If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Requests for reprints should be addressed to Allison J. Kasmari, MD, Department of Internal Medicine, Penn State Hershey Medical Center, College of Medicine, 500 University Drive, PO Box 850, Mail Code H039, Hershey, PA 17033.
Hepatocellular carcinoma is the most common primary liver malignancy, commonly a sequelae of hepatitis C infection, but can complicate cirrhosis of any cause. Whether metabolic syndrome and its components, type II diabetes, hypertension, and hyperlipidemia increase the risk of hepatocellular carcinoma independent of cirrhosis is unknown.
A retrospective cohort study was conducted using the MarketScan insurance claims database from 2008-2012. Individuals with hepatocellular carcinoma aged 19-64 years and age and sex-matched controls were included. Multivariate analysis of hepatocellular carcinoma risk factors was performed.
Hepatitis C (odds ratio [OR] 2.102) was the largest risk factor for hepatocellular carcinoma. Other independent risk factors were type II diabetes (OR 1.353) and hypertension (OR 1.229). Hyperlipidemia was protective against hepatocellular carcinoma (OR 0.885). The largest risk increase occurred with hypertension with type II diabetes and hepatitis C (OR 4.580), although hypertension and type II diabetes without hepatitis C still incurred additional risk (OR 3.399). Type II diabetes and hyperlipidemia had a similar risk if hepatitis C was present (OR 2.319) or not (OR 2.395). Metformin (OR 0.706) and cholesterol medications (OR 0.645) were protective in diabetics. Insulin (OR 1.640) increased the risk of hepatocellular carcinoma compared with the general type II diabetes population.
In the absence of cirrhosis, type II diabetes and hypertension were independent risk factors for hepatocellular carcinoma. Hyperlipidemia and medical management of type II diabetes with metformin and cholesterol medication appeared to reduce the incidence of hepatocellular carcinoma. In contrast, insulin was associated with a higher risk of hepatocellular carcinoma.
Another study from the Department of Veterans Affairs (VA) showed that diabetes served as a risk factor only in the setting of other well-established risk factors for hepatocellular carcinoma (ie, alcoholic cirrhosis or hepatitis B or C).
Metabolic syndrome includes a combination of type II diabetes, obesity, hypertension, and hyperlipidemia. The suggestion of type II diabetes as a potential risk factor for hepatocellular carcinoma raises the question of whether other metabolic syndrome components, for example, hypertension, hyperlipidemia, and obesity, may also increase hepatocellular carcinoma risk.
Recently, hepatocellular carcinoma has been associated with components of metabolic syndrome in the setting of cirrhosis. However, with the magnitude of hepatocellular carcinoma and its projected increase in the US, meaningful prevention and screening may need to begin prior to the detection of cirrhosis. For this reason, we set out to determine if type II diabetes, metabolic syndrome, or its components are risk factors for hepatocellular carcinoma independent of cirrhosis.
Materials and Methods
The MarketScan Commercial Claims and Encounters Database (Truven Health Analytics, Bethesda, Md) captures clinical utilization, expenditures, and prescription drug claims data from a selection of large employers and health plans, including commercial insurance companies, Blue Cross and Blue Shield plans, and third-party administrators. The database includes data from approximately 100 payers, and represents insured employees and their dependents, early retirees, Consolidated Omnibus Budget Reconciliation Act (COBRA) individuals, and Medicare-eligible retirees with employer-provided Medicare Supplemental plans. Data from 2008 to 2012 were used, which include information on over 56 million covered lives annually.
The study population included adults >18 years of age represented in the MarketScan Database between 2008 and 2012 who had an outpatient visit with a primary or secondary diagnosis code of hepatocellular carcinoma (International Classification of Diseases, Ninth Revision [ICD-9] code: 155.0). In order to minimize the error that can occur with insurance coding, the diagnosis had to be present on 2 separate occasions to be included. The total number of patients with hepatocellular carcinoma was 17,446. Patients with concomitant diagnoses of hepatitis B (ICD-9: 070.2, 070.22, 070.3, 070.32), alcoholic liver damage (ICD-9: 571.0, 571.1, 571.3), hereditary hemochromatosis (ICD-9: 275.0), nonalcoholic fatty liver disease (ICD-9: 571.8), nonalcoholic steatohepatitis (ICD-9: 571.8), cirrhosis (ICD-9: 571.2, 571.5, 571.6), alpha-1 antitrypsin deficiency (ICD-9: 273.4), autoimmune hepatitis (ICD-9: 571.42), and Wilson disease (ICD-9: 275.1) were excluded. Because hepatitis C is the largest known risk factor for hepatocellular carcinoma in the US, these patients were included in the analysis to evaluate the contribution of hepatitis C with and without type II diabetes or metabolic syndrome. There were 7473 patients remaining after exclusions were applied. Controls were age and sex matched in a 1:3 fashion, creating a control group of 22,110 individuals that complied with the same exclusion criteria. The Figure describes how the study sample was constructed. The mean age was 57.7 years (SD ± 9.08 years), with a range of 19-64 years (Table 1). There was an equal distribution of males and females between cases and controls; confirming age and sex match at a rate of 1:3 (99%) was effective. Cases were categorized based on geographic region of health care consumption. Most cases were from the South at 39.72%, with the Northeast, North-central, and West regions being nearly equal in distribution (19.4%, 20.62%, and 19.19%). The geographic region was unknown in 1.06% of cases.
