Skip to main content
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2019 Mar 15.
Published in final edited form as: Int J Cancer. 2017 Nov 24;142(6):1202–1208. doi: 10.1002/ijc.31148

Autoimmune diseases and breast cancer risk by tumor hormone-receptor status among elderly women

Catherine Schairer 1,2, Ruth M Pfeiffer 1, Shahinaz M Gadalla 1
PMCID: PMC5773407  NIHMSID: NIHMS925089  PMID: 29144542

Abstract

The female preponderance of many autoimmune diseases suggests a possible hormonal etiology. Little research exists on systemic and organ-specific autoimmune diseases and risk of breast cancer by tumor estrogen receptor (ER)- and progesterone receptor (PR)- status. Here we evaluate associations between selected systemic and organ-specific autoimmune diseases and breast cancer risk overall and by tumor ER- and PR-status. We used linked Surveillance, Epidemiology and End Results (SEER)-Medicare data, with first female breast cancer cases ages ≥66 years identified by SEER registries (years 1992–2011) (N = 209,929). We selected female controls (N = 200,000) from a stratified 5% random sample of Medicare recipients who were alive and breast cancer-free. We assessed exposures until 12 months before breast cancer diagnosis/selection using Medicare claims data. We estimated odds ratios (OR) and 99.9% confidence intervals (CI) using unconditional and multinomial logistic regression. We found reduced breast cancer risk among those with rheumatoid arthritis (OR=0.84; 99.9% CI 0.79–0.89), systemic lupus erythematosus (OR=0.82; 99.9% CI 0.70–0.97), and pernicious anemia (OR=0.90; 99.9% CI 0.83–0.97), and increased risk among those with psoriasis (OR=1.16; 99.9% CI 1.06–1.27). Statistically significant alterations in risk for rheumatoid arthritis were limited to ER-positive (+) breast cancer, whereas those for the other three conditions were further limited to ER+/PR+ breast cancer. However, only differences for rheumatoid arthritis by ER-status were statistically significant (p-heterogeneity=.0001). The reasons for these associations need to be investigated in future studies accounting for host characteristics and autoimmune disease treatment.

Keywords: breast cancer, autoimmune disease, estrogen receptor, progesterone receptor

Introduction

Autoimmune diseases are a heterogeneous group of diseases, but all are characterized by the presence of autoantibodies and autoreactive T cells that attack “self” antigens, either systemically or organ-specifically (1, 2). Some evidence suggests that some parts of the immune system may protect against cancer, while other parts may promote cancer development (3). A previous analysis of systemic autoimmune rheumatic diseases (SARDs) and breast cancer risk in the female Surveillance, Epidemiology, and End Results (SEER)-Medicare population reported that women diagnosed with rheumatoid arthritis were less likely to be diagnosed with breast cancer; reductions in risk were not statistically significantly different for estrogen receptor (ER)-positive (+) and ER-negative (−) breast cancer (4). Breast cancer risk was not associated with any other SARDs, except for a reduction in risk of ER− breast cancer among those with systemic lupus erythematosus. To our knowledge, associations between organ-specific autoimmune diseases and breast cancer risk have not been previously reported in the SEER-Medicare population, except for a reduction in risk after pernicious anemia (5).

The female preponderance of many autoimmune diseases suggests a possible hormonal etiology (6). Therefore, it is of interest to examine associations with breast cancer by the hormone-receptor status of the tumors. While others have examined the risk of breast cancer associated with both SARDs and organ-specific autoimmune diseases (e.g. reference 1), to our knowledge, no studies other than the earlier one in the SEER-Medicare database (4) have examined associations between autoimmune diseases and the risk of breast cancer by ER-status of the tumor and none have accounted for both ER- and progesterone receptor (PR)-status.

In this manuscript, we expand on the previous analysis of SARDs and breast cancer risk in the SEER-Medicare database (4) by including additional years of breast cancer case ascertainment (through 2011), organ-specific autoimmune diseases, and analyses by tumor ER- and PR-status.

Materials and Methods

We designed a case-control study within the SEER-Medicare linked database (https://healthcaredelivery.cancer.gov/seermedicare/), with cases identified by SEER registries and information on medical exposures obtained from Medicare claims data. Details on the design have been published previously (7).

