Abstract
Clonal Hematopoiesis of Indeterminant Potential (CHIP) is an age-related phenomenon associated with increased risk of hematologic malignancy. Preclinical studies have shown that infection is a driver of CHIP; clinical studies in people living with HIV suggest a relationship between chronic infection and CHIP, but the association between infection frequency and incident CHIP in the general population remains unknown. We leveraged the Atherosclerosis Risk in Communities (ARIC) study to design a closed prospective cohort study. CHIP was determined based on whole-exome sequencing at two time points 20 years apart. Included were 3,367 individuals without cancer or CHIP at Time 1 and without hematologic malignancy by Time 2. The 3,367 study participants had an average age of 55.3 years at Time 1; 59.1% were female, 40.9% were male; 24% were Black, and 76% were White. Documented infection was assessed from routinely collected hospital discharge summaries. Frequency was categorized as no documented infection, 1 infection, 2 infections, or ≥ 3 infections. Of the participants, 19.7% had incident CH, 6.9% had large CHIP, and 5.2% had large non-DNMT3A CHIP. Participants with ≥ 3 documented infections had an increased odds of incident CHIP (OR 1.41, p = 0.03), especially large CHIP (OR 1.83, P = 0.008) and large non-DNMT3A CHIP (OR 1.81; p = 0.02). This study is the first to demonstrate an association between infection and incident CHIP in a general population, highlighting a modifiable risk factor for CHIP. Further work is required to describe the mutation-specific impact underlying this observed relationship.
Keywords: Infection, clonal hematopoiesis, CHIP, clonal hematopoiesis of indeterminate potential, infection frequency, incident CHIP
Category: Clinical Investigations
Teaser Abstract
Clonal Hematopoiesis of Indeterminant Potential (CHIP) is a malignancy precursor state that has been shown in preclinical models to be accelerated by infection. We leveraged an ongoing epidemiology study to determine the relationship between infection frequency and incident CHIP. CHIP was determined via whole-exome sequencing at two time points. Documented infection was assessed from routinely collected hospital discharge summaries. Participants with ≥ 3 documented infections had an increased odds of incident CHIP, especially large non-DNMT3A CHIP. This study is the first to demonstrate an association between infection and incident CHIP in a general population, highlighting a modifiable risk factor for CHIP.
Introduction
Clonal hematopoiesis of Indeterminant Potential (CHIP) is an age-related phenomenon characterized by clonal expansion of hematopoietic stem cells (HSCs)1. CHIP results from a fitness advantage conferred by naturally acquired somatic mutations, leading to an overrepresented clonal population in the bone marrow which can be assessed by sequencing peripheral blood leukocytes. The conventional definition of CHIP is the presence of a driver mutation with a variant allele frequency (VAF) of ≥ 2%.1 The genes implicated in CHIP, such as DNMT3A and TET2, are previously described driver mutations for hematologic malignancy.2,3 CHIP itself is considered a malignancy precursor state: individuals with CHIP have a 4-fold increased relative risk of hematologic malignancy.4 In addition, CHIP confers an increased risk of cardiovascular disease, other diseases of aging, and all-cause mortality.1,4–9 Recent studies demonstrate that CHIP clone size correlates with degree of risk: individuals with large CH (VAF ≥ 10%) have greater risk of cardiovascular disease and hematologic malignancy as compared to those with small CHIP (VAF 2–10%).4,6,7 Given the significant health impact of CHIP, defining factors that predispose individuals to CHIP is of significant interest.
Mechanistic studies highlight that CHIP is a product of both genetics and environmental factors.10–13 Clones expand in response to mutation-specific environmental drivers and selective pressures.14,15 Inflammatory states have been shown to drive expansion of several common clones.14–16 For example, chronic mycobacterial infection selects for DNMT3A-mutant hematopoietic stem cell (HSC) expansion relative to wild type HSCs in a competitive murine transplant model.12 Similarly, TET2-mutant HSCs expand after exposure to bacterial lipopolysaccharides.13 This relationship between infection and CHIP holds in clinical studies: there is an increased prevalence of CHIP in people living with HIV (PLWH) as compared to age-matched controls without HIV.17–23 These studies indicate that infection may be a powerful driver of CHIP. However, the relationship between infection and incident CHIP has not yet been determined as few clinical studies have serial assessments of CHIP.
We leveraged longitudinal Atherosclerosis Risk in Communities (ARIC) study data, which contains assessment of CHIP at two time points 20 years apart, to assess the association between infection frequency and incident CHIP and investigated whether the observed associations vary on the basis of driver mutation or clone size. We utilized hospital-documented and self-reported infection data to assess whether the site or number of different infections influenced CHIP risk.
