Tumor Incidence in Patients with Non-Alcoholic Fatty Liver Disease
; ; ; ; ; ;
Background: The incidence of cancer is increasing worldwide. The role of comorbidities in this development is debated. The aim of this study was to investigate the significance of non-alcoholic fatty liver disease (NAFLD) for the incidence of cancer of various kinds in Germany.
Methods: Between 2000 and 2015, data on 31 587 patients with established NAFLD were collected for analysis. A control group (n = 31 587) assembled for comparison was matched for sex, age, treating physician, and Charlson Comorbidity Index (CCI).
Results: By 10 years after the index date, 15.3% of patients with NAFLD and 13.4% of patients in the control group had been diagnosed with cancer (p <0.001). Patients with NAFLD exhibited significantly higher rates of male genital cancers (HR 1.26; 95% confidence interval [1.06; 1.5]; p = 0.008), skin cancer (HR 1.22 [1.07; 1.38]; p = 0.002) and breast cancer (HR 1.2 [1.01; 1.43]; p = 0.036). In this analysis, the rate of hepatocellular carcinoma did not differ between patients with NAFLD and patients without NAFLD (0.19% vs. 0.12%; p = 0.204).
Conclusion: NAFLD slightly increases the risk of breast cancer in women, genital cancer in men, and skin cancer irrespective of sex. Thus, NAFLD can be considered a marker of increased cancer risk.
Non-alcoholic fatty liver disease (NAFLD) is one of the major causes of chronic liver disease, with an estimated prevalence of 24% in the general population (1, 2). Metabolically burdened patients with obesity and type 2 diabetes mellitus, in particular, are at increased risk. Based on the growing prevalence of these risk factors, increasing incidence of NAFLD in the USA and the European countries France, Germany, Italy, Spain and the UK has been predicted (3, 4). The spectrum of NAFLD ranges from simple fatty liver (hepatic steatosis [NAFL]) over inflammatory non-alcoholic steatohepatitis (NASH) all the way to liver cirrhosis with associated complications and hepatocellular carcinoma. NASH is the most rapidly growing indication for liver transplantation in the USA (5). The most common cause of death in NAFLD is cardiovascular disease followed by extrahepatic malignancy (6, 7, 8). Patients with advanced NAFLD (histological fibrosis stage F3 and F4) have a sevenfold increased risk of developing hepatocellular carcinoma compared with people without chronic liver disease (9). Moreover, NAFLD is the second leading for liver transplantation in patients with hepatocellular carcinoma in the USA (10).
Worldwide, both the incidence and the mortality of cancer are growing rapidly (11). For 2018, 18.1 million new cancer cases and 9.6 million cancer deaths have been estimated (11). In Germany, nearly 480 000 incident cases of cancer occurred in 2014, causing 22% of all deaths in women and 28% in men (12). Based on demographic trends, the number of new cancer cases in Germany is expected to increase by more than 20% between 2010 and 2030 (12).
In patients with NAFLD, metabolic inflammation could be a relevant factor for increased risk of developing cancer (13). Little is known about the specific types of cancer in patients with NAFLD.
The aim of this study was to investigate the risk and incidence of the most common cancer types in NAFLD. To this end, we conducted a comparative analysis of data from a database of outpatient care in Germany that contains information on 7.49 million patients collected over a 15-year period (14).
Materials and methods
The study data are based on the Disease Analyzer database (IQVIA) (14). A data set containing 7.49 million patients over a period of 15 years was explored (Figure 1). For a 1-year period, this corresponds to 1.5–2 million patients, corresponding to approximately 3% of all outpatient (primary and specialist) medical practices in Germany. More detailed information about the database can be found in the eMethods.
This retrospective cohort study included adult patients (≥ 18 years) with newly diagnosed NAFLD (ICD-10: K75.8, K76.0) in 1262 outpatient medical practices in Germany between January 2000 and December 2015 (index date; Figure 1). Only cases with an observation time of at least 12 months prior to the index date were included. Patients with a cancer diagnosis (ICD-10: C00–C99) prior to index date were excluded. Deviations from this diagnostic pattern are accounted for by matching for the medical practices. NAFLD patients were matched to non-NAFLD patients by age, sex, physician, index year, and Charlson Comorbidity Index (CCI). Each NAFLD patient was paired with a non-NAFLD patient based on age, sex, CCI, and treatment in the same primary care practice in the same year. CCI was based on the diagnoses documented within 12 months prior to or on the index date.
