Platelet counts in HFE p.C282Y/p.C282Y and wt/wt post-screening clinical evaluation participants

Abstract Our aim was to document the effects of genotype HFE p.C282Y/p.C282Y and hemochromatosis-associated laboratory and clinical manifestations on platelet counts (PC). We compiled genotype (p.C282Y/p.C282Y or HFE wt/wt (absence of p.C282Y and p.H63D (rs1799945)), age, sex, body mass index, presence/absence of chronic fatigue, swelling/tenderness of second/third metacarpophalangeal joints, and hyperpigmentation, transferrin saturation (TS), serum ferritin (SF), hemoglobin levels, absolute neutrophil, lymphocyte, and monocyte counts, C-reactive protein levels, and PC of non-Hispanic white participants in a hemochromatosis and iron overload post-screening clinical examination. There were 171 men and 254 women (141 p.C282Y/p.C282Y, 284 wt/wt) of median age 53 y. Median TS and SF were higher in p.C282Y/p.C282Y than wt/wt participants grouped by sex (p < .0001, all comparisons). Median PC by genotype was lower in men than women (p < .0001, both comparisons). Regression on PC using 14 independent variables identified these significant positive associations: absolute neutrophil, lymphocyte, and monocyte counts and C-reactive protein levels and these significant inverse associations: age, TS, and hemoglobin levels. We conclude that PC is significantly associated with absolute neutrophil, lymphocyte, and monocyte counts, and C-reactive protein (positive) and age, TS, and hemoglobin (inverse), after adjustment for other variables. HFE genotypes we studied were not significantly associated with PC. Plain Language Summary What is the context? Hemochromatosis is typically associated with inheritance of two copies of p.C282Y, a common mutation of the HFE gene on chromosome 6p that regulates iron absorption. Platelet counts, age, and serum levels of liver enzymes have been used to estimate risks of cirrhosis in adults with hemochromatosis. Lower platelet counts in Europeans are significantly associated with a mutation in CARMIL1, a gene on chromosome 6p close to HFE. Clinical and laboratory associations of normal platelet counts in adults with HFE p.C282Y/p/C282Y and wt/wt uncomplicated by cirrhosis are unreported. What is new? We studied normal platelet counts in 425 white adults who participated in a primary care-based hemochromatosis screening program. These participants did not have cirrhosis or other conditions that often influence platelet counts. Our analyses of 14 variables identified these significant positive associations with platelet counts, after adjustment for other variables: absolute neutrophil, lymphocyte, and monocyte counts and C-reactive protein levels; and these significant inverse associations: age, TS, and hemoglobin levels. What is the impact? Laboratory and clinical factors significantly associated with platelet counts in adults with HFE p.C282/p.C282Y or wt/wt are similar to those in persons unselected for HFE genotypes or hemochromatosis. It is unlikely that genes that influence platelet counts are closely linked to HFE on chromosome 6p. Adults with hemochromatosis and HFE p.C282/p.C282Y who have abnormal platelet counts should be evaluated for cirrhosis or non-iron platelet disorders.

• Platelet counts, age, and serum levels of liver enzymes have been used to estimate risks of cirrhosis in adults with hemochromatosis.• Lower platelet counts in Europeans are significantly associated with a mutation in CARMIL1, a gene on chromosome 6p close to HFE. • Clinical and laboratory associations of normal platelet counts in adults with HFE p.C282Y/ p/C282Y and wt/wt uncomplicated by cirrhosis are unreported.
What is new?
• We studied normal platelet counts in 425 white adults who participated in a primary carebased hemochromatosis screening program.These participants did not have cirrhosis or other conditions that often influence platelet counts.• Our analyses of 14 variables identified these significant positive associations with platelet counts, after adjustment for other variables: absolute neutrophil, lymphocyte, and monocyte counts and C-reactive protein levels; and these significant inverse associations: age, TS, and hemoglobin levels.