The study was limited to females, aged 66 years of age or older (persons of age 65 were excluded because they did not have sufficient time to accrue exposure information) with a minimum of 13 months of Part A, Part B non-health maintenance organization Medicare coverage preceding cancer diagnosis, as assessed using Medicare claims data. First invasive breast cancers (ICD-O-3 site codes C50.0-C50.9 where behavior = 3 (malignant)) diagnosed between January 1, 1992 and December 31, 2011 from 13 SEER registries and from January 1, 2000 to December 31, 2011 from the additional 5 registries in SEER 18 were included. A total of 313,159 cases of first invasive breast cancer among women aged 66 years and older were identified. We then excluded cases diagnosed at only autopsy or death (n = 3,431), without a known month of cancer diagnosis (n = 1,636), not living in the included SEER areas at the time of diagnosis (n = 2,408), without a Medicare claim between their entry date and 1 year prior to breast cancer diagnosis (n = 71,166), without 13 months of Medicare Part A, Part B excluding HMO claims (n = 24,574), or whose death date was prior to cancer diagnosis date (n = 15), leaving 209,929 invasive breast cancer cases included in the analysis. Information on ER- and PR-status of the tumors was available from the SEER registry data for the years of case ascertainment. Fifteen percent of tumors overall had unknown ER-status, with the percentage declining from 26% in 1992 to 5% in 2011. The corresponding percentages for PR-status were 16%, declining from 28% to 5%.

A total of 200,000 female controls were selected from a file created from two subsets of a 5% random sample of Medicare recipients: the Summarized Denominator (SUMDENOM) file, which is made of 5% random sample of Medicare recipients who never developed cancer and the Patient Entitlement and Diagnosis Summary File (PEDSF) 5% file, which is a 5% random sample of all people with cancer. Controls were selected in strata by calendar year and 5-year age group and were frequency-matched to the distribution among breast cancer cases. To be eligible for selection as a control in a particular calendar year, recipients had to be alive and breast cancer-free as of July 1 in that year.

We identified SARDs and organ-specific autoimmune diseases, the exposures of interest, as well as diabetes and dyslipidemia, adjustment factors, using codes shown in Table 1. ICD-9-CM diagnosis codes were considered if there was at least one inpatient or two outpatient/physician claims with a minimum of 30 days between claims. We defined medical conditions as present if they met the claims definition at least 12 months prior to cancer diagnosis or selection. This 12-month window was chosen to minimize the potential effects of breast cancer on the diagnosis of exposures, while still ensuring that all participants had at least one month out of the 13 months of Medicare coverage required for entrance into the study to assess exposure.

Table 1.

Codes used to define autoimmune diseases and other medical conditions that were adjusted for in the analyses

Autoimmune diseases Codes
Systemic autoimmune rheumatic diseases (SARDs) rheumatoid arthritis (ICD-9 714.0, 714.1, 714.2, 714.3, 714.81, or V82.1), systemic lupus erythematosus (ICD-9 710.0), Sjogren’s syndrome (ICD-9 710.2), dermatomyositis (ICD-9 710.3), systemic sclerosis (scleroderma) (ICD-9 710.1), sarcoidosis (ICD-9 135), Goodpasture’s syndrome (ICD-9 446.21), and Behcet’s syndrome ICD-9 136.1), other (ICD-9 codes 446.5, 710.4, 725.x).
Organ-specific autoimmune diseases Hashimoto’s thyroiditis (ICD-9 code 245.2), pernicious anemia (ICD-9-CM 281.0), vitiligo (ICD-9 code 709.01), multiple sclerosis (ICD-9 340), Grave’s disease (ICD-9 codes 242.00, 242.01), glomerulonephritis (ICD-9 code 580.4, 582.x), ankylosing spondylitis (ICD-9 code 720.0), Celiac disease (ICD-9 code 579.0), Chagas disease (ICD-9 code 086.0, 086.2), Crohn’s disease (ICD-9 code 555. x),inflammatory bowel disease (ICD-9 codes 555.0, 555.1, 555.2, 555.9, 556.0, 556.1, 556.2, 556.3, 556.4, 556.5, 556.6, 556.8, 556.9,), psoriasis (ICD-9 code 696.x).
Medical conditions adjusted for in the analyses
Diabetes Icd-9-CM codes 250, 357.2x, 362.0x, 366.41, 249.xx
Dyslipidemia ICD-9 272

We calculated ORs and 99.9% confidence intervals (CIs) (to account for multiple comparisons) using unconditional logistic regression analyses. All statistical tests were two-sided with an α of .0009. We performed statistical analyses using SAS version 9.3 (SAS Institute). Unless otherwise noted, all analyses were adjusted for the matching factors of age (continuous) and year of diagnosis/selection (continuous), and additionally for race/ethnicity (white, black, mixed, Asian, Hispanic, North American Indian, other/unknown), grouped SEER region categorized as western (San Francisco, Hawaii, New Mexico, Seattle, Utah, San Jose, Los Angeles, greater California), northeastern (Connecticut, New Jersey), north-central (Detroit, Iowa), and southern (Atlanta, rural Georgia, Kentucky, Louisiana, greater Georgia), months between study entry and diagnosis/selection date (continuous), and dyslipidemia and diabetes, which were previously identified risk factors in breast cancer subgroups (7), both coded yes/no. We used multinomial logistic regression with nominal outcomes for analyses by tumor hormone-receptor status compared to controls with the Wald test for assessing statistical significance of effect differences.