The CCI is a weighted index that accounts for the number and severity of comorbidities and includes a wide range of comorbidities including macrovascular diseases, pulmonary diseases, gastrointestinal, and renal diseases, diabetes, tumors, and HIV infection (15). For the controls, the index date was that of a randomly selected visit between January 2000 and December 2015 (Figure 1).
Study outcomes and covariates
The main outcome of the study was the incidence of cancer (ICD 10: C00–C99) in total and cancer of different organs including liver (ICD 10: C22), lip, oral cavity and pharynx (ICD 10: C00–C14), digestive organs excluding liver (ICD 10: C15–C26 excl. C22), respiratory organs (ICD 10: C30–C39), skin (ICD 10: C43, C44), breast (ICD 10: C50), female genital organs (ICD 10: C51–C58), male genital organs (ICD 10: C60–C63), urinary tract (ICD 10: C64–C68), and lymphoid and hematopoietic tissue (ICD 10: C81–C96) as a function of NAFLD.
The Disease Analyzer uses anonymized electronic medical records. Patient data was analyzed as aggregates with no individual health data available. Individual informed consent or IRB approval was not obtained.
Differences in the sample characteristics between those with and those without NAFLD were tested using chi-squared tests for categorical variables and Wilcoxon tests for continuous variables. Cox regression analyses were conducted to study the association between the NAFLD and cancer incidence. These models were performed separately for different cancers and adjusted for the presence of hypertension (ICD-10: I10), dyslipidemia (ICD-10: E78), diabetes mellitus (ICD-10: E10–E14), and obesity (ICD-10: E66). Because regression models were calculated for cancer in total as well as for 10 different cancer types, a Bonferroni correction for p-value was performed, and a p-value of <0.004 (calculated as < 0.05 / 11) was considered statistically significant. The analyses were carried out using SAS version 9.4.
Basic characteristics of the study sample
This study included 31 587 patients with NAFLD and 31 587 patients without NAFLD. The baseline characteristics of the study patients are shown in Table 1. Their mean age was 58.0 (± 14.0) years; 48.0% were women. The mean CCI was 0.9 (± 1.0) in both cohorts with no significant difference. Compared with controls, higher proportions of individuals with NAFLD had obesity (5.0% vs. 2.4%), diabetes mellitus (11.4% vs. 6.9%), dyslipidemia (19.8% vs. 12.8%), and hypertension (26.8% vs. 18.4%) (Table 1).
Association of NAFLD and incidence of cancer
Within 10 years of the index date, cancer was diagnosed in 15.3% of patients with NAFLD and 13.4% of patients without NAFLD (log-rank p <0.001) (Figure 2). In total, 2431 incident cancers (1597 in NAFLD and 834 in controls) were recorded. The three most common types of cancer in NAFLD and controls were skin cancer, cancer of the digestive organs (excluding hepatocellular carcinoma), and cancer of lymphoid and hematopoietic tissue. On regression analyses NAFLD was significantly associated with the incidence of cancer (hazard ratio [HR] 1.15, 95% confidence interval [1.08; 1.21]; p <0.001). The highest risk in NAFLD was observed for male genital cancer (HR 1.26 [1.06; 1.5]; p = 0.008), followed by skin (HR 1.22 [1.07; 1.38]; p = 0.002) and breast cancer (HR 1.20 [1.01; 1.43]; p = 0.036). The association with skin cancer was independent of gender, and women exhibited a higher HR (HR 1.42 [1.03; 1.95]; p = 0.0300) than men [HR 1.35 [1.06; 1.73]; p = 0.0146); the difference was, however, not statistically significant. The incidence of liver cancer was coded at 0.19% in NAFLD and 0.12% in non-NAFLD patients with no statistically significant difference between the two groups (p = 0.204) (Table 2).
This is the first analysis to report the incidence of cancer in NAFLD in a well-matched cohort using a large dataset derived from outpatient medical practices in Germany. We found that the risk of developing cancer was 15% higher after the diagnosis of NAFLD than in a matched control cohort. This is in agreement with data from a large longitudinal cohort study in the USA in which 4722 NAFLD patients and 14 441 age- and sex-matched control persons were followed up for 8 years. In that study, NAFLD was associated with a nearly twofold increase in the risk of developing cancer, particularly cancers of the gastrointestinal tract, liver, and uterus (6). In our NAFLD patients, we observed increased incidence of male genital cancer (including penis, prostate, and testis), skin cancer including malignant cutaneous melanoma, and breast cancer. In a recent analysis of an Italian population-based cohort of 23 358 individuals with diabetes and 383 799 without diabetes, taking no account of comorbidities, the overall cancer incidence rate in patients with diabetes mellitus type 2 was 1.22 [1.15; 1.29], significantly higher than in those without diabetes (16). A meta-analysis on the relation between diabetes and cancer risk included 121 cohort studies and reported a pooled adjusted relative risk for all cancer entities of 1.27 [1.21; 1.32] in women and 1.19 [(1.13; 1.25] in men (17). Interestingly, women with diabetes exhibited an approximately 6% greater risk than men with diabetes (17). Our study did not specifically match for diabetes. However, after controlling for the CCI we found that skin cancer, breast cancer, and genital cancer were independently associated with NAFLD.