Introduction
Hemochromatosis in whites of western European descent is associated with homozygosity for p.C282Y (exon 4, c.845 G>A; rs1800562), a common missense allele of the HFE gene (homeostatic iron regulator, chromosome 6p22.2). 1,2 HFE, a non-classical class I major histocompatibility complex protein, is an upstream regulator of hepcidin and thus of iron homeostasis. 3Laboratory phenotypes of many adults at diagnosis of HFE p.C282Y/p.C282Y include elevated levels of transferrin saturation (TS) and serum ferritin (SF). 4Adults with p.C282Y/p.C282Y have increased risks to develop iron overload and resulting complications, including arthropathy, diabetes mellitus, hypogonadotropic hypogonadism, hepatic cirrhosis, and cardiomyopathy. 4Severe iron overload and cirrhosis due to hemochromatosis occur predominantly in men. 4,5on-HFE alleles and environmental factors modify iron loading in adults with hemochromatosis. 2,4,6here are no published data documenting the effects of HFE p. C282Y/p.C282Y and hemochromatosis-associated laboratory and clinical manifestations on blood platelet counts (PC). 7The combination of PC, age, and serum aminotransferase levels in adults with hemochromatosis has been used to estimate risks of cirrhosis, 8-10 a complication that occurs in < 10% of p.C282Y homozygotes identified in either screening or non-screening cohorts published in the 21st C. 11 Lower PC in Europeans are significantly associated with rs12526480, 12 an intronic polymorphism in CARMIL1, a gene close to HFE (chromosome 6p22.2). 13 Platelet plasma membranes express HFE protein demonstrable with Western blots and immunocytochemistry. 14 Iron is required for in vitro production of biuret-reactive protein in rabbit platelets. 15Human megakaryocyte progenitor cells express transferrin receptor-1 in vitro. 16Thus, it is plausible but unproven that PC of adults with p.C282Y/p.C282Y differ from those of adults with HFE wt/wt (absence of p.C282Y and p.H63D (rs1799945)).
The aim of this study was to document the effects of genotype HFE p.C282Y/p.C282Y and hemochromatosis-associated laboratory and clinical manifestations on PC.We studied all 425 self-reported non-Hispanic adult white participants in post-screening clinical evaluations of the primary care-based, cross-sectional Hemochromatosis and Iron Overload Screening (HEIRS) Study who had p.C282Y/p.C282Y or HFE wt/wt, normal PC, no common acquired conditions that affect PC, and complete data.Using univariate methods, we compared age, body mass index (BMI), prevalences of chronic fatigue, swelling/tenderness of second/third metacarpophalangeal (MP) joints, hyperpigmentation, TS, SF, hemoglobin (Hb), absolute neutrophil count (ANC), absolute lymphocyte count (ALC), absolute monocyte count (AMC), C-reactive protein level (CRP), and PC of participants grouped by HFE genotypes and sex.Using multivariable analysis, we determined significant associations of PC with 14 laboratory and clinical variables.We discuss the present results in the context of PC previously reported in adults not selected for TS and SF phenotypes or HFE genotypes.

Ethical approval
The HEIRS Study, conducted by the National Heart, Lung, and Blood Institute/National Human Genome Research Institute in accordance with principles of the Declaration of Helsinki, evaluated diverse aspects of hemochromatosis and iron overload in a primary care-based sample of 101,168 adults enrolled during the interval 2001-2003 at four Field Centers in the United States and one in Canada. 17Participants ≥ 25 y of age were recruited from outpatient facilities associated with five Field Centers and gave written informed consent for initial screening and post-screening clinical evaluation.Local Institutional Review Boards of the HEIRS Study Coordinating Center, Central Laboratory, and each Field Center approved the Study protocol that is described in detail elsewhere. 18

