Hemochromatosis, Iron Overload–Related Diseases, and Pancreatic Cancer Risk in the Surveillance, Epidemiology, and End Results (SEER)-Medicare

Although pancreatic cancer (pancreatic cancer) only accounts for 3% of all incident cancers, it is highly fatal with a 5-year survival rate of 10% and ranks third for cancer mortality in the United States (1). Experimental studies of iron overload suggest that iron accumulation in pancreatic islets impairs insulin secretion and β-cell function and accelerates pancreatic β-cell death (2). Higher serum iron has also been found to be associated with pancreatic cancer, although not consistently (3–5). Previous epidemiologic studies have identified associations of red meat consumption and heme iron with type 2 diabetes mellitus and pancreatic cancer risk (6). Hemochromatosis is a disease characterized by iron overload, and patients with sideroblastic anemia, congenital dyserythropoietic anemia, and chronic liver disease (CLD) are prone to iron overload (7). We investigated the association of hemochromatosis, sideroblastic and congenital dyserythropoietic anemias, and non–alcoholic-related CLD (NACLD) with pancreatic cancer in a large U.S. population-based case–control study in older people. We hypothesize that iron overload–related diseases are associated with increased pancreatic cancer risk.

We conducted a population-based, nested case–control study within the Surveillance Epidemiology, and Ends Results Program (SEER)-Medicare with details described elsewhere (8). Pancreatic cancer cases were defined as individuals with primary malignant disease [first and later (occurring after another cancer diagnosis other than pancreatic cancer) primaries; International Classification of Disease for Oncology, ICD-O codes C25.0-C25.9, malignant code 3], identified from the SEER-Medicare Patient Entitlement Diagnosis Summary File (PEDSF). Cases were diagnosed at ages 66 to 99 years, between 1992 and 2015. Individuals diagnosed with cancer by death certificate and at autopsy were excluded. To avoid surveillance bias (i.e., diagnosis of iron overload diseases led to increased surveillance for cancer), we excluded the 12-month period prior to case diagnosis from the ascertainment of iron-overload diseases. pancreatic cancer cases and controls were eligible they were if they had at least one Medicare claim (MEDPAR, NCH, or OUTPATIENT) more than 12 months prior to diagnosis.

A total of 320,296 population-based controls (4:1 ratio for each case) were randomly selected from a 5% subcohort of Medicare-enrolled beneficiaries. Controls were selected from the Summarized Denominator (SUMDENOM) file that contains demographic and claims data from a 5% random sample of Medicare recipients residing in SEER areas who never developed cancer and from a 5% sample of recipients in the PEDSF who developed cancer. Pancreatic cancer cases and matched controls were required to have a minimum of 13 months of Medicare Part A and B, health maintenance organization coverage prior to diagnosis/selection, making the minimum age at diagnosis 66 years. Controls were alive and either free of cancer or pancreatic cancer (for first primary or later primaries, respectively) as of July 1 of the calendar year that the case was diagnosed and frequency matched to the pancreatic cancer cases by age (66–69, 70–74, 75–79, 80–84, 85–89, 90–94, 95–99), sex, and calendar year of diagnosis.

We used the ICD-9 codes from the Medicare claim files to identify hemochromatosis (275.0, 275.01, and 275.03), sideroblastic (285.0), and congenital dyserythropoietic (285.8) anemias and NACLD (571.4, 571.41, 571.42, 571.49, 571.5, 571.6, 571.8, and 571.9). An individual was classified as having a specific condition if there were at least 1 inpatient claim or 2 physician/outpatient claims at least 30 days apart. We required the conditions to be present at least 12 months prior to case diagnosis or control selection to avoid surveillance bias and differential assessment of exposure status.

Demographic characteristics for cases and controls were compared using t tests for continuous variables and χ² tests for categorical values. We used unconditional logistic regression models to calculate ORs and 95% confidence intervals (CI) for the association of iron overload–related diseases with first, later, and all primary pancreatic cancer, adjusting for the a priori–selected covariates that are known to be associated with pancreatic cancer listed in Table 1 and for SEER registries.

In total, 80,074 (61,081 first and 18,993 later primaries) incident pancreatic cancer cases were identified. Compared with controls, cases were more often non-Hispanic black and overweight/obese, and had chronic obstructive pulmonary disease, smoking, alcohol, pancreatitis, and type 2 diabetes–related diagnoses (Table 1). Hemochromatosis or iron-overload anemias were not associated with pancreatic cancer (Table 2). We observed a significant 30% increased risk for later primary pancreatic cancer among patients diagnosed with sideroblastic anemia. NACLD was significantly associated a 10%–17% elevated risk for first, later, and all pancreatic cancer (Table 2).

Overall, we did not observe associations between diagnosed hemochromatosis, iron-overload anemias, and pancreatic cancer in this large, nested case-control study within the SEER-Medicare population. We observed significant elevated pancreatic cancer risks for NACLD, particularly for nonalcoholic fatty liver disease (NAFLD) and cirrhosis (NAC). As adiposity and diabetes are risk factors for NAFLD, NAC, and pancreatic cancer, residual confounding is likely present because we use claims data for adjustment. We speculate that the higher later primary pancreatic cancer risk among participants with sideroblastic anemia might reflect effects of therapy for the earlier cancer(s). Limitations of our study include the low proportion of participants with hemochromatosis or iron-overload anemias diagnoses, incomplete ascertainment of the conditions, and the older age of our study population. Our results partly support the hypothesis that diagnosed iron-overload diseases are associated with pancreatic cancer in older people.

M.J. Barrett reports other support from NCI during the conduct of the study. No disclosures were reported by the other authors.

S. Julián-Serrano: Data curation, writing–original draft, writing–review and editing. F. Yuan: Writing–review and editing. M.J. Barrett: Data curation, formal analysis, methodology, writing–review and editing. R.M. Pfeiffer: Software, formal analysis, supervision, methodology, writing–review and editing. R.Z. Stolzenberg-Solomon: Conceptualization, resources, data curation, supervision, funding acquisition, investigation, methodology, writing–original draft, project administration, writing–review and editing.

This work was supported by the Intramural Research Program of the NCI, NIH. The collection of cancer incidence data used in this study was supported by the California Department of Public Health as part of the statewide cancer reporting program mandated by California Health and Safety Code Section 103885; the National Cancer Institute's Surveillance, Epidemiology and End Results Program under contract HHSN261201000140C awarded to the Cancer Prevention Institute of California, contract HHSN261201000035C awarded to the University of Southern California, and contract HHSN261201000034C awarded to the Public Health Institute; and the Centers for Disease Control and Prevention's National Program of Cancer Registries, under agreement # U58DP003862–01 awarded to the California Department of Public Health. The ideas and opinions expressed herein are those of the author(s) and endorsement by the State of California Department of Public Health, the NCI, and the Centers for Disease Control and Prevention or their Contractors and Subcontractors is not intended nor should be inferred. The authors acknowledge the efforts of the National Cancer Institute; the Office of Research, Development and Information, CMS; Information Management Services (IMS), Inc.; and the Surveillance, Epidemiology, and End Results (SEER) Program tumor registries in the creation of the SEER-Medicare database.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.