Methods
Study Population and Design
The ARIC (Atherosclerosis Risk in Communities, RRID:SCR_021769) study is an ongoing prospective study designed to assess risk factors for cardiovascular disease and incidence of cardiovascular outcomes in adults recruited from 4 US communities (Forsyth County, North Carolina; Jackson, Mississippi; Minneapolis, Minnesota; and Washington County, Maryland). A total of 15,792 female and male, primarily Black and White participants aged 45 to 64 years old were recruited. Study enrollment (Visit 1) occurred in 1987–1989 and participants have completed subsequent study visits every few years. At each visit, participants underwent a physical examination, were interviewed, and had blood collected. Information collected included date of birth, race, and education at Visit 1; height at Visit 1 and 3–5; and smoking, weight, diabetes status, and dyslipidemia at each of Visits 1 to 5. Annually, participants were called by telephone to update their medical and exposure histories. As part of the ARIC protocol, hospital discharge summaries were routinely abstracted. The ARIC study protocol was approved by the institutional review boards of all participating centers and all participants provided written informed consent at each visit.
For this study, we developed a closed cohort of ARIC participants who had CHIP status ascertained by whole exome sequencing (WES) at two time points separated by approximately 20 years as previously described.24 For the majority (76%), DNA from peripheral blood leukocytes collected at Visit 2 (1990 – 1992) was used for the first time point; for the remainder of participants, DNA from blood collected at another time point prior to Visit 5 was used. DNA from Visit 5 (2011–2013) was used for the second time point (Figure 1). We excluded participants who were not Black or White (due to small sample size), who had a cancer history of any type including hematologic malignancy prior to the first time point, who had a hematologic malignancy between the first and second time point, and who did not provide full consent. To be able to study incident CHIP, we also excluded participants with CHIP at the first time point. Ultimately, 3,367 participants were included in the closed cohort (Figure 2).
Figure 1.
Study Timeline
Figure 2.
Schema for Inclusion of Study Participants in Analysis
* Two Black ARIC study participants (1 from the Minneapolis, Minnesota field center and 1 from the Washington County, Maryland field center) were not included in this analysis due to race/field center discordance.
Assessment of Infection Site and Frequency
The primary exposure was hospital-documented infection history. Hospital discharge summaries are collected and abstracted for ARIC study participants following participant response to annual telephone calls that included questions about hospitalization since the last call and routine surveillance of hospitals in the areas where ARIC participants were recruited. Infections included in our analysis were identified by International Statistical Classification of Diseases and Related Health Problems, Tenth Revision (ICD-10) codes which have been previously described and provided in Tables S1 and S2.25 We included infections from Visit 1 up to the date of Visit 5 in our analysis. Infection sites included upper respiratory tract infections, lower respiratory tract infections, gastrointestinal tract infections, urinary tract infections, skin and soft-tissue infections, nervous system infections, sepsis, musculoskeletal system infections, and cardiovascular system infections. For frequency, we summed the number of documented infections within and among the hospitalizations. For sensitivity analysis, we separately considered the effects of infections 1) between Visit 1 and the median date of hospitalization with a documented infection and 2) between the median date of hospitalization with a documented infection and Visit 5. This sensitivity analysis allowed us to consider the effect of infection recency on CHIP incidence.
Assessment of CHIP Status, VAF, and Mutation Type
CHIP status was ascertained at two time points by whole-exome sequencing performed on DNA extracted from peripheral blood leukocytes at the Broad Institute (Time 1; HiSeq 2000) and Baylor College of Medicine (Time 2; NovaSeq 6000). CHIP mutations were identified with the GATK MuTect2 (Broad Institute) somatic variant caller26 based on 50 previously identified genes variants that promote clonal expansion of hematopoietic stem cells.3,27,28 Per convention, variant allele frequency (VAF) ≥ 2% was used to identify CHIP, and VAF ≥ 10% was considered large CHIP. Participants with VAF <2% were considered not to have CHIP. Concordance between the two platforms has been reported (83% for VAF≥2%, 93% for VAF≥10%).24
Statistical Analysis
Multivariable logistic regression models were used to estimate odds ratios (OR) and 95% confidence intervals (CI) for the association of infection frequency (0 [reference], 1, 2, ≥ 3) and site, with incident CHIP, large CHIP, and mutation type (DNMT3A, TET2, non-DNMT3A). Models were run separately for documented and self-reported infections. Models were adjusted for potential confounders including age (continuous), terms for race and center (Black from Jackson, Black from Forsyth, White from Forsyth County, White from Washington County, White from Minneapolis [reference]), sex (male [reference], female), education level (basic [reference], intermediate, advanced), cigarette smoking status (current, former, never [reference]), alcohol drinking (current, former, never [reference]), BMI (continuous), diabetes status (diagnosed diabetes, undiagnosed diabetes, at risk for diabetes, normoglycemic [reference]), and dyslipidemia (yes, no [reference]). To account for confounding by age, we modeled each descriptive variable as the outcome and CHIP status as the predictor, adjusting for age. For categorical variables, logistic regression (PROC LOGISTIC) was used to estimate age-adjusted proportions by CHIP status. For continuous variables, generalized linear mixed models (PROC GLIMMIX) were used to estimate age-adjusted means by CHIP status. Models were adjusted for race and study recruitment site, sex, education level, cigarette smoking status, alcohol use, BMI, and comorbidities (diabetes and dyslipidemia). We tested for trend across number of infections by entering into the model an ordinal term with values of 0, 1, 2, and 3, and evaluating the coefficient using Wald test. All analyses were run using SAS v. 9.4 and reported p-values are from 2-sided tests with statistical significance set at p < 0.05.