Results

Basic information on the study population is shown in Table 2. As expected, cases and controls were similar on study design matching factors. The distributions of mammography screening, dyslipidemia and diabetes were also similar in cases and controls. Cases were slightly more likely to be Caucasian (87% vs. 83%).

Table 2.

Characteristics of the study population

Characteristic Cases (N = 209,929) Controls (N = 200,000)

Age in years (mean) 76 76

Year of diagnosis (mean) 2003 2003

Duration of coverage in months (mean) 66 66

SEER regiona (%)
  Western 45.2 45.4
  Northeastern 19.0 18.2
  North-central 17.4 17.5
  Southern 18.4 19.0

Race/ethnicity (%)
  White 87.3 83.2
  Black 7.5 8.0
  Mixed 1.6 2.1
  Asian 2.1 3.8
  Hispanic 1.2 2.2
  North American Indian 0.2 0.4
  Other/Unknown 0.2 0.2

Mammogram between 1 and 4 years before diagnosis (%) 35.2 34.5

Dyslipidemia (%) 51.2 50.9

Diabetes (%) 22.2 22.4
a

Western (San Francisco, Hawaii, New Mexico, Seattle, Utah, San Jose, Los Angeles, greater California), Northeastern (Connecticut, New Jersey), North-central (Detroit, Iowa), and Southern (Atlanta, rural Georgia, Kentucky, Louisiana, greater Georgia).

Nearly 6% of controls were diagnosed with SARDs; 64% of controls with SARDs had rheumatoid arthritis. There were very few subjects with Goodpasture’s syndrome (7 total) or Behcets syndrome (14 total), and none with Chagas disease. Approximately 5% of controls were diagnosed with organ-specific autoimmune diseases, but individual conditions were very rare; fewer than 1% had each disease, except for pernicious anemia (1.9%) and psoriasis (1.2%). Among controls with SARDs, 11.6% also had organ-specific autoimmune diseases; of those with organ-specific autoimmune diseases, 13.6% also had SARDs.

Dyslipidemia and diabetes were more prevalent in controls with SARDs (62.6% and 31.3%, respectively) and organ-specific autoimmune diseases (67.5% and 32.8%) than in controls without SARDs (50.2% and 21.8%) or organ-specific autoimmune diseases (52% and 21.8%).

Unadjusted ORs for breast cancer associated with SARDs and organ-specific autoimmune diseases were 0.86 (99.9% CI 0.82–0.90) and 0.99 (99.9% CI 0.94–1.04), respectively. The corresponding adjusted estimates, including mutual adjustment for each other, were 0.86 (99.9% CI 0.82–0.90) and 0.98 (99.9% CI 0.94–1.03) (Table 3). The reduction in risk with SARDs was primarily attributable to rheumatoid arthritis (OR = 0.84; 99.9% CI 0.79–0.89) and systemic lupus erythematosus (OR = 0.82; 99.9% CI 0.70–0.97). The overall null association with organ-specific autoimmune diseases obscured the reduced risk associated with pernicious anemia (OR =0.90; 99.9% CI 0.83–0.97) and the increased risk associated with psoriasis (OR = 1.16; 99.9% CI 1.06–1.27).

Table 3.

ORs (99.9% CI) for breast cancer overall and by estrogen (ER)-status (positive (+) or negative (−)) associated with systemic autoimmune rheumatic diseases (SARDs) and organ-specific autoimmune diseases

Autoimmune disease Controls
N=200,000
Total breast cancer
N = 209,929
ER+ breast cancer
N=149,306
ER− breast cancer
N = 28,140
p-value for
ER+ vs. ER−

#
exposed
#
exposed
ORa (99.9% CI)
vs. unexposed
#
exposed
ORa (99.9% CI)
vs. unexposed
#
exposed
ORa (99.9% CI)
vs. unexposed

Total SARDs 11769 10720 0.86 (0.82–0.90) 7570 0.84 (0.80–0.88) 1543 0.93 (0.85–1.02) .0004
Rheumatoid arthritis 7477 6474 0.84 (0.79–0.89) 4480 0.80 (0.75–0.86) 972 0.92 (0.82–1.04) .0001
Sjogren’s syndrome 782 721 0.94 (0.79–1.11) 528 0.93 (0.77–1.13) 95 0.90 (0.62–1.29)
Systemic sclerosis 262 238 0.90 (0.67–1.21) 183 0.94 (0.68–1.30) 26 0.76 (0.39–1.47)
Systemic lupus erythematosus 958 777 0.82 (0.70–0.97) 552 0.79 (0.66–0.95) 128 0.95 (0.69–1.31) .07
Sarcoidosis 288 287 0.94 (0.71–1.24) 210 0.94 (0.70–1.28) 41 0.89 (0.51–1.55)

