Background: Elevated keratinocyte carcinoma risk is present with several immune-related conditions, e.g., solid organ transplantation and non-Hodgkin lymphoma. Because many immune-related conditions are rare, their relationships with keratinocyte carcinoma have not been studied.

Methods: We used Medicare claims to identify cutaneous squamous cell carcinoma (SCC) and basal cell carcinoma (BCC) cases in 2012, and controls matched on sex and age. All subjects were aged 65 to 95 years, of white race, and had attended ≥1 dermatologist visit in 2010–2011. Immune-related conditions were identified during 1999–2011 using Medicare claims. Associations were estimated with logistic regression, with statistical significance determined after Bonferroni correction for multiple comparisons.

Results: We included 258,683 SCC and 304,903 BCC cases. Of 47 immune-related conditions, 21 and 9 were associated with increased SCC and BCC risk, respectively. We identified strongly elevated keratinocyte carcinoma risk with solid organ transplantation (SCC OR = 5.35; BCC OR = 1.94) and non-Hodgkin lymphoma (SCC OR = 1.62; BCC OR = 1.25). We identified associations with common conditions, e.g., rheumatoid arthritis [SCC OR = 1.06, 95% confidence interval (95% CI), 1.04–1.09] and Crohn's disease (SCC OR = 1.33, 95% CI, 1.27–1.39; BCC OR = 1.10, 95% CI, 1.05–1.15), and rare or poorly characterized conditions, e.g., granulomatosis with polyangiitis (SCC OR = 1.88; 95% CI, 1.61–2.19), autoimmune hepatitis (SCC OR = 1.81; 95% CI, 1.52–2.16), and deficiency of humoral immunity (SCC OR = 1.51, 95% CI, 1.41–1.61; BCC OR = 1.22, 95% CI, 1.14–1.31). Most conditions were more positively associated with SCC than BCC. Associations were generally consistent regardless of prior keratinocyte carcinoma history.

Conclusions: Many immune-related conditions are associated with elevated keratinocyte carcinoma risk and appear more tightly linked to SCC.

Impact: Immunosuppression or immunosuppressive treatment may increase keratinocyte carcinoma risk, particularly SCC. Cancer Epidemiol Biomarkers Prev; 26(7); 998–1007. ©2017 AACR.

This article is featured in Highlights of This Issue, p. 983

Approximately 5.4 million keratinocyte carcinomas (also known as nonmelanoma skin cancers) are diagnosed in the United States annually (1). Squamous cell carcinoma (SCC) and basal cell carcinoma (BCC) are the primary keratinocyte carcinoma types, and risk for both increases with ultraviolet radiation (UVR) exposure, fair skin pigmentation, and older age (2). Risk is also increased among people with certain medical conditions affecting the immune system. Most notably, solid organ transplant recipients have markedly increased risk of cutaneous SCC and BCC compared with the general population (3–5). Elevated risk of cutaneous SCC and BCC has also been identified in human immunodeficiency virus (HIV)–infected people and non-Hodgkin lymphoma patients (6, 7), and there is evidence of high skin cancer risk among bone marrow transplant recipients (8).

Increased risk may result from effects of immune-related disease processes or treatments (e.g., immunosuppressive or immunomodulatory medications). Among transplant recipients, elevated keratinocyte carcinoma risk may be driven by specific immunosuppressant medications, with particularly high risk observed among people taking azathioprine or cyclosporine (9–11). Immunomodulatory therapy has also been linked to keratinocyte carcinoma incidence among people with rheumatoid arthritis or inflammatory bowel disease, two common autoimmune conditions (12).

Many other rare immune-related conditions have not been studied, but further research could provide insights into the relationship between immune function and keratinocyte carcinoma risk. Clinical populations with substantially elevated risk may benefit from screening, as currently suggested in guidelines for transplant recipients (13–15). In the current study, we utilize Medicare claims data to assess associations between a wide spectrum of immune-related conditions and incidence of cutaneous SCC and BCC in the U.S. elderly population.

We conducted a case-control study among people aged 65 to 95 years in the U.S. Medicare population in 2012. Medicare is the U.S. federal health insurance program for people age 65 and older, and Medicare claims provide information on clinical visits, medical diagnoses, and procedures. We limited the population to non-Hispanic whites, as only small proportions of keratinocyte carcinomas were diagnosed outside this group. The first year during which International Classification of Diseases (version 9, ICD-9) diagnosis codes differentiated between SCC and BCC was 2012 (16). We included people with non-health maintenance organization (HMO) Medicare coverage (because claims for individual diagnoses and procedures are not submitted by HMOs) that began before 2012 and ended after January 1, 2012, and who had at least 1 Medicare claim prior to 2012. Use of Medicare claims for this study was exempt from Institutional Review Board review.

For our study, we identified Medicare patients with SCC or BCC in 2012, based on a physician claim with both an ICD-9 diagnosis code of 173.x2 (for SCC cases) or 173.x1 (for BCC cases) and a Healthcare Common Procedure Coding System (HCPCS) code indicating a skin cancer treatment: 11600-11606, 11620-11626, 11640-11646, 17260-17266, 17270-17276, 17280-17286, 17304, 17311, 17313, as described previously (1). For people with multiple diagnoses in 2012, only the first diagnosis was included. If a person was diagnosed with BCC and SCC on the same day, one diagnosis was chosen at random to categorize the case. Controls were selected from people enrolled in Medicare without SCC or BCC diagnoses in 2012. Controls were selected with a 10:1 ratio, frequency matched to cases on sex and age (in 5-year categories).

For cases and controls, we used Medicare claims during January 1, 1999, to December 31, 2011, to identify 47 different immune-related conditions, including 3 primary immunodeficiency conditions, 3 transplant-related conditions, 32 autoimmune diseases, 6 hematologic malignancies and related conditions, and 3 allergic diseases (listed in Table 1). The included conditions were identified based on review of prior articles that assessed numerous immune-related conditions (17, 18) and through consultation with physician experts. The prevalence of the immune-related conditions ranged widely, from allergic rhinitis (diagnosed in 18.1% of initially selected controls, Table 1) and rheumatoid arthritis (3.9%) to Behçet's disease (<0.1%) and deficiency of cell-mediated immunity (<0.1%). Ten conditions that are not immune-related and lack known associations with skin cancer (“control conditions”) were also assessed for comparison. Each condition was identified based on the presence of one hospital claim, or two outpatient/provider claims more than 30 days apart (codes listed in Table 1). Procedure and diagnosis-related group codes were used in addition to diagnosis codes to identify solid organ transplantation and bone marrow transplantation.

Table 1.

ICD-9 codes used to identify immune-related and control conditions and frequency of conditions in source population for controls