Results
Study Participant Characteristics
Baseline characteristics are summarized in Table 1 and compared to characteristics at study end in Table 2. Study participants had a mean age of 55.3 years at the first sequencing time point; the second sequencing time point was approximately 20 years later. Of the 3,367 participants, 59.1% were female and 40.9% were male; 76% were White and 24% were Black. At study conclusion, 19.7% had incident CHIP, of which 6.9% had large CHIP, and 5.2% had non-DNMT3A large CHIP.
Table 1.
Age-Adjusted Baseline Characteristics of Participants at Initial Sequencing Visit by Subsequent Incident CHIP status in ARIC
| Characteristics | All Participants (n = 3,367) |
Participants without CHIP (n = 2,704) |
Participants with Incident CHIP (n = 663) |
|---|---|---|---|
| Age, mean (years) | 55.3 | 55.1 | 56.1 |
| Female | 59.1% | 60.6% | 53.1% |
| Male | 40.9% | 39.4% | 46.9% |
| Black | 24% | 24% | 24% |
| White | 76% | 76% | 76% |
| BMI, mean (kg/m2) | 27.8 | 27.9 | 27.5 |
| Study Recruitment Site | |||
| Forsyth County, North Carolina | 17.7% | 17.6% | 18.0% |
| Jackson, Mississippi | 22.3% | 22.1% | 23.0% |
| Minneapolis, Minnesota | 36.0% | 35.8% | 37.0% |
| Washington County, Maryland | 24.1% | 24.6% | 22.0% |
| Education Level * | |||
| Basic | 13.3% | 13.0% | 14.7% |
| Intermediate | 41.4% | 41.7% | 40.1% |
| Advanced | 45.3% | 45.3% | 45.2% |
| Alcohol Use | |||
| Current | 61.1% | 61.4% | 60.1% |
| Former | 17.7% | 17.4% | 18.9% |
| Never | 21.2% | 21.2% | 21.0% |
| Tobacco Use | |||
| Current | 14.4% | 13.9% | 16.5% |
| Former | 41.2% | 41.1% | 41.8% |
| Never | 44.4% | 45.0% | 41.7% |
| Smoking Pack Years (ever-smokers), years | 19.5 | 19.5 | 19.5 |
| Co-morbidities | |||
| Coronary Heart Disease | 3.2% | 3.1% | 3.4% |
| Diabetes Mellitus | |||
| Diagnosed Diabetes | 4.8% | 4.7% | 5.0% |
| Undiagnosed Diabetes | 4.9% | 4.9% | 4.9% |
| At-Risk Diabetes | 46.9% | 47.3% | 45.4% |
| Normoglycemic | 43.5% | 43.2% | 44.7% |
| Hypertension | 23.6% | 23.4% | 24.2% |
| Dyslipidemia | 51.0% | 50.9% | 51.4% |
Categories of education are defined as follows: 1) Basic = less than completion of high school; 2) Intermediate = completed high school or equivalent; 3) Advanced = completed at least some college (even if just one year)
Abbreviations: ARIC = Atherosclerosis Risk in Communities; BMI = body mass index; CHIP = Clonal Hematopoiesis of Indeterminant Potential; kg = kilogram; m = meters; n = number
Table 2.