Table 1Sociodemographic Data and Study Group Description
A univariate analysis was performed with type II diabetes (ICD-9: 250.xx), hypertension (ICD-9: 401.0, 401.1, 401.9), and hyperlipidemia (ICD-9: 272.4), as well as with hepatitis C (ICD-9: 070.54, 070.70, 070.71). The first 3 variables were chosen as they represent significant components of the metabolic syndrome. Hepatitis C was analyzed as a variable, as it currently represents the largest known risk factor for hepatocellular carcinoma. A conditional multivariable logistic regression analysis was then performed to determine the association of type II diabetes hypertension, hyperlipidemia, and hepatitis C with the development of hepatocellular carcinoma. In addition, medications used to treat diabetes and hyperlipidemia were included in the primary multivariable analysis. Diabetes medications included: insulin, metformin, thiazolidinediones, and sulfonylureas (glipizide, glyburide, glimepiride). They were identified in the analysis as being present or not. There were no data included on length of use or whether there were multiple medications used in the same patient. Hyperlipidemia medications included: lovastatin, pravastatin, simvastatin, fluvastatin, atorvastatin, rosuvastatin, pitavastatin, gemfibrozil, fenofibrate, nicotinic acid, cholestyramine, colestipol, colesevelam, and ezetimibe. These medications were consolidated into one category, “cholesterol medications,” for analysis and were designated as being present or not. The medication analysis was constructed as described based on prior studies' identification of medication group impact on metabolic syndrome and hepatocellular carcinoma risk rather than individual medications.
In an effort to evaluate metabolic syndrome and type II diabetes independent impact on hepatocellular carcinoma risk, a sub-analysis was performed, which separately evaluated patients with and without hepatitis C. Cases and controls were divided into 2 groups: type II diabetes with hepatitis C and type II diabetes without hepatitis C. A conditional multivariable logistic regression analysis was performed with hypertension and hyperlipidemia as covariants to create groups that met criteria for metabolic syndrome. Obesity is an important factor in metabolic syndrome; however, it is seldom coded as a diagnosis on insurance claims. Therefore, the metabolic syndrome sub-analysis was used as a surrogate for the lack of obesity data. Attempts to evaluate the diagnosis of metabolic syndrome by diagnosis code (ICD-9: 277.7) were met with similar difficulty to that of obesity, as it is seldom coded on insurance claims.
This study was funded by the Department of Gastroenterology and Hepatology at Penn State Hershey Medical Center. There were no external funding sources. The study was approved by the Institutional Review Board at Penn State Hershey Medical Center.
A univariate analysis was performed to evaluate risk factors for hepatocellular carcinoma individually (Table 2). There were 2451 patients with hepatocellular carcinoma who had type II diabetes, 4764 had hypertension, 2931 had hyperlipidemia, and 2303 had hepatitis C. Type II diabetes, hypertension, and hepatitis C all had a P value of <.001, indicating a significant association with hepatocellular carcinoma. Interestingly, when analyzed in a univariate fashion, hyperlipidemia was found not to have a significant association with hepatocellular carcinoma (P value = .6421).
Table 2Univariate Analysis of Risk Factors for Hepatocellular Carcinoma
Analysis of patients with hepatocellular carcinoma in a multivariable analysis demonstrates that patients with type II diabetes are 1.353 times more likely to develop hepatocellular carcinoma than those without type II diabetes (95% confidence interval [CI], 1.249-1.465) (Table 3). Hepatitis C represented the largest individual risk factor, with a 2.102-increased risk of developing hepatocellular carcinoma (95% CI, 1.964-2.250). Hypertension was associated with an increased risk of 1.229 (95% CI, 1.155-1.308). Conversely, hyperlipidemia was associated with a 12% hepatocellular carcinoma risk reduction (odds ratio [OR] 0.885; 95% CI, 0.831-0.942).
