CHIP Happens: Clonal Hematopoiesis of Indeterminate Potential and Its Relationship to Solid Tumors

As a person ages, somatic mutations accumulate in blood and bone marrow cells. Some of these mutations, particularly those in leukemia-associated genes, can lead to the clonal expansion of hematopoietic stem cells (HSC). When detected in the blood of people with normal hematologic parameters, this phenomenon is called clonal hematopoiesis of indeterminate potential (CHIP). Commonly related to smoking and aging, CHIP is associated with an increased risk for hematologic malignancy, cardiovascular disease (CVD), and all-cause mortality (1–3). This is summarized in Fig. 1. Recent interest in CHIP has focused on how such mutations are acquired, how they are associated with poor outcomes, and how patients with CHIP can be clinically managed. Research has begun to uncover key roles for self-renewal, differentiation, DNA damage repair pathways, and inflammatory signaling in CHIP development and resulting outcomes.

Despite this progress, the role of CHIP in solid tumor progression and patient outcomes is less understood. Given the importance of immune surveillance and the knowledge that patients treated with chemotherapy and radiation for solid tumors have higher prevalence of CHIP and particular CHIP mutations (4), it is critical to understand how these alterations in immune cells impact tumor immunology. This review will summarize what is known about CHIP more broadly, review the existing literature about CHIP in solid tumors, and discuss the consequences of CHIP on non-hematologic malignancies.

Clonal hematopoiesis in nonmalignant conditions was first characterized in observations of X-inactivation skewing in the peripheral blood of healthy elderly women with normal blood counts (5–7). Females are expected to have approximately equal inactivation of each X chromosome; therefore, any skew in this ratio is indicative of preferential expansion of a population of cells with the same X-inactivation pattern. Advances in sequencing technologies have allowed for identification of mutations in individual genes associated with X-inactivation skewing (8) and with expansion of nonmalignant HSC populations (9).

In 2014, three landmark papers (1–3) showed that CHIP is a common entity defined by the expansion of subclonal populations of hematopoietic cells with mutations in genes associated with myeloid malignancies in otherwise healthy people. Furthermore, these and more recent studies identified a strong association between aging and the prevalence of CHIP, with the phenomenon affecting at least 10% of people over 70 years old and nearly 20% of people over the age of 90 (10–14). These estimations are likely an underestimation, with true prevalence varying by disease state, detection limit, and variant-calling method. Patients with CHIP have a 0.5% to 1% risk each year for developing hematologic malignancy, compared with < 0.1% for people without CHIP (1). This risk is associated with both the abundance of the subclonal population and the number of CHIP mutations present (15–17). Patients with CHIP also have an increased risk of all-cause mortality, coronary heart disease, and ischemic stroke (1, 2, 13).

Currently, CHIP is defined by the presence of a somatic leukemia-associated mutation at a variant allele frequency (VAF) of at least 2% in the blood or bone marrow of a person without a hematologic malignancy (18, 19). CHIP mutations are generally heterozygous; therefore, a VAF of 2% indicates that 4% of leukocytes carry the alteration. The term clonal hematopoiesis is used more broadly to encompass any nonmalignant or nonimmune reactive expansion of a hematopoietic cell clone. It can be defined by multiple parameters, including the total number of somatic mutations in the hematopoietic compartment (13), overrepresentation of a hematopoietic cell clone even in the absence of an identifiable mutation (1), or the presence of mosaic chromosomal alterations (20–23). While these entities are clinically and scientifically important, this review focuses specifically on CHIP as defined by somatic single-nucleotide variants, insertions, and deletions in myeloid neoplasia-associated genes.

