Advertisement

Hepatocellular Carcinoma in the Absence of Cirrhosis in United States Veterans Is Associated With Nonalcoholic Fatty Liver Disease

  • Sahil Mittal
    Correspondence
    Reprint requests Address requests for reprints to: Sahil Mittal, MD, MS, Michael E. DeBakey VA Medical Center, 2002 Holcombe Boulevard (152), Houston, Texas 77030. fax: (713) 873-3505.
    Affiliations
    Center of Innovation, Effectiveness and Quality, Michael E. DeBakey Veterans Affairs Medical Center, Houston, Texas

    Baylor College of Medicine, Houston, Texas
    Search for articles by this author
  • Hashem B. El-Serag
    Affiliations
    Center of Innovation, Effectiveness and Quality, Michael E. DeBakey Veterans Affairs Medical Center, Houston, Texas

    Baylor College of Medicine, Houston, Texas

    Sections of Health Services Research, Section of Gastroenterology and Hepatology, Michael E. DeBakey Veterans Affairs Medical Center, Houston, Texas
    Search for articles by this author
  • Yvonne H. Sada
    Affiliations
    Center of Innovation, Effectiveness and Quality, Michael E. DeBakey Veterans Affairs Medical Center, Houston, Texas

    Baylor College of Medicine, Houston, Texas
    Search for articles by this author
  • Fasiha Kanwal
    Affiliations
    Center of Innovation, Effectiveness and Quality, Michael E. DeBakey Veterans Affairs Medical Center, Houston, Texas

    Baylor College of Medicine, Houston, Texas

    Sections of Health Services Research, Section of Gastroenterology and Hepatology, Michael E. DeBakey Veterans Affairs Medical Center, Houston, Texas
    Search for articles by this author
  • Zhigang Duan
    Affiliations
    Center of Innovation, Effectiveness and Quality, Michael E. DeBakey Veterans Affairs Medical Center, Houston, Texas

    Sections of Health Services Research, Section of Gastroenterology and Hepatology, Michael E. DeBakey Veterans Affairs Medical Center, Houston, Texas
    Search for articles by this author
  • Sarah Temple
    Affiliations
    Center of Innovation, Effectiveness and Quality, Michael E. DeBakey Veterans Affairs Medical Center, Houston, Texas

    Sections of Health Services Research, Section of Gastroenterology and Hepatology, Michael E. DeBakey Veterans Affairs Medical Center, Houston, Texas
    Search for articles by this author
  • Sarah B. May
    Affiliations
    Center of Innovation, Effectiveness and Quality, Michael E. DeBakey Veterans Affairs Medical Center, Houston, Texas

    Sections of Health Services Research, Section of Gastroenterology and Hepatology, Michael E. DeBakey Veterans Affairs Medical Center, Houston, Texas
    Search for articles by this author
  • Jennifer R. Kramer
    Affiliations
    Center of Innovation, Effectiveness and Quality, Michael E. DeBakey Veterans Affairs Medical Center, Houston, Texas

    Sections of Health Services Research, Section of Gastroenterology and Hepatology, Michael E. DeBakey Veterans Affairs Medical Center, Houston, Texas
    Search for articles by this author
  • Peter A. Richardson
    Affiliations
    Center of Innovation, Effectiveness and Quality, Michael E. DeBakey Veterans Affairs Medical Center, Houston, Texas

    Sections of Health Services Research, Section of Gastroenterology and Hepatology, Michael E. DeBakey Veterans Affairs Medical Center, Houston, Texas
    Search for articles by this author
  • Jessica A. Davila
    Affiliations
    Center of Innovation, Effectiveness and Quality, Michael E. DeBakey Veterans Affairs Medical Center, Houston, Texas

    Sections of Health Services Research, Section of Gastroenterology and Hepatology, Michael E. DeBakey Veterans Affairs Medical Center, Houston, Texas
    Search for articles by this author

      Background & Aims

      Hepatocellular carcinoma (HCC) can develop in individuals without cirrhosis. We investigated risk factors for development of HCC in the absence of cirrhosis in a U.S. population.

      Methods

      We identified a national cohort of 1500 patients with verified HCC during 2005 to 2010 in the U.S. Veterans Administration (VA) and reviewed their full VA medical records for evidence of cirrhosis and risk factors for HCC. Patients without cirrhosis were assigned to categories of level 1 evidence for no cirrhosis (very high probability) or level 2 evidence for no cirrhosis (high probability), which were based on findings from histologic analyses, laboratory test results, markers of fibrosis from noninvasive tests, and imaging features.

      Results

      A total of 43 of the 1500 patients with HCC (2.9%) had level 1 evidence for no cirrhosis, and 151 (10.1%) had level 2 evidence for no cirrhosis; the remaining 1203 patients (80.1%) had confirmed cirrhosis. Compared with patients with HCC in presence of cirrhosis, greater proportions of patients with HCC without evidence of cirrhosis had metabolic syndrome, nonalcoholic fatty liver disease (NAFLD), or no identifiable risk factors. Patients with HCC without evidence of cirrhosis were less likely to have abused alcohol or have hepatitis C virus infection than patients with cirrhosis. Patients with HCC and NAFLD (unadjusted odds ratio, 5.4; 95% confidence interval, 3.4–8.5) or metabolic syndrome (unadjusted odds ratio, 5.0; 95% confidence interval, 3.1–7.8) had more than 5-fold risk of having HCC in the absence of cirrhosis, compared with patients with HCV-related HCC.

      Conclusions

      Approximately 13% of patients with HCC in the VA system do not appear to have cirrhosis. NAFLD and metabolic syndrome are the main risk factors for HCC in the absence of cirrhosis.