One striking finding of the current analysis is the lack of an increased risk of hepatic malignancy. In absolute numbers, 24 patients in the NAFLD cohort versus 11 patients in the non-NAFLD cohort developed liver cancer (Table 2). This corresponds to a cumulative incidence of 0.19% vs. 0.12% using the Kaplan–Meier method (eFigure 1), with no significant differences between the two groups.
In a Korean analysis that included patients with non-cirrhotic NAFLD, a 15 times higher incidence of hepatocellular carcinoma, a twofold incidence of colorectal cancer in men, and a 1.9 times higher incidence of breast cancer in women were observed (18). While the risk of developing breast cancer in our study is comparable to the Korean study, the findings for hepatocellular carcinoma and colorectal cancer are not repeated in the present analysis. This may be related to differences in the study population and to the stringent matching for comorbidities in our study. Moreover, differences in the intensity of screening for liver cancer are likely: while in the Korean study patients were recruited in a tertiary-level hospital, our cohort was recruited from outpatient care data.
Worldwide, the incidence of hepatocellular carcinoma is 0.44 per 1000 person-years in NAFLD patients (19) and 5.29 per 1,000 person-years in NASH (1); significantly lower than that of chronic hepatitis B (1). The stage of liver disease affects the incidence of both hepatocellular carcinoma and cholangiocarcinoma. Compared to persons without liver disease, NAFLD patients with advanced disease—defined as histological fibrosis stages F3 or F4—have a 7 times higher risk of developing hepatocellular carcinoma (9). In the present study cohort, however, no data were available on the stage of fibrosis.
Another difference from previously published data is that our study did not replicate the described increase in the incidence of colorectal cancer (20). Again, this may relate to the greater degree of matching for comorbidities in our study. Published data give the annual age-standardized incidence rate of advanced colorectal cancer per 100 000 in Germany in 2014 as 21.5 for men and 14.9 for women (21). Additionally, no data on lifestyle are recorded in the Disease Analyzer database, so the exposure to carcinogenic factors, e.g., tobacco use, is unknown. As respiratory cancers did not vary between the two groups, it can be speculated that no essential difference with regard to tobacco use existed.
NAFLD is frequently diagnosed in patients with diabetes, and there is a large overlap between these two populations. We matched for the comorbidities that are captured in the CCI and thus were able to specifically address the contribution of NAFLD.
Coding for obesity in the Disease Analyzer database was lower than the prevalence of 23 to 35% stated in the literature (22). This may be explained by the lack of a reimbursement-relevant code in outpatient care and the failure to define obesity as a stand-alone disease (23). Nonetheless, visceral obesity is a well recognized risk factor in the development of various cancers, and a wealth of data is available on the association with NAFLD and cancers (20). A recent study found that NAFLD is associated with a higher risk of cancer while obesity by itself is not The authors concluded that NAFLD is a critical factor in obesity-associated cancer (6).
Our analysis has limitations attributable to the nature of a database of diagnoses based on ICD-10 codes. Misclassification related to incorrect coding or incomplete coding may occur. Because NAFLD is an asymptomatic disease, there is a risk that the diagnosis will be missed and thus not coded. Furthermore, a clear distinction between predominantly alcoholic versus non-alcoholic liver disease is difficult to make and overlap can exist. Therefore, the outpatient cohort explored here is likely to differ from cohorts in which disease stage and activity have been defined using liver biopsy (24). Nonetheless, the data from the present analysis show that the NAFLD population exhibits a slightly increased risk of cancer. Moreover, the Disease Analyzer database does not capture detailed laboratory data. Information regarding disease severity, and especially the stage of fibrosis, is therefore lacking. Consequently, we could not assess a potential association between advanced liver disease and the incidence of cancer.
While the retrospective, observational nature of the analysis introduces a potential bias, prospective registries are needed to validate the current findings. Lastly, the Disease Analyzer database, recently validated, is an outpatient database (14). This may lead to systematic underestimation of patients whose cancer is at a late stage and who do not leave the hospital after diagnosis. On the other hand, this is the largest population that has been analyzed in relation to the incidence of cancer risk associated with NAFLD in Germany and thus offers useful guidance in determining the best practice patterns for patients suffering from NAFLD (e.g., urgently advising NAFLD patients to participate in cancer screening programs). The diagnosis of cancer is prominent in the patient history and thus the coding is presumably precise.