The most common CHIP-associated mutations occur in the epigenetic modifiers DNMT3A, TET2, and ASXL1 (24, 25) and confer loss of protein function (1–3). Many of these mutations, which can occur in all hematopoietic lineages (12), are often acquired early in the development of acute myeloid leukemia (AML; ref. 25) and have been shown to increase the self-renewal capacity of preleukemic HSCs (26). While the effect of each specific mutation on CHIP development and progression is still being elucidated, research to date has highlighted the role of epigenetic regulation in the self-renewal, differentiation, and oncogenic transformation of HSCs (27). These concepts are extensively reviewed by previous authors (28, 29). For the purpose of this review, a brief summary is provided below.

DNMT3A is a DNA methyltransferase that participates in de novo methylation (30, 31). Multiple groups have demonstrated that hematopoietic loss of Dnmt3a in mice can drive the expansion of HSCs while decreasing their differentiation capacity, particularly in the setting of progressive bone marrow transplantation (31–33). The mechanism of increased self-renewal in Dnmt3a-null HSCs has been demonstrated to be due, in part, to changes in epigenetic regulation, including region-specific changes in 5-methylcytosine (5mC) levels and histone methylation (31–34).

TET2 is a methylcytosine dioxygenase that oxidizes 5mC to 5-hydroxymethylcytosine (27). Loss of TET2 function is often caused by truncating frameshift or nonsense mutations, or missense mutations in the catalytic regions, and is associated with DNA hypermethylation (8, 14, 27, 35). Research suggests that TET2 mutations promote expansion of the hematopoietic stem and progenitor compartment and alter blood cell differentiation (27, 35, 36). Tet2 loss in hematopoietic cells in mice also results in the premature differentiation of immature myeloid progenitor cells towards the monocyte/macrophage lineage, particularly in older animals and following bone marrow transplantation (35, 37, 38).

Finally, ASXL1, a member of the polycomb repressive complex, is known to be involved in multiple posttranslational modifications (29, 39–41). Mouse models of Asxl1 mutation have revealed expansion of mutant Asxl1 HSCs and derived cells in native hematopoiesis and aging settings (39, 42). However, other work has observed opposite effects of Asxl1 mutations, including the decreased ability for hematopoietic repopulation and increased HSC apoptosis (41, 43). The differential impact of Asxl1 mutations on the hematopoietic system reinforces the fact that though CHIP carriers may have similar clinical phenotypes, their mutation-specific mechanisms of action can vary significantly.

While the genes discussed here are most common overall, CHIP-associated mutations also occur in genes that are not epigenetic regulators, including JAK2, TP53, PPM1D, and SF3B1. In particular, mutations in genes related to DNA damage repair are increasingly being recognized for their interactions with solid tumors. Though many CHIP mutations are loss-of-function, some variants such as JAK2 V617F are gain-of-function variants (44). Some of these mutations, such as those in TP53, are considered preleukemic due to their established high risk of transformation to AML (45). However, the contributions of these mutations to the CHIP phenotype have been less studied and require further investigation. In addition to mutation-specific biology, the context of CHIP development and the presence of selective pressure on the hematopoietic system must be considered.

There is increasing evidence that patients with solid tumors have higher rates of CHIP than the general population. A study of over 10,000 Chinese patients with non-hematologic cancers found that 9.1% of patients under 40 and 23.1% of patients over 80 had CHIP-associated mutations detectable in their cell-free DNA (46). This rate of CHIP is higher than what has been reported in untreated populations (47), likely reflecting an enrichment for CHIP in patients with solid tumors. Similarly, analysis of nearly 9,000 American patients with non-hematologic cancers showed that 25.1% of patients had at least one clonal somatic mutation in hematopoietic cells (48). In this cohort, the majority of CHIP mutations were also present in the tumor at a much lower VAF, suggesting that mutation-bearing immune cells may infiltrate the tumor and contaminate tumor sequencing data. These estimated high rates of CHIP among patients treated for solid tumors warrant further investigation. Moreover, many genes implicated in CHIP are also implicated in solid tumors, underscoring the need for careful variant calling pipelines. Sequencing blood from patients with cancer may uncover immune cell–specific somatic mutations, cancer-specific mutations from circulating tumor DNA, and germline mutations.