      Keywords

      Abbreviations used in this paper:

      AFP (alpha-fetoprotein), APRI (aspartate aminotransferase to platelet ratio), BCLC (Barcelona Clinic Liver Cancer), CI (confidence interval), EMR (electronic medical record), HBV (hepatitis B virus), HCC (hepatocellular carcinoma), HCV (hepatitis C virus), ICD-9-CM (International Classification of Diseases, 9th Revision-Clinical Modification), MELD (Model for End-Stage Liver Disease), NAFLD (nonalcoholic fatty liver disease), OR (odds ratio), VA (Veterans Administration)
      Hepatocellular carcinoma (HCC) is the most common primary malignancy of the liver.
      • Mittal S.
      • El-Serag H.B.
      Epidemiology of hepatocellular carcinoma: consider the population.
      Cirrhosis is the precursor lesion for most HCC cases. Nevertheless, HCC is known to occur in the absence of cirrhosis. Among the major etiologic risk factors for HCC, hepatitis B virus (HBV) is known to be involved directly in liver mutagenesis. A recent study of patients treated with entecavir reported that up to 44.6% of HBV-related HCC developed in the absence of cirrhosis.
      • Wong G.L.
      • Chan H.L.
      • Chan H.Y.
      • et al.
      Accuracy of risk scores for patients with chronic hepatitis B receiving entecavir treatment.
      Outside the United States, a multicenter study conducted in Italy among all HCC patients diagnosed during a 1-year period reported that only 6.9% of patients developed HCC in the absence of cirrhosis.
      • Stroffolini T.
      • Andreone P.
      • Andriulli A.
      • et al.
      Characteristics of hepatocellular carcinoma in Italy.
      Although some studies have reported that hepatitis C virus (HCV) and alcohol can be associated with development of HCC in the absence of cirrhosis, the causal relationship between HCV, alcohol, and development of HCC in the absence of cirrhosis remains controversial.
      • Albeldawi M.
      • Soliman M.
      • Lopez R.
      • et al.
      Hepatitis C virus-associated primary hepatocellular carcinoma in non-cirrhotic patients.
      • Bralet M.P.
      • Regimbeau J.M.
      • Pineau P.
      • et al.
      Hepatocellular carcinoma occurring in nonfibrotic liver: epidemiologic and histopathologic analysis of 80 French cases.
      • Grazi G.L.
      • Cescon M.
      • Ravaioli M.
      • et al.
      Liver resection for hepatocellular carcinoma in cirrhotics and noncirrhotics: evaluation of clinicopathologic features and comparison of risk factors for long-term survival and tumour recurrence in a single centre.
      Recent literature has documented nonalcoholic fatty liver disease (NAFLD)-associated HCC in the absence of cirrhosis or advanced hepatic fibrosis.
      • Guzman G.
      • Brunt E.M.
      • Petrovic L.M.
      • et al.
      Does nonalcoholic fatty liver disease predispose patients to hepatocellular carcinoma in the absence of cirrhosis?.
      There is a lack of systematic studies examining the epidemiologic and etiologic risk profile of HCC in the absence of cirrhosis. The majority of published studies on HCC in the absence of cirrhosis included patients undergoing either resection or liver transplantation.
      • Bralet M.P.
      • Regimbeau J.M.
      • Pineau P.
      • et al.
      Hepatocellular carcinoma occurring in nonfibrotic liver: epidemiologic and histopathologic analysis of 80 French cases.
      • Lerut J.
      • Mergental H.
      • Kahn D.
      • et al.
      Place of liver transplantation in the treatment of hepatocellular carcinoma in the normal liver.
      • Paradis V.
      • Zalinski S.
      • Chelbi E.
      • et al.
      Hepatocellular carcinomas in patients with metabolic syndrome often develop without significant liver fibrosis: a pathological analysis.
      • Smoot R.L.
      • Nagorney D.M.
      • Chandan V.S.
      • et al.
      Resection of hepatocellular carcinoma in patients without cirrhosis.
      Furthermore, most studies reported data from tertiary referral centers
      • Yeh M.M.
      • Daniel H.D.
      • Torbenson M.
      Hepatitis C-associated hepatocellular carcinomas in non-cirrhotic livers.
      • Ertle J.
      • Dechene A.
      • Sowa J.P.
      • et al.
      Non-alcoholic fatty liver disease progresses to hepatocellular carcinoma in the absence of apparent cirrhosis.
      and may not be representative of the actual prevalence and risk factors of HCC in the absence of cirrhosis in the general population. Because most strategies for preventing HCC, including surveillance and chemoprevention, have targeted individuals with cirrhosis, it is important to obtain a better understanding about the development of HCC in the absence of cirrhosis because of its potential implications on current clinical paradigms.
      By using automated as well as medical record data obtained from the national Veterans Health Administration (VA) system, we conducted a retrospective cohort study among HCC patients diagnosed during 2005 to 2011. The aims of our study were (1) to estimate the frequency of HCC that developed in the absence of cirrhosis and (2) to examine and compare the etiologic risk factor distribution between HCC patients without cirrhosis with those with cirrhosis.

      Methods

       Study Population

      Data were obtained from VA administrative data files combined with review of patient electronic medical records (EMRs), and relevant data were abstracted by using structured data abstraction tools by trained medical record abstractors (S.T., S.M.) (Supplementary Material).

      VA information resource center. Available at: http://www.virec.reserach.va.gov/VSF/overview.htm. Accessed April 24, 2014.

      We identified a cohort of 10,695 patients who had a HCC diagnosis across VA hospitals between October 1, 2004 and September 30, 2011 (fiscal years 2005–2010). Patients with possible HCC were initially identified on the basis of the presence of International Classification of Diseases, 9th Revision-Clinical Modification (ICD-9-CM) code 155.0 (malignant neoplasm of liver) in the absence of code 155.1 (intrahepatic cholangiocarcinoma).
      • Davila J.A.
      • Weston A.
      • Smalley W.
      • et al.
      Utilization of screening for hepatocellular carcinoma in the United States.
      On the basis of a desired sample size of 1500, we selected a random computer-generated sample of patients for chart review to determine the study eligibility criteria. We included patients in the study if they had a diagnosis of HCC confirmed either by histopathology or imaging criteria according to the 2005 American Association for the Study of Liver Disease or European Association for the Study of Liver Disease guidelines.
      • Bruix J.
      • Sherman M.
      American Association for the Study of Liver Diseases. Management of hepatocellular carcinoma: an update.
      We reviewed charts of 2719 patients with possible HCC to arrive at 1500 study subjects. We excluded 830 patients because of insufficient evidence for HCC diagnosis. We further excluded 389 patients without recent VA healthcare utilization (defined as at least 1 inpatient or outpatient encounter at any VA facility within 1 year before the date of HCC diagnosis), cases with HCC recurrence and first diagnosis of HCC before study period, or those who received treatment before establishing guideline-based diagnosis. Thus, our final study cohort included 1500 patients with verified HCC.

