known to cause steatosis such as amiodarone, tamoxifen, methotrexate, tetracycline or corticosteroids | ||
| France, 2008–2013 | Patients with chronic hepatitis B infection in in Metropolitan France from January 2008 to December 2013 (largest Western cohort study on the prognosis of patients with CHB to date) | Nationwide, observational cohort study: | No information recorded about CHB treatment |
| Mallet et al. 201730 | • Adults discharged from hospital with a primary or associated diagnosis of CHB (ICD-10 B18.0 or B18.1 codes) | ||
| • Recipients of solid organ transplant or allogeneic stem cell transplant before January 2008 were excluded |
ALD: alcohol-related liver disease; BMI: body mass index; CHB: chronic hepatitis B; DM: diabetes mellitus; DNA; deoxyribonucleic acid; HbA1c: glycosylated hemoglobin; HBeAg: hepatitis B e antigen; HBsAg: hepatitis B surface antigen; HBV: hepatitis B virus; HCC: hepatocellular carcinoma; HCV: hepatitis C virus; HepB: hepatitis B; HIV: human immunodeficiency virus; ICD: International Classification of Diseases; KPNC: Kaiser Permanente Northern California; LC: liver cirrhosis; LB: liver biopsy; LS: Liver stiffness; TE: Transient elastography; NAFLD: Non-alcoholic fatty liver disease; NHIRD: National Health Insurance Research Database; NR: not reported; US: United States