In summary, NAFLD contributes to a small increase in the risk of breast cancer in women, genital cancer in men, and skin cancer irrespective of sex in a large outpatient care population in Germany after controlling for comorbidities.
Conflict of interest statement
PD Wörns has received consultancy payments from Abbvie, BMS, Bayer, Ipsen, Roche, and Eisai; reimbursement of congress attendance fees or training course costs from Abbvie, BMS, Bayer, Gilead Sciences, and Ipsen; and payments for preparing training courses from Abbvie, BMS, Bayer, Gilead Sciences, Ipsen, and MSD.
Prof. Galle has received consultancy payments from Bayer, BMS, MSD, Lilly, Roche, Astra Zeneca, Sirtex, and Ipsen, as well as honoraria for conducting clinical studies from Intercept, Gilead, Novartis, Galmed, Madrigal, Celgene, Enanta, NGMbio, Allergan, AbbVie, BMS, Dr. Falk Pharma, Sanofi, Pilantes, Genfi, Inventiva, Novo Nordisk, Jansen Easthorn, Cyamabay, Tobira Therapeutics, Böhringer Ingelheim, Roche, and Takeda.
Prof. Schattenberg has received consultancy payments from BMS, Echosens, Genfit, Gilead Sciences, Intercept Pharmaceuticals, Madrigal, Novartis, Pfizer, and Roche, as well as reimbursement of congress attendance fees and travel and accommodation costs from. Gilead Sciences.
Dr. Labenz has received publication honoraria from Norgine and Nordmark; reimbursement of congress attendance fees and travel and accommodation costs from. Merz; and research grants from Merz and Norgine.
Prof. Kostev is employed by IQVIA.
The remaining authors declare that no conflict of interest exists.
Manuscript received on 18 December 2019, revised version accepted on 27 May 2020
Prof. Dr. med. Jörn M. Schattenberg
Schwerpunkt Metabolische Lebererkrankungen
I. Medizinische Klinik, Universitätsmedizin Mainz
55131 Mainz, Germany
Cite this as:
Huber Y, Labenz C, Michel M, Wörns M-A, Galle PR, Kostev K, Schattenberg JM: Tumor incidence in patients with non-alcoholic fatty liver disease. Dtsch Arztebl Int 2020; 117: 719–24.
eMethods, eFigure, eTable:
PD Dr. med. Marcus-A. Wörns, Prof. Dr. med. Peter R. Galle
Metabolic Liver Diseases, Department of Medicine I, University Medical Center Mainz: Dr. med. Yvonne Huber, Dr. med. Christian Labenz, Maurice Michel, Prof. Dr. med. Jörn M. Schattenberg
Epidemiology, IQVIA, Frankfurt: Prof. Dr. rer. med. Karel Kostev
|1.||Younossi ZM, Koenig AB, Abdelatif D, Fazel Y, Henry L, Wymer M: Global epidemiology of nonalcoholic fatty liver disease-meta-analytic assessment of prevalence, incidence, and outcomes. Hepatology 2016; 64: 73–84 CrossRef MEDLINE|
|2.||Younossi Z, Tacke F, Arrese M, et al.: Global perspectives on nonalcoholic fatty liver disease and nonalcoholic steatohepatitis. Hepatology 2019; 69: 2672–82 CrossRef MEDLINE|
|3.||Estes C, Anstee QM, Arias-Loste MT, et al.: Modeling NAFLD disease burden in China, France, Germany, Italy, Japan, Spain, United Kingdom, and United States for the period 2016–2030. J Hepatol 2018; 69: 896–904 CrossRef MEDLINE|
|4.||Estes C, Razavi H, Loomba R, Younossi Z, Sanyal AJ: Modeling the epidemic of nonalcoholic fatty liver disease demonstrates an exponential increase in burden of disease. Hepatology 2018; 67: 123–33 CrossRefMEDLINE PubMed Central|
|5.||Siddique O, Joseph-Talreja M, Yoo ER, et al.: Rising rate of liver transplantation in the baby boomer generation with non-alcoholic steatohepatitis in the United States. J Clin Transl Hepatol 2017; 5: 193–6 CrossRef MEDLINE PubMed Central|
|6.||Allen AM, Hicks SB, Mara KC, Larson JJ, Therneau TM: The risk of incident extrahepatic cancers is higher in non-alcoholic fatty liver disease than obesity – a longitudinal cohort study. J Hepatol 2019; 71: 1229–36 CrossRef MEDLINE|