ICD-9 codesN (%) in control source populationa
Immunosuppressive conditions 
 Deficiency of humoral immunity 279.0 20,256 (0.2) 
 HIV infection V08, 42.X-44.X 5,464 (0.1) 
 Deficiency of cell-mediated immunity 279.1 749 (<0.1) 
Transplant conditions 
 Solid organ transplantb V42.0-1, V42.6-7, V42.83-84, 996.81-84, 996.86-87 16,444 (0.2) 
 Bone marrow transplantc V42.81-V42.82, 996.85, 279.5 4,866 (<0.1) 
 Graft vs. host diseased 279.5 436 (<0.1) 
Autoimmune conditions 
 Rheumatoid arthritis 714.0 414,564 (3.9) 
 Pernicious anemia 281.0 320,960 (3.0) 
 Psoriasisd 696.0, 696.1 258,266 (2.4) 
 Polymyalgia rheumatica 725.X 200,019 (1.9) 
 Ulcerative colitis 556.X 104,833 (1.0) 
 Crohn's disease 555.X 61,599 (0.6) 
 Graves' disease 242.0 48,866 (0.5) 
 Sclerodermad 701.0, 710.1 51,178 (0.5) 
 Giant cell arteritis 446.5 48,866 (0.5) 
 Hashimoto's thyroiditis 245.2 47,529 (0.4) 
 Sjögren's syndrome 710.2 47,529 (0.4) 
 Systemic lupus erythematosusd 710.0 45,675 (0.4) 
 Addison's disease 255.41 25,810 (0.2) 
 Uveitis 364.3 24,550 (0.2) 
 Celiac disease 579.0 24,716 (0.2) 
 Immune thrombocytopenic purpurad 287.31 24,716 (0.2) 
 Multiple sclerosis 340.X 25,202 (0.2) 
 Myasthenia gravis 358.0 22,018 (0.2) 
 Sarcoidosis 135.X 19,481 (0.2) 
 Ankylosing spondylitis 720.0 18,455 (0.2) 
 Scleritis 379.0 18,057 (0.2) 
 Discoid lupus erythematosusd 695.4 16,422 (0.2) 
 Vitiligod 374.53, 709.01 14,850 (0.1) 
 Polymyositis/dermatomyositisd 710.3, 710.4 13,278 (0.1) 
 Guillain-Barré syndrome 357.0 11,946 (0.1) 
 Autoimmune hemolytic anemia 283.0 8,635 (0.1) 
 Primary biliary cirrhosis 571.6 7,734 (0.1) 
 Granulomatosis with polyangiitis 446.4 4,478 (<0.1) 
 Autoimmune hepatitis 571.42 3,450 (<0.1) 
 Polyarteritis nodosa 446.0 2,886 (<0.1) 
 Reactive arthritis 99.3 1,534 (<0.1) 
 Behçet's disease 136.1 375 (<0.1) 
Hematologic malignancies and related conditions 
 Non-Hodgkin lymphoma 200.X, 202.X, 204.1 171,963 (1.6) 
 Aplastic anemia 284.1, 284.8, 284.9 120,871 (1.1) 
 Paraproteinemias/related disorders 273.0–273.3 89,029 (0.8) 
 Multiple myeloma 203.0, 203.1 35,022 (0.3) 
 Leukemia 204.0, 204.2, 204.8, 205.X-207.X 22,537 (0.2) 
 Hodgkin lymphoma 201.X 12,789 (0.1) 
Allergic condition   
 Allergic rhinitis 477.X 1,929,244 (18.1) 
 Asthma 493.X 1,163,028 (10.9) 
 Atopic dermatitis/eczemad 691.8 164,642 (1.5) 
Control conditions 
 Esophageal reflux 530.81 3,386,297 (31.8) 
 Glaucoma 365.X 2,084,295 (19.6) 
 Acute sinusitis 461.X 1,553,964 (14.6) 
 Accidental fall E880-E888 787,340 (7.4) 
 Calculus of kidney 592.0 607,205 (5.7) 
 Dysthymic disorder 300.4 463,889 (4.4) 
 Tinea pedisd 110.4 301,930 (2.8) 
 Cholecystitis 574.1 199,896 (1.9) 
 Acute appendicitis 540.X 66,179 (0.6) 
 Motor vehicle traffic accident E810–819 27,919 (0.3) 
ICD-9 codesN (%) in control source populationa
Immunosuppressive conditions 
 Deficiency of humoral immunity 279.0 20,256 (0.2) 
 HIV infection V08, 42.X-44.X 5,464 (0.1) 
 Deficiency of cell-mediated immunity 279.1 749 (<0.1) 
Transplant conditions 
 Solid organ transplantb V42.0-1, V42.6-7, V42.83-84, 996.81-84, 996.86-87 16,444 (0.2) 
 Bone marrow transplantc V42.81-V42.82, 996.85, 279.5 4,866 (<0.1) 
 Graft vs. host diseased 279.5 436 (<0.1) 
Autoimmune conditions 
 Rheumatoid arthritis 714.0 414,564 (3.9) 
 Pernicious anemia 281.0 320,960 (3.0) 
 Psoriasisd 696.0, 696.1 258,266 (2.4) 
 Polymyalgia rheumatica 725.X 200,019 (1.9) 
 Ulcerative colitis 556.X 104,833 (1.0) 
 Crohn's disease 555.X 61,599 (0.6) 
 Graves' disease 242.0 48,866 (0.5) 
 Sclerodermad 701.0, 710.1 51,178 (0.5) 
 Giant cell arteritis 446.5 48,866 (0.5) 
 Hashimoto's thyroiditis 245.2 47,529 (0.4) 
 Sjögren's syndrome 710.2 47,529 (0.4) 
 Systemic lupus erythematosusd 710.0 45,675 (0.4) 
 Addison's disease 255.41 25,810 (0.2) 
 Uveitis 364.3 24,550 (0.2) 
 Celiac disease 579.0 24,716 (0.2) 
 Immune thrombocytopenic purpurad 287.31 24,716 (0.2) 
 Multiple sclerosis 340.X 25,202 (0.2) 
 Myasthenia gravis 358.0 22,018 (0.2) 
 Sarcoidosis 135.X 19,481 (0.2) 
 Ankylosing spondylitis 720.0 18,455 (0.2) 
 Scleritis 379.0 18,057 (0.2) 
 Discoid lupus erythematosusd 695.4 16,422 (0.2) 
 Vitiligod 374.53, 709.01 14,850 (0.1) 
 Polymyositis/dermatomyositisd 710.3, 710.4 13,278 (0.1) 
 Guillain-Barré syndrome 357.0 11,946 (0.1) 
 Autoimmune hemolytic anemia 283.0 8,635 (0.1) 
 Primary biliary cirrhosis 571.6 7,734 (0.1) 
 Granulomatosis with polyangiitis 446.4 4,478 (<0.1) 
 Autoimmune hepatitis 571.42 3,450 (<0.1) 
 Polyarteritis nodosa 446.0 2,886 (<0.1) 
 Reactive arthritis 99.3 1,534 (<0.1) 
 Behçet's disease 136.1 375 (<0.1) 
Hematologic malignancies and related conditions 
 Non-Hodgkin lymphoma 200.X, 202.X, 204.1 171,963 (1.6) 
 Aplastic anemia 284.1, 284.8, 284.9 120,871 (1.1) 
 Paraproteinemias/related disorders 273.0–273.3 89,029 (0.8) 
 Multiple myeloma 203.0, 203.1 35,022 (0.3) 
 Leukemia 204.0, 204.2, 204.8, 205.X-207.X 22,537 (0.2) 
 Hodgkin lymphoma 201.X 12,789 (0.1) 
Allergic condition   
 Allergic rhinitis 477.X 1,929,244 (18.1) 
 Asthma 493.X 1,163,028 (10.9) 
 Atopic dermatitis/eczemad 691.8 164,642 (1.5) 
Control conditions 
 Esophageal reflux 530.81 3,386,297 (31.8) 
 Glaucoma 365.X 2,084,295 (19.6) 
 Acute sinusitis 461.X 1,553,964 (14.6) 
 Accidental fall E880-E888 787,340 (7.4) 
 Calculus of kidney 592.0 607,205 (5.7) 
 Dysthymic disorder 300.4 463,889 (4.4) 
 Tinea pedisd 110.4 301,930 (2.8) 
 Cholecystitis 574.1 199,896 (1.9) 
 Acute appendicitis 540.X 66,179 (0.6) 
 Motor vehicle traffic accident E810–819 27,919 (0.3) 

NOTE: Conditions are sorted within categories by frequency.

aThe source population for controls includes white, non-Hispanic individuals in the U.S. Medicare population who were frequency matched by age and sex to the initial 467,948 SCC and 597,882 BCC cases in 2012 (N = 10,658,300). This population only included individuals with non-HMO coverage who had at least 1 Medicare claim in 2012.

bSolid organ transplants were also identified if any of the following claims were present: HCPCS codes 44136, 47135, 47136, 48160, 50360, 50365, 32851-32854, 33935, or 33945; DRG codes 302, 480, 495, 512, or 513 before January 1, 2007; or DRG codes 005-008, 010, or 652 after December 31, 2007.

cBone marrow transplants were also identified if any of the following claims were present: HCPCS codes of 38240 or 38241; DRG code 481 before January 1, 2007; or DRG code 009 after December 31, 2007.

dConditions characterized by skin manifestations.

We also assessed factors that could influence skin cancer risk or ascertainment. Frequencies of dermatologist and nondermatologist physician visits were calculated for 2010–2011 as measures of medical engagement before cancer diagnosis. Average annual ambient UVR was determined by linking the zip code of residence in 2012 to daily estimates of cloud-adjusted noon-time ultraviolet B radiation from a national database (19).

Associations of immune-related conditions with SCC and BCC were estimated using multivariable logistic regression. In preliminary analyses stratified by the number of dermatologist visits in 2010–2011 (not shown), many associations with immune-related conditions differed in direction and/or magnitude across strata, with stronger positive associations among people with no dermatologist visits, particularly for skin conditions. This could indicate differential surveillance of the skin during nondermatologist medical visits based on the presence or absence of an immune-related condition involving the skin. By contrast, individuals seen by a dermatologist would be expected to have some type of skin examination, and as a result more uniform ascertainment of skin cancers, regardless of whether an immune-related condition was present. To reduce the possibility of surveillance bias, we limited the population to people with at least one dermatologist visit in 2010–2011. In this population, we estimated associations adjusting for age, sex, UVR decile category, frequency of dermatologist visits (continuous variable), and frequency of nondermatologist medical visits (continuous variable). Associations were also estimated separately among people with and without a history of keratinocyte carcinoma diagnosis in Medicare before 2012.

Statistical significance was determined after Bonferroni correction for multiple comparisons. As 57 conditions (47 immune-related and 10 control conditions) were assessed with 2 outcomes (BCC and SCC), the Bonferroni-adjusted alpha level was 0.05/114 = 0.000439. All tests were two-sided.