Age-Adjusted Baseline Characteristics of Participants at Second Sequencing Visit by Subsequent Incident CHIP status in ARIC
| Characteristics | All Participants (n = 3,367) |
Participants without CHIP (n = 2,704) |
Participants with Incident CHIP (n = 663) |
|---|---|---|---|
| Age, mean (years) | 75.5 | 75.2 | 76.5 |
| BMI, mean (kg/m2) | 28.8 | 28.8 | 28.7 |
| Alcohol Use | |||
| Current | 50.7% | 51.1% | 49.1% |
| Former | 28.1% | 27.6% | 30.3% |
| Never | 21.2% | 21.3% | 20.6% |
| Tobacco Use | |||
| Current | 3.6% | 3.6% | 3.5% |
| Former | 55.3% | 54.6% | 58.2% |
| Never | 41.2% | 41.9% | 38.4% |
| Smoking Pack Years (ever-smokers), years | 20.8 | 20.7 | 21.2 |
| Co-morbidities | |||
| Coronary Heart Disease | 13.6% | 13.1% | 15.4% |
| Diabetes Mellitus | |||
| Diagnosed Diabetes | 31.1% | 31.8% | 27.9% |
| Undiagnosed Diabetes | 1.4% | 1.4% | 1.3% |
| At-Risk Diabetes | 67.6% | 66.8% | 70.8% |
| Normoglycemic | 0.0% | 0.0% | 0.0% |
| Hypertension | 67.5% | 66.7% | 70.8% |
| Dyslipidemia | 42.7% | 42.6% | 43.3% |
Abbreviations: ARIC = Atherosclerosis Risk in Communities; BMI = body mass index; CHIP = Clonal Hematopoiesis of Indeterminant Potential; kg = kilogram; m = meters; n = number
Association between Infection Frequency and Incident CH
Compared to study participants without documented infections, those with ≥ 3 documented infections had increased odds of incident CHIP (OR = 1.41, p = 0.03). This association was even stronger when considering ≥ 3 documented infections and large CHIP (OR = 1.81, p = 0.008). These observed associations varied when stratified by CHIP mutation (Table 3). The strongest association observed was between 1 documented infection (OR = 1.51, p = 0.01) and DNMT3A CHIP (Table 3). Of the 282 individuals with DNMT3A CHIP, 21 had a R882 hotspot mutation. The association between infection and R882 DNMT3A CHIP was stronger (OR = 2.28, p = 0.07) than for non-R882 DNMT3A CHIP (OR = 1.17, p = 0.30). We observed a stepwise positive trend between 1 (OR = 0.91), 2 (OR = 1.11), and ≥ 3 (OR = 1.44) infections and TET2 CH (Table 3) across clone size (Table S3). As the other CHIP mutations in the cohort were less frequent and could not be evaluated independently, we assessed the relationship between infection and all CHIP mutations excluding DNMT3A. We also observed a stepwise positive trend between 1 (OR = 1.10), 2 (OR = 1.18), and ≥ 3 (OR = 1.47) hospital documented infections and non-DNMT3A CHIP (Table 3), of which the strongest association was between ≥ 3 infections and large non-DNMT3A CHIP (Figure 3). Altogether, the data showed a strong association between infection and CHIP, including both DNMT3A and non-DNMT3A CHIP.
Table 3.
Association between Hospital Documented Infection Frequency and Incident CHIP in ARIC
| Any CHIP | DNMT3A CHIP | TET2 CHIP | non-DNMT3A CHIP | ||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Number of infections | Study participants, n | Events | p-value | Odds Ratio (95% CI) | p-trend | Events | p-value | Odds Ratio (95% CI) | p-trend | Events | p-value | Odds Ratio (95% CI) | p-trend | Events | p-value | Odds Ratio (95% CI) | p-trend |
| 0 | 2,452 | 450 | --- | 199 | --- | 104 | --- | 293 | --- | ||||||||
| 1 | 459 | 105 | 0.06 | 1.26 (0.99, 1.61) | 0.04 | 51 | 0.01 | 1.52 (1.09, 2.12) | 0.5 | 20 | 0.7 | 0.91 (0.56, 1.50) | 0.3 | 64 | 0.5 | 1.10 (0.81, 1.49) | 0.03 |
| 2 | 207 | 36 | 0.9 | 0.98 (0.67, 1.42) | 11 | 0.3 | 0.70 (0.37, 1.32) | 10 | 0.8 | 1.11 (0.56, 2.18) | 28 | 0.5 | 1.18 (0.77, 1.80) | ||||
| ≥ 3 | 249 | 63 | 0.03 | 1.41 (1.03, 1.92) | 21 | 0.5 | 1.18 (0.73, 1.91) | 17 | 0.2 | 1.44 (0.83, 2.49) | 47 | 0.04 | 1.47 (1.03, 2.10) | ||||
The reference group, highlighted in grey, for these analyses are individuals without any hospital documented infections.
Abbreviations: ARIC = Atherosclerosis Risk in Communities; CHIP = Clonal Hematopoiesis of Indeterminant Potential; n = number; CI = confidence interval
Figure 3.
Association between Hospital Documented Infection Frequency and non-DNMT3A CHIP by Clone Size
Association between Infection Site and Incident CHIP
Documented infection was positively associated with incident CHIP regardless of the number of different infection sites (Table S4); the strongest observed association was for individuals with ≥ 3 infection sites and large CH (OR = 1.91, p = 0.04). There were no significant associations between infection site and incident CH (Table 4).
Table 4.
