AS628 - Clonal hematopoiesis of indeterminate potential (CHIP)

Investigator Names and Contact Information

Alex Reiner (apreiner@uw.edu) Eric Whitsel (eric_whitsel@med.unc.edu)

Introduction/Intent

Clonal hematopoiesis is a common, age-related condition in which hematopoietic stem cells in the bone marrow undergo somatic mutations that lead to overgrowth (“clones”) of a genetically distinct subpopulation of blood cells. Large-scale next-generation sequencing studies have found such somatic mutations in hematologic malignancy-associated genes (e.g., DNMT3A, TET2, and ASXL1) in blood from healthy subjects, a condition known as clonal hematopoiesis of indeterminate potential (CHIP). The prevalence of CHIP increases markedly with age, from <1% of the population at age <40; to nearly 10%, 15% and 25% at ages 40-49, 50-59, and 60-69 [1] and to possibly >50% at ages >85. Despite its name, CHIP has been linked to multiple health conditions, including >10-fold higher risk of leukemia/blood cancers, approximately double the risk of cardiovascular disease (CVD), and ~40% higher risk of all-cause mortality. However, most studies of CHIP are cross-sectional and limited research is available on risk factors for the development, incidence/progression, and consequences of CHIP over long-term follow up. WHI is particularly well-positioned to address these limitations because of its longitudinal design, imbedded clinical trials/ancillary studies, and ongoing surveillance of incident disease/mortality among aging women. To this end, WHI’s large size, racial/ethnic diversity, repeated blood collections, and repeated characterization of putative risk factors for CHIP progression over fifteen years in the Long Life Study (LLS-1) and over 25-30 years in the Long-Life Study 2 (LLS-2) are exceptional resources for CHIP research. Such inherited and acquired risk factors include e.g. aging-related measures (telomere length; epigenetic age), randomized interventions (hormone therapy [HT]; calcium/vitamin D [CaD] supplementation; dietary modification [DM]), and environmental exposures (air/noise pollution; radiation). Indeed, WHI provides an excellent platform for assessing exposure-CHIP-disease associations, including those with incident coronary heart disease (CHD), stroke, heart failure, venous thromboembolism (VTE), diabetes, dementia, cancers (total; site-specific), and mortality (all-cause; cause-specific). Notably, the proposed competing renewal application will leverage an independently-funded blood collection and clinical assessment at the LLS-2 exam to be completed in 2022-23. We will therefore:

Aim 1. Use 7,800 LLS samples at baseline and approximately fifteen years later to estimate exposure-CHIP associations. More specifically, we will:

(1a) Estimate associations between prevalent CHIP (at baseline), incidence or progression of CHIP (between baseline and LLS), and putative risk factors for CHIP.

(1b) Assess CHIP heritability and identify novel germline genomic factors that underlie both prevalence and incidence of CHIP.

Modification to Aim 1. In our competing renewal, measure CHIP using blood collected at the LLS-2 exam in ~5,000 surviving LLS-2 participants. In combination with the CHIP measurements performed during the original AS628, these additional CHIP measurements generated during the competing renewal will allow us to extend analyses of age-related CHIP progression in up to ~9,000 WHI participants over a 30-year period spanning baseline through LLS-2. These longitudinal CHIP data will be used to assess the association or modification of CHIP growth trajectories by intercurrent CVD and related risk factors, environmental, and lifestyle/behavioral factors (including physical activity and change in body weight); and extend analyses of CHIP association with important aging-related physical performance/frailty and cognitive function phenotypes at the LLS-1 and LLS-2 exams.

Aim 2. Use the LLS cohort, additional baseline CHIP data generated in a total of ~17,000 WHI participants, and a nested case-cohort design to estimate CHIP-outcome associations. More specifically, we will:

(2a) Estimate associations between CHIP and incident clinical CVD outcomes, dementia, and mortality

(2b) Estimate associations between CHIP, quantitative hematological (erythrocyte, leukocyte, and platelet) traits, and benign hematologic disorders.

Modification to Aim 2. In our competing renewal, measure CHIP at baseline among an additional 5,000 rigorously adjudicated and subtyped incident heart failure cases and 5,000 controls matched on age, race/ethnicity, and follow-up time, to assess whether CHIP is differentially associated with heart failure with preserved vs. reduced ejection fraction.

Aim 3. Informed by results from Aims 1 and 2, we will use Mendelian randomization approaches, mediation analyses, and polygenic risk scores to assess causal mediation of exposure-outcome associations by CHIP. In secondary analyses, we also will explore:

(1) Modification of CHIP-CVD associations estimated in Aim 2a by exposures examined in Aim 1a.

(2) Associations of individual CHIP driver mutations with outcomes in Aims 2a-b.

(3) Differential expression of genes and pathways associated with CHIP in a subset of 1400 participants with whole blood transcriptomic RNAseq data at LLS.