Study population

We identified 467,948 SCC and 597,882 BCC cases among elderly Medicare beneficiaries in 2012. After limiting to people with at least 1 dermatologist visit in 2010–2011, 258,683 SCC cases and 304,903 BCC cases remained (Table 2, BCC:SCC ratio of 1.2:1). Among SCC cases, 61.7% were male and 33.2% were younger than 75 years of age. Among BCC cases, 63.1% were male and 38.7% were younger than 75 years of age. Among the controls frequency matched to SCC and BCC cases, 1,012,520 and 1,284,902, respectively, had at least 1 dermatologist visit in 2010–2011 and were included in analyses. The age and sex distributions of controls were similar to the distributions for cases (Table 2). In 1999, the first year of claims data evaluated for immune-related conditions, 50.6% of SCC cases and 50.1% of SCC controls were enrolled in Medicare, whereas 45.2% of BCC cases and 45.1% of BCC controls were in Medicare. About a third of SCC and BCC cases resided in areas with the highest quintile of annual UVR exposure, whereas only about a quarter of controls lived in these areas. Cases more frequently visited a dermatologist, with 73.1% of SCC cases and 69.5% of BCC cases having more than 1 dermatologist visit during 2010–2011, compared with 59.0% of SCC controls and 58.6% of BCC controls. Cases also had slightly more frequent nondermatologist physician visits (Table 2).

Table 2.

Characteristics of SCC cases, BCC cases, and corresponding controls in the Medicare population

SCCBCC
Cases (N = 258,683)Controls (N = 1,012,520)Cases (N = 304,903)Controls (N = 1,284,902)
N (%)N (%)N (%)N (%)
Sex 
 Male 159,715 (61.7) 640,174 (63.2) 192,354 (63.1) 813,954 (63.4) 
 Female 98,968 (38.3) 372,346 (36.8) 112,549 (36.9) 470,948 (36.7) 
Age on January 1, 2012, in years 
 65–69 33,741 (13.0) 121,145 (12.0) 50,046 (16.4) 195,051 (15.2) 
 70–74 52,206 (20.2) 206,047 (20.4) 68,030 (22.3) 289,745 (22.6) 
 75–79 57,778 (22.3) 233,710 (23.1) 68,921 (22.6) 299,601 (23.3) 
 80–84 58,452 (22.6) 236,169 (23.3) 64,065 (21.0) 278,012 (21.6) 
 85–89 40,402 (15.6) 156,078 (15.4) 39,687 (13.0) 165,579 (12.9) 
 90–94 16,104 (6.2) 59,371 (5.9) 14,154 (4.6) 56,914 (4.4) 
First year of available Medicare claims 
 1999 130,827 (50.6) 506,968 (50.1) 137,875 (45.2) 580,053 (45.1) 
 2000–2001 28,549 (11.0) 121,486 (12.0) 33,798 (11.1) 151,445 (11.8) 
 2002–2003 27,290 (10.6) 100,976 (10.0) 33,329 (10.9) 135,208 (10.5) 
 2004–2005 21,671 (8.4) 88,592 (8.8) 27,968 (9.2) 122,996 (9.6) 
 2006–2007 20,705 (8.0) 82,424 (8.1) 28,447 (9.3) 119,727 (9.3) 
 2008–2009 18,436 (7.1) 69,339 (6.9) 26,671 (8.8) 108,486 (8.4) 
 2010–2011 11,205 (4.3) 42,735 (4.2) 16,815 (5.5) 66,987 (5.2) 
Ambient UVR, mW/m2* 
 ≤24.77 30,594 (11.8) 166,659 (16.5) 43,605 (14.3) 209,845 (16.3) 
 24.78–27.92 40,306 (15.6) 198,456 (19.6) 50,301 (16.5) 249,967 (19.5) 
 27.93–34.73 40,813 (15.8) 177,549 (17.5) 50,979 (16.7) 226,237 (17.6) 
 34.74–43.64 53,034 (20.5) 212,608 (21.0) 63,943 (21.0) 271,764 (21.2) 
 43.65+ 93,936 (36.3) 257,248 (25.4) 96,075 (31.5) 327,089 (25.5) 
Dermatologist visits 
 1 69,500 (26.9) 415,052 (41.0) 92,968 (30.5) 532,382 (41.4) 
 2+ 189,183 (73.1) 597,468 (59.0) 211,935 (69.5) 752,520 (58.6) 
Nondermatologist physician visits 
 ≤3 28,079 (10.9) 124,579 (12.3) 39,509 (13.0) 167,164 (13.0) 
 4–7 52,679 (20.4) 225,610 (22.3) 68,922 (22.6) 293,277 (22.8) 
 8–11 48,564 (18.8) 202,580 (20.0) 59,747 (19.6) 257,314 (20.0) 
 12–19 56,916 (22.0) 222,327 (22.0) 65,937 (21.6) 277,824 (21.6) 
 20+ 72,445 (28.0) 237,424 (23.5) 70,788 (23.2) 289,323 (22.5) 
SCCBCC
Cases (N = 258,683)Controls (N = 1,012,520)Cases (N = 304,903)Controls (N = 1,284,902)
N (%)N (%)N (%)N (%)
Sex 
 Male 159,715 (61.7) 640,174 (63.2) 192,354 (63.1) 813,954 (63.4) 
 Female 98,968 (38.3) 372,346 (36.8) 112,549 (36.9) 470,948 (36.7) 
Age on January 1, 2012, in years 
 65–69 33,741 (13.0) 121,145 (12.0) 50,046 (16.4) 195,051 (15.2) 
 70–74 52,206 (20.2) 206,047 (20.4) 68,030 (22.3) 289,745 (22.6) 
 75–79 57,778 (22.3) 233,710 (23.1) 68,921 (22.6) 299,601 (23.3) 
 80–84 58,452 (22.6) 236,169 (23.3) 64,065 (21.0) 278,012 (21.6) 
 85–89 40,402 (15.6) 156,078 (15.4) 39,687 (13.0) 165,579 (12.9) 
 90–94 16,104 (6.2) 59,371 (5.9) 14,154 (4.6) 56,914 (4.4) 
First year of available Medicare claims 
 1999 130,827 (50.6) 506,968 (50.1) 137,875 (45.2) 580,053 (45.1) 
 2000–2001 28,549 (11.0) 121,486 (12.0) 33,798 (11.1) 151,445 (11.8) 
 2002–2003 27,290 (10.6) 100,976 (10.0) 33,329 (10.9) 135,208 (10.5) 
 2004–2005 21,671 (8.4) 88,592 (8.8) 27,968 (9.2) 122,996 (9.6) 
 2006–2007 20,705 (8.0) 82,424 (8.1) 28,447 (9.3) 119,727 (9.3) 
 2008–2009 18,436 (7.1) 69,339 (6.9) 26,671 (8.8) 108,486 (8.4) 
 2010–2011 11,205 (4.3) 42,735 (4.2) 16,815 (5.5) 66,987 (5.2) 
Ambient UVR, mW/m2* 
 ≤24.77 30,594 (11.8) 166,659 (16.5) 43,605 (14.3) 209,845 (16.3) 
 24.78–27.92 40,306 (15.6) 198,456 (19.6) 50,301 (16.5) 249,967 (19.5) 
 27.93–34.73 40,813 (15.8) 177,549 (17.5) 50,979 (16.7) 226,237 (17.6) 
 34.74–43.64 53,034 (20.5) 212,608 (21.0) 63,943 (21.0) 271,764 (21.2) 
 43.65+ 93,936 (36.3) 257,248 (25.4) 96,075 (31.5) 327,089 (25.5) 
Dermatologist visits 
 1 69,500 (26.9) 415,052 (41.0) 92,968 (30.5) 532,382 (41.4) 
 2+ 189,183 (73.1) 597,468 (59.0) 211,935 (69.5) 752,520 (58.6) 
Nondermatologist physician visits 
 ≤3 28,079 (10.9) 124,579 (12.3) 39,509 (13.0) 167,164 (13.0) 
 4–7 52,679 (20.4) 225,610 (22.3) 68,922 (22.6) 293,277 (22.8) 
 8–11 48,564 (18.8) 202,580 (20.0) 59,747 (19.6) 257,314 (20.0) 
 12–19 56,916 (22.0) 222,327 (22.0) 65,937 (21.6) 277,824 (21.6) 
 20+ 72,445 (28.0) 237,424 (23.5) 70,788 (23.2) 289,323 (22.5) 

*Ambient UVR is based on annual daily average noon-time ultraviolet B irradiance as measured by the National Aeronautics and Space Administration's Total Ozone Mapping Spectrometer.

Physician visits were counted during time covered by Medicare during 2010–2011.