Association between Infection Site and Incident CHIP in ARIC
| Infection Site | Study participants, n | p-value | Odds Ratio (95% CI) |
|---|---|---|---|
| Upper Respiratory Tract Infection | 42 | 0.7 | 1.16 (0.56, 2.40) |
| Lower Respiratory Tract Infection | 654 | 0.8 | 1.03 (0.82, 1.30) |
| Gastrointestinal Tract Infection | 264 | 0.2 | 1.23 (0.90, 1.68) |
| Urinary Tract Infection | 171 | 0.8 | 1.05 (0.71, 1.55) |
| Prostatitis | 18 | 0.8 | 0.85 (0.27, 2.63) |
| Vaginal Infection | 24 | 0.8 | 1.17 (0.43, 3.21) |
| Skin and Soft Tissue Infection | 51 | 0.3 | 1.36 (0.72, 2.57) |
| Nervous System Infection | 14 | 0.5 | 0.59 (0.13, 2.70) |
| Sepsis | 56 | 0.5 | 0.79 (0.40, 1.56) |
| Musculoskeletal Infection | 90 | >0.9 | 0.99 (0.59, 1.68) |
| Cardiovascular Infection | 14 | 0.6 | 0.65 (0.14, 2.97) |
The reference group for these analyses are individuals without infection.
Abbreviations: ARIC = Atherosclerosis Risk in Communities; n = number; CH = Clonal Hematopoiesis of Indeterminant Potential; CI = confidence interval
Association between Infection Recency and Incident CHIP
Infection is a known driver of clonal expansion in preclinical models. Although clinical epidemiology studies in PLWH suggest that infection impacts CH acquisition in humans, other studies reveal that CH inherently increases risk of infection acquisition.29–32 As our two sequencing timepoints are separated by twenty years, the relative timing of CH acquisition and hospitalization for infection is uncertain. To assess the relationship between infection timing and incident CHIP and mitigate reverse causation, we performed a recency analysis. We hypothesized that if infections drive CHIP, then earlier infections would have a stronger positive association with CHIP. Conversely, if infections are a result of CHIP, then we would observe a stronger association between recent infection and CHIP. We therefore stratified the study cohort by the median date of infection (4/10/2008). There was a significant association between a single infection documented before the median date and large CHIP (OR = 2.65, p = 0.01), providing evidence in support of infection as a driver of CHIP (Table S5).
Discussion
Prior preclinical studies demonstrate a relationship between infection and CHIP, while clinical studies have shown an association between HIV and CHIP.12,13,17–23 However, it is unknown whether there is a broad association between infections other than HIV and the development of CHIP in human populations. In this longitudinal study in a general population of middle-aged and older adults, we assessed the association between infection history and incident CHIP over an approximately 20-year time frame. We have leveraged the ARIC study – the only large cohort study with serial sequencing and infection data – and utilized epidemiologic approaches to determine the impact of infection on incident CH. We found that people with ≥ 3 infections documented in hospitalization records since cohort recruitment were more likely to develop incident CHIP compared with those without such history. We conclude that infection frequency may influence the development of CHIP.
Previous reports highlight that CHIP can increase likelihood of subsequent infection, such as sepsis, or more severe infection.29–32 This work is the first to assess the reverse relationship: the association between specific acute infections and subsequent CHIP. Our analysis suggests that the specific infection is not as important as the number of infectious challenges, consistent with preclinical data demonstrating that inflammation triggered by infection provides a selective advantage for mutant clones. Given the small numbers of individuals with specific infections, additional analyses in pooled cohorts would be ideal to determine if particular types of infections or specific pathogens are stronger drivers of incident CHIP. The limitations of our work despite the data available for the ARIC cohort highlights the need for additional cohorts with multiple sequencing timepoints and comprehensive clinical data. We assessed the relationship between the number of different infection sites and CHIP. We observed positive associations between hospital documented infection sites and incident CHIP. This was unsurprising given those with more infection sites invariably have increased infection frequency.
Our study design, specifically use of a cohort with two sequencing timepoints separated by 20 years, afforded us the unique opportunity to assess the effect of infection frequency on incident CHIP. However, our approach does not allow us to exclude the known effect of CHIP on infection susceptibility as we are unable to determine the exact time point of CHIP acquisition relative to the timing of CHIP incidence. To address this limitation, we stratified our cohort by infection frequency and recency. We observed strong associations between a single infection prior to the median date of hospitalization for infection and overall CHIP as well as large CHIP which makes clear that there is some impact of infection on CHIP development. However, we also noted that of those with ≥ 3 infections, individuals who had a recent infection (after the median date of hospitalization for infection in the cohort) had increased likelihood of incident CHIP. This association is stronger than what we observed for those with ≥ 3 infections without an infection following the median date. This difference could be a product of sample size as there were more individuals in the cohort with infections following the median date as compared to prior to the median date. Alternatively, this could indeed reflect an effect of CHIP on infection susceptibility.
Our study design is necessary to answer the study question but introduces survivor bias: as study participants were required to survive to the second sequencing timepoint, those who succumbed to infection were not included in our analysis. This could result in underestimation of the impact of infection on CHIP incidence. Additionally, 30% of ARIC participants alive at the second sequencing timepoint were unwilling or unable to attend the study visit, further limiting our study population. Moreover, we are unable to capture infections diagnosed in the outpatient setting which limits our ability to assess the relationship between infection severity and CHIP. We utilized ICD codes to determine infection frequency which is not as sensitive or specific as blood cultures or other diagnostic investigations; additionally, we were unable to determine the pathogenic agent for each documented infection. We hope that these questions can be addressed in future studies.