The rate of CHIP varies significantly across cancer types and based on patient variables. For example, one study identified the highest incidence of CHIP among patients with thyroid and bladder cancer, with the lowest incidence among patients with germ cell tumors (48). Conversely, a separate study observed high rates of CHIP in patients with melanoma and bladder cancer and noted a low incidence of CHIP in patients with thyroid cancer (46). A third study found an increased association between CHIP and incident lymphoma, lung, and kidney cancers (49). Finally, a recent study identified that patients with a DNMT3A CHIP mutation at a VAF > 10% have an increased risk of lung and prostate cancer (medRxiv 2021.12.29.21268342).

It is important to note that most studies about CHIP and solid tumors involve patients who received at least one line of therapy prior to sample collection. Research has linked genotoxic stress, including cytotoxic chemotherapy and radiotherapy, to CHIP, particularly in genes associated with DNA damage response such as TP53, PPM1D, and CHEK2 (50–57). Examination of CHIP in a large cohort of patients with solid tumors showed that PPM1D and TP53 mutations in clonally expanded hematopoietic cells are significantly associated with prior chemotherapy (48).

When exposed to cytotoxic chemotherapy, HSCs with mutations in PPM1D and TP53 can outcompete wild-type HSCs (58–61). These findings have further been corroborated by research showing that mutations in these two genes are enriched in therapy-related myelodysplastic syndrome (MDS) compared with non-induced MDS (62). Additional research suggests that the growth advantage conferred by mutations in these DNA damage response pathways is present only in the face of the selective pressure of cytotoxic chemotherapy or radiation (4, 59). Thus, treatment for solid tumors appears to be responsible for at least some of the higher rates of CHIP in these populations.

Despite these associations, the majority of CHIP-associated mutations in patients with solid tumors still appear to occur in the epigenetic regulators DNMT3A, TET2, and ASXL1 (46, 48). Multiple groups have observed associations between CHIP and aging (1–3, 8, 14, 63–65) or smoking (1, 13, 48), both of which are also observed in solid tumor studies. Consequently, though genotoxic stress is an important factor for CHIP in patients with solid tumors, other influences such as aging and smoking continue to be relevant in this context.

A study of over 24,000 patients with cancer confirmed many of the above findings. Researchers found that nearly 50% of CHIP mutations in participant's peripheral blood were in DNMT3A, TET2, and ASXL1 (4). Incorporating clinical records, researchers observed that ASXL1 mutations were associated with smoking, while TP53, PPM1D, and CHEK2 mutations were associated with cytotoxic chemotherapy and radiation. In addition, they detected a greater number of CHIP mutations in patients exposed to more lines of therapy. As with other studies (48, 66, 67), the authors did not identify immune checkpoint blockade as a driver of CHIP mutations. They also did not identify an association between targeted therapies and CHIP. Additional studies have reinforced the association between smoking and ASXL1 CHIP mutations (49, medRxiv 2021.12.29.21268342), which needs to be carefully considered as a confounding variable in future work.

The development of CHIP is the result of multiple pathways that may be mutation-specific, as well as environment and cancer-type dependent. Additional research is needed to better understand treatment-specific associations (68) while accounting for germline risks and other confounding factors. A complicating factor in understanding the incidence and cause of CHIP-associated mutations in patients with cancer is the high variability of treatment regimens. Quantifying the presence of CHIP in pretreatment samples from patients with solid tumors will further elucidate the relationship between CHIP and solid tumors. These investigations will help illuminate how treatment of solid tumors contributes to the risk of CHIP, and thus to the risk of CVD, hematologic malignancy, and all-cause mortality.