Associations of immune-related conditions with SCC and BCC

After adjustment for age, sex, dermatologist and nondermatologist physician visit frequency, and annual UVR exposure, 21 (45%) conditions were significantly associated with elevated SCC risk, including all primary immunodeficiency conditions, transplant conditions, and hematologic malignancies (Table 3; Fig. 1). Only 28% of autoimmune conditions and no allergic conditions were associated with elevated risk. The conditions associated with the largest increases in SCC risk were solid organ transplantation [odds ratio = 5.35, 95% confidence interval (CI), 5.07–5.65; Table 3], graft versus host disease (OR = 2.58; 95% CI, 1.77–3.76), deficiency of cell-mediated immunity (OR = 2.09; 95% CI, 1.52–2.88), granulomatosis with polyangiitis (also known as Wegener's granulomatosis, OR = 1.88; 95% CI, 1.61–2.19), bone marrow transplant (OR = 1.84; 95% CI, 1.60–2.13), and autoimmune hepatitis (OR = 1.81; 95% CI, 1.52–2.16). The most prevalent conditions associated with increased SCC risk included rheumatoid arthritis (OR = 1.06; 95% CI, 1.04–1.09), non-Hodgkin lymphoma (OR = 1.62; 95% CI, 1.58–1.66), and aplastic anemia (OR =1.27; 95% CI, 1.22–1.31). Two of the three allergic conditions and 28% of autoimmune conditions (N = 9) were associated with lower risk (Fig. 1). Of the 7 autoimmune conditions characterized by skin manifestations (marked in Table 1), 5 were associated with lower SCC risk. Only one control condition (10%), acute appendicitis, was associated with elevated risk, and the elevation was smaller than for most of the immune-related conditions (OR = 1.10; Fig. 1; Table 3). Half of the control conditions were associated with lower SCC risk (Fig. 1).

Table 3.

Associations of immune-related conditions and control conditions with cutaneous SCC and BCC

