Erythrocytosis in the general population: clinical characteristics and association with clonal hematopoiesis

Abstract

Erythrocytosis is a common reason for referral to hematology services and is usually secondary in origin. The aim of this study was to assess clinical characteristics and clonal hematopoiesis (CH) in individuals with erythrocytosis in the population-based Lifelines cohort (n = 147 167). Erythrocytosis was defined using strict (World Health Organization [WHO] 2008/British Committee for Standards in Hematology) and wide (WHO 2016) criteria. Individuals with erythrocytosis (strict criteria) and concurrent leukocytosis and/or thrombocytosis were 1:2 matched with individuals with isolated erythrocytosis and analyzed for somatic mutations indicative of CH (≥5% variant allele frequency). One hundred eighty five males (0.3%) and 223 females (0.3%) met the strict criteria, whereas 4868 males (7.6%) and 309 females (0.4%) met the wide criteria. Erythrocytosis, only when defined using strict criteria, was associated with cardiovascular morbidity (odds ratio [OR], 1.8; 95% confidence interval [CI], 1.2-2.6), cardiovascular mortality (hazard ratio [HR], 2.2; 95% CI, 1.0-4.6), and all-cause mortality (HR, 1.7; 95% CI, 1.2-2.6), independent of conventional risk factors. Mutations were detected in 51 of 133 (38%) evaluable individuals, with comparable frequencies between individuals with and without concurrent cytosis. The JAK2 V617F mutation was observed in 7 of 133 (5.3%) individuals, all having concurrent cytosis. The prevalence of mutations in BCOR/BCORL1 (16%) was high, suggesting aberrant epigenetic regulation. Erythrocytosis with CH was associated with cardiovascular morbidity (OR, 9.1; 95% CI, 1.2-68.4) in a multivariable model. Our data indicate that only when defined using strict criteria erythrocytosis is associated with cardiovascular morbidity (especially in the presence of CH), cardiovascular mortality, and all-cause mortality.

Graphical abstract

Figure 1.

Prevalence of erythrocytosis (strict criteria) as function of sex and age and the association of erythrocytosis with CVD and survival. (A) Prevalence of erythrocytosis according to sex and age categories. (B) Forest plot for the risk of CVD and all-cause mortality. Logistic regression analyses and Cox proportional hazards regression included age, sex, BMI, systolic and diastolic blood pressure, smoking status, medical history of diabetes, number of medications used, androgen drug use, drugs for obstructive airway disease, antihypertensive drug use, family history of CVD, and LDL cholesterol as covariates. Absence of erythrocytosis was used as reference. Circles indicate the OR/HR, with horizontal lines corresponding to 95% CIs. (C) Kaplan-Meier curve for overall survival, stratified according to the presence of erythrocytosis.

Figure 2.

Spectrum of clonal hematopoiesis for individuals with erythrocytosis with and without concurrent leuko- and/or thrombocytosis. (A) Mutational landscape for all somatic mutations detected in sequenced individuals with erythrocytosis (n = 133). (B) Proportion of individuals with 1 or more mutations according to the presence of concurrent leuko- and/or thrombocytosis. (C) Proportion of individuals with mutated JAK2, BCOR/BCORL1, or other mutations (including CHIP genes) according to the presence of concurrent leuko- and/or thrombocytosis. (D) Contribution of individual mutated genes to the spectrum of clonal hematopoiesis detected in individuals with isolated erythrocytosis (red) and individuals with erythrocytosis with concurrent leuko- and/or thrombocytosis (blue). Bars indicate the proportion of individuals with a gene mutation.