The most well-studied consequence of somatic mutations in hemopoietic cells in patients with solid tumors is therapy-related myeloid neoplasms (tMN), in which patients who receive chemotherapy or radiation may develop secondary hematologic malignancies. tMNs are a challenging and devastating complication of solid tumor management and carry a median overall survival (OS) of less than 1 year (69–72). As with the transformation of non–treatment-related CHIP to myeloid malignancies, research has shown that the presence of CHIP following therapy for solid tumors increases the risk of developing a tMN (4, 63, 64, 73, 74). The risk of tMNs is not equal across all types of CHIP, but rather appears to vary by mutation status and treatment history. In a study of over 1,000 patients with high-grade ovarian cancer treated with a PARP inhibitor, the presence of TP53 CHIP prior to treatment was associated with an increased rate of tMN posttreatment (54). This association was not observed for DNMT3A, TET2, or ASXL1 CHIP.

Yet it is becoming clear that tMNs are not the only associated poor outcome in solid tumor patients with CHIP. Similar to patients without cancer, patients with solid tumors are at risk for CVD (73). Even after accounting for known CHIP-associated risks, a handful of studies have found inferior OS in patients with solid tumor who have CHIP. Using medical records and sequencing data from patients with solid tumors, researchers observed that the presence of CHIP mutations with a VAF of at least 10% was associated with statistically significant inferior OS (48) even after adjusting for age, gender, and smoking status. Shorter OS could not be explained by deaths from hematologic disease or CVD, and the most common cause of death in this population was progression of the primary cancer. In another study, CHIP was found to be associated with worse OS in patients being treated with immunotherapy for multiple types of solid tumors (75). Whether this association is due to underlying germline risk of both CHIP and solid tumors (13, 76, 77), increased resistance to therapy and thus increased exposure to CHIP-associated genotoxic stress, or alterations in tumor immunosurveillance remains to be known. An important confounder to account for is the fact that exposure to a greater burden of chemotherapy and radiation may be a proxy for more aggressive tumors that are harder to manage therapeutically, resulting in increased toxicity and worse outcomes independent of CHIP.

CHIP in the setting of metastatic disease remains even less studied. In a trial of 237 patients with metastatic colorectal cancer, researchers observed a trend towards increased OS for CHIP carriers (56). This finding was strongest among patients with DNMT3A CHIP, who displayed earlier tumor shrinkage compared with non-CHIP patients. Interestingly, a study of patients being treated with PD-1/PD-L1 blockade measured worse OS in CHIP carriers across many types of solid tumors with the notable exception of colorectal cancer (75). Even after excluding patients with mismatch repair deficiency, patients with colorectal cancer and CHIP had better OS following immunotherapy. Together, these studies underscore the importance of tumor-specific biology and disease stage on the interaction between CHIP and solid tumors.

Evidence that CHIP clones are present in the tumor microenvironment comes from both retrospective computational analyses and direct sequencing of tumor infiltrating leukocytes. One study determined that at least 8% of reported mutations in tumor-only sequencing of solid neoplasms were likely due to the presence of CHIP (78). These findings, along with others (79), suggest that CHIP mutations can appear in tumor-only sequencing due to immune cell infiltration. Congruently, one group identified CHIP mutations in CD45⁺ cells in 7 of 15 tumor samples from patients with primary untreated breast cancer (80). These mutations, which were not detected in microdissected tumor cells, had VAFs in the range of 5% to 20% and were highly enriched in tumor-infiltrating leukocytes compared with peripheral blood mononuclear cells. Interestingly, this suggests that CHIP clones may preferentially infiltrate the tumor compared with other leukocytes. Other groups have since observed CHIP mutations in tumor infiltrating leukocytes from both untreated and pretreated tumors (81). The presence of CHIP-containing tumor-infiltrating immune cells reinforces the need for future investigations on how these clones might impact the tumor immune microenvironment.

Although the intricacies of CHIP-related inflammatory processes are outside the immediate scope of this review, understanding the role of CHIP on inflammatory programs and the tumor microenvironment is essential for providing a comprehensive review for solid tumors. Herein we will provide a brief overview of CHIP influences on inflammatory programs in solid tumors. Research has linked inflammation to many aspects of solid tumors, including tumor initiation, tumor progression, and response to therapy. For example, IL6 and IL1β are widely considered pro-tumorigenic cytokines that are frequently upregulated in the tumor microenvironment (82, 83). However, the role of inflammation on solid tumor biology is complex and dependent on context, and studies have only begun to explore how the presence of CHIP mutations may alter immune cell signaling and function.