Despite the overall large size of the study, the final sample size does not provide sufficient power for sub analyses by specific infection or specific mutation. For example, due to the rarity of some CHIP mutations in this cohort, we had insufficient power to investigate the relationship between infection and specific mutations aside from DNMT3A and TET2. While this observational study cannot determine causation, this work builds upon and corroborates prior experimental research. Furthermore, we used the conventional definition of VAF ≥ 2% to define CHIP and so rarer clones were not included in this analysis. As previously described, only 50 genes implicated in CHIP were included in this analysis, thus quite a number of rare CHIP associated mutations were not represented in this study. Despite these limitations, our work suggests that the relationship between infection frequency and incident CH varies according to the CHIP driver mutation. A single hospital-documented infection is positively associated with incident DNMT3A CHIP, with weaker associations observed for 2 and ≥ 3 infections. In contrast, we observe a positive trend between infection frequency and incident TET2 CHIP. We surmise that this trend did not reach statistical significance because of the small number of individuals with incident TET2 CHIP. A significant stepwise positive trend hold when we assess the relationship between infection frequency and non-DNMT3A CHIP. Recent work suggest that the mechanism of clonal expansion may vary based on the mutation. Previous work has implicated TCL1A as a promoter of clonal expansion: TCL1A is not typically expressed in DNMT3A-mutated or wildtype HSCs as compared to TET2.33 It is possible that repeat infectious challenges may promote TCL1A expression. Mechanistic work to assess the relationship between infection, TCLA1 expression, and CH would be of great value. Alternatively, TET2 clones may simply be more sensitive to inflammatory cues compared to DNMT3A clones.
Importantly, this is the first study to show a positive association between infection and incident CHIP. Using a unique resource that includes assessment of CHIP at two time points separated by 20 years along with extensive medical records, we were able to document a positive association between infection history and CHIP incidence. Recency analysis indicates that this association is not explained by an inverse association between CHIP and infection susceptibility. Additional work is necessary to determine if the relationship between infection frequency and CHIP is affected by infection type or treatment.
Supplementary Material
Highlights.
Infection is associated with incident CH in the general population
Individuals with ≥ 3 documented infection have increased odds of CH acquisition
The relationship between infection frequency and incident CH may vary with mutation.
Acknowledgments:
This work was supported by grants from the American Society of Hematology and the National Institutes of Health. A.L. is supported by the ASH Hematology Inclusion Pathway Fellow Award, T32 in Infection and Immunity (AI 055413–19), and NIH Loan Repayment Award (1L70HL175832–01). K.Y.K is supported by NIH R35(HL155672); K.Y.K and E.A.P are supported by P01(CA265748). The Atherosclerosis Risk in Communities study has been funded in whole or in part with federal funds from the National Heart, Lung, and Blood Institute, National Institutes of Health, Department of Health and Human Services, under contract nos. (75N92022D00001, 75N92022D00002, 75N92022D00003, 75N92022D00004, 75N92022D00005). This work is also supported by R01 HL148050. Funding support for “Building on GWAS for NHLBI-diseases: the U.S. CHARGE consortium” was provided by the NIH through the American Recovery and Reinvestment Act of 2009 (ARRA) (5RC2HL102419). Sequencing was carried out at the Baylor College of Medicine Human Genome Sequencing Center (U54 HG003273 and R01 HL086694). Studies on cancer in ARIC are also supported by the National Cancer Institute (U01 CA164975). The authors thank the staff and participants of the ARIC study for their important contributions. Cancer data was provided by the Maryland Cancer Registry, Center for Cancer Prevention and Control, Maryland Department of Health, with funding from the State of Maryland and the Maryland Cigarette Restitution Fund. The collection and availability of cancer registry data are also supported by the Cooperative Agreement NU58DP006333, funded by the Centers for Disease Control and Prevention. Its contents are solely the responsibility of the authors and do not necessarily represent the official views of the Centers for Disease Control and Prevention or the Department of Health and Human Services.
