Genetic modifiers of penetrance to liver endpoints in HFE hemochromatosis: Associations in a large community cohort

Abstract

Background: The iron overload condition hereditary hemochromatosis (HH) can cause liver cirrhosis and cancer, diabetes, and arthritis. Males homozygous for the p.C282Y missense mutation in the Homeostatin Iron Regulator (HFE) gene have greatest risk; yet, only a minority develop these conditions. We aimed to determine whether common genetic variants influencing iron levels or liver disease risk in the general population also modify clinical penetrance in HFE p.C282Y and p.H63D carriers.

Methods: We studied 1294 male and 1596 female UK Biobank HFE p.C282Y homozygous participants of European ancestry with medical records up to 14 years after baseline assessment. Polygenic scores quantified genetic effects of blood iron biomarkers and relevant diseases (identified in the general population). Analyses were also performed in other HFE p.C282Y/p.H63D genotype groups.

Results: In male p.C282Y homozygotes, a higher iron polygenic score increased the risk of liver fibrosis or cirrhosis diagnoses (odds ratio for the top 20% of iron polygenic score vs. the bottom 20% = 4.90: 95% confidence intervals, 1.63-14.73; p = 0.005), liver cancer, and osteoarthritis but not diabetes. A liver cirrhosis polygenic score was associated with liver cancer diagnoses. In female p.C282Y homozygotes, the osteoarthritis polygenic score was associated with increased osteoarthritis diagnoses and type-2 diabetes polygenic score with diabetes. However, the iron polygenic score was not robustly associated with diagnoses in p.C282Y female homozygotes or in other p.C282Y/p.H63D genotypes.

Conclusions: HFE p.C282Y homozygote penetrance to clinical disease in a large community cohort was partly explained by common genetic variants that influence iron and risks of related diagnoses in the general population, including polygenic scores in HH screening and diagnosis, may help in estimating prognosis and treatment planning.

FIGURE 1

Linear associations among four iron status biomarker polygenic scores and hereditary hemochromatosis comorbidities in Homeostatin Iron Regulator (HFE) p.C282Y homozygous male participants. Results are from logistic regression models adjusted for age, assessment center, and principal components of ancestry 1 to 10. Percentage (%) of 1294 HFE p.C282Y homozygous male participants of European genetic ancestry who ever received a diagnosis in the available data (up to September 2021). See Table S4 for details, including associations in women. CI, confidence interval; TIBC, total iron‐binding capacity; TranSat, transferrin saturation. *p < 0.05 (false discovery rate adjusted using the Benjamini‐Hochberg method). Arrows indicate where the CIs go beyond the x axis limits.

FIGURE 2

Iron polygenic score association with diagnosis of liver fibrosis or cirrhosis in Homeostatin Iron Regulator (HFE) p.C282Y homozygous male participants. Iron polygenic score is stratified into five equally sized groups (quintiles). Results are from logistic regression models adjusted for age, assessment center, and principal components of ancestry 1 to 10. N = HFE p.C282Y homozygous male participants of European genetic ancestry in quintile. N cases = participants in quintile who ever received a diagnosis of liver fibrosis or cirrhosis in the available data (up to September 2021). Iron polygenic score is the score for total iron levels. See Table S5 for details including polygenic score cut points. CI, confidence interval.

FIGURE 3

Linear increases in iron polygenic scores and hereditary hemochromatosis comorbidities in other Homeostatin Iron Regulator (HFE) genotype groups. Analysis performed in UK Biobank male participants of European ancestry. Results are from logistic regression models adjusted for age, assessment center, and principal components of ancestry 1 to 10. Percentage (%) of HFE genotype group who ever received a diagnosis in the available data (up to September 2021). See Table S4 for details, including associations in women. CI, confidence interval. HMZ = homozygotes. HET = heterozygotes. comp.het = compound heterozygotes. *p < 0.05 (false discovery rate adjusted using the Benjamini‐Hochberg method). Arrows indicate where the CIs go beyond the x axis limits.

FIGURE 4

Linear increases in polygenic scores for hereditary hemochromatosis comorbidities affect likelihood of corresponding diagnosis in Homeostatin Iron Regulator (HFE) p.C282Y homozygotes. Results are from logistic regression models adjusted for age, assessment center, and principal components of ancestry 1 to 10. Liver cirrhosis polygenic score was tested against diagnoses of liver fibrosis or cirrhosis and separately against diagnoses of liver cancer. Osteoarthritis polygenic score was tested against diagnoses of osteoarthritis and type‐2 diabetes (T2D) and polygenic score against T2D. See Table S6 for details. CI, confidence interval. Arrows indicate where the CIs go beyond the x axis limits.