Primary care-based screening
The HEIRS Study recruited participants from a health maintenance organization, diagnostic blood collection centers, and public and private primary care offices in ambulatory clinics associated with five Field Centers. 17Ninety-eight percent of self-reported non-Hispanic whites were recruited at Field Centers in Alabama, California, Ontario, and Oregon/Hawaii. 19Participation rates of non-Hispanic white HFE p.C282Y homozygotes in primary carebased screening were significantly greater than estimated by Hardy-Weinberg proportions at four of the five Field Centers. 17aboratory testing at screening included only TS and SF phenotyping and HFE p.C282Y and p.H63D allele-specific genotyping. 17EIRS Study recruiters solicited participation of men and women equally.The greater proportion of women than men in the present study (59.8% vs. 40.2%;p < .0001) is representative of the greater proportion of women than men who participated in HEIRS Study initial primary care-based screening (62.9% vs. 37.1%, respectively; p < .0001). 17This is consistent with previous reports that women use more health care services than men. 4 The prevalence of HFE p.C282Y/p.C282Y in non-Hispanic white men and women who participated in the HEIRS Study did not differ significantly. 19ll participants reported their race/ancestry categories defined by the HEIRS Study and approved by the National Heart, Blood, and Lung Institute/National Human Genome Research Institute. 17,18Of 299 HFE p.C282Y homozygotes detected, 94.0% reported non-Hispanic white ancestry.Prevalences (95% confidence intervals) of p.C282Y homozygotes in race/ethnicity groups included: non-Hispanic whites 0.44 (0.42-0.47),Native Americans 0.11 (0.0061-0.20),Hispanics 0.027 (0.022-0.032), blacks/African Americans 0.014 (0.012-0.017),Pacific Islanders 0.012 (0.0043-0.032), and Asians 0.000039 (0.000015, 0.00010). 17

Post-screening clinical evaluations
Invitations to participate in post-screening clinical evaluations were extended to the following primary care-based screening participants, regardless of self-reported race/ethnicity: 1) participants with HFE p.C282Y/p.C282Y (regardless of screening TS and SF); 2) participants whose screening TS and SF exceeded study thresholds (TS > 55%, SF >300 µg/L (M); TS > 45%, SF >200 µg/L (F), regardless of HFE genotype); and 3) participants who had normal screening TS and SF and HFE wt/wt who were in the 25th-75th percentile of sex-specific distributions and who were frequency-matched for age (25-45 y, >45-65 y, and > 65 y) and sex in a 1:1 ratio with all participants with p.C282Y/p.C282Y identified at each of the five Field Centers. 18,20A total of 2265 participants were invited to post-screening clinical evaluations, of whom 1687 (74.5%) were evaluated. 20Median interval between primary-care-based screening and post-screening clinical evaluations was eight months. 18,20ost-screening clinical evaluations included the following: 1) questionnaires completed by participants that addressed medical history and medications; 2) focused physical examinations performed by HEIRS Study physicians; and 3) laboratory testing of blood specimens. 20

Laboratory testing in post-screening clinical evaluations
A morning blood sample was obtained from each post-screening clinical evaluation participant after an overnight fast of ≥8 h to minimize diurnal variation of TS and SF 21 and blood cell counts.- 22,23Testing measurements included the following: serum TS and SF (Hitachi 911 Analyzer, Roche Applied Science, Indianapolis, IN, USA), complete blood counts (Beckman Coulter GenS, Beckman/Coulter, Fullerton, CA, USA), and confirmation of HFE genotype. 20HFE p.C282Y and p.H63D alleles were determined from spots of whole blood using a modification of the Invader assay (Third Wave Technologies, Madison, WI, USA) that increases the allele-specific fluorescent signal by including 12 cycles of locus-specific polymerase chain reaction before the cleavase reaction. 17Platelet function was not evaluated.

Post-screening clinical evaluation participants included
The present cohort includes all non-Hispanic whites with either HFE p.C282Y/p.C282Y (n = 141) or wt/wt (n = 285) who attended post-screening clinical evaluations, fasted overnight ≥8 h before attending post-screening clinical evaluations, and had complete blood counts with normal PC.This cohort represents 49.5% of non-Hispanic white p.C282Y/p.C282Y post-screening clinical evaluation participants (141/285) and 54.3% of non-Hispanic white wt/wt post-screening clinical evaluation participants (284/523) (p = .2101).

Post-screening clinical evaluation participants excluded
We excluded post-screening clinical evaluation participants who had any of the following: iron deficiency (TS < 10% and SF <30 µg/L (M) and TS < 10% and SF <20 µg/L (F)); common acquired conditions that affect PC; viral hepatitis B or C 25,26 ; self-reported cirrhosis 27 ; self-reported diagnosis of malignancy 28 ; self-reported anti-cancer therapy 29 ; self-reported chronic inflammatory conditions (chronic inflammation, chronic infection, autoimmune disease, or lupus) 30 ; and self-reported pregnancy. 31

Study variables
We documented effects of these variables on PC: HFE genotype, age, sex, BMI, reports of chronic fatigue, 20 swelling/tenderness of the second/third MP joints, 20 hyperpigmentation, 20 TS, SF, Hb, ANC, ALC, AMC, and CRP.