|7.||Chalasani N, Younossi Z, Lavine JE, et al.: The diagnosis and management of nonalcoholic fatty liver disease: practice guidance from the American Association for the Study of Liver Diseases. Hepatology 2018; 67: 328–57 CrossRef MEDLINE|
|8.||Labenz C, Huber Y, Michel M, et al.: Impact of NAFLD on the incidence of cardiovascular diseases in a primary care population in Germany. Dig Dis Sci 2019; 10.1007/s10620–019–05986–9 CrossRef MEDLINE|
|9.||Younossi ZM: Non-alcoholic fatty liver disease – a global public health perspective. J Hepatol 2019; 70: 531–44 CrossRef MEDLINE|
|10.||Khan FZ, Perumpail RB, Wong RJ, Ahmed A: Advances in hepatocellular carcinoma: nonalcoholic steatohepatitis-related hepatocellular carcinoma. World J Hepatol 2015; 7: 2155–61 CrossRef MEDLINE PubMed Central|
|11.||Bray F, Ferlay J, Soerjomataram I, Siegel RL, Torre LA, Jemal A: Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin 2018; 68: 394–424 CrossRef MEDLINE|
|12.||Robert Koch Institute and the association of population-based cancer registries in Germany. Cancer in Germany 2013/2014. Berlin: RKI 2018 .|
|13.||Gehrke N, Schattenberg JM: Metabolic inflammation – a role for hepatic inflammatory pathways as driver of comorbidities in non-alcoholic fatty liver disease? Gastroenterology 2020; 158: 1929–47 CrossRef MEDLINE|
|14.||Rathmann W, Bongaerts B, Carius HJ, Kruppert S, Kostev K: Basic characteristics and representativeness of the German Disease Analyzer database. Int J Clin Pharmacol Ther 2018; 56: 459–66 CrossRef MEDLINE|
|15.||Quan H, Sundararajan V, Halfon P, et al.: Coding algorithms for defining comorbidities in ICD-9-CM and ICD-10 administrative data. Med Care 2005; 43: 1130–9 CrossRef MEDLINE|
|16.||Ballotari P, Vicentini M, Manicardi V, et al.: Diabetes and risk of cancer incidence: results from a population-based cohort study in northern Italy. BMC Cancer 2017; 17: 703 CrossRef MEDLINE PubMed Central|
|17.||Ohkuma T, Peters SAE, Woodward M: Sex differences in the association between diabetes and cancer: a systematic review and meta-analysis of 121 cohorts including 20 million individuals and one million events. Diabetologia 2018; 61: 2140–54 CrossRef MEDLINE PubMed Central|
|18.||Kim G-A, Lee HC, Choe J, et al.: Association between non-alcoholic fatty liver disease and cancer incidence rate. J Hepatol 2018; 68: 140–6 CrossRef MEDLINE|
|19.||Younossi Z, Stepanova M, Ong JP, et al.: Nonalcoholic Steatohepatitis Is the Fastest Growing Cause of Hepatocellular Carcinoma in Liver Transplant Candidates. Clin Gastroenterol Hepatol 2019; 17: 748–55 CrossRef MEDLINE|
|20.||Sanna C, Rosso C, Marietti M, Bugianesi E: Non-alcoholic fatty liver disease and extra-hepatic cancers. Int J Mol Sci 2016; 17: 717 CrossRef MEDLINE PubMed Central|
|21.||Oppelt KA, Luttmann S, Kraywinkel K, Haug U: Incidence of advanced colorectal cancer in Germany: comparing claims data and cancer registry data. BMC Med Res Methodol 2019; 19: 142 CrossRef MEDLINE PubMed Central|
|22.||Mensink G, Schienkiewitz A, Haftenberger M, Lampert T, Ziese T, Scheidt-Nave C: Overweight and obesity in Germany (DEGS1). Bundesgesundheitsblatt Gesundheitsforschung Gesundheitsschutz 2013; 56: 786–94 CrossRef MEDLINE|
|23.||Hebebrand J, Dabrock P, Lingenfelder M, Mand E, Rief W, Voit W: Ist Adipositas eine Krankheit? Interdisziplinäre Perspektiven. Dtsch Arztebl 2004; 101: A 2468–74 VOLLTEXT|
|24.||Labenz C, Huber Y, Kalliga E, et al.: Predictors of advanced fibrosis in non-cirrhotic non-alcoholic fatty liver disease in Germany. Aliment Pharmacol Ther 2018; 48: 1109–16 CrossRef MEDLINE|