       Categorization of Cirrhosis Status

      Patients with HCC were categorized into 3 mutually exclusive groups as (1) level 1 evidence of no cirrhosis (very high probability), (2) level 2 evidence of no cirrhosis (high probability), or (3) confirmed cirrhosis (Table 1).
      Table 1Study Definition for Classification of Cirrhosis Categories
      Cirrhosis categoryDefinition
      Level 1 evidence of no cirrhosis (very high probability), n = 43No evidence of cirrhosis on resection specimen or liver biopsy performed within 1 year before or at time of HCC diagnosis

      AND

      No features suggestive of cirrhosis on abdominal imaging available nearest to HCC diagnosis within 3 years before HCC diagnosis
      Level 2 evidence of no cirrhosis (high probability), n = 151APRI <1 based on laboratory results available nearest to HCC diagnosis within 6 months before and 4 weeks after HCC diagnosis

      AND

      No features suggestive of cirrhosis on abdominal imaging performed nearest to HCC diagnosis within 3 years before HCC diagnosis

      AND

      Two of 3 test values in normal range based on laboratory results available nearest to HCC diagnosis within 6 months before and 4 weeks after HCC diagnosis (albumin >3.5 g/L, platelets >200,000/μL, or international normalized ratio <1.1)
      Confirmed cirrhosis, n = 1201Documented cirrhosis on resection specimen or liver biopsy performed any time before or at time of HCC diagnosis

      OR

      Features suggestive of cirrhosis on abdominal imaging performed nearest to HCC diagnosis within 3 years before HCC diagnosis

      OR

      Documented presence of ascites, varices, or hepatic encephalopathy

      OR

      Abnormal values on 2 of 3 laboratory tests available nearest to HCC diagnosis within 6 months before and 4 weeks after HCC diagnosis (albumin <3.0 g/L, platelets <200,000 μL, international normalized ratio >1.1)
      Unclassified, n = 105Insufficient information to classify in any cirrhosis category

       Risk Factors for Hepatocellular Carcinoma

      HCV status was determined by the presence of positive anti-HCV or HCV RNA tests detected any time before or after HCC diagnosis. HBV was defined by a positive surface antigen detected any time before or after HCC diagnosis. Alcohol abuse was defined as history of more than 3 drinks a day, documentation of alcoholism/alcohol abuse in a physician’s progress notes, enrollment in a substance abuse treatment program, or history of alcoholic hepatitis. NAFLD was determined on the basis of documented evidence of hepatic steatosis on liver biopsy or, in the absence of liver biopsy, by the presence of metabolic syndrome in the absence of other causes of chronic liver disease (HCV, HBV, alcohol abuse, and no documentation of primary biliary cirrhosis, primary sclerosing cholangitis, autoimmune hepatitis, hemochromatosis, or Wilson disease) before HCC diagnosis. Metabolic syndrome was defined by using U.S. National Cholesterol Education Program Adult Treatment Panel III guidelines,
      Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults
      Executive summary of the third report of the national cholesterol education program (NCEP) expert panel on detection, evaluation, and treatment of high blood cholesterol in adults (adult treatment panel III).
      except for replacing elevated waist circumference with body mass index >28.8 kg/m2 in both men and women.
      • Ascha M.S.
      • Hanouneh I.A.
      • Lopez R.
      • et al.
      The incidence and risk factors of hepatocellular carcinoma in patients with nonalcoholic steatohepatitis.
      Less common causes of HCC such as hemochromatosis, Wilson disease, alpha1-antitrypsin deficiency, or autoimmune hepatitis were captured when diagnostic laboratory test results were positive (eg, homozygosity for C282Y) or diagnoses were listed in the problem list or progress notes. Patients having none of the above risk factors were classified as idiopathic HCC.

       Patient Characteristics

      Patient characteristics such as demographics, Model for End-Stage Liver Disease (MELD) score liver disease complications (ascites, encephalopathy, varices), performance status (Eastern Cooperative Oncology Group performance status 0–5), Barcelona Clinic Liver Cancer (BCLC) HCC stage at diagnosis (A–D), and presence of portal vein thrombosis, medical comorbidities, and mental health disorders were manually extracted from EMRs.

       Statistical Analysis

      Demographic features, HCC risk factors, clinical factors, and tumor characteristics were compared among the 3 groups of HCC patients: level 1 evidence of no cirrhosis (very high probability), level 2 evidence of no cirrhosis (high probability), and confirmed cirrhosis by using χ2 test for discrete variables and t test for continuous variables. Logistic regression models were used to identify variables independently associated with HCC in the absence of cirrhosis. Patients with level 1 and level 2 evidence of no cirrhosis (very high probability or high probability) were combined for these analyses. Patient and clinical factors that were significant at P value <.10 in the univariate analysis were retained in the final multivariable model, except for comorbidities that were excluded to avoid multicollinearity because they are part of the definition of NAFLD and metabolic syndrome. Odds ratios (ORs) and their accompanying 95% confidence intervals (CIs) were calculated. Analyses were performed by using SAS version 9.1 (SAS Institute, Cary, NC).