Statistics
The dataset for analyses consisted of complete post-screening clinical evaluation observations in 141 participants with HFE p. C282Y/p.C282Y and 284 participants with wt/wt.Data for age, TS, SF, and PC (x 10 9 /L) are displayed to the nearest integer.Descriptive data are displayed as enumerations, percentages, or medians (range).Kolmogorov-Smirnov testing demonstrated that age, BMI, TS, SF, Hb, ANC, ALC, AMC, CRP, and PC data differed significantly from those that are normally distributed.Thus, we display these data as medians (range) and compared them using Mann-Whitney U tests (two-tailed).Percentages were compared using Fisher's exact test (two-tailed).
We performed backward stepwise regression on PC using these 14 independent variables: HFE genotype, sex, age, BMI, chronic fatigue, swelling/tenderness of second/third MP joints, and hyperpigmentation, TS, SF, Hb, ANC, ALC, AMC, and CRP.We used elimination of the last independent variable with p ≥ .05 as the stopping rule for the regression.
We defined p < .05 to be significant, although a Bonferroni correction for multiple univariate comparisons was applied to control the type 1 error rate.We used Excel® 2000 (Microsoft Corp., Redmond, WA, USA) and GraphPad Prism 8® (2018; GraphPad Software, San Diego, CA, USA).

Comparisons of participants by HFE genotype
Men with HFE p.C282Y/p.C282Y were younger and had higher median TS and SF, higher prevalence of chronic fatigue, and higher median Hb than men with wt/wt (Table I).Median PC in men with p.C282Y/p.C282Y was lower than that of men with wt/wt, although this difference was not significant (Table I).
Women with HFE p.C282Y/p.C282Y had higher median TS, SF, and Hb than women with wt/wt (Table II).Median PC in women with p.C282Y/p.C282Y was lower than that of women with wt/wt, although this difference was not significant (Table II).

Regression on platelet counts
We performed backward stepwise regression on PC using 14 independent variables.Regression identified these four significant positive associations: ANC, ALC, AMC, and CRP (Table III).
Regression identified these three significant inverse associations: age, TS, and Hb (Table III).This regression accounted for 26.5% of the variance of PC (ANOVA p of regression < 0.0001).HFE genotype, sex, and five other independent variables we studied in this regression were not significantly associated with PC, after adjustment for other variables.