Footnotes
Disclosure of Conflicts of Interest:
C.M.B. reports grant/research support (through his institution) from Abbott Diagnostic, Akcea, Amgen, Arrowhead, Eli Lilly, Ionis, Merck, New Amsterdam, Novartis, Novo Nordisk, Roche Diagnostic, NIH, AHA, ADA, and consulting fees from 89Bio, Abbott Diagnostics, Amgen, Arrowhead, Astra Zeneca, Denka Seiken, Eli Lilly, Esperion, Genentech, Ionis, Merck, New Amsterdam, Novartis, Novo Nordisk, Roche Diagnostic. P.N. reports research grants from Allelica, Amgen, Apple, Boston Scientific, Cleerly, Genentech / Roche, Ionis, Novartis, and Silence Therapeutics, personal fees from AIRNA, Allelica, Apple, AstraZeneca, Bain Capital, Blackstone Life Sciences, Bristol Myers Squibb, Creative Education Concepts, CRISPR Therapeutics, Eli Lilly & Co, Esperion Therapeutics, Foresite Capital, Foresite Labs, Genentech / Roche, GV, HeartFlow, Magnet Biomedicine, Merck, Novartis, Novo Nordisk, TenSixteen Bio, and Tourmaline Bio, equity in Bolt, Candela, Mercury, MyOme, Parameter Health, Preciseli, and TenSixteen Bio, royalties from Recora for intensive cardiac rehabilitation, and spousal employment at Vertex Pharmaceuticals, all unrelated to the present work. The remaining authors declare no competing interests.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- 1.Jaiswal S, Fontanillas P, Flannick J, et al. Age-Related Clonal Hematopoiesis Associated with Adverse Outcomes. New England Journal of Medicine. 2014;371(26):2488–2498. doi: 10.1056/nejmoa1408617 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Bernstein N, Spencer Chapman M, Nyamondo K, et al. Analysis of somatic mutations in whole blood from 200,618 individuals identifies pervasive positive selection and novel drivers of clonal hematopoiesis. Nat Genet. 2024;56(6):1147–1155. doi: 10.1038/s41588-024-01755-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Bick AG, Weinstock JS, Nandakumar SK, et al. Inherited causes of clonal haematopoiesis in 97,691 whole genomes. Nature. 2020;586(7831):763–768. doi: 10.1038/s41586-020-2819-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Singh J, Li N, Ashrafi E, et al. Clonal hematopoiesis of indeterminate potential as a prognostic factor: a systematic review and meta-analysis. Blood Adv. 2024;8(14):3771–3784. doi: 10.1182/bloodadvances.2024013228 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Jaiswal S Clonal hematopoiesis and nonhematologic disorders. Blood. 2020;136(14):1606–1614. doi: 10.1182/blood.2019000989 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Weeks LD, Ebert BL. Causes and consequences of clonal hematopoiesis. Blood. 2023;142(26):2235–2246. doi: 10.1182/blood.2023022222 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Saadatagah S, Uddin MM, Weeks LD, et al. Clonal Hematopoiesis Risk Score and All-Cause and Cardiovascular Mortality in Older Adults. JAMA Netw Open. 2024;7(1):E2351927. doi: 10.1001/jamanetworkopen.2023.51927 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Saadatagah S, Naderian M, Uddin M, et al. Atrial Fibrillation and Clonal Hematopoiesis in TET2 and ASXL1. JAMA Cardiol. 2024;9(6):497–506. doi: 10.1001/jamacardio.2024.0459 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.van Zeventer IA, Salzbrunn JB, de Graaf AO, et al. Prevalence, predictors, and outcomes of clonal hematopoiesis in individuals aged $80 years. Blood Adv. 2021;5(8):2115–2122. doi: 10.1182/BLOODADVANCES.2020004062 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Pasupuleti SK, Ramdas B, Burns SS, et al. Obesity-induced inflammation exacerbates clonal hematopoiesis. Journal of Clinical Investigation. 2023;133(11). doi: 10.1172/JCI163968 [DOI] [Google Scholar]
- 11.Levin MG, Nakao T, Zekavat SM, et al. Genetics of smoking and risk of clonal hematopoiesis. Sci Rep. 2022;12(1). doi: 10.1038/s41598-022-09604-z [DOI] [Google Scholar]
- 12.Hormaechea-Agulla D, Matatall KA, Le DT, et al. Chronic infection drives Dnmt3a-loss-of-function clonal hematopoiesis via IFNγ signaling. Cell Stem Cell. 2021;28(8):1428–1442.e6. doi: 10.1016/j.stem.2021.03.002 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Meisel M, Hinterleitner R, Pacis A, et al. Microbial signals drive pre-leukaemic myeloproliferation in a Tet2-deficient host. Nature. 2018;557(7706):580–584. doi: 10.1038/s41586-018-0125-z [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Florez MA, Tran BT, Wathan TK, DeGregori J, Pietras EM, King KY. Clonal hematopoiesis: Mutation-specific adaptation to environmental change. Cell Stem Cell. 2022;29(6):882–904. doi: 10.1016/j.stem.2022.05.006 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.King KY, Huang Y, Nakada D, Goodell MA. Environmental influences on clonal hematopoiesis. Exp Hematol. 