Our data suggests increased risk of hepatocellular carcinoma (OR 1.640; 95% CI, 1.482-1.814) in diabetic patients using insulin. Metformin was associated with a 33% lower risk of hepatocellular carcinoma (OR 0.706; 95% CI, 0.632-0.788) (Table 3). There was no significant change in risk associated with sulfonylureas. Cholesterol medications were associated with a lower risk than metformin, with 44% hepatocellular carcinoma risk reduction (OR 0.645; 95% CI, 0.601-0.692).
To evaluate metabolic syndrome, we looked at hypertension and hyperlipidemia in the setting of type II diabetes with and without hepatitis C. Hypertension had a significant impact on hepatocellular carcinoma risk, compounding the established risk of hepatitis C and type II diabetes (Table 4). Hypertension patients with both type II diabetes and hepatitis C were associated with a 4.580-increased risk of hepatocellular carcinoma (95% CI, 3.705-5.662), whereas those with hypertension and type II diabetes without hepatitis C had a 3.399 risk (95% CI, 3.11-3.708). Patients with hyperlipidemia and type II diabetes had a similar risk, whether hepatitis C was present or not. Hyperlipidemia with type II diabetes and hepatitis C carried a 2.319 risk (95% CI, 1.978-2.718), while those with hyperlipidemia and type II diabetes without hepatitis C had a 2.395 risk (95% CI, 2.215-2.590).
Table 4Secondary Conditional Multivariable Sub-analysis for HCC
DM with HCV
DM without HCV
CI = confidence interval; DM = diabetes mellitus; HCC = hepatocellular carcinoma; HCV = hepatitis C virus; HLP = hyperlipidemia; HTN = hypertension; OR = odds ratio.
Hepatitis C cirrhosis carries a 39.89-fold (95% CI, 36.29-43.84) increased risk of hepatocellular carcinoma. Hepatocellular carcinoma occurs in these individuals secondary to hepatic inflammation and fibrosis, as well as malignant transformation of hepatitis C-infected cells.
There are limited data evaluating the risk of hepatocellular carcinoma associated with type II diabetes and metabolic syndrome independent of cirrhosis. Type II diabetes or obesity, or both, had a 2.47-fold (95% CI, 2.34-2.61) increased risk of hepatocellular carcinoma in the setting of cirrhosis according to a SEER-Medicare database study conducted in 1991 to 2007.
Further, a recent VA study showed that 13% of patients with hepatocellular carcinoma had no evidence of cirrhosis. Of those without cirrhosis that had hepatocellular carcinoma, a higher proportion had metabolic syndrome.
These studies are less generalizable to the US population due to male predominance and inclusion of other causes of cirrhosis in the first mentioned study. Another VA prospective study identified a 1.5 hazard ratio [HR] (95% CI, 0.9-2.5) for hepatocellular carcinoma with type II diabetes, which was not statistically significant.
A meta-analysis of studies that have looked at type II diabetes and hepatocellular carcinoma showed a summary relative risk of 1.56 (95% CI, 1.30-1.87) for type II diabetes associated with hepatocellular carcinoma.
In our study, the risk of hepatocellular carcinoma in type II diabetes was 1.353 (95% CI, 1.249-1.465). Unlike the VA study, we had equal sex and age distribution. Our study differed in design from those previously reported because we controlled for all causes of cirrhosis except hepatitis C. Our study population, however, did not include subjects over the age of 64 years, whereas the VA had a mean age of 62 years and included older subjects.
Hypertension and its association with hepatocellular carcinoma risk has not been well studied. However, hypertension has been associated with an increased hepatocellular carcinoma mortality.
Our study shows an independent risk association between hypertension and hepatocellular carcinoma (OR 1.229; 95% CI, 1.155-1.308). One Medicare database study showed a 2.22-adjusted OR for hepatocellular carcinoma risk with hypertension (95% CI, 2.04-2.42).
Unlike our study, causes of cirrhosis were not excluded and the population was older. Hepatocellular carcinoma occurring with metabolic syndrome has been thought to occur at an older age than hepatocellular carcinoma from other causes, which could explain the differences in findings.
Consistent with these data, we demonstrated that hyperlipidemia was associated with decreased hepatocellular carcinoma risk without cirrhosis (OR 0.885; 95% CI, 0.831-0.942). In the Medicare population, the adjusted OR of hepatocellular carcinoma risk with hyperlipidemia in the setting of cirrhosis was 1.35 (95% CI, 1.26-2.45).
Metabolic syndrome promotes hepatocellular carcinoma development in multiple ways. It creates insulin resistance that leads to an increase in insulin-like growth factor-1, the most powerful activator of cellular proliferation.
study using a Medicare claims database demonstrated that metabolic syndrome occurred more commonly in patients with hepatocellular carcinoma (37.1%) compared with those without hepatocellular carcinoma (17.1%), and metabolic syndrome was associated with increased risk of hepatocellular carcinoma (OR 2.13; 95% CI, 1.96-2.31). Although this population is older than the one analyzed in our study, a relationship still exists between hepatocellular carcinoma and metabolic syndrome.