      Results

      The mean age of the study cohort at the time of HCC diagnosis was 63.7 years (standard deviation, 9.5), and the vast majority were men (99.8%). The greatest proportions of patients were white non-Hispanic (59.8%), followed by blacks (26.1%) and Hispanics (11.9%). As compared with race distribution in the whole VA population, a higher proportion of HCC patients in our cohort were either black or Hispanic, and lesser proportion were white. HCV testing was performed in 96.1%, HBV testing in 89.5%, and liver biopsy in 52.4% of the study cohort. Among patients in the cohort, 67.5% had HCV, 4.6% had HBV, 80.6% had alcoholic liver disease, 8% had NAFLD, 1.7% had other risk factors (hemochromatosis, alpha1-antitrypsin deficiency, or autoimmune hepatitis), and 2.6% were idiopathic with no identifiable risk factor.
      A total of 43 (2.9%) had level 1 evidence of no cirrhosis (very high probability), 151 (10.1%) had level 2 evidence of no cirrhosis (high probability), 1203 (80.1%) had confirmed cirrhosis, and in 105 (6.7%) there was insufficient information to classify these patients into any of the above categories. Among HCC patients with level 1 evidence of no cirrhosis, the hepatic fibrosis stage was 0 in 4 (9.3%), 1 in 13 (30.2%), 2 in 21 (48.9%), and 3 in 5 patients (11.6%) according to the Metavir scoring system.
      HCC risk factor distribution stratified by cirrhosis status is presented in Table 2. Among patients with NAFLD-related HCC, 34.6% had level 1 or level 2 evidence of no cirrhosis, and only 65.4% patients had confirmed cirrhosis at time of HCC diagnosis. Among HCC patients with metabolic syndrome irrespective of presence of other risk factors, 88.9% of them had confirmed cirrhosis, and only 19% had no evidence of cirrhosis. However, among patients with HCC and metabolic syndrome as the only risk factor (excluding HCV, HBV, and alcohol abuse), 32.7% had level 1 or level 2 evidence of no cirrhosis, and only 67.3% had confirmed cirrhosis at time of HCC diagnosis. On the other hand, among HCC patients with HCV, only 8.9% had level 1 or level 2 evidence of no cirrhosis, and most (91.1%) had confirmed cirrhosis at time of HCC diagnosis. Similarly, among HCC patients with HBV or alcohol abuse, 92.3% and 88.9%, respectively, had confirmed cirrhosis at time of HCC diagnosis.
      Table 2Risk Factor Distribution in HCC Cohort by Cirrhosis Category (N = 1395)
      Out of 1500 patients, 105 had insufficient information to classify in any cirrhosis category.
      Risk factor, n (%)Level 1 evidence of no cirrhosis, n = 43Level 2 evidence of no cirrhosis, n= 151Confirmed cirrhosis, n = 1201P value
      NAFLD6 (5.6)31 (28.9)70 (65.4)<.01
      HCV18 (1.9)67 (7.0)867 (91.1)<.01
      HBV2 (3.1)3 (4.6)60 (92.3).25
      Alcohol abuse29 (2.6)96 (8.5)1008 (88.9)<.01
      Metabolic syndrome14 (3.9)54 (15.1)289 (80.9)<.01
      Others
      Others: hemochromatosis, autoimmune hepatitis, alpha1-antitrypsin deficiency.
      2 (8.0)3 (12.0)20 (80.0).45
      Idiopathic
      None of the above.
      2 (5.9)11 (32.4)21 (61.8)<.01
      NOTE. Risk factor distribution is not mutually exclusive because many patients had more than 1 risk factor.
      a Out of 1500 patients, 105 had insufficient information to classify in any cirrhosis category.
      b Others: hemochromatosis, autoimmune hepatitis, alpha1-antitrypsin deficiency.
      c None of the above.
      HCC patients with level 1 or level 2 evidence of no cirrhosis were more likely to be older as compared with those with confirmed cirrhosis (Table 3). The mean MELD scores were significantly lower in HCC patients with level 1 (mean score, 8.8; standard deviation, 3.2) and level 2 evidence of no cirrhosis (mean score, 9.0; standard deviation, 3.6) compared with HCC patients with confirmed cirrhosis (mean score, 12.2; standard deviation, 4.8) (P < .01). No significant differences were observed by race. Patients with HCC in the presence of level 1 evidence of no cirrhosis were more likely to have alpha-fetoprotein (AFP) <20 ng/mL compared with HCC in presence of confirmed cirrhosis (Table 3), whereas the prevalence of portal vein thrombosis was significantly lower in HCC patients with very high (2.3%) or high (11.9%) probability of no cirrhosis as compared with HCC in presence of cirrhosis (18.5%) (P < .01). At the time of diagnosis, HCC patients with level 1 evidence of no cirrhosis were more likely to have BCLC stage B tumor (vs C or D) as compared with HCC with cirrhosis. Tumor differentiation was not significantly different by cirrhosis status.
      Table 3Comparison of Demographics, Comorbidities, and Tumor Characteristics Among HCC Patients According to Cirrhosis Status
      VariablesLevel 1 evidence of no cirrhosis, n = 43Level 2 evidence of no cirrhosis, n = 151Confirmed cirrhosis, n = 1201P value
      Mean age, y (standard deviation)65.5 (8.5)69.7 (10.7)62.6 (9.0)<.01
      Sex (%)
       Male43 (100)150 (99.3)1199 (99.8).44
      Race (%).05
       White23 (53.5)95 (62.9)711 (59.2)
       Black17 (39.5)44 (29.1)301 (25.1)
       Hispanic3 (6.9)9 (5.9)160 (13.3)
       Other03 (1.9)29 (2.4)
      Body mass index, kg/m2 (%).74
       <2510 (23.3)39 (25.8)261 (21.7)
       25–29.914 (32.6)56 (37.1)443 (36.9)
       30+19 (44.2)56 (37.1)497 (41.4)
      Mean MELD score (standard deviation)8.8 (3.2)9.0 (3.6)12.2 (4.8)<.01
      Medical comorbidities (%)
       Diabetes23 (53.5)68 (45.0)477 (39.7).10
       Hypertension38 (88.4)131 (86.8)872 (72.6)<.01
       Human immunodeficiency virus1 (2.3)7 (4.6)37 (3.1).56
       Myocardial infarction5 (11.6)28 (18.5)90 (7.5)<.01
       Peripheral vascular disease5 (11.6)30 (19.9)114 (9.5)<.01
      AFP <20 ng/mL (%)26 (60.5)53 (35.1)437 (36.4)<.01
      Portal vein thrombosis (%)1 (2.3)18 (11.9)222 (18.5)<.01
      BCLC stage (%)<.01
       A5 (11.6)6 (3.9)177 (14.7)
       B21 (48.8)42 (27.8)265 (22.1)
       C11 (25.6)72 (47.7)444 (36.9)
       D015 (9.9)231 (19.2)
       Missing6 (13.9)16 (10.6)84 (6.9)
      Tumor differentiation (%).63
       Well15 (37.5)27 (25)148 (26)
       Moderate9 (22.5)24 (22)122 (21)
       Poorly4 (10)8 (7)55 (10)
       Missing12 (30)50 (46)244 (43)