Total organ-specific 10262 10667 0.98 (0.94–1.03) 7739 0.97 (0.92–1.02) 1432 1.01 (0.91–1.11) 277
Crohn’s disease 623 582 0.91 (0.70–1.17) 421 0.91 (0.69–1.21) 83 0.90 (0.53–1.51)
Multiple sclerosis 395 406 0.95 (0.75–1.20) 293 0.92 (0.71–1.19) 64 1.09 (0.70–1.71)
Pernicious anemia 3841 3608 0.90 (0.83–0.97) 2510 0.87 (0.80–0.95) 495 0.96 (0.82–1.13) .04
Psoriasis 2368 2899 1.16 (1.06–1.27) 2174 1.18 (1.07–1.30) 343 1.04 (0.86–1.26) .03
Inflammatory bowel disease 1340 1322 0.96 (0.81–1.15) 961 0.93 (0.77–1.13) 189 1.04 (0.74–1.47)
Vitiligo 95 108 1.17 (0.73–1.86) 80 1.15 (0.69–1.92) 17 1.26 (0.52–3.02)
Glomerulonephritis 323 272 0.83 (0.63–1.09) 186 0.81 (0.60–1.11) 38 0.81 (0.46–1.43)
Celiac disease 156 127 0.75 (0.51–1.12) 98 0.76 (0.50–1.18) 16 0.73 (0.31–1.75)
Ankylosing spondylitis 136 123 0.91 (0.60–1.38) 76 0.78 (0.48–1.26) 23 1.26 (0.60–2.68)
Grave’s disease 756 849 1.07 (0.91–1.27) 643 1.10 (0.92–1.31) 121 1.13 (0.81–1.56)
Hashimoto thyroiditis 908 975 1.00 (0.85–1.16) 744 1.01 (0.86–1.19) 124 0.96 (0.70–1.32)
a

Variables included in the models: age, race/ethnicity, region, year of diagnosis/selection, months between entry and diagnosis/selection, mammogram, diabetes, dyslipidemia, autoimmune conditions included in the table (for analysis of total SARDs and total organ-specific autoimmune diseases, only these conditions were included in the model; for analyses of the individual conditions, the total variables were not included in the models, only the individual conditions). Statistically significant ORs are bolded.

All of the statistically significant associations mentioned above were limited to ER+ breast cancer, although only differences for rheumatoid arthritis by ER-status were significant at α = .0009. Of note, the percentage of cases with missing ER-status was similar for those with and without SARDs and for those with and without organ-specific autoimmune diseases.

Reductions in risk of ER+ breast cancer associated with total SARDs and rheumatoid arthritis were statistically significant regardless of tumor PR-status (Table 4). The associations with systemic lupus erythematosus, pernicious anemia and psoriasis were statistically significant only for ER+/PR+ tumors. In general, associations were stronger for ER+/PR+ than ER+/PR− breast cancer, but p-values for heterogeneity were .01 (all SARDs), .02 (rheumatoid arthritis), .06 (systemic lupus erythematosus), .06 (pernicious anemia), and .53 (psoriasis), and did not meet the .0009 criterion for statistical significance. There were no statistically significant associations for any cross-classification of ER/PR status with the other autoimmune conditions.

Table 4.

ORs (99.9% CIs) for breast cancer cross-classified by estrogen receptor (ER)- and progesterone receptor (PR)-status (both classified as positive (+) or negative (−)) associated with selected autoimmune diseases

ER+/PR+ breast cancer
N=121,349
ER+/PR− breast cancer
N = 25,256
ER−/PR+ breast cancer
N = 1,946
ER−/PR− breast cancer
N = 25,837
#
exposed
cases
ORa (99.9% CI) #
exposed
cases
ORa (99.9% CI) #
exposed
cases
ORa (99.9% CI) #
exposed
cases
ORa (99.9% CI)
Systemic autoimmune rheumatic diseases (SARDs)
Total SARDS 6090 0.83 (0.72–0.87) 1356 0.89 (0.81–0.99) 124 1.17 (0.86–1.61) 1405 0.91 (0.83–1.01)
Rheumatoid arthritis 3582 0.79 (0.74–0.85) 811 0.87 (0.76–0.98) 72 1.02 (0.68–1.53) 895 0.92 (0.82–1.04)
Systemic lupus erythematosus 437 0.77 (0.63–0.93) 109 0.94 (0.67–1.33) 15 1.62 (0.66–3.95) 113 0.91 (0.65–1.27)
Organ-specific autoimmune diseases
Total 6270 0.96 (0.91–1.02) 1345 1.02 (0.92–1.12) 103 1.15 (0.81–1.61) 1311 0.99 (0.90–1.10)
Pernicious anemia 2012 0.86 (0.78–0.94) 463 0.95 (0.80–1.12) 45 1.36 (0.82–2.26) 447 0.94 (0.79–1.11)
Psoriasis 1789 1.19 (1.07–1.32) 352 1.14 (0.95–1.39) 20 0.93 (0.44–1.97) 317 1.04 (0.85–1.27)
a

Variables included in the models: age, race/ethnicity, region, year of diagnosis/selection, months between entry and diagnosis/selection, mammogram, diabetes, dyslipidemia, autoimmune conditions included in the table (for analysis of total SARDs and total organ-specific autoimmune diseases, only these conditions were included in the mode l; for analyses of the individual conditions, the total variables were not included in the models, only the individual conditions). Statistically significant ORs are bolded.