SCCBCC
OR (95% CI)P valueOR (95% CI)P value
Immunosuppressive conditions 
 Deficiency of humoral immunity 1.51 (1.41–1.61) 2.93E–32* 1.22 (1.14–1.31) 7.34E–09* 
 HIV infection 1.34 (1.15–1.57) 2.28E–04* 1.31 (1.14–1.51) 2.10E–04* 
 Deficiency of cell-mediated immunity 2.09 (1.52–2.88) 6.72E–06* 1.25 (0.89–1.75) 1.95E–01 
Transplant conditions 
 Solid organ transplant 5.35 (5.07–5.65) <1.0E–100* 1.94 (1.82–2.06) 2.16E–99* 
 Bone marrow transplant 1.84 (1.60–2.13) 5.88E–17* 1.38 (1.20–1.58) 5.25E–06* 
 Graft vs. host disease 2.58 (1.77–3.76) 7.41E–07* 1.44 (0.97–2.14) 6.76E–02 
Autoimmune conditions 
 Rheumatoid arthritis 1.06 (1.04–1.09) 3.50E–09* 0.99 (0.97–1.01) 2.40E–01 
 Pernicious anemia 1.02 (1.00–1.05) 6.70E–02 0.95(0.92–0.97) 4.62E–06* 
 Psoriasis 0.85 (0.84–0.87) 5.48E–51* 0.72 (0.71–0.74) <1.0E–100* 
 Polymyalgia rheumatica 1.04 (1.02–1.07) 2.45E–03 1.03 (1.00–1.06) 5.08E–02 
 Ulcerative colitis 1.07 (1.03–1.11) 3.37E–04* 1.03 (0.99–1.07) 1.06E–01 
 Crohn's disease 1.33 (1.27–1.39) 1.19E–32* 1.10 (1.05–1.15) 5.47E–05* 
 Graves' disease 0.84 (0.79–0.90) 2.38E–08* 0.87 (0.83–0.92) 1.55E–06* 
 Scleroderma 0.84 (0.80–0.89) 4.84E–10* 0.78 (0.74–0.82) 1.26E–21* 
 Giant cell arteritis 1.02 (0.97–1.08) 4.07E–01 1.02 (0.96–1.07) 5.38E–01 
 Hashimoto's thyroiditis 0.79 (0.75–0.84) 4.60E–16* 0.86 (0.82–0.91) 6.62E–09* 
 Sjögren's syndrome 0.90 (0.85–0.95) 2.30E–04* 0.86 (0.81–0.90) 1.91E–08* 
 Systemic lupus erythematosus 0.89 (0.83–0.94) 8.00E–05* 0.77 (0.73–0.82) 1.61E–17* 
 Addison's disease 1.16 (1.07–1.25) 4.09E–04* 0.96 (0.88–1.04) 2.65E–01 
 Uveitis 1.00 (0.92–1.09) 9.28E–01 0.95 (0.88–1.03) 1.87E–01 
 Celiac disease 0.95 (0.88–1.03) 2.17E–01 0.98 (0.91–1.06) 6.32E–01 
 Immune thrombocytopenic purpura 1.23 (1.14–1.33) 5.01E–08* 1.01 (0.93–1.09) 8.86E–01 
 Multiple sclerosis 0.82 (0.74–0.91) 1.77E–04* 0.80 (0.73–0.88) 3.03E–06* 
 Myasthenia gravis 1.16 (1.07–1.25) 4.91E–04 1.00 (0.92–1.08) 9.63E–01 
 Sarcoidosis 1.12 (1.02–1.21) 1.14E–02 1.02 (0.94–1.10) 7.17E–01 
 Ankylosing spondylitis 1.04 (0.95–1.14) 4.02E–01 1.04 (0.96–1.13) 3.25E–01 
 Scleritis 0.97 (0.89–1.06) 5.46E–01 1.00 (0.92–1.08) 9.96E–01 
 Discoid lupus erythematosus 0.64 (0.58–0.71) 7.42E–19* 0.63 (0.57–0.69) 2.52E–22* 
 Vitiligo 0.83 (0.77–0.90) 8.84E–06* 0.66 (0.60–0.71) 6.30E–23* 
 Polymyositis/dermatomyositis 1.00 (0.90–1.10) 9.25E–01 0.95 (0.86–1.05) 3.40E–01 
 Guillain-Barré syndrome 1.01 (0.90–1.14) 8.75E–01 0.90 (0.80–1.01) 7.26E–02 
 Autoimmune hemolytic anemia 1.64 (1.47–1.84) 5.34E–18* 1.18 (1.04–1.33) 9.63E–03 
 Primary biliary cirrhosis 1.16 (1.01–1.33) 3.55E–02 0.86 (0.74–0.99) 3.35E–02 
 Granulomatosis with polyangiitis 1.88 (1.61–2.19) 9.13E–16* 1.30 (1.11–1.53) 1.14E–03 
 Autoimmune hepatitis 1.81 (1.52–2.16) 4.77E–11* 1.08 (0.88–1.32) 4.51E–01 
 Polyarteritis nodosa 1.54 (1.25–1.89) 4.50E–05* 1.22 (0.99–1.49) 5.71E–02 
 Reactive arthritis 1.11 (0.84–1.48) 4.55E–01 1.21 (0.93–1.58) 1.47E–01 
 Behçet's disease 0.95 (0.50–1.82) 8.85E–01 0.86 (0.47–1.57) 6.16E–01 
Hematologic malignancies and related conditions     
 Non-Hodgkin lymphoma 1.62 (1.58–1.66) <1.0E–100* 1.25 (1.22–1.28) 2.62E–69* 
 Aplastic anemia 1.27 (1.22–1.31) 6.67E–38* 1.02 (0.98–1.06) 3.25E–01 
 Paraproteinemias/related disorders 1.15 (1.11–1.20) 1.72E–13* 1.11 (1.07–1.15) 4.95E–08* 
 Multiple myeloma 1.30 (1.22–1.38) 8.30E–17* 1.04 (0.98–1.11) 2.08E–01 
 Leukemia 1.59 (1.48–1.70) 3.16E–37* 1.15 (1.06–1.23) 2.94E–04* 
 Hodgkin lymphoma 1.61 (1.47–1.76) 5.41E–24* 1.43 (1.31–1.57) 2.11E–15* 
Allergic condition     
 Allergic Rhinitis 0.91 (0.90–0.92) 3.31E–63* 0.93 (0.92–0.94) 1.41E–42* 
 Asthma 0.98 (0.97–0.99) 2.36E–03 0.92 (0.91–0.93) 8.54E–38* 
 Atopic dermatitis/eczema 0.83 (0.81–0.85) 6.87E–46* 0.73 (0.72–0.75) <1.0E–100* 
Control conditions 
 Esophageal reflux 0.97 (0.96–0.98) 6.89E–09* 0.94 (0.93–0.95) 2.73E–45* 
 Glaucoma 0.99 (0.98–1.00) 4.38E–02 0.99 (0.98–1.00) 3.97E–02 
 Acute sinusitis 0.98 (0.97–0.99) 2.50E–04* 0.97 (0.96–0.98) 1.56E–07* 
 Accidental fall 0.96 (0.94–0.98) 8.02E–06* 0.89 (0.88–0.91) 1.43E–36* 
 Calculus of kidney 1.02 (1.00–1.04) 1.44E–02 1.04 (1.02–1.05) 4.50E–06* 
 Dysthymic disorder 0.86 (0.84–0.88) 2.56E–36* 0.85 (0.83–0.87) 8.87E–46* 
 Tinea pedis 0.91 (0.89–0.93) 1.41E–17* 0.87 (0.85–0.88) 2.95E–44* 
 Cholecystitis 0.97 (0.94–1.00) 4.37E–02 0.98 (0.95–1.01) 2.26E–01 
 Acute appendicitis 1.10 (1.05–1.16) 1.89E–04* 1.05 (1.00–1.10) 6.63E–02 
 Motor vehicle traffic accident 0.89 (0.82–0.97) 1.12E–02 1.00 (0.92–1.08) 9.73E–01 
SCCBCC
OR (95% CI)P valueOR (95% CI)P value
Immunosuppressive conditions 
 Deficiency of humoral immunity 1.51 (1.41–1.61) 2.93E–32* 1.22 (1.14–1.31) 7.34E–09* 
 HIV infection 1.34 (1.15–1.57) 2.28E–04* 1.31 (1.14–1.51) 2.10E–04* 
 Deficiency of cell-mediated immunity 2.09 (1.52–2.88) 6.72E–06* 1.25 (0.89–1.75) 1.95E–01 
Transplant conditions 
 Solid organ transplant 5.35 (5.07–5.65) <1.0E–100* 1.94 (1.82–2.06) 2.16E–99* 
 Bone marrow transplant 1.84 (1.60–2.13) 5.88E–17* 1.38 (1.20–1.58) 5.25E–06* 
 Graft vs. host disease 2.58 (1.77–3.76) 7.41E–07* 1.44 (0.97–2.14) 6.76E–02 
Autoimmune conditions 
 Rheumatoid arthritis 1.06 (1.04–1.09) 3.50E–09* 0.99 (0.97–1.01) 2.40E–01 
 Pernicious anemia 1.02 (1.00–1.05) 6.70E–02 0.95(0.92–0.97) 4.62E–06* 
 Psoriasis 0.85 (0.84–0.87) 5.48E–51* 0.72 (0.71–0.74) <1.0E–100* 
 Polymyalgia rheumatica 1.04 (1.02–1.07) 2.45E–03 1.03 (1.00–1.06) 5.08E–02 
 Ulcerative colitis 1.07 (1.03–1.11) 3.37E–04* 1.03 (0.99–1.07) 1.06E–01 
 Crohn's disease 1.33 (1.27–1.39) 1.19E–32* 1.10 (1.05–1.15) 5.47E–05* 
 Graves' disease 0.84 (0.79–0.90) 2.38E–08* 0.87 (0.83–0.92) 1.55E–06* 
 Scleroderma 0.84 (0.80–0.89) 4.84E–10* 0.78 (0.74–0.82) 1.26E–21* 
 Giant cell arteritis 1.02 (0.97–1.08) 4.07E–01 1.02 (0.96–1.07) 5.38E–01 
 Hashimoto's thyroiditis 0.79 (0.75–0.84) 4.60E–16* 0.86 (0.82–0.91) 6.62E–09* 
 Sjögren's syndrome 0.90 (0.85–0.95) 2.30E–04* 0.86 (0.81–0.90) 1.91E–08* 
 Systemic lupus erythematosus 0.89 (0.83–0.94) 8.00E–05* 0.77 (0.73–0.82) 1.61E–17* 
 Addison's disease 1.16 (1.07–1.25) 4.09E–04* 0.96 (0.88–1.04) 2.65E–01 
 Uveitis 1.00 (0.92–1.09) 9.28E–01 0.95 (0.88–1.03) 1.87E–01 
 Celiac disease 0.95 (0.88–1.03) 2.17E–01 0.98 (0.91–1.06) 6.32E–01 
 Immune thrombocytopenic purpura 1.23 (1.14–1.33) 5.01E–08* 1.01 (0.93–1.09) 8.86E–01 
 Multiple sclerosis 0.82 (0.74–0.91) 1.77E–04* 0.80 (0.73–0.88) 3.03E–06* 
 Myasthenia gravis 1.16 (1.07–1.25) 4.91E–04 1.00 (0.92–1.08) 9.63E–01 
 Sarcoidosis 1.12 (1.02–1.21) 1.14E–02 1.02 (0.94–1.10) 7.17E–01 
 Ankylosing spondylitis 1.04 (0.95–1.14) 4.02E–01 1.04 (0.96–1.13) 3.25E–01 
 Scleritis 0.97 (0.89–1.06) 5.46E–01 1.00 (0.92–1.08) 9.96E–01 
 Discoid lupus erythematosus 0.64 (0.58–0.71) 7.42E–19* 0.63 (0.57–0.69) 2.52E–22* 
 Vitiligo 0.83 (0.77–0.90) 8.84E–06* 0.66 (0.60–0.71) 6.30E–23* 
 Polymyositis/dermatomyositis 1.00 (0.90–1.10) 9.25E–01 0.95 (0.86–1.05) 3.40E–01 
 Guillain-Barré syndrome 1.01 (0.90–1.14) 8.75E–01 0.90 (0.80–1.01) 7.26E–02 
 Autoimmune hemolytic anemia 1.64 (1.47–1.84) 5.34E–18* 1.18 (1.04–1.33) 9.63E–03 
 Primary biliary cirrhosis 1.16 (1.01–1.33) 3.55E–02 0.86 (0.74–0.99) 3.35E–02 
 Granulomatosis with polyangiitis 1.88 (1.61–2.19) 9.13E–16* 1.30 (1.11–1.53) 1.14E–03 
 Autoimmune hepatitis 1.81 (1.52–2.16) 4.77E–11* 1.08 (0.88–1.32) 4.51E–01 
 Polyarteritis nodosa 1.54 (1.25–1.89) 4.50E–05* 1.22 (0.99–1.49) 5.71E–02 
 Reactive arthritis 1.11 (0.84–1.48) 4.55E–01 1.21 (0.93–1.58) 1.47E–01 
 Behçet's disease 0.95 (0.50–1.82) 8.85E–01 0.86 (0.47–1.57) 6.16E–01 
Hematologic malignancies and related conditions     
 Non-Hodgkin lymphoma 1.62 (1.58–1.66) <1.0E–100* 1.25 (1.22–1.28) 2.62E–69* 
 Aplastic anemia 1.27 (1.22–1.31) 6.67E–38* 1.02 (0.98–1.06) 3.25E–01 
 Paraproteinemias/related disorders 1.15 (1.11–1.20) 1.72E–13* 1.11 (1.07–1.15) 4.95E–08* 
 Multiple myeloma 1.30 (1.22–1.38) 8.30E–17* 1.04 (0.98–1.11) 2.08E–01 
 Leukemia 1.59 (1.48–1.70) 3.16E–37* 1.15 (1.06–1.23) 2.94E–04* 
 Hodgkin lymphoma 1.61 (1.47–1.76) 5.41E–24* 1.43 (1.31–1.57) 2.11E–15* 
Allergic condition     
 Allergic Rhinitis 0.91 (0.90–0.92) 3.31E–63* 0.93 (0.92–0.94) 1.41E–42* 
 Asthma 0.98 (0.97–0.99) 2.36E–03 0.92 (0.91–0.93) 8.54E–38* 
 Atopic dermatitis/eczema 0.83 (0.81–0.85) 6.87E–46* 0.73 (0.72–0.75) <1.0E–100* 
Control conditions 
 Esophageal reflux 0.97 (0.96–0.98) 6.89E–09* 0.94 (0.93–0.95) 2.73E–45* 
 Glaucoma 0.99 (0.98–1.00) 4.38E–02 0.99 (0.98–1.00) 3.97E–02 
 Acute sinusitis 0.98 (0.97–0.99) 2.50E–04* 0.97 (0.96–0.98) 1.56E–07* 
 Accidental fall 0.96 (0.94–0.98) 8.02E–06* 0.89 (0.88–0.91) 1.43E–36* 
 Calculus of kidney 1.02 (1.00–1.04) 1.44E–02 1.04 (1.02–1.05) 4.50E–06* 
 Dysthymic disorder 0.86 (0.84–0.88) 2.56E–36* 0.85 (0.83–0.87) 8.87E–46* 
 Tinea pedis 0.91 (0.89–0.93) 1.41E–17* 0.87 (0.85–0.88) 2.95E–44* 
 Cholecystitis 0.97 (0.94–1.00) 4.37E–02 0.98 (0.95–1.01) 2.26E–01 
 Acute appendicitis 1.10 (1.05–1.16) 1.89E–04* 1.05 (1.00–1.10) 6.63E–02 
 Motor vehicle traffic accident 0.89 (0.82–0.97) 1.12E–02 1.00 (0.92–1.08) 9.73E–01 

NOTE: The model is restricted to subjects with 1+ dermatologist visits in 2010–2011, and is adjusted for age, sex, ultraviolet B decile, dermatologist visits, and nondermatologist physician visits.

*Statistically significant P value < 0.000439.

Figure 1.

Associations of immune-related conditions and controls conditions with SCC of the skin. Each point represents the association between a condition and SCC risk among subjects with at least 1 dermatology visit, adjusted for age, sex, dermatologist visits, nondermatologist physician visits, and ultraviolet B exposure based on residence. The dotted horizontal line represents P = 0.000439, the Bonferroni threshold for statistical significance. As the y-axis shows P values plotted on the –log10 scale, associations appearing above the dotted line are statistically significant.

Figure 1.

Associations of immune-related conditions and controls conditions with SCC of the skin. Each point represents the association between a condition and SCC risk among subjects with at least 1 dermatology visit, adjusted for age, sex, dermatologist visits, nondermatologist physician visits, and ultraviolet B exposure based on residence. The dotted horizontal line represents P = 0.000439, the Bonferroni threshold for statistical significance. As the y-axis shows P values plotted on the –log10 scale, associations appearing above the dotted line are statistically significant.

Close modal

In adjusted models, 9 (19%) conditions were associated with higher BCC risk, including 2 (67%) primary immunodeficiency conditions, 2 (67%) transplant conditions, 4 (67%) hematologic malignancies, and 1 (3%) autoimmune conditions (Fig. 2). Conditions associated with the largest increases in BCC risk were solid organ transplantation (OR = 1.94; 95% CI, 1.82–2.06), Hodgkin lymphoma (OR = 1.43; 95% CI, 1.31–1.57), bone marrow transplant (OR = 1.38; 95% CI, 1.20–1.58), HIV (OR, 1.31; 95% CI, 1.14–1.51), and non-Hodgkin lymphoma (OR = 1.25; 95% CI, 1.22–1.28; Table 3). The most prevalent conditions associated with increased BCC risk included non-Hodgkin lymphoma, paraproteinemias/related disorders (OR = 1.11; 95% CI, 1.07–1.15), and Crohn's disease (OR = 1.10; 95% CI, 1.05–1.15). All allergic conditions and 31% of autoimmune conditions were significantly associated with lower risk (Fig. 2). Of the 7 autoimmune conditions largely characterized by skin manifestations, 5 were associated with lower BCC risk. One control condition (10%), calculus of the kidney, was associated with elevated risk, but the elevation was smaller than for many immune-related conditions (OR = 1.04, Table 3). Half of control conditions were associated with lower BCC risk (Fig. 2).

Figure 2.

Associations of immune-related conditions and controls conditions with BCC of the skin. Each point represents the association between a condition and BCC risk among subjects with at least 1 dermatology visit, adjusted for age, sex, dermatologist visits, nondermatologist physician visits, and ultraviolet B exposure based on residence. The dotted horizontal line represents P = 0.000439, the Bonferroni threshold for statistical significance. As the y-axis is plotted on the –log10 scale, associations appearing above this dotted line are statistically significant.

Figure 2.

Associations of immune-related conditions and controls conditions with BCC of the skin. Each point represents the association between a condition and BCC risk among subjects with at least 1 dermatology visit, adjusted for age, sex, dermatologist visits, nondermatologist physician visits, and ultraviolet B exposure based on residence. The dotted horizontal line represents P = 0.000439, the Bonferroni threshold for statistical significance. As the y-axis is plotted on the –log10 scale, associations appearing above this dotted line are statistically significant.

Close modal

Overall, 40 immune-related conditions (85%) manifested associations with SCC and BCC in the same direction, though associations with SCC were frequently more positive (Fig. 3). Among 24 immune-related conditions positively associated with BCC (ORs greater than 1.00), 22 (92%) had stronger associations with SCC. Among the 23 immune-related conditions negatively associated with BCC (ORs less than 1.00), 18 (78%) had associations with SCC that were more positive, i.e., closer to the null or ORs greater than 1.00.

Figure 3.

Comparison of ORs for associations of immune-related conditions with SCC and BCC. Each point represents a condition and its associations with SCC and BCC risk based on models for people with at least one dermatology visit in 2010–2011, adjusted for age, sex, dermatologist visits, nondermatologist physician visits, and ultraviolet B exposure based on residence. ORs are depicted on logarithmic scales. The solid diagonal line represents where conditions would be if associations with SCC and BCC were equal. Conditions above and to the left of the solid line have a more positive association with SCC than BCC. Conditions below and to the right of the solid line have a more positive relationship with BCC than SCC.

Figure 3.

Comparison of ORs for associations of immune-related conditions with SCC and BCC. Each point represents a condition and its associations with SCC and BCC risk based on models for people with at least one dermatology visit in 2010–2011, adjusted for age, sex, dermatologist visits, nondermatologist physician visits, and ultraviolet B exposure based on residence. ORs are depicted on logarithmic scales. The solid diagonal line represents where conditions would be if associations with SCC and BCC were equal. Conditions above and to the left of the solid line have a more positive association with SCC than BCC. Conditions below and to the right of the solid line have a more positive relationship with BCC than SCC.

Close modal

Associations among people with and without a personal history of keratinocyte carcinoma

When we stratified analyses by history of a prior keratinocyte carcinoma diagnosis, most associations were similar in magnitude (Supplementary Table S1). Of the 21 immune-related conditions associated with elevated SCC risk and the 9 immune-related conditions associated with elevated BCC risk in the primary analysis, 20 (95%) and 8 (89%) of these conditions, respectively, had ORs greater than 1.00 among people with and without a prior keratinocyte carcinoma diagnosis, though many were not statistically significant in both strata. Associations of ulcerative colitis with SCC and paraproteinemia with BCC were observed only among people with a prior keratinocyte carcinoma.

In our study, we found that many immune-related conditions were associated with elevated risk of keratinocyte carcinoma, especially SCC. Transplant conditions, primary immunodeficiency conditions, and hematologic malignancies were most consistently associated with higher risks. Most immune-related conditions were associated with greater elevations in SCC than BCC risk, suggesting that immune status may be especially important in the etiology of SCC. Moreover, associations with immune-related conditions were stronger and more frequent than those for control conditions, pointing to unique contributions of immune system dysfunction in keratinocyte carcinoma etiology. Mechanisms underlying associations with immune-related conditions could include effects of disease processes, specific treatments, behavioral changes in response to disease, or common risk factors.

Solid organ transplant recipients are the most well-known example of an immunosuppressed population with high skin cancer risk, with prior studies estimating SCC risk 40 to 120 times higher and BCC risk 6 to 10 times higher than in the general population (3–5). These elevations are larger than we observed, but these studies did not account for differences in medical visit attendance and included younger individuals who have low keratinocyte carcinoma risk in the absence of transplantation. Transplant recipients use immunosuppressant medications to prevent organ rejection (3, 5). Some immunosuppressants have documented photosensitizing properties, and so elevations in keratinocyte carcinoma risk may not be solely due to immunosuppression (9, 10, 20, 21).

Non-Hodgkin lymphoma (including chronic lymphocytic leukemia) has been associated with keratinocyte carcinoma risk in prior studies (7, 22, 23), and keratinocyte carcinoma diagnosis is also associated with subsequent non-Hodgkin lymphoma risk (23, 24). This reciprocal pattern suggests a common underlying cause for both cancer types, such as a shared genetic predisposition or underlying immune dysfunction. We observed strong increases in keratinocyte carcinoma risk for bone marrow transplant recipients and specifically recipients with graft versus host disease, an immunologic complication of bone marrow transplantation. Importantly, some immunosuppressants used in solid organ transplantation are also used to treat graft versus host disease as well as autoimmune conditions (25–29). As a result, effects of these medications, either through increases in immune dysfunction or photosensitization, could explain the elevated keratinocyte carcinoma risk associated with a number of different conditions.

Of note, we identified a number of novel associations with elevated keratinocyte carcinoma risk for conditions that had not previously been studied. For instance, granulomatosis with polyangiitis, autoimmune hepatitis, and polyarteritis nodosa were strongly associated with SCC risk. All are severe autoimmune diseases that require treatment with immunosuppressants (30–32). SCC risk was also strongly elevated in people with autoimmune hemolytic anemia, an autoimmune disorder that can require immunosuppressive treatment or splenectomy, and which can be secondary to another autoimmune disease (33). Deficiencies of cell-mediated and humoral immunity were associated with keratinocyte carcinoma risk. As captured by ICD-9 codes, these conditions are somewhat vague but would presumably include inherited defects in the immune system as well as immunodeficiency secondary to another disease, such as HIV. Associations with these conditions, in which immunodeficiency is caused by the disease itself rather than by treatment, support an etiologic mechanism driven by immune dysfunction independent of photosensitizing medication effects.

For clinical populations with heightened keratinocyte carcinoma risk, an increased emphasis on skin cancer screening and prevention may be appropriate. Importantly, we found consistent elevations regardless of individuals' prior history of keratinocyte carcinoma. It may be particularly important to target patients without a prior history, who might not be recognized as having a high risk for keratinocyte carcinoma. For solid organ transplant recipients, skin cancer prevention and screening measures are already considered in long-term management (13–15). Although elevations in keratinocyte carcinoma risk for other conditions were not as high as those observed for solid organ transplantation, an evaluation of the costs and benefits of increased skin cancer screening would be appropriate for some of these other clinical populations, especially for people who might not otherwise see a dermatologist.

We also observed conditions associated with decreased keratinocyte carcinoma risk, particularly autoimmune conditions with dermatologic manifestations and allergic conditions. Autoimmune and allergic diseases are characterized by an overly reactive immune system. For allergic diseases, this overreactivity is typically not severe enough to require use of systemic immunosuppressant medications (34), and so the immune system can largely maintain normal functioning. Instead, a highly reactive immune system might help eliminate developing tumor cells (35). Although research on these relationships is limited, there is some prior biologic and epidemiologic evidence indicating that allergic responses may be protective against keratinocyte carcinoma development (35, 36).

Behavioral differences related to UVR exposure may affect some associations with keratinocyte carcinoma risk. Many of the inverse associations were for conditions with primary skin manifestations, such as scleroderma and discoid lupus. Patients with these conditions might more frequently cover their skin to hide lesions. Other diseases, such as systemic lupus erythematosus and allergic rhinitis, might be exacerbated by sunlight exposure or time outdoors (37, 38). Multiple sclerosis, which was associated with lower keratinocyte carcinoma risk, can considerably decrease mobility (39) and consequently reduce the amount of time outdoors.

Overall, many associations appeared stronger for SCC than BCC, which may indicate that mechanisms specific to SCC development are preferentially influenced by immune dysfunction. Even for conditions associated with lower risk, most associations with SCC were closer to the null. This pattern could reflect a combination of behavioral effects that reduce risk of both SCC and BCC, with immune defects that most strongly increase SCC risk.

This study has some limitations. Information was not available on factors that influence keratinocyte carcinoma risk, such as sunburn history, time spent outdoors, and sunscreen use. While we accounted for differences in geographic UVR exposure based on zip codes of residence in 2012, these may not represent long-term residences and individuals may have spent substantial time in areas with different levels of UVR exposure prior to 2012. Also, we were only able to capture conditions that appeared in claims when people were 65 years of age or older, and these findings may not generalize to younger populations. Finally, many keratinocyte carcinomas go undiagnosed or untreated, and these were not captured in our study. However, specificity influences the bias observed in relative effect measures, such as ORs, more strongly than sensitivity (40). Our keratinocyte carcinoma definition should provide high specificity because it would be unlikely that individuals without a keratinocyte carcinoma would receive a keratinocyte carcinoma diagnosis and treatment. While unlikely to bias our ORs, our use of this definition could explain the low ratio of BCCs to SCCs, as BCCs are perhaps more likely to go untreated among elderly adults. A prior study of keratinocyte carcinoma incidence in the Medicare population observed a similar BCC:SCC ratio, noting that the ratio could reflect different patterns across calendar time and age groups (1).

This study also has important strengths. We included an exceptionally large population of cases and controls sampled in a representative manner from the Medicare population. Consequently, we were able to examine a wide array of conditions, including very rare conditions. The level of statistical significance for many associations was very strong (i.e., with P values less than 10−10). We differentiated between keratinocyte carcinoma types, which allowed us to demonstrate that many associations were stronger for SCC than BCC. We also accounted for the frequency of dermatologist and nondermatologist physician visits to reduce differences in surveillance that affect keratinocyte carcinoma detection. Finally, we assessed 10 control conditions and demonstrated that the strong associations with immune-related conditions did not merely reflect an artifact of our study design.

In conclusion, our findings strongly support an etiologic role for immunosuppression in keratinocyte carcinoma development across a broad spectrum of immune-related conditions. Almost all conditions associated with elevated risk were either directly immunosuppressive or treated with immunosuppressive medications. Associations were consistently stronger for SCC, indicating that immunologic dysfunction is likely of particular importance for this keratinocyte carcinoma type. These findings support prioritizing research focused on understanding the role of immune dysfunction in the etiology of keratinocyte carcinoma. We also found a few conditions associated with lower keratinocyte carcinoma risk, which could be due to protective effects of an overactive immune system or behaviors that reduce UVR exposure. For novel associations identified here, future research could help elucidate mechanisms by examining how keratinocyte carcinoma risk is influenced by severity of disease, medication use, and UVR exposure. Ultimately, research along these lines will help inform skin cancer screening and prevention for high risk clinical populations.

S.T. Arron reports receiving commercial research grant from Genentech and is a consultant/advisory board member for Gerson Lehrman Group and Portola Pharma. No potential conflicts of interest were disclosed by the other authors.

Conception and design: E.L. Yanik, E.A. Engels

Development of methodology: E.L. Yanik, S.T. Arron, E.A. Engels

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): E.K. Cahoon, M.K. Connolly, E.A. Engels

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): E.L. Yanik, R.M. Pfeiffer, M.A. Weinstock, S.T. Arron, M. Chaloux, M.K. Connolly, P. Nagarajan, E.A. Engels

Writing, review, and/or revision of the manuscript: E.L. Yanik, R.M. Pfeiffer, M.A. Weinstock, E.K. Cahoon, S.T. Arron, M. Chaloux, M.K. Connolly, P. Nagarajan, E.A. Engels

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): M. Chaloux

Study supervision: E.L. Yanik, E.A. Engels

We would like to thank individuals in the Keratinocyte Carcinoma Consortium (Keracon) for useful discussions and input.

E.L. Yanik, R.M. Pfeiffer, D.M. Freedman, E.K. Cahoon, and E.A. Engels were supported by the Intramural Research Program of the NCI.

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.

1.
Rogers
HW
,
Weinstock
MA
,
Feldman
SR
,
Coldiron
BM
. 
Incidence estimate of nonmelanoma skin cancer (keratinocyte carcinomas) in the U.S. population, 2012
.
JAMA Dermatol
2015
;
151
:
1081
6
.
2.
Gloster
HM
 Jr.
,
Neal
K
. 
Skin cancer in skin of color
.
J Am Acad Dermatol
2006
;
55
:
741
60
;
quiz 61–4
.
3.
Jensen
AO
,
Svaerke
C
,
Farkas
D
,
Pedersen
L
,
Kragballe
K
,
Sorensen
HT
. 
Skin cancer risk among solid organ recipients: A nationwide cohort study in Denmark
.
Acta Derm Venereol
2010
;
90
:
474
9
.
4.
Bannon
FJ
,
McCaughan
JA
,
Traynor
C
,
O'Brien
K
,
Gavin
AT
,
Maxwell
AP
, et al
Surveillance of nonmelanoma skin cancer incidence rates in kidney transplant recipients in Ireland
.
Transplantation
2014
;
98
:
646
52
.
5.
Krynitz
B
,
Edgren
G
,
Lindelof
B
,
Baecklund
E
,
Brattstrom
C
,
Wilczek
H
, et al
Risk of skin cancer and other malignancies in kidney, liver, heart and lung transplant recipients 1970 to 2008–a Swedish population-based study
.
Int J Cancer
2013
;
132
:
1429
38
.
6.
Silverberg
MJ
,
Leyden
W
,
Warton
EM
,
Quesenberry
CP
 Jr.
,
Engels
EA
,
Asgari
MM
. 
HIV infection status, immunodeficiency, and the incidence of non-melanoma skin cancer
.
J Natl Cancer Inst
2013
;
105
:
350
60
.
7.
Brewer
JD
,
Habermann
TM
,
Shanafelt
TD
. 
Lymphoma-associated skin cancer: Incidence, natural history, and clinical management
.
Int J Dermatol
2014
;
53
:
267
74
.
8.
DePry
JL
,
Vyas
R
,
Lazarus
HM
,
Caimi
PF
,
Gerstenblith
MR
,
Bordeaux
JS
. 
Cutaneous malignant neoplasms in hematopoietic cell transplant recipients: A systematic review
.
JAMA Dermatol
2015
;
151
:
775
82
.
9.
Jiyad
Z
,
Olsen
CM
,
Burke
MT
,
Isbel
NM
,
Green
AC
. 
Azathioprine and risk of skin cancer in organ transplant recipients: Systematic review and meta-analysis
.
Am J Transplant
2016
May 10. [Epub ahead of print].
10.
Yanik
EL
,
Siddiqui
K
,
Engels
EA
. 
Sirolimus effects on cancer incidence after kidney transplantation: A meta-analysis
.
Cancer Med
2015
;
4
:
1448
59
.
11.
Herman
M
,
Weinstein
T
,
Korzets
A
,
Chagnac
A
,
Ori
Y
,
Zevin
D
, et al
Effect of cyclosporin A on DNA repair and cancer incidence in kidney transplant recipients
.
The J Lab Clin Med
2001
;
137
:
14
20
.
12.
Scott
FI
,
Mamtani
R
,
Brensinger
CM
,
Haynes
K
,
Chiesa-Fuxench
ZC
,
Zhang
J
, et al
Risk of nonmelanoma skin cancer associated with the use of immunosuppressant and biologic agents in patients with a history of autoimmune disease and nonmelanoma skin cancer
.
JAMA Dermatol
2016
;
152
:
164
72
.
13.
Kasiske
BL
,
Zeier
MG
,
Chapman
JR
,
Craig
JC
,
Ekberg
H
,
Garvey
CA
, et al
KDIGO clinical practice guideline for the care of kidney transplant recipients: A summary
.
Kidney international
2010
;
77
:
299
311
.
14.
Lucey
MR
,
Terrault
N
,
Ojo
L
,
Hay
JE
,
Neuberger
J
,
Blumberg
E
, et al
Long-term management of the successful adult liver transplant: 2012 practice guideline by the American Association for the Study of Liver Diseases and the American Society of Transplantation
.
Liver Transpl
2013
;
19
:
3
26
.
15.
Costanzo
MR
,
Dipchand
A
,
Starling
R
,
Anderson
A
,
Chan
M
,
Desai
S
, et al
The International Society of Heart and Lung Transplantation Guidelines for the care of heart transplant recipients
.
J Heart Lung Transplant
2010
;
29
:
914
56
.
16.
National Center for Health Statistics
.
2011 Addenda
.
Atlanta, GA:
Centers for Disease Control and Prevention; 2013.
Available from:
http://www.cdc.gov/nchs/icd/icd9cm_addenda_guidelines.htm.
17.
Engels
EA
,
Cerhan
JR
,
Linet
MS
,
Cozen
W
,
Colt
JS
,
Davis
S
, et al
Immune-related conditions and immune-modulating medications as risk factors for non-Hodgkin's lymphoma: A case-control study
.
Am J Epidemiol
2005
;
162
:
1153
61
.
18.
Lanoy
E
,
Engels
EA
. 
Skin cancers associated with autoimmune conditions among elderly adults
.
Br J Cancer
2010
;
103
:
112
4
.
19.
National Aeronautics and Space Administration
.
Total ozone mapping spectrometer data product: Erythemal UV exposure
. In:
Goddard Space Flight Center, editor
.
Greenbelt, MD;
2004
.
20.
O'Donovan
P
,
Perrett
CM
,
Zhang
X
,
Montaner
B
,
Xu
Y-Z
,
Harwood
CA
, et al
Azathioprine and UVA light generate mutagenic oxidative DNA damage
.
Science
2005
;
309
:
1871
4
.
21.
Han
W
,
Soltani
K
,
Ming
M
,
He
YY
. 
Deregulation of XPC and CypA by cyclosporin A: An immunosuppression-independent mechanism of skin carcinogenesis
.
Cancer Prev Res
2012
;
5
:
1155
62
.
22.
Levi
F
,
Randimbison
L
,
Te
V-C
,
La Vecchia
C
. 
Non-Hodgkin's lymphomas, chronic lymphocytic leukaemias and skin cancers
.
Br J Cancer
1996
;
74
:
1847
50
.
23.
Adami
J
,
Frisch
M
,
Yuen
J
,
Glimelius
B
,
Melbye
M
. 
Evidence of an association between non-Hodgkin's lymphoma and skin cancer
.
BMJ
1995
;
310
:
1491
5
.
24.
Wheless
L
,
Black
J
,
Alberg
AJ
. 
Nonmelanoma skin cancer and the risk of second primary cancers: A systematic review
.
Cancer Epidemiol Biomarkers Prev
2010
;
19
:
1686
95
.
25.
Paul
CF
,
Ho
VC
,
McGeown
C
,
Christophers
E
,
Schmidtmann
B
,
Guillaume
JC
, et al
Risk of malignancies in psoriasis patients treated with cyclosporine: A 5 y cohort study
.
J Invest Dermatol
2003
;
120
:
211
6
.
26.
Marcil
I
,
Stern
RS
. 
Squamous-cell cancer of the skin in patients given PUVA and ciclosporin: Nested cohort crossover study
.
The Lancet
2001
;
358
:
1042
5
.
27.
Fiorino
G
,
Danese
S
. 
Adalimumab and azathioprine combination therapy for Crohn's disease: A shining diamond?
J Crohns Colitis
2016
;10:1257–8.
28.
Pedersen
EG
,
Pottegard
A
,
Hallas
J
,
Friis
S
,
Hansen
K
,
Jensen
PE
, et al
Risk of non-melanoma skin cancer in myasthenia patients treated with azathioprine
.
Eur J Neurol
2014
;
21
:
454
8
.
29.
Curtis
RE
,
Metayer
C
,
Rizzo
JD
,
Socie
G
,
Sobocinski
KA
,
Flowers
ME
, et al
Impact of chronic GVHD therapy on the development of squamous-cell cancers after hematopoietic stem-cell transplantation: An international case-control study
.
Blood
2005
;
105
:
3802
11
.
30.
Furuta
S
,
Jayne
D
. 
Emerging therapies in antineutrophil cytoplasm antibody-associated vasculitis
.
Curr Opin Rheumatol
2014
;
26
:
1
6
.
31.
Liberal
R
,
Krawitt
EL
,
Vierling
JM
,
Manns
MP
,
Mieli-Vergani
G
,
Vergani
D
. 
Cutting edge issues in autoimmune hepatitis
.
J Autoimmun
2016
;75:6–19.
32.
De Virgilio
A
,
Greco
A
,
Magliulo
G
,
Gallo
A
,
Ruoppolo
G
,
Conte
M
, et al
Polyarteritis nodosa: A contemporary overview
.
Autoimmun Rev
2016
;
15
:
564
70
.
33.
Jaime-Perez
JC
,
Rodriguez-Martinez
M
,
Gomez-de-Leon
A
,
Tarin-Arzaga
L
,
Gomez-Almaguer
D
. 
Current approaches for the treatment of autoimmune hemolytic anemia
.
Arch Immunol Ther Exp (Warsz)
2013
;
61
:
385
95
.
34.
Muraro
A
,
Lemanske
RF
 Jr.
,
Hellings
PW
,
Akdis
CA
,
Bieber
T
,
Casale
TB
, et al
Precision medicine in patients with allergic diseases: Airway diseases and atopic dermatitis-PRACTALL document of the European Academy of Allergy and Clinical Immunology and the American Academy of Allergy, Asthma & Immunology
.
J Allergy Clin Immunol
2016
;
137
:
1347
58
.
35.
Hoste
E
,
Cipolat
S
,
Watt
FM
. 
Understanding allergy and cancer risk: What are the barriers?
Nat Rev Cancer
2015
;
15
:
131
2
.
36.
Hwang
CY
,
Chen
YJ
,
Lin
MW
,
Chen
TJ
,
Chu
SY
,
Chen
CC
, et al
Cancer risk in patients with allergic rhinitis, asthma and atopic dermatitis: a nationwide cohort study in Taiwan
.
Int J Cancer
2012
;
130
:
1160
7
.
37.
National Institute of Arthritis and Musculoskeletal and Skin Diseases
.
Systemic lupus erythematosus.
Bethesda, MD:
National Institutes of Health
; 
2015.
Available from: http://www.niams.nih.gov/health_info/Lupus/default.asp#link_dd
.
38.
American College of Allergy, Asthma, and Immunology
.
Allergic rhinitis.
Arlington Heights, IL: American College of Allergy, Asthma, and Immunology
; 
2016
.
Available from
: http://acaai.org/allergies/types/hay-fever-rhinitis.
39.
Sutliff
MH.
Contribution of impaired mobility to patient burden in multiple sclerosis
.
Curr Med Res Opin
2010
;
26
:
109
19
.
40.
Copeland
KT
,
Checkoway
H
,
Mcmichael
AJ
,
Holbrook
RH
. 
Bias due to misclassification in the estimation of relative risk
.
Am J Epidemiol
1977
;
105
:
488
95
.