       Association Between Etiologic Risk Factors and Risk of Hepatocellular Carcinoma in the Absence of Cirrhosis

      We performed multiple logistic regression to examine the association of etiologic risk factors and risk of HCC in absence of cirrhosis. For this analysis we combined HCC with very high or high probability of no cirrhosis into single group in 1 level of the outcome variable (vs cirrhosis). Model 1 examined NAFLD as the main exposure variable, whereas model 2 examined metabolic syndrome (Table 4); we did not combine these 2 variables in 1 model because of overlap (metabolic syndrome was used to define NAFLD in the absence of liver biopsy). In the unadjusted analysis, HCC patients with NAFLD had more than 5-fold risk (OR, 5.4; 95% CI, 3.4–8.5) of having HCC in the absence of cirrhosis compared with patients with HCV-related HCC. Patients with HCC and underlying metabolic syndrome as the only risk factor for liver disease were also 5-fold times more likely (OR, 5.0; 95% CI, 3.1–7.8) to develop HCC in absence of cirrhosis as compared with patients with HCV-related HCC. Patients with HCC with alcohol abuse as the only risk factor had greater than 2-fold risk of having HCC in absence of cirrhosis as compared with HCC with underlying HCV. HCC patients with HBV infection did not have a higher risk of HCC in absence of cirrhosis as compared with HCV-related HCC. Adjusting for age, race, and BCLC stage, MELD score, AFP, and portal vein thrombus did not change the magnitude or direction of the association between etiologic risk factors and risk of HCC in absence of cirrhosis.
      Table 4Association Between Etiologic Risk Factors and Risk of HCC in Absence of Cirrhosis
      Model 1Model 2
      VariableUnadjusted OR (95% CI)Adjusted OR
      Adjusted for age, race, BCLC stage, MELD score, AFP, and portal vein thrombus.
      (95% CI)
      VariableUnadjusted OR (95% CI)Adjusted OR
      Adjusted for age, race, BCLC stage, MELD score, AFP, and portal vein thrombus.
      (95% CI)
      EtiologyEtiology
      HCV1.01.0HCV1.01.0
      NAFLD5.4 (3.4–8.5)3.9 (2.1–7.3)Metabolic syndrome5.0 (3.1–7.8)3.4 (1.9–6.4)
      Alcohol abuse2.6 (1.8–3.8)2.5 (1.6–3.9)Alcohol abuse2.6 (1.8–3.8)2.5 (1.5–3.9)
      HBV1.6 (0.4–7.1)1.3 (0.2–7.2)HBV1.6 (0.4–7.1)1.3 (0.2–7.2)
      Idiopathic4.7 (2.4–9.5)3.6 (1.6–8.3)Idiopathic6.0 (3.0–11.9)5.1 (2.2–11.7)
      NOTE. Results of logistic regression analysis examining 2 models (model 1 contains NAFLD, model 2 contains metabolic syndrome).
      a Adjusted for age, race, BCLC stage, MELD score, AFP, and portal vein thrombus.