Due to limited studies of CHIP in the tumor immune microenvironment, researchers have used the impact of CHIP on other disease states to elucidate the role of CHIP in solid tumors. Several studies have suggested that inflammation may be the underlying cause for the association between CHIP and all-cause mortality, CVD, and stroke (2, 13, 14). In a hypercholesterolemia-prone mouse model, Tet2 knockout in the bone marrow resulted in accelerated atherosclerosis and increased atherosclerotic plaque size compared with control mice (65, 84). These findings were myeloid-lineage specific and associated with increased chemokine and cytokine gene expression from affected macrophages, including IL6 and IL1β (65, 84). Furthermore, in both mouse and human cells of monocyte lineage, deficiency of Tet2/TET2 has been shown to upregulate IL6 and other pro-inflammatory mediators (85).

A clinical study demonstrated that patients with CHIP have an increased risk of coronary heart disease and early-onset myocardial infarction when compared with patients without CHIP (65). Another study analyzed outcomes in patients with a SNP in the IL6 receptor (IL6R) that decreases IL6 signaling. Patients with both CHIP and this SNP have a decreased risk of CVD events compared with CHIP carriers with wild-type IL6R (86). For individuals with the IL6R variant but without CHIP, there was no change in CVD risk. In both studies, risk of incident CVD was associated with CHIP clone size, indicating a possible dose-dependent relationship. These results suggest that changes to inflammatory signaling, particularly in myeloid cells, are responsible for the increased risk of CVD observed in patients with CHIP.

The inflammatory phenotype of CHIP is also associated with potential changes in response to cell-based therapies. While multiple studies have noted possible indications of differential response to CAR T-cell therapy in patients with CHIP, the direction and strength of effect has been variable. A study of patients with multiple myeloma and non-Hodgkin lymphoma found the presence of CHIP to be associated with increased cytokine release syndrome (CRS) in response to CAR T-cell therapy (57). However, a different study of patients with relapsed/refractory B-cell non-Hodgkin lymphoma with and without CHIP did not observe differences in the rate or severity of CRS or immune effector cell-associated neurotoxicity syndrome (ICANS; ref. 87). Unlike the previous study, patients with CHIP were found to have improved OS despite a comparable response rate to CAR T-cell therapy. In contrast, another study on patients with relapsed/refractory large B-cell lymphoma treated with CAR T cells observed a higher incidence of high-grade ICANS in CHIP carriers, primarily driven by DNMT3A, TET2, and ASXL1 CHIP (74).

Recent experiments have begun to demonstrate how CHIP mutations may alter CAR T-cell biology. A case study of a patient with delayed but good response to therapy demonstrated that disruption of TET2 in CAR T cells was accompanied by preferential antigen-driven expansion of the mutated clone. This resulted in high-grade CRS and elevated pro-inflammatory cytokines including IFNγ, TNFα, and IL6 (88). Furthermore, the study demonstrated knockdown of TET2 in CD8⁺ CAR T cells increased granzyme B and perforin expression upon CAR-specific stimulation. Another group observed that knockout of DNMT3A in CAR T cells increases the expansion and antitumor response of these cells during prolonged antigen-specific stimulation (89). This effect appears to be supported by the upregulation of IL10. Despite these intriguing findings, current studies are limited, and additional research will help increase our understanding of how CHIP mutations influence inflammatory signaling and T-cell function and may provide guidance on the use of these cell-based therapies in CHIP carriers.