Discussion

We found reductions in risk of breast cancer among elderly women with rheumatoid arthritis, systemic lupus erythematosus, and pernicious anemia, and increases in risk with psoriasis. Statistically significant alterations in risk for rheumatoid arthritis were limited to ER+ breast cancer, whereas those for the other three conditions were further limited to ER+/PR+ breast cancer. However, only differences for rheumatoid arthritis by ER-status were statistically significant.

To our knowledge, no studies have conducted similar analyses considering both tumor ER- and PR-status. In the only study of which we are aware that considered ER-status, also in the SEER-Medicare database but with many fewer cases and controls, risk reduction associated with rheumatoid arthritis was statistically significant for ER+, but not ER− breast cancer, although this difference was not statistically significant (4). Contrary to the current analysis, risk of ER− but not ER+ breast cancer was reduced in those with systemic lupus erythematosus, but this association was based on only 10 cases (4) and may likely be due to chance.

Risk of breast cancer in those with autoimmune diseases was most comprehensively examined in a Swedish study among 200,000 patients hospitalized with any of 33 different autoimmune diseases (1). Our findings for rheumatoid arthritis, systemic lupus erythematosus, and pernicious anemia are consistent with results from this study. Rheumatoid arthritis has also been associated with reduced breast cancer risk in two meta-analyses (8, 9), systemic lupus erythematosus with reduced risk in a meta-analysis (10) but not in other analyses in the Medicare population (4,11), and pernicious anemia with reduced risk in the SEER-Medicare population (5). Psoriasis has been linked to increased breast cancer risk in a meta-analysis (12), but not in the Swedish study (1) or a more recent study (13). Notably, effect sizes and number of cases for this association are small.

Our findings for systemic sclerosis and sarcoidosis are also consistent with the Swedish study and two meta-analyses (15, 16); those for multiple sclerosis agree with the Swedish study (1), but variable results have been reported elsewhere (17). Sjögren’s syndrome, Celiac disease, Crohn’s disease, and Hashimoto/hypothyroidism were all associated with statistically significantly reduced risk in the Swedish study (1), but not in our study. Inconsistent results have also been reported elsewhere for Crohn’s disease (18). Our results also differ for Grave’s disease/hyperthyroidism and inflammatory bowel disease, which were associated with small but statistically significant increases in risk in the Swedish study (1). We note that the Swedish study included those of any age hospitalized for autoimmune disorders, and was likely skewed to more severe disease. This contrasts with our study which included older women with autoimmune diagnoses from inpatient or outpatient records. Moreover, it is likely that the relatively small number of breast cancer cases in the Swedish study (N = 4607), sometimes small effect sizes, and the differing levels of statistical significance account for some of the differences in the results of the two studies.

The associations with ER+ but not ER− breast cancer in the current analysis suggest that hormonal factors might be related to the observed associations. At least seven of the autoimmune diseases we examined have an excess in women (19, 20). Notably, we found reductions in risk of ER+ breast cancer with only two (rheumatoid arthritis, systemic lupus erythematosus) of the seven, increased risk of ER+ breast cancer for one (psoriasis) of five with no known female excess, and reduced risk with another with no known female excess (pernicious anemia) (21). Although it is thought that sex hormones contribute to the female excess in autoimmune diseases, levels of these hormones have not been found to be significantly different in women with and without autoimmune diseases, except possibly for higher levels in patients with systemic lupus erythematosus (19). Thus, our findings of a reduced risk of ER+ breast cancer with systemic lupus erythematosus are puzzling given that higher estrogen levels may also be associated with increased risk of ER+ breast cancer (22). In spite of the female excess of rheumatoid arthritis, the peak incidence in women occurs at the time of menopause, which is associated with decreased production of estrogens; in many women with rheumatoid arthritis, disease activity diminishes during pregnancy, when levels of female sex hormones are high. Furthermore, there is some evidence that menopausal hormone therapy decreases disease activity in postmenopausal rheumatoid arthritis patients (23, 24). Thus, lower hormone levels might be associated with increased risk of rheumatoid arthritis but decreased risk of hormone sensitive breast cancers. The relationships between pernicious anemia and psoriasis and hormonal factors are unclear (21, 25).