There is limited data about hepatocellular carcinoma associated with metabolic syndrome in noncirrhotic patients. A case series by Perumpail et al
In this cohort of 44 patients, 6 patients developed hepatocellular carcinoma with either nonalcoholic steatohepatitis, nonalcoholic fatty liver disease, or with 2 components of metabolic syndrome and no underlying liver disease. All 6 patients had at least 2 components of metabolic syndrome; all 6 had hypertension, and type II diabetes or insulin resistance, and 5 had hyperlipidemia. All of these patients had both hypertension and type II diabetes, which is the combination in our analysis that represented the largest risk (OR 3.399; 95% CI, 3.11-3.708).
Our study suggests that diabetes, hypertension, and metabolic syndrome are associated with hepatocellular carcinoma independent of cirrhosis. In order to combat these risks, it is important to understand how the management of these conditions influences hepatocellular carcinoma risk. A Taiwanese study showed a protective effect with metformin (HR 0.49; 95% CI, 0.37-0.84); likewise, metformin had an OR of 0.706 (95% CI, 0.632-0.788) in our study.
In our analysis, sulfonylureas and thiazolidinediones were not significant. Interestingly, we found insulin to be associated with an increased risk of hepatocellular carcinoma. This may serve as a marker of disease severity rather than representing an association with the medication itself, as patients with more severe diabetes are typically treated with insulin. The differences seen between the Taiwanese study and our results may be caused by differences in underlying population risk factors. In addition, the Taiwanese study was able to analyze the duration of therapy and its effect on risk, which was not available in our analysis.
Medications, such as statins, also seem to be associated with lower hepatocellular carcinoma risk in prior studies,
A population-based case-control Taiwanese study found that statin use was associated with reduced risk of hepatocellular carcinoma (OR 0.53; 95% CI, 0.41-0.69), which is similar to the reduced risk suggested in our analysis with the use of any cholesterol medication, including statins (OR 0.645, 95% CI, 0.601-0.692).
This study found that statins were protective against hepatocellular carcinoma, demonstrating greater protection occurring in individuals that had been treated with statins for longer periods of time. Our study design did not include time course for medications, but simply that the medication had been used, requiring further analysis to evaluate a dose–response relationship with statins.
Metformin has antitumoral properties through the activation of the 5′adenosine monophosphate-activated protein kinase pathway. This inhibits mTOR, downregulates c-Myc,
Based on several reported mechanistic and epidemiologic studies, metformin and statins have been advocated as chemoprotective agents for hepatocellular carcinoma. Larger randomized clinical trials need to be performed to help this become standard practice.
Several limitations of this study must be acknowledged. The patient population is relatively young, with ages ranging from 19-64 years. The MarketScan Database includes claims from private insurers and does not include the Medicare population. Insurance claims data may under-represent conditions when compared with other objective research methods. Importantly, in this study there is the chance that cirrhosis is not identified by insurance claim. A recent study revealed that only 54% of patients with biopsy-proven cirrhosis actually had an ICD-9 code for cirrhosis in claims data.
Additionally, conditions such as nonalcoholic fatty liver disease may be present in these patients but not coded for as a diagnosis. The time period analyzed during our study is relatively short (2008-2012), making it difficult to determine how the duration of risk-factor exposure influences the development of hepatocellular carcinoma. Finally, a critical component of metabolic syndrome is obesity. As the purpose of diagnosis codes is for billing, obesity is not commonly documented as a diagnosis on insurance claims. Therefore, obesity was not able to be included as a variable. More prospective studies, including obesity as a variable, are necessary to fully investigate this critical component of metabolic syndrome. While associations between diabetes and metabolic syndrome with hepatocellular carcinoma were seen in this study, further analysis is required to determine the true association in the absence of cirrhosis, as insurance claims data cannot guarantee that no undiagnosed cirrhosis or nonalcoholic fatty liver disease was included in the study population.
Success in treatment of hepatocellular carcinoma is dependent on early detection. Therefore, understanding the risk factors is important for adequate treatment. Type II diabetes and hypertension appear to be associated with hepatocellular carcinoma, in the absence of cirrhosis. The medical management of diabetes and hyperlipidemia, particularly with metformin and cholesterol medications, appears to be associated with reduced risk of hepatocellular carcinoma. On the contrary, insulin use was associated with increased risk of hepatocellular carcinoma in diabetic patients without cirrhosis. These findings may translate into targets for screening and therapies to improve survival in patients with hepatocellular carcinoma.
The study was approved by the Institutional Review Board at Penn State Hershey Medical Center.
Epidemiology of hepatocellular carcinoma in the United States: where are we? Where do we go?.