      Discussion

      In a large national cohort of randomly identified U.S. veterans with HCC, we conducted a comprehensive medical record review to confirm diagnosis of cirrhosis and underlying risk factors for HCC and found that approximately 13% of HCC cases developed in patients without any evidence of cirrhosis. Compared with HCC in the presence of cirrhosis, these patients were more likely to have metabolic syndrome or NAFLD or no identifiable risk factor and less likely to have alcohol abuse or HCV infection. Among etiologic risk factors for HCC, patients with metabolic syndrome or NAFLD-related HCC had the highest risk, followed by alcohol-related HCC, of developing HCC in absence of cirrhosis compared with HCV-related HCC. HCC patients with HBV infection did not have a higher risk of HCC in absence of cirrhosis as compared with HCV-related HCC.
      The risk of HCC developing in the absence of cirrhosis varies according to etiology of the liver disease. A systemic review of studies published between 1998 and 2009 reported that among patients with HBV-related HCC who received antiviral treatment, 9.5% of patients had no evidence of cirrhosis.
      • Papatheodoridis G.V.
      • Lampertico P.
      • Manolakopoulos S.
      • et al.
      Incidence of hepatocellular carcinoma in chronic hepatitis B patients receiving nucleos(t)ide therapy: a systematic review.
      However, a study that examined long-term outcomes of chronic hepatitis B among white patients reported that HCC risk was mostly limited to patients with cirrhosis, findings similar to our study results.
      • Fattovich G.
      • Olivari N.
      • Pasino M.
      • et al.
      Long-term outcome of chronic hepatitis B in Caucasian patients: mortality after 25 years.
      In the Hepatitis C Antiviral Long Term Treatment against Cirrhosis study conducted in the United States, the annual risk of HCV-related HCC among non-cirrhotic patients was 0.8% compared with 2%–8% per year in cirrhotic patients.
      • Bruix J.
      • Sherman M.
      American Association for the Study of Liver Diseases. Management of hepatocellular carcinoma: an update.
      • Lok A.S.
      • Seeff L.B.
      • Morgan T.R.
      • et al.
      Incidence of hepatocellular carcinoma and associated risk factors in hepatitis C-related advanced liver disease.
      Among patients with alcohol abuse–related HCC, estimates of HCC without evidence of cirrhosis varies widely from 11.5 to 49%, depending on geographical location of the study and definition of alcohol abuse.
      • Trevisani F.
      • Frigerio M.
      • Santi V.
      • et al.
      Hepatocellular carcinoma in non-cirrhotic liver: a reappraisal.
      In our study, approximately 10% of those with alcohol abuse or HCV-related HCC and 7% of those with HBV-related HCC did not have evidence of cirrhosis at time of HCC diagnosis. In contrast, 35% of NAFLD-related HCC patients and 33.8% of idiopathic HCC patients did not have evidence of cirrhosis at the time of HCC diagnosis (Figure 1). However, it is possible that we have missed diagnosis of NAFLD among some patients with idiopathic HCC because of absence of liver biopsy or undiagnosed components of metabolic syndrome.
      Figure thumbnail gr1
      Figure 1Proportion of HCC patients with or without evidence of cirrhosis by risk factor.
      Studies from other parts of the world have also reported high proportion of NAFLD-related HCC occurring in absence of cirrhosis. In a study from Japan, only 51% of 87 cases of NAFLD-related HCC had cirrhosis at time of HCC diagnosis.
      • Yasui K.
      • Hashimoto E.
      • Komorizono Y.
      • et al.
      Characteristics of patients with nonalcoholic steatohepatitis who develop hepatocellular carcinoma.
      Similarly, among 31 HCC patients with metabolic syndrome as the only risk factor undergoing liver resection in a French hospital, only 35% had advanced hepatic fibrosis.
      • Paradis V.
      • Zalinski S.
      • Chelbi E.
      • et al.
      Hepatocellular carcinomas in patients with metabolic syndrome often develop without significant liver fibrosis: a pathological analysis.
      In a study from Germany, only 58% of 36 NAFLD-related HCC patients had cirrhosis at time of HCC diagnosis.
      • Ertle J.
      • Dechene A.
      • Sowa J.P.
      • et al.
      Non-alcoholic fatty liver disease progresses to hepatocellular carcinoma in the absence of apparent cirrhosis.
      Another study from United States on HCC patients undergoing curative treatment found that only about 73% of 52 NAFLD-HCC patients had underlying cirrhosis.
      • Reddy S.K.
      • Steel J.L.
      • Chen H.W.
      • et al.
      Outcomes of curative treatment for hepatocellular cancer in nonalcoholic steatohepatitis versus hepatitis C and alcoholic liver disease.
      A study that used health system database found only 46% of patients with NAFLD/nonalcoholic steatohepatitis and HCC had underlying cirrhosis.
      • Sanyal A.
      • Poklepovic A.
      • Moyneur E.
      • et al.
      Population-based risk factors and resource utilization for HCC: US perspective.
      However, in this study, authors relied only on nonvalidated ICD codes to capture diagnosis of NAFLD. In our study we performed manual chart review to confirm diagnosis of HCC, presence or absence of underlying cirrhosis, and risk factor. Our study findings extend the findings of these previous studies and highlight that the sequence of events from steatohepatitis to cirrhosis and finally HCC may not be linear in a substantial proportion of patients with NAFLD-related or metabolic syndrome–related HCC. It seems that the increased risk of HCC in these patients is due to both progression to cirrhosis and oncogenic potential of NAFLD or metabolic syndrome per se.
      We used findings on liver biopsy performed within 1 year of HCC diagnosis to evaluate for diagnosis of cirrhosis. Liver biopsy is prone to sampling error and can miss cirrhosis.
      • Rockey D.C.
      • Caldwell S.H.
      • Goodman Z.D.
      • et al.
      Liver biopsy.
      This is especially true for cases with stage 3 hepatic fibrosis.
      • Ratziu V.