Treatments for autoimmune diseases, particularly if they differentially affect ER+ and ER− breast cancer, also need to be considered in interpreting our findings. Non-steroidal anti-inflammatory drugs (NSAIDs) are among current treatment options (26) that have been associated with a greater reduction in risk of ER+ breast cancer than breast cancer altogether (27). Thus, NSAID use may contribute to the reduced risk we found for ER+ but not ER− breast cancer in those with rheumatoid arthritis. On the other hand, reduced breast cancer risk was not evident for other autoimmune diseases that might be treated by NSAIDs. Glucocorticoids, potent anti-inflammatory drugs, have not been associated with breast cancer risk (28). Some disease-modifying-antirheumatic drugs (DMARDs) used to treat rheumatoid arthritis, appear to increase risk of specific cancer types, such as bladder cancer, but few appear to increase the overall cancer risk (29). There is no evidence that the increased risk of breast cancer associated with psoriasis is due to treatment (30).

Assessing the effects of possible confounding factors is also important in evaluating our results. Low to moderate alcohol consumption has been associated with reduced risk of rheumatoid arthritis (31) and possibly systemic lupus erythematosus (32), and increased risk of ER+ breast cancer (33). Had we been able to adjust for alcohol consumption, our associations with these two conditions and breast cancer risk would most likely be closer to the null (34). Obesity has been associated with increased risk of both rheumatoid arthritis (35), psoriasis (35) and ER+ breast cancer (22); thus, adjustment for body mass index would likely have resulted in estimates associated with rheumatoid arthritis further from the null and estimates for psoriasis closer to the null (34). Estimated prevalence of oral estrogen-progestin use is very low in the age group included in this analysis (36), minimizing concerns about lacking this information.

Strengths of this study include a larger number of breast cancer cases following most autoimmune diseases than available in previously published studies. In addition, we have information on both ER- and PR-status of the tumors. Other strengths include the population-based nature of the study, claim-based data on autoimmune conditions, and information on frequency of medical visits and mammographic screening to address the intensity of medical surveillance. The study design ensured that there was no bias in recall of autoimmune conditions. We have accounted for the potential influence of reverse causality and detection bias around the time of cancer diagnosis by ascertaining exposures only until one year prior to cancer diagnosis or selection of controls.

Limitations include the relatively short period of exposure ascertainment, but given that these are chronic health conditions, their capture in the medical records may be adequate. However, we did not have detailed information on the duration and severity of the conditions. As treatment data were not available, we could not untangle the effects of disease and treatment. As indicated above, we were also unable to adjust for several potential confounding factors, such as alcohol and obesity. Although approximately 15% of tumors had missing hormone-receptor status, the degree of missing data did not vary by exposures.

In summary, this study suggests that tumor hormone receptor status is important to consider when assessing the relationship between autoimmune diseases and breast cancer risk. Future research that accounts for such tumor characteristics as well as treatment for the autoimmune diseases and other patient characteristics will be important in untangling the complex relationship between these diseases and breast cancer risk.

Novelty and Impact.

This is the first study to examine associations of systemic and organ-specific autoimmune diseases with breast cancer according to tumor estrogen receptor (ER)- and progesterone receptor (PR)-status. Our findings that statistically significant alterations in risk with rheumatoid arthritis, systemic lupus erythematosus, pernicious anemia, and psoriasis varied according to tumor ER- and PR-status offer important clues to the complex relationship between certain autoimmune diseases and breast cancer risk.

Acknowledgments

We would like to acknowledge the programming support of Winnie Ricker at Information Management Services, Inc.