2020;83:66–73. doi: 10.1016/j.exphem.2019.12.005 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Jakubek YA, Reiner AP, Honigberg MC. Risk factors for clonal hematopoiesis of indeterminate potential and mosaic chromosomal alterations. Translational Research. 2023;255:171–180. doi: 10.1016/j.trsl.2022.11.009 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Bhattacharya R, Uddin MM, Patel AP, et al. Risk factors for clonal hematopoiesis of indeterminate potential in people with HIV: a report from the REPRIEVE trial. Blood Adv. 2024;8(4):959–967. doi: 10.1182/bloodadvances.2023011324 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Van Der Heijden WA, Van Deuren RC, Van De Wijer L, et al. Clonal Hematopoiesis Is Associated With Low CD4 Nadir and Increased Residual HIV Transcriptional Activity in Virally Suppressed Individuals With HIV. Journal of Infectious Diseases. 2022;225(8):1339–1347. doi: 10.1093/infdis/jiab419 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Wang S, Pasca S, Post WS, et al. Clonal hematopoiesis in men living with HIV and association with subclinical atherosclerosis. AIDS. 2022;36(11):1521–1531. doi: 10.1097/QAD.0000000000003280 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Knudsen AD, Eskelund CW, Benfield T, et al. Clonal hematopoiesis of indeterminate potential in persons with HIV. AIDS. 2024;38(4):487–495. doi: 10.1097/QAD.0000000000003788 [DOI] [PubMed] [Google Scholar]
- 21.Vorri SC, Christodoulou I, Karanika S, Karantanos T. Human Immunodeficiency Virus and Clonal Hematopoiesis. Cells. 2023;12(5). doi: 10.3390/cells12050686 [DOI] [Google Scholar]
- 22.Bick AG, Popadin K, Thorball CW, et al. Increased prevalence of clonal hematopoiesis of indeterminate potential amongst people living with HIV. Sci Rep. 2022;12(1). doi: 10.1038/s41598-021-04308-2 [DOI] [Google Scholar]
- 23.Dharan NJ, Yeh P, Bloch M, et al. HIV is associated with an increased risk of age-related clonal hematopoiesis among older adults. Nat Med. 2021;27(6):1006–1011. doi: 10.1038/s41591-021-01357-y [DOI] [PubMed] [Google Scholar]
- 24.Uddin MM, Saadatagah S, Niroula A, et al. Long-term longitudinal analysis of 4,187 participants reveals insights into determinants of clonal hematopoiesis. Nature Communications. 2024;15(1). doi: 10.1038/s41467-024-52302-9 [DOI] [Google Scholar]
- 25.Bohn B, Lutsey PL, Misialek JR, et al. Incidence of Dementia Following Hospitalization With Infection Among Adults in the Atherosclerosis Risk in Communities (ARIC) Study Cohort. JAMA Netw Open. 2023;6(1):e2250126. doi: 10.1001/jamanetworkopen.2022.50126 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Benjamin D, Sato T, Cibulskis K, Getz G, Stewart C, Lichtenstein L. Calling Somatic SNVs and Indels with Mutect2. Published online December 2, 2019. doi: 10.1101/861054 [DOI] [Google Scholar]
- 27.Jaiswal S, Natarajan P, Silver AJ, et al. Clonal Hematopoiesis and Risk of Atherosclerotic Cardiovascular Disease. New England Journal of Medicine. 2017;377(2):111–121. doi: 10.1056/nejmoa1701719 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Gibson CJ, Lindsley RC, Tchekmedyian V, et al. Clonal hematopoiesis associated with adverse outcomes after autologous stem-cell transplantation for lymphoma. Journal of Clinical Oncology. 2017;35(14):1598–1605. doi: 10.1200/JCO.2016.71.6712 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Zekavat SM, Lin SH, Bick AG, et al. Hematopoietic mosaic chromosomal alterations increase the risk for diverse types of infection. Nat Med. 2021;27(6):1012–1024. doi: 10.1038/s41591-021-01371-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Schenz J, Rump K, Siegler BH, et al. Increased prevalence of clonal hematopoiesis of indeterminate potential in hospitalized patients with COVID-19. Front Immunol. 2022;13. doi: 10.3389/fimmu.2022.968778 [DOI] [Google Scholar]
- 31.Ronchini C, Caprioli C, Tunzi G, et al. High-sensitivity analysis of clonal hematopoiesis reveals increased clonal complexity of potential-driver mutations in severe COVID-19 patients. PLoS One. 2024;19(1 January). doi: 10.1371/journal.pone.0282546 [DOI] [Google Scholar]
- 32.Bolton KL, Koh Y, Foote MB, et al. Clonal hematopoiesis is associated with risk of severe Covid-19. Nat Commun. 2021;12(1). doi: 10.1038/s41467-021-26138-6 [DOI] [Google Scholar]
- 33.Weinstock JS, Gopakumar J, Burugula BB, et al. Aberrant activation of TCL1A promotes stem cell expansion in clonal haematopoiesis. Nature. 2023;616(7958):755–763. doi: 10.1038/s41586-023-05806-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.