In addition to CVD, CHIP has been associated with differential incidence or outcomes in many other age-related and inflammatory conditions including heart failure (90–95), chronic obstructive pulmonary disease (96), severe COVID-19 infection (97), chronic kidney disease (98), gout (99, medRxiv 2021.12.29.21268342), osteoporosis (100), antineutrophil cytoplasmic antibody—associated autoimmune vasculidities (67), unprovoked pulmonary embolism (101), and hemophagocytic lymphohistiocytosis (102). Many of the above studies have identified associations with increased risk or adverse outcomes driven by mutations in specific CHIP genes, in connection with alterations in inflammatory signaling. For example, loss of Tet2 in mouse HSCs accelerated development of emphysema (96), while deletion of Dnmt3a in mouse HSCs increased osteoclastogenesis and decreased bone mass, driven in part by pro-inflammatory cytokines (100). Given the enrichment of CHIP among patients with solid tumors, these diseases are important considerations for patients with solid tumors and their clinicians. Combined with the known increased risks of infection and thromboembolism in patients with cancer, these associations may further contribute to morbidity and mortality in these populations.

In addition to disease-specific associations, current research has focused on highlighting potential influences of CHIP on systemic inflammatory signaling. One study found that people with CHIP have higher serum levels of IL6 and TNFα when compared with people without CHIP (103). In a large population study, CHIP was associated with increased IL6 in peripheral circulation. Furthermore, patients with TET2 CHIP had increased IL1β and patients with JAK2 or SF3B1 CHIP had increased IL18 (76). Additional studies have reinforced the finding that people and mice with CHIP have increased levels of circulating cytokines, including IL6 (74, 88, 91, 92, 104, 105).

Deletion of Tet2 in mouse HSCs and macrophages resulted in higher levels of IL1β via signaling through the NLRP3 inflammasome (99), which has been shown to have roles both pro- and antitumor immune signaling (106). The NLRP3 inflammasome has been further implicated in mice with gain-of-function mutations in Ppm1d, resulting in the upregulation of IL1β and IL18 (95). In a study exploring the functional effects of this signaling pathway in HSCs with CHIP, inhibition of the NLRP3 inflammasome in mice with Tet2 deficiency was protective in a model of heart failure (93). In addition, among patients treated with an anti-IL1β antibody, those with CHIP trended towards greater relative risk reduction of major adverse cardiovascular events compared with those without CHIP (105). Knockout of CHIP-associated genes in murine T cells has also been shown to alter inflammatory signaling in both pro-tumor and antitumor pathways (107, 108). These findings suggest that CHIP may have systemic-level effects on the immune system and support the observation that CHIP can interact with a variety of pathologies to influence clinical outcomes.

Despite these recent studies, further work is needed to better understand the inflammatory nature of CHIP across mutations and disease contexts. The effects of CHIP-associated inflammatory signaling are relevant for patients with solid tumors for multiple reasons. First, these patients have higher rates of CHIP and thus are at higher risk for these disease states. Understanding how CHIP influences pathogenesis and the outcomes for myriad conditions will be critical in anticipating and caring for these patients. Second, the mechanisms by which CHIP contributes to dysregulated immune cells and inflammation provide insight into how CHIP biology and solid tumor biology interact, either within the tumor immune microenvironment or more systemically.

Studies have begun to explore the impact of common CHIP-associated mutations on tumor progression. One group found that in mice susceptible to adenomas, myeloid-specific loss of p53 results in significantly higher numbers of intestinal adenomas (104). In parallel, these mice have higher levels of mRNA for pro-inflammatory cytokines, including IL6 and IL1β, and have upregulation of genes associated with M2-like tumor-associated macrophages (TAM) that tend to be pro-tumorigenic.

Leveraging multiple mouse models of cancer and multiple tissue-specific knock-ins, another group found that subcutaneous tumor models grew faster when mutant Asxl1 was expressed in T cells but not in myeloid cells (109). In a spontaneous tumor model, however, mutation of Asxl1 promoted tumor progression even when expressed in all blood cells. By characterizing the immune cell infiltrate in tumors, the authors observed exhaustion and reduced representation of Asxl1-mutant T cells compared with wild-type T cells. They further showed that mutant Asxl1 promoted a more pro-inflammatory phenotype in myeloid cells, highlighting its lineage and context dependent effects on tumor progression.