Abbreviations

ER

estrogen receptor

+

positive

negative

PR

progesterone receptor

SARD

systemic autoimmune rheumatic disease

SEER

Surveillance, Epidemiology, and End Results

NSAIDs

non-steroidal anti-inflammatory drugs

References

  • 1.Hemminki K, Liu X, Ji J, Forsti A, Sundquist J, Sundquist Effect of autoimmune diseases on risk and survival in female cancers. Gynecol Oncol. 2012;127:180–5. doi: 10.1016/j.ygyno.2012.07.100. [DOI] [PubMed] [Google Scholar]
  • 2.Zhernakova A, Withoff S, Wijmenga C. Clinical implications of shared genetics and pathogenesis in autoimmune diseases. Nat Rev Endocrinol. 2013;9:646–59. doi: 10.1038/nrendo.2013.161. [DOI] [PubMed] [Google Scholar]
  • 3.Franks AL, Slansky JE. Multiple associations between a broad spectrum of autoimmune diseases, chronic inflammatory diseases and cancer. Anticancer Res. 2012;32:1119–36. [PMC free article] [PubMed] [Google Scholar]
  • 4.Gadalla SM, Amr S, Langenberg P, Baumgarten M, Davidson WF, Schairer C, Engels EA, Pfeiffer RM, Goedert JJ. Breast cancer risk in elderly women with systemic autoimmune rheumatic diseases: a population-based case-control study. Br J Cancer. 2009;100:817–21. doi: 10.1038/sj.bjc.6604906. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Murphy G, Dawsey SM, Engels EA, Ricker W, Parsons R, Etemadi A, Lin S-W, Abnet CC, Freedman ND. Cancer risk after pernicious anemia in the US elderly population. Clin Gastroenterol Hepatol. 2015;13:2282–9. doi: 10.1016/j.cgh.2015.05.040. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Quintero OL, Amador-Patarroyo MJ, Montoya-Ortiz G, Rojas-Villarraga A, Anaya J-M. Autoimmune disease and gender: plausible mechanisms for the female predominance of autoimmunity. J Autoimmun. 2012;38:J109–19. doi: 10.1016/j.jaut.2011.10.003. [DOI] [PubMed] [Google Scholar]
  • 7.Schairer C, Gadalla SM, Pfeiffer RM, Moore SC, Engels EA. Diabetes, abnormal glucose, dyslipidemia, hypertension, and risk of inflammatory and other breast cancer. Cancer Epidemiol Biomarkers Prev. 2017;26:862–8. doi: 10.1158/1055-9965.EPI-16-0647. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Smitten AL, Simon TA, Hochberg MC, Suissa S. A meta-analysis of the incidence of malignancy in adult patients with rheumatoid arthritis. Arthritis Res Ther. 2008;0:R45. doi: 10.1186/ar2404. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Tian G, Liang J-N, Wang Z-Y, Zhou D. Breast cancer risk in rheumatoid arthritis: an update meta-analysis. Biomed Res Int. 2014;2014:9. doi: 10.1155/2014/453012. Article ID 453012, http://dx.doi.org/10.1155/2014/453012. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Bernatsky S, Ramsey-Goldman R, Foulkes WD, Gordon C, Clarke AE. Breast, ovarian, and endometrial malignancies in systemic lupus erythematosus: a meta-analysis. Br J Cancer. 2011;104:1478–81. doi: 10.1038/bjc.2011.115. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Khaliq W, Qayyum R, Clough J, Vaidya D, Wolff AC, Becker DM. Comparison of breast cancer risk in women with and without systemic lupus erythematosus in a Medicare population. Breast Cancer Res Treat. 2015;151:465–74. doi: 10.1007/s10549-015-3412-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Pouplard C, Brenaut E, Horreau C, Barnetche T, Misery L, Richard M-A, Aractingi S, Aubin F, Cribier B, Joly P, Jullien D, Le Maitre M, Ortonne J-P, Paul C. Risk of cancer in psoriasis: a systematic review and meta-analysis of epidemiologic studies. JEADV. 2013;27:36–46. doi: 10.1111/jdv.12165. [DOI] [PubMed] [Google Scholar]
  • 13.Chiesa Fuxench ZC, Shin DB, Beatty AO, Gelfand JM. The risk of cancer in patients with psoriasis: a population-based cohort study in the Health Improvement Network. JAMA Dermatol. 2016;152:282–90. doi: 10.1001/jamadermatol.2015.4847. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Liang Y, Yang Z, Qin B, Zhong R. Primary Sjogren’s syndrome and malignancy risk: a systematic review and meta-analysis. Ann Rheum Dis. 2014;73:1151–56. doi: 10.1136/annrheumdis-2013-203305. [DOI] [PubMed] [Google Scholar]
  • 15.Bonifazi M, Tramacere I, Pomponio G, Gabrielli B, Avvedimento EV, La Vecchia C, Negri E, Gabrielli A. Systemic schlerosis (scleroderma) and cancer risk: systemic review and meta-analysis of observational studies. Rheumatology (Oxford) 2013;52:143–54. doi: 10.1093/rheumatology/kes303. [DOI] [PubMed] [Google Scholar]
  • 16.Bonifazi M, Bravi F, Gasparini S, La Vecchia C, Gabrielli A, Wells AU, Renzoni EA. Sarcoidosis and cancer risk: systematic review and meta-analysis of observational studies. Chest. 2015;147:778–91. doi: 10.1378/chest.14-1475. [DOI] [PubMed] [Google Scholar]