Discussion
A novel aspect of this study is documentation of the effects of HFE p.C282Y/p.C282Y and other hemochromatosis-associated laboratory and clinical manifestations on PC.We compared PC in non-Hispanic white adults who participated in post-primary care-based screening clinical evaluations who had HFE genotypes p.C282Y/p.C282Y or wt/wt and normal PC.
][10] In this study, we excluded subjects who had abnormal PC.We also excluded subjects with cirrhosis, an acquired complication of adults with hemochromatosis and p. C282Y/p.C282Y significantly associated with age, severe iron overload, alcohol consumption, and diabetes. 5Cirrhosis was diagnosed in < 10% of adults with p.C282Y/p.C282Y in either screening or non-screening cohorts published in the 21st C. 11 Lower PC in Europeans in another study were significantly associated with rs12526480, 12 an intronic polymorphism in CARMIL1, a gene close to HFE (chromosome 6p22.2) which encodes capping protein regulator and myosin 1 linker 1. 13 Although the allele frequency of rs12526480 in 286,160 Europeans was 0.3343, 32 the clinical importance of this singlenucleotide polymorphism, if any, has not been reported.The present results support our postulate that no locus on chromosome 6p in linkage disequilibrium with HFE p.C282Y influences PC. 12,13 We observed a significant inverse association of age with PC in the present study, after adjustment for other variables.In studies of adult population cohorts unselected for TS and SF phenotypes or HFE genotypes in the U.S., 33,34 Italy, 35,36 and France, 37 age was inversely associated with PC.Decrements of PC occur in healthy adults aged > 70 y. 33,36Aging is associated with decreasing quantities of bone marrow hematopoietic tissue and increasing apoptosis of bone marrow cells, 38,39 although agerelated decrements of PC may not occur in synchrony with decreasing marrow cellularity. 405][46] After adjustment for other variables, sex was not significantly associated with PC in the present study.Thrombopoietin levels do not explain differences in PC between healthy women and men. 46,47In women, hormone replacement therapy increased megakaryocytes in one study, 48 increased PC in another report, 49 did not change PC in two other studies, 50,51 and decreased PC in a fifth study. 52In men, testosterone replacement therapy increased PC from baseline values in one of two trials. 53osaic loss of Y chromosome, the most frequently detected somatic copy number alteration in men, is associated with higher PC independent of age. 54Although PC are strongly influenced by heritable factors, 12,35,45,55 the present observations indicate that HFE p.C282Y is not one of those factors.
We observed a significant inverse association of TS with PC in this study, after adjustment for other variables.][59] In another study, oral iron supplements decreased PC of volunteer whole-blood donors with hypoferritinemia although oral iron supplements had no significant effect on PC of donors who had normal or stable SF levels. 60edian Hb values were higher in men and women with HFE p.C282Y/p.C282Y than corresponding values in men and women with HFE wt/wt in this study, confirming previous reports. 61,62e also observed a significant inverse association of Hb with PC, after adjustment for other variables.
We observed significant positive associations of ANC, ALC, and AMC with PC, after adjustment for other variables.In another study of HEIRS Study screening program participants, there were positive associations of HFE p.C282Y/p.C282Y with lymphocyte and basophil counts. 63There was also a significant positive association of PC and white blood cells in a large population-based study of adults unselected for TS and SF phenotypes or HFE genotypes. 35RP was directly associated with PC in this study after adjustment for other variables.In a large population-based study of adults unselected for TS and SF phenotypes or HFE genotypes, there was also a positive association of CRP and PC. 35In another report, pre-phlebotomy plasma levels of CRP and interleukin-6 did not differ significantly in adults with HFE p.C282Y/p.C282Y who had high or low iron stores. 64 strength of the present study is the analysis of a dataset of post-screening clinical evaluation observations of non-Hispanic white adults with HFE p.C282Y/p.C282Y and wt/wt, normal PC, no common acquired conditions that affect PC, and complete data.A limitation of the present study is that we did not evaluate effects of smoking reports, systolic/diastolic blood pressure, serum cholesterol, triglyceride, glucose, or d-dimer levels, or anti-platelet drug use on PC, although these variables were relatively minor determinants of PC variance in a general adult population. 35eviewing medical records of participants, obtaining liver specimens by biopsy, estimating liver iron content, providing therapy to achieve iron depletion, detecting non-HFE alleles that modify iron loading, measuring thrombopoietin levels, mean platelet volume, platelet crit, and platelet function, and assessing effects of platelet HFE protein on thrombopoiesis and iron homeostasis were beyond the scope of HEIRS Study post-screening clinical examinations. 18e conclude that PC is significantly associated with ANC, ALC, AMC, and CRP (positive) and age, TS, and Hb (inverse), after adjustment for other variables.HFE genotypes we studied were not significantly associated with PC.

Table II .
Characteristics of women in a post-primary care-based screening clinical evaluation.a a HFE, homeostatic iron regulator gene; wt/wt, absence of HFE p.C282Y or p.H63D; BMI, body mass index; MP, metacarpophalangeal; TS, transferrin saturation; SF, serum ferritin; Hb, hemoglobin; ANC, absolute neutrophil count; ALC, absolute lymphocyte count; AMC, absolute monocyte count; CRP, C-reactive protein.bBonferroni correction (significant p <0.0038) did not change significance of these uncorrected values.
a b stepwise regression, associations of platelet counts with HFE genotype, sex, body mass index, chronic fatigue, swelling/tenderness of second/ third metacarpophalangeal joints, hyperpigmentation, and serum ferritin were not significant.