      • Charlotte F.
      • Heurtier A.
      • et al.
      Sampling variability of liver biopsy in nonalcoholic fatty liver disease.
      However, among HCC patients with no evidence of cirrhosis on liver biopsy in our study, only a small fraction (<12%) had stage 3 fibrosis. Moreover, we labeled HCC patients as having no evidence of cirrhosis only when biopsy evidence was supplemented by absence of features of cirrhosis on abdominal imaging. Furthermore, we used biopsy findings within 1 year of HCC diagnosis, thus minimizing the possibility of fibrosis progression in the time interval between biopsy and HCC diagnosis. In absence of liver biopsy, we relied on noninvasive marker aspartate aminotransferase to platelet ratio (APRI) to rule out cirrhosis. A recent meta-analysis on APRI with cutoff of 1.0 had summary receiver operating characteristic curve of 0.83 with negative predictive value of 69% in excluding cirrhosis.
      • Lin Z.H.
      • Xin Y.N.
      • Dong Q.J.
      • et al.
      Performance of the aspartate aminotransferase-to-platelet ratio index for the staging of hepatitis C-related fibrosis: an updated meta-analysis.
      A diagnostic tool with summary receiver operating characteristic curve of 100% is considered perfect, and anything above 80% is considered good. To further reduce the chances of misclassification of cirrhosis status in the absence of liver biopsy, we supplemented APRI score with no evidence of cirrhosis on imaging and absence of any laboratory parameters suggestive of cirrhosis. Thus, a patient had to fulfill all 3 criteria to qualify for level 2 evidence of no cirrhosis. HCC patients with no evidence of cirrhosis had significantly lower MELD score and lower prevalence of portal vein thrombosis, thus providing internal validity of our findings. We observed that HCC patients without evidence of cirrhosis were older as compared with patients with HCC in presence of cirrhosis. This may be due to the higher prevalence of NAFLD among non-cirrhotic HCC patients, whereas cirrhotic HCC patients had a higher proportion of HCV-related or alcohol abuse–related HCC.
      Our findings have important implications for HCC surveillance practices and paradigms in patients with NAFLD. Current guidelines do not recommend surveillance in patients with NAFLD who do not have evidence of cirrhosis.
      • Bruix J.
      • Sherman M.
      American Association for the Study of Liver Diseases. Management of hepatocellular carcinoma: an update.
      Studies have shown that HCC screening is under-utilized among cirrhotic patients who are at the highest risk of developing HCC and most likely to benefit from HCC surveillance.
      • Davila J.A.
      • Henderson L.
      • Kramer J.R.
      • et al.
      Utilization of surveillance for hepatocellular carcinoma among hepatitis C virus-infected veterans in the United States.
      It will be logistically impractical to expand the risk pool by including the large NAFLD population. Although evidence suggests that a substantial proportion of NAFLD patients can develop HCC in absence of cirrhosis, the absolute risk of HCC in non-cirrhotic patients is not known, and the strategies and benefits of screening for HCC in NAFLD patients without cirrhosis have not been examined. Chemoprevention may also be a feasible strategy if an intervention has low toxicity and high efficacy. Because of the epidemiologic association of diabetes mellitus and obesity with HCC, it seems reasonable to seek and treat concomitant metabolic conditions in patients with nonalcoholic steatohepatitis to reduce the risk of HCC. There is evidence to suggest that metformin reduces the risk of HCC among diabetics.
      • Lai S.W.
      • Chen P.C.
      • Liao K.F.
      • et al.
      Risk of hepatocellular carcinoma in diabetic patients and risk reduction associated with anti-diabetic therapy: a population-based cohort study.
      • Hassan M.M.
      • Curley S.A.
      • Li D.
      • et al.
      Association of diabetes duration and diabetes treatment with the risk of hepatocellular carcinoma.
      Studies of these and other risk factors of HCC among NAFLD patients with and without cirrhosis are needed.
      Our study has several limitations. The study population was predominantly male, which may limit generalizability of results. Large liver specimen obtained during liver transplantation or hepatic resection for HCC treatment provides for more accurate staging of fibrosis and severity of underlying liver disease. However, limiting the study to such patients would have introduced considerable selection bias by overestimating the number of patients with preserved liver function and consequently less advanced liver disease and defeated our primary objective to study the prevalence of HCC in absence of cirrhosis and its risk factors among the general population. We used all available VA EMRs to identify risk factors including a search of all progress notes for any evidence of alcohol abuse before HCC diagnosis. However, there is still chance of occult alcohol use and therefore misclassification of risk category between alcohol use and NAFLD. It is also plausible that HCC that we attributed to NAFLD because of the presence of metabolic syndrome in the absence of other risk factors is due to other currently unknown genetic or metabolic causes. However, this definition of NAFLD has been successfully used in several NAFLD studies.
      • Ascha M.S.
      • Hanouneh I.A.
      • Lopez R.
      • et al.
      The incidence and risk factors of hepatocellular carcinoma in patients with nonalcoholic steatohepatitis.
      • Marrero J.A.
      • Fontana R.J.
      • Su G.L.
      • et al.
      NAFLD may be a common underlying liver disease in patients with hepatocellular carcinoma in the United States.
      In conclusion, we found that up to 13.0% of HCC patients had no evidence of cirrhosis at the time of HCC diagnosis. The main risk factors for this entity were NAFLD or metabolic syndrome (vs HCV, HBV, or alcohol abuse). Future research is needed to identify actionable risk factors and/or biomarkers to predict NAFLD patients at higher risk of developing HCC.