Research on the effects of TET2 loss in HSCs and tumor progression has shown conflicting results. One group found that loss of Tet2 in myeloid cells slowed melanoma growth in mice in a T cell–dependent manner (110). In addition, they found increased expression of proinflammatory M1 signatures and decreased expression of immunosuppressive M2 signatures in Tet2-null TAMs compared with Tet2 wild-type TAMs. However, a different group found the opposite effect. In this study, both pan-hematopoietic and myeloid-specific Tet2 loss promoted lung cancer growth that was driven by increased angiogenesis (111). These contradictory findings emphasize the challenge of uncovering the complex and multifaceted interactions between specific mutations, the cells they are found in, and their tissue environment.

Altogether, the presence of CHIP clones in the tumor microenvironment, the inflammatory dysregulation associated with CHIP, and experiments demonstrating altered tumor progression in animals with CHIP point to exciting hypotheses about the potentially complex interplay between CHIP and solid tumors.

Given that CHIP is a phenomenon of hematopoietic cells, it is logical to hypothesize that CHIP may also have an impact on the immune response to solid tumors. An important next step in the field is building a deeper understanding of the association between chemotherapy or radiation and the development of CHIP. While much of this association is likely due to selective pressure from treatment and induction of mutations, changes in epigenetic regulation and stochastic processes should also be investigated. Deeper questions about this association include whether the biology of certain tumors drives CHIP development or the biology of CHIP alters cancer development. Further research is needed to understand why patients with solid tumors who have CHIP have worse OS, even after accounting for hematologic malignancy and CVD. Important factors to consider include changes in immune cell function and tumor microenvironment driven by CHIP mutations, accumulation of genotoxic stress in settings of therapy resistance, and underlying germline risk.

As seen in the mixed results of existing data, the impact of CHIP is likely specific to the mutation, tumor characteristics, immune cell types, and treatment context. Though additional evidence is needed before changing clinical practice, understanding the pro- or anti-inflammatory actions of different CHIP clones may ultimately help guide therapy choice, identify new targets for patient-specific treatment, and predict outcomes in patients with solid tumors. For example, incorporating a patient's risk factors for developing CHIP and the tendency for a particular therapy to promote expansion of a CHIP clone could help steer the patient's treatment plan to manage the risk of developing a tMN. Furthermore, new or existing drugs may be able to exploit CHIP-related alterations in the tumor immune microenvironment. The effects of CHIP in solid tumors appear to be complex and are currently not fully understood. Until these effects are elucidated, clinicians should be careful not to undertreat a solid tumor due to concerns for CHIP risks. However, the promise of integrating information about CHIP to help steer solid tumor management represents an exciting future potential.

Many questions remain beyond the scope of this review. Other forms of clonal hematopoiesis, particularly mosaic chromosomal alterations, are increasingly recognized for their roles in inflammation and disease (20–22, 112, 113). Furthermore, the functional effects of CHIP mutations on other immune cell types such as B cells and natural killer cells are poorly characterized. A deeper understanding of the widespread and possibly reciprocal effects of clonal hematopoiesis on solid tumors represents a promising opportunity to impact the health of many cancer patients.

S.C. Reed reports grants from NIH during the conduct of the study. B.H. Park reports other support from Tempus during the conduct of the study as well as other support from Horizon Discovery; personal fees from Jackson Labs, Hologic, Celcuity Inc, and EQRx; and grants from Lilly outside the submitted work. No disclosures were reported by the other author.

This work was supported by NIH T32GM007347 (to S.C. Reed) and NIH/NCI SPORE P50CA098131 (to S.C. Reed and B.H. Park).

The publication costs of this article were defrayed in part by the payment of publication fees. Therefore, and solely to indicate this fact, this article is hereby marked “advertisement” in accordance with 18 USC section 1734.