  • 17.Marrie RA, Reider N, Cohen J, Stuve O, Trojano M, Sorensen PS, Reingold SC, Cutter G. A systematic review of the incidence and prevalence of cancer in multiple sclerosis. Mult Scler. 2015;21:294–304. doi: 10.1177/1352458514564489. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Hemminki K, Li X, Sundquist J, Sundquist K. Cancer risk in Crohn disease patients. Ann Oncol. 2009;20:574–80. doi: 10.1093/annonc/mdn595. [DOI] [PubMed] [Google Scholar]
  • 19.Nussinovitch U, Shoenfeld Y. The role of gender and organ specific autoimmunity. Immunity Reviews. 2012;11:A377–85. doi: 10.1016/j.autrev.2011.11.001. [DOI] [PubMed] [Google Scholar]
  • 20.LLeo A, Battezzati PM, Selmi C, Gershwin ME, Podda M. Is autoimmunity a matter of sex? Immunity Reviews. 2008;7:626–30. doi: 10.1016/j.autrev.2008.06.009. [DOI] [PubMed] [Google Scholar]
  • 21.Osborne D, Sobczynska-Malefora A. Autoimmune mechanisms in pernicious anaemia & thyroid disease. Autoimmunity Reviews. 2015;14:763–8. doi: 10.1016/j.autrev.2015.04.011. [DOI] [PubMed] [Google Scholar]
  • 22.Munsell MF, Sprague BL, Berry DA, Chisolm G, Trentham-Dietz A. Body mass index and breast cancer risk according to postmenopausal estrogen-progestin use and hormone receptor status. Epidemiol Rev. 2014;36:114–36. doi: 10.1093/epirev/mxt010. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Islander U, Jochems C, Lagerquist MK, Forsblad-d’Elia H, Carlsten H. Estrogens in rheumatoid arthritis; the immune system and bone. Mol Cell Endocrinol. 2011;335:14–29. doi: 10.1016/j.mce.2010.05.018. [DOI] [PubMed] [Google Scholar]
  • 24.Khan WA, Khan MWA. Cancer morbidity in rheumatoid arthritis: role of estrogen metabolites. Biomed Res Int. 2013;2013:748178. doi: 10.1155/2013/748178. doi: 1155/2013/748178. Epub 2013 Sep 17. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Wu S, Cho E, Li W-Q, Grodstein F, Qureshi AA. Hormonal factors and risk of psoriasis in women: a cohort study. Acta Derm Venereol. 2016;96:927–31. doi: 10.2340/00015555-2312. [DOI] [PubMed] [Google Scholar]
  • 26.Bays AM, Gardner G. Pharmacologic therapies for rheumatologic and autoimmune conditions. Med Clin North Am. 2016;100:719–31. doi: 10.1016/j.mcna.2016.03.001. [DOI] [PubMed] [Google Scholar]
  • 27.De Pedro M, Baeza S, Excudero MT, Dierssen-Sotos T, Gómez-Acebo I, Pollán M, Llorca J. Effect of COX-2 inhibitors and other non-steroidal inflammatory drugs on breast cancer risk: a meta-analysis. Breast Cancer Res Treat. 2015;149:525–36. doi: 10.1007/s10549-015-3267-9. [DOI] [PubMed] [Google Scholar]
  • 28.Sørensen GV, Cronin-Fenton DP, Sørensen HT, Ulrichsen SP, Pedersen L, Lash TL. Use of glucocorticoids and risk of breast cancer: a Danish population-based case-control study. Breast Cancer Research. 2012;14:R21. doi: 10.1186/bcr3106. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Wilton KM, Matteson EL. Malignancy incidence, management and prevention in patients with rheumatoid arthritis. Rheumatol Ther. 2017 May 15; doi: 10.1007/s40744-017-0064-4. [Epub ahead of print] [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Patel RV, Clark LN, Lebwohl M, Weinberg JM. Treatments for psoriasis and the risk of malignancy. J Am Acad Dermatol. 2009;60:1001–17. doi: 10.1016/j.jaad.2008.12.031. [DOI] [PubMed] [Google Scholar]
  • 31.Jin Z, Xiang C, Cai Q, Wei X, He J. Alcohol consumption as a preventive factor for developing rheumatoid arthritis: a dose-response meta-analysis of prospective studies. Ann Rheum Dis. 2014;73:1962–7. doi: 10.1136/annrheumdis-2013-203323. [DOI] [PubMed] [Google Scholar]
  • 32.Takvorian SU, Merola JF, Costenbader KH. Cigarette smoking, alcohol consumption and risk of systemic lupus erythematosus. Lupus. 2014;23:537–44. doi: 10.1177/0961203313501400. [DOI] [PubMed] [Google Scholar]
  • 33.Suzuki R, Orsini N, Mignone L, Saji S, Wolk A. Alcohol intake and risk of breast cancer defined by estrogen and progesterone receptor status--a meta-analysis of epidemiological studies. Int J Cancer. 2008;122:1832–41. doi: 10.1002/ijc.23184. [DOI] [PubMed] [Google Scholar]
  • 34.Mehio-Sibai A, Feinleig M, Sibai TA, Armenian HK. A positive and negative confounding variable? A simple teaching aid for clinicians and students. Ann Epidemiol. 2005;15:421–3. doi: 10.1016/j.annepidem.2004.10.004. [DOI] [PubMed] [Google Scholar]
  • 35.Versini M, Jeandel PY, Rosenthal E, Shoenfeld Y. Obesity in autoimmune diseases: not a passive bystander. Autoimmun Rev. 2014;13:981–1000. doi: 10.1016/j.autrev.2014.07.001. [DOI] [PubMed] [Google Scholar]
  • 36.Jewett PI, Gangnon RE, Trentham-Dietz A, Sprague BL. Trends of postmenopausal estrogen plus progestin prevalence in the United States between 1970 and 2010. Obstet Gynecol. 2014;124:727–33. doi: 10.1097/AOG.0000000000000469. [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES