Nonmelanoma skin cancer in patients with hereditary hemochromatosis: A case-control study

To the Editor: Hereditary hemochromatosis (HH) is a genetic disease of altered iron homeostasis with known complications such as hepatocellular carcinoma and cutaneous hyperpigmentation. Risk of hepatocellular carcinoma in patients with HH is hypothesized to increase due to iron-catalyzed oxidative damage.1Toyokuni S. Role of iron in carcinogenesis: cancer as a ferrotoxic disease.Cancer Sci. 2009; 100: 9-16https://doi.org/10.1111/J.1349-7006.2008.01001.XCrossref PubMed Scopus (0) Google Scholar,2Fu Y. Chung F.-L. Oxidative stress and hepatocarcinogenesis.Hepatoma Res. 2018; 4: 39https://doi.org/10.20517/2394-5079.2018.29Crossref PubMed Google Scholar Although oxidative damage has also been implicated in nonmelanoma skin cancer (NMSC) pathogenesis, its link to HH is unknown.3Sander C.S. Hamm F. Elsner P. Thiele J.J. Oxidative stress in malignant melanoma and non-melanoma skin cancer.Br J Dermatol. 2003; 148: 913-922https://doi.org/10.1046/J.1365-2133.2003.05303.XCrossref PubMed Google Scholar,4Tilli C.M.L.J. Van Steensel M.A.M. Krekels G.A.M. Neumann H.A.M. Ramaekers F.C.S. Molecular aetiology and pathogenesis of basal cell carcinoma.Br J Dermatol. 2005; 152: 1108-1124https://doi.org/10.1111/J.1365-2133.2005.06587.XCrossref PubMed Google Scholar We sought to assess the association between HH and NMSC and whether phlebotomy treatment modifies this relationship. We retrospectively reviewed data on patients with HH and malignant skin neoplasms treated at Massachusetts General Hospital, Brigham and Women's Hospital, and Faulkner Hospital from 1980 to 2021. Controls without HH (1 per case) with skin neoplasms were matched by age, sex, and race. HH carriers and/or those without pathology-proven skin neoplasms were excluded. Fisher's exact test, univariable regression analyses, and multivariable regression analyses (adjusted for age, sex, skin phototype, smoking history, family history of NMSC, and NMSC risk factors, such as tanning bed usage) were conducted using R. P values were two-tailed and Benjamini-Hochberg false discovery rate correction was performed, with P ≤ .05 determining statistical significance. This study was approved by the institutional review board of Mass General Brigham. Among 448 patients with HH and 448 control patients extracted, 321 patients with HH and 374 controls were included after chart review (Table I). A positive association between HH and NMSC diagnosis was observed (adjusted odds ratio, 2.02; 95% confidence interval, 0.865-4.718; P = .083), though the difference was not statistically significant after false discovery rate correction (Table II). When specifically examining incidence of tumor subtypes, we observed that the association between HH and basal cell carcinoma (BCC) was 111% greater than that of controls (adjusted odds ratio, 2.11; 95% confidence interval, 1.288-3.450; P = .005) (Table II). In a subsequent analysis of patients with HH by phlebotomy treatment status, receiving or having received phlebotomy was significantly associated with NMSC diagnosis (adjusted odds ratio, 5.73; 95% confidence interval, 1.293-25.367; P = .015) (Table II).Table IUnivariable regression of patients with HH and controls with subset analysis by phlebotomy treatment statusPatients with HHControlUnadjusted odds ratio/mean differenceP value§Analyses were repeated in a subgroup of patients with genetically confirmed HH. There were no statistical differences from present findings, suggesting that the observed associations were unlikely driven by misclassification of non-HH cases.NMSC diagnosis∗Denotes the number of patients diagnosed with any NMSCs. Among HH cohort, 30.3% (n = 89) of patients had concurrent diagnoses of BCC and SCC. Among control cohort, 16.3% (n = 51) of patients had concurrent diagnoses of BCC and SCC.294 (91.6%)312 (83.4%)OR 2.164 (95% CI, 1.225, 3.822).005BCC diagnosis235 (73.2%)220 (58.8%)OR 1.913 (95% CI, 1.306, 2.802).005SCC diagnosis145 (45.2%)142 (38.0%)OR 1.346 (95% CI, 0.939, 1.929).086Age at NMSC diagnosis (mean)61.561.6MD −0.078 (95% CI, −2.350, 2.194).936Number of NMSCs (mean)∗Denotes the number of patients diagnosed with any NMSCs. Among HH cohort, 30.3% (n = 89) of patients had concurrent diagnoses of BCC and SCC. Among control cohort, 16.3% (n = 51) of patients had concurrent diagnoses of BCC and SCC.1.671.56MD 0.109 (95% CI, −0.030, 0.248).086Subgroup analysisPatients with HH receiving phlebotomyPatients with HH not receiving phlebotomyNMSC diagnosis†Among patients with HH receiving phlebotomy, 29.8% (n = 79) had concurrent diagnoses of BCC and SCC. Among patients with HH not receiving phlebotomy, 34.5% (n = 10) had concurrent diagnoses of BCC and SCC.265 (94.0%)29 (74.4%)OR 5.375 (95% CI, 1.713, 16.865).005BCC diagnosis208 (73.8%)27 (69.2%)OR 1.249 (95% CI, 0.479, 3.260).689SCC diagnosis133 (47.2%)12 (30.8%)OR 2.008 (95% CI, 0.780, 5.168).143Age at NMSC diagnosis (mean)61.363.4MD −2.111 (95% CI, −7.871, 3.649).573Number of NMSCs (mean)‡Number of NMSCs was recorded in intervals of five as follows: 1, 2-5, 6-10, 11+.1.6721.621MD 0.051 (95% CI, −0.354, 0.456).745Bolded values indicate significance P < .05.BCC, Basal cell carcinoma; CI, confidence interval; HH, hereditary hemochromatosis; MD, mean difference; NMSC, non-melanoma skin cancer; OR, odds ratio; SCC, squamous cell carcinoma.∗ Denotes the number of patients diagnosed with any NMSCs. Among HH cohort, 30.3% (n = 89) of patients had concurrent diagnoses of BCC and SCC. Among control cohort, 16.3% (n = 51) of patients had concurrent diagnoses of BCC and SCC.† Among patients with HH receiving phlebotomy, 29.8% (n = 79) had concurrent diagnoses of BCC and SCC. Among patients with HH not receiving phlebotomy, 34.5% (n = 10) had concurrent diagnoses of BCC and SCC.‡ Number of NMSCs was recorded in intervals of five as follows: 1, 2-5, 6-10, 11+.§ Analyses were repeated in a subgroup of patients with genetically confirmed HH. There were no statistical differences from present findings, suggesting that the observed associations were unlikely driven by misclassification of non-HH cases. Open table in a new tab Table IIMultivariable regression of patients with HH and controls with subset analysis by phlebotomy treatment statusPatients with HHControlAdjusted odds ratio/mean differenceP value§Analyses were repeated in a subgroup of patients with genetically confirmed HH. There were no statistical differences from present findings, suggesting that the observed associations were unlikely driven by misclassification of non-HH cases.NMSC diagnosis∗Denotes the number of patients diagnosed with any NMSCs. Among HH cohort, 30.3% (n = 89) of patients had concurrent diagnoses of BCC and SCC. Among control cohort, 16.3% (n = 51) of patients had concurrent diagnoses of BCC and SCC.294 (91.6%)312 (83.4%)aOR, 2.020 (95% CI, 0.865, 4.718).083BCC diagnosis235 (73.2%)220 (58.8%)aOR, 2.108 (95% CI, 1.288, 3.450).005SCC diagnosis145 (45.2%)142 (38.0%)aOR, 1.196 (95% CI, 0.761, 1.881).309Age at NMSC diagnosis (mean)61.561.6aMD, −0.863 (95% CI, −2.443, 0.717).199Number of NMSCs (mean)∗Denotes the number of patients diagnosed with any NMSCs. Among HH cohort, 30.3% (n = 89) of patients had concurrent diagnoses of BCC and SCC. Among control cohort, 16.3% (n = 51) of patients had concurrent diagnoses of BCC and SCC.1.671.56aMD, 0.107 (95% CI, −0.058, 0.272).157Subgroup analysisPatients with HH receiving phlebotomyPatients with HH not receiving phlebotomyNMSC diagnosis†Among patients with HH receiving phlebotomy, 29.8% (n = 79) had concurrent diagnoses of BCC and SCC. Among patients with HH not receiving phlebotomy, 34.5% (n = 10) had concurrent diagnoses of BCC and SCC.265 (94.0%)29 (74.4%)aOR, 5.726 (95% CI, 1.293, 25.367).015BCC diagnosis208 (73.8%)27 (69.2%)aOR, 0.940 (95% CI, 0.299, 2.959).890SCC diagnosis133 (47.2%)12 (30.8%)aOR, 1.754 (95% CI, 0.628, 4.899).265Age at NMSC diagnosis (mean)61.363.4aMD, −2.940 (95% CI, −6.596, 0.716).095Number of NMSCs (mean)‡Number of NMSCs were recorded in intervals of five as follows: 1, 2-5, 6-10, 11+.1.6721.621aMD, 0.041 (95% CI, −0.372, 0.454).890Bolded values indicate significance P < .05.aMD, Adjusted mean difference; aOR, adjusted odds ratio; BCC, basal cell carcinoma; CI, confidence interval; HH, hereditary hemochromatosis; NMSC, non-melanoma skin cancer; SCC, squamous cell carcinoma.∗ Denotes the number of patients diagnosed with any NMSCs. Among HH cohort, 30.3% (n = 89) of patients had concurrent diagnoses of BCC and SCC. Among control cohort, 16.3% (n = 51) of patients had concurrent diagnoses of BCC and SCC.† Among patients with HH receiving phlebotomy, 29.8% (n = 79) had concurrent diagnoses of BCC and SCC. Among patients with HH not receiving phlebotomy, 34.5% (n = 10) had concurrent diagnoses of BCC and SCC.‡ Number of NMSCs were recorded in intervals of five as follows: 1, 2-5, 6-10, 11+.§ Analyses were repeated in a subgroup of patients with genetically confirmed HH. There were no statistical differences from present findings, suggesting that the observed associations were unlikely driven by misclassification of non-HH cases. Open table in a new tab Bolded values indicate significance P < .05. BCC, Basal cell carcinoma; CI, confidence interval; HH, hereditary hemochromatosis; MD, mean difference; NMSC, non-melanoma skin cancer; OR, odds ratio; SCC, squamous cell carcinoma. Bolded values indicate significance P < .05. aMD, Adjusted mean difference; aOR, adjusted odds ratio; BCC, basal cell carcinoma; CI, confidence interval; HH, hereditary hemochromatosis; NMSC, non-melanoma skin cancer; SCC, squamous cell carcinoma. Findings from our study suggest that HH may be associated with increased risk of BCCs. While the underlying pathophysiology has not been previously explored, this may be consistent with prior work that identified oxidative stress as a key component in both HH and BCC pathogeneses.3Sander C.S. Hamm F. Elsner P. Thiele J.J. Oxidative stress in malignant melanoma and non-melanoma skin cancer.Br J Dermatol. 2003; 148: 913-922https://doi.org/10.1046/J.1365-2133.2003.05303.XCrossref PubMed Google Scholar,4Tilli C.M.L.J. Van Steensel M.A.M. Krekels G.A.M. Neumann H.A.M. Ramaekers F.C.S. Molecular aetiology and pathogenesis of basal cell carcinoma.Br J Dermatol. 2005; 152: 1108-1124https://doi.org/10.1111/J.1365-2133.2005.06587.XCrossref PubMed Google Scholar One hypothesis is that excess iron may react with reactive oxygen species induced by ultraviolet radiation through Fenton reactions, leading to formation of damaging hydroxyl radicals.5Wright J.A. Richards T. Srai S.K.S. The role of iron in the skin and cutaneous wound healing.Front Pharmacol. 2014; 5: 156https://doi.org/10.3389/FPHAR.2014.00156/BIBTEXCrossref PubMed Scopus (0) Google Scholar Additionally, the finding that patients who received phlebotomy had higher rates of NMSC diagnosis suggests that phlebotomy may not be protective against NMSC development and that the severity of hemochromatosis—as indicated by phlebotomy requirement—may correlate with NMSC risk. This study's limitations include its retrospective nature and its exclusion of other NMSC risk factors, such as alcohol use. Generalizability may also be limited to regional populations as data were extracted from 1 health care system. Risk of cutaneous neoplasms among patients with HH remains understudied. This study is a first exploration of possible association between HH and BCCs and aims to provide a preliminary direction for future work. Further investigation through prospective studies and tissue sampling studies is warranted. None disclosed.