      Supplementary Material

       Data Sources

      Administrative data included the Medical SAS Outpatient and Inpatient files and the VA Vital Status File. The Medical SAS files contain patient demographic data as well as diagnoses according to ICD-9-CM and procedures according to Common Procedural Terminology codes. We determined date of death, if any, in the Vital Status File that uses an algorithm to select the most accurate date of death by using the VA Medical SAS Inpatient file, Beneficiary Identification & Records Locator System Death File, Medicare Vital Status file, and Social Security Administration death file.

      VA information resource center. Available at: http://www.virec.reserach.va.gov/VSF/overview.htm. Accessed April 24, 2014.

      Patient EMR information was obtained by accessing the Compensation and Pension Records Interchange, which is a VA application that provides access to the EMRs found in the Computerized Patient Record System at any VA facility nationwide.

      References

        • Mittal S.
        • El-Serag H.B.
        Epidemiology of hepatocellular carcinoma: consider the population.
        J Clin Gastroenterol. 2013; 47: S2-S6
        • Wong G.L.
        • Chan H.L.
        • Chan H.Y.
        • et al.
        Accuracy of risk scores for patients with chronic hepatitis B receiving entecavir treatment.
        Gastroenterology. 2013; 144: 933-944
        • Stroffolini T.
        • Andreone P.
        • Andriulli A.
        • et al.
        Characteristics of hepatocellular carcinoma in Italy.
        J Hepatol. 1998; 29: 944-952
        • Albeldawi M.
        • Soliman M.
        • Lopez R.
        • et al.
        Hepatitis C virus-associated primary hepatocellular carcinoma in non-cirrhotic patients.
        Dig Dis Sci. 2012; 57: 3265-3270
        • Bralet M.P.
        • Regimbeau J.M.
        • Pineau P.
        • et al.
        Hepatocellular carcinoma occurring in nonfibrotic liver: epidemiologic and histopathologic analysis of 80 French cases.
        Hepatology. 2000; 32: 200-204
        • Grazi G.L.
        • Cescon M.
        • Ravaioli M.
        • et al.
        Liver resection for hepatocellular carcinoma in cirrhotics and noncirrhotics: evaluation of clinicopathologic features and comparison of risk factors for long-term survival and tumour recurrence in a single centre.
        Aliment Pharmacol Ther. 2003; 17: 119-129
        • Guzman G.
        • Brunt E.M.
        • Petrovic L.M.
        • et al.
        Does nonalcoholic fatty liver disease predispose patients to hepatocellular carcinoma in the absence of cirrhosis?.
        Arch Pathol Lab Med. 2008; 132: 1761-1766
        • Lerut J.
        • Mergental H.
        • Kahn D.
        • et al.
        Place of liver transplantation in the treatment of hepatocellular carcinoma in the normal liver.
        Liver Transpl. 2011; 17: S90-S97
        • Paradis V.
        • Zalinski S.
        • Chelbi E.
        • et al.
        Hepatocellular carcinomas in patients with metabolic syndrome often develop without significant liver fibrosis: a pathological analysis.
        Hepatology. 2009; 49: 851-859
        • Smoot R.L.
        • Nagorney D.M.
        • Chandan V.S.
        • et al.
        Resection of hepatocellular carcinoma in patients without cirrhosis.
        Br J Surg. 2011; 98: 697-703
        • Yeh M.M.
        • Daniel H.D.
        • Torbenson M.
        Hepatitis C-associated hepatocellular carcinomas in non-cirrhotic livers.
        Mod Pathol. 2010; 23: 276-283
        • Ertle J.
        • Dechene A.
        • Sowa J.P.
        • et al.
        Non-alcoholic fatty liver disease progresses to hepatocellular carcinoma in the absence of apparent cirrhosis.
        Int J Cancer. 2011; 128: 2436-2443
      1. VA information resource center. Available at: http://www.virec.reserach.va.gov/VSF/overview.htm. Accessed April 24, 2014.

        • Davila J.A.
        • Weston A.
        • Smalley W.
        • et al.
        Utilization of screening for hepatocellular carcinoma in the United States.
        J Clin Gastroenterol. 2007; 41: 777-782
        • Bruix J.
        • Sherman M.
        American Association for the Study of Liver Diseases. Management of hepatocellular carcinoma: an update.
        Hepatology. 2011; 53: 1020-1022
        • Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults
        Executive summary of the third report of the national cholesterol education program (NCEP) expert panel on detection, evaluation, and treatment of high blood cholesterol in adults (adult treatment panel III).
        JAMA. 2001; 285: 2486-2497
        • Ascha M.S.
        • Hanouneh I.A.
        • Lopez R.
        • et al.
        The incidence and risk factors of hepatocellular carcinoma in patients with nonalcoholic steatohepatitis.
        Hepatology. 2010; 51: 1972-1978
        • Papatheodoridis G.V.
        • Lampertico P.
        • Manolakopoulos S.
        • et al.
        Incidence of hepatocellular carcinoma in chronic hepatitis B patients receiving nucleos(t)ide therapy: a systematic review.
        J Hepatol. 2010; 53: 348-356
        • Fattovich G.
        • Olivari N.
        • Pasino M.
        • et al.
        Long-term outcome of chronic hepatitis B in Caucasian patients: mortality after 25 years.
        Gut. 2008; 57: 84-90
        • Lok A.S.
        • Seeff L.B.
        • Morgan T.R.
        • et al.
        Incidence of hepatocellular carcinoma and associated risk factors in hepatitis C-related advanced liver disease.
        Gastroenterology. 2009; 136: 138-148
        • Trevisani F.
        • Frigerio M.
        • Santi V.
        • et al.
        Hepatocellular carcinoma in non-cirrhotic liver: a reappraisal.
        Dig Liver Dis. 2010; 42: 341-347
        • Yasui K.
        • Hashimoto E.
        • Komorizono Y.
        • et al.
        Characteristics of patients with nonalcoholic steatohepatitis who develop hepatocellular carcinoma.
        Clin Gastroenterol Hepatol. 2011; 9: 428-433
        • Reddy S.K.
        • Steel J.L.
        • Chen H.W.
        • et al.
        Outcomes of curative treatment for hepatocellular cancer in nonalcoholic steatohepatitis versus hepatitis C and alcoholic liver disease.
        Hepatology. 2012; 55: 1809-1819
        • Sanyal A.
        • Poklepovic A.
        • Moyneur E.
        • et al.
        Population-based risk factors and resource utilization for HCC: US perspective.
        Curr Med Res Opin. 2010; 26: 2183-2191
        • Rockey D.C.
        • Caldwell S.H.
        • Goodman Z.D.
        • et al.
        Liver biopsy.
        Hepatology. 2009; 49: 1017-1044
        • Ratziu V.
        • Charlotte F.
        • Heurtier A.
        • et al.
        Sampling variability of liver biopsy in nonalcoholic fatty liver disease.
        Gastroenterology. 2005; 128: 1898-1906
        • Lin Z.H.
        • Xin Y.N.
        • Dong Q.J.
        • et al.
        Performance of the aspartate aminotransferase-to-platelet ratio index for the staging of hepatitis C-related fibrosis: an updated meta-analysis.
        Hepatology. 2011; 53: 726-736
        • Davila J.A.
        • Henderson L.
        • Kramer J.R.
        • et al.
        Utilization of surveillance for hepatocellular carcinoma among hepatitis C virus-infected veterans in the United States.
        Ann Intern Med. 2011; 154: 85-93
        • Lai S.W.
        • Chen P.C.
        • Liao K.F.
        • et al.
        Risk of hepatocellular carcinoma in diabetic patients and risk reduction associated with anti-diabetic therapy: a population-based cohort study.
        Am J Gastroenterol. 2012; 107: 46-52
        • Hassan M.M.
        • Curley S.A.
        • Li D.
        • et al.
        Association of diabetes duration and diabetes treatment with the risk of hepatocellular carcinoma.
        Cancer. 2010; 116: 1938-1946
        • Marrero J.A.
        • Fontana R.J.
        • Su G.L.
        • et al.
        NAFLD may be a common underlying liver disease in patients with hepatocellular carcinoma in the United States.
        Hepatology. 2002; 36: 1349-1354