CHIP-ing Away at Post-Transplant Outcomes: the Role of Somatic Mutations in Heart Transplant Outcomes

A traditional view of genetics in cardiovascular disease was that of germline variation, that is, genetic variation that is present at conception in all cells, including germ cells, is static over a lifetime, and can confer risk of heart failure (HF) and other cardiovascular diseases. Somatic mutations, that is, changes in the DNA sequence of a cell that occur after conception, have resided primarily in the domain of oncology and tumor pathobiology. Recent advances, however, have uncovered the role of somatic mutation in cardiovascular disease. Most notably, the accumulation of somatic mutations in hematopoietic stem and progenitor cells (HSPCs)¹ resulting in mutant clones, so-called clonal hematopoiesis of indeterminate potential (CHIP), has been identified as a risk factor for cardiovascular disease.

CHIP is an age-related phenomenon and is defined as the presence of CHIP-driven genetic mutations present in circulating blood cells at a variant allele frequency (VAF) ≥ 2%, in the absence of hematological malignancy.¹ These mutated clones in HSPCs can be detected by sequencing peripheral blood DNA. CHIP is uncommon in young patients under the age of 40 (< 1%), whereas prevalence increases to 10% or more in individuals aged 65 or older. CHIP emerged as a novel cardiovascular disease (CVD) biomarker in 2017, first associated with risk of atherosclerotic cardiovascular disease and stroke and now, subsequently, with atrial fibrillation and HF risk.^2–5 Patients with HF have a high prevalence of CHIP and a 2-fold increase in mortality risk compared to patients with HF but without CHIP.^6,7 Individuals with large CHIP clones (VAF ≥ 10%) seem to have greater risk of both hematological and cardiovascular outcomes. CHIP occurs most commonly in the epigenetic modulators DNMT3A, TET2 and ASXL1 and is associated with heightened levels of systemic inflammation, believed to be a key mechanism of the association of CHIP with a variety of cardiovascular diseases.

Orthotopic heart transplantation (OHT) remains the therapy of choice for patients with end-stage HF. Despite improvements in transplant care, long-term complications, including cardiac allograft vasculopathy (CAV), persist as challenges in this population of patients. At present, gold-standard surveillance for CAV relies on invasive testing with coronary angiography; therefore, novel noninvasive biomarkers for post-OHT risk are desperately needed. In the United States, the median age of OHT recipients has been increasing; therefore, the impact of CHIP in this population has been raised as a potential comorbidity.^8,9 The small handful of prior studies of CHIP in OHT have had discordant results, leaving open the question of the role of CHIP as an OHT outcome risk factor, given its known association with HF.^10,11

In the current issue of the Journal of Cardiac Failure, Simitsis and coauthors describe their study of the association of CHIP with CAV, graft failure, malignancy, retransplantation, and mortality outcomes in a single-center cohort of 95 OHT recipients at Brigham and Women’s Hospital.¹² With a median age of 61 years, the prevalence of CHIP (VAF ≥ 2%) in this population was 31.6%, higher than previously reported in OHT populations. Of the cohort, CHIP was assessed prior to transplantation in 5 individuals, whereas in the other 90 subjects, CHIP was assessed after OHT. Similar to larger epidemiological studies of CHIP in cardiovascular disease, DNMT3A variants were found most commonly; however, they were followed by PPM1D variants as the second most commonly mutated CHIP gene (in contrast to TET2 in epidemiological studies), an observation also seen by Amancherla et al.¹¹ The number of CHIP mutations present, unsurprisingly, increased with both age and time since transplant. After adjusting for age, baseline renal dysfunction and immunosuppression with sirolimus were independent predictors of the presence of CHIP in this population; however, a multivariable model including these factors and histories of grade 2R or greater antibody-mediated rejection did not outperform age alone for CHIP prediction.

The primary composite outcomes of the first occurrence of grade ≥ 2 CAV, graft failure, malignancy, retransplantation, or all-cause mortality were ascertained both retrospectively and prospectively from the date of CHIP assessment. The primary composite outcome occurred in 44 of 95 subjects studied (46.3%). Despite a long follow-up (median 9.7 years), there were no significant differences in the primary composite outcome in those harboring a CHIP mutation, in those with larger clones (VAF ≥ 5%) or in those with non-DNTM3A CHIP after multivariable adjustment. Similarly, the authors found no association between CHIP and the individual components of the composite outcome, including mortality. Overall CAV rates in this population were 22.1%, compared to over 40% in the work by Amancherla et al.¹¹ and only 3% compared to that from Scolari et al.¹⁰ Notably, rates of ischemic cardiomyopathy were relatively low (13.8%) compared to either of the prior studies and to the overall OHT population in the United States (> 20%), which may influence the findings for CHIP and its atherosclerotic ties in the present work.

Interestingly, CHIP actually displayed a protective effect for graft failure (HR 0.19 [95% CI 0.06—0.65]; P = 0.008) in multivariable models, but these findings were not significant in univariable analyses, have wide confidence intervals and did not show these effects for VAF ≥ 5% or non-DNTM3A CHIP. If this were a clinically significant finding, we might expect to see the strongest association at higher VAF, as has been seen for other cardiovascular associations with CHIP. The work by Silver et al. found a nonsignificant but numerically lower rate of graft complications in DNMT3A and TET2 CHIP across all solid-organ transplants studied.¹³ A potential hypothesis that has been proposed for this signal is that inflammatory TET2 clones are suppressed by transplant immunosuppressive therapies. This may be somewhat analogous to the benefit of canakinumab in TET2 CHIP carriers from CANTOS (Canakinumab Anti-Inflammatory Thrombosis Outcome Study).¹⁴

This study has multiple strengths and brings new light and clarification to the role of CHIP and OHT. It is important to highlight the strengths and also certain limitations. The use of targeted sequencing here increases the sensitivity and specificity for detecting CHIP compared to the whole-exome sequencing typically performed in epidemiological studies. A key strength of this work is the inclusion of a long length of follow-up; the median time from transplant to enrollment was 8.5 years, and the combined retrospective and prospective follow-up was a median of 9.7 years after OHT. However, we must also consider the ascertainment and survival bias inherent in this study design; patients who died due to early transplant complications, including primary graft dysfunction (PGD), particularly given the known pathophysiology of inflammatory cytokines in PGD, were not included. Importantly, diagnostic modalities for CAV phenotyping do not uniformly include the use of intravascular ultrasound, creating heterogeneity in outcomes assessed across various transplant centers. Although the study includes a relatively large sample size for a single center, as with many studies of OHT, the small sample size limits power and more focused gene-specific analyses in this population. This highlights the importance of more coordinated consortia studies for studying OHT outcomes overall and for the evaluation of the molecular drivers of these outcomes. For example, the recently formed International Consortium on PGD is an example of the power of multicenter studies for this relatively small population of patients.¹⁵

Importantly, in the current study, CHIP ascertainment was asynchronous with the time of OHT; therefore, there is temporal uncertainty about when patients acquired CHIP, which impacts the assumptions about the length of CHIP exposure and the OHT outcomes studied. The authors assumed stable CHIP clone dynamics over a period of 5 years based on prior work concerning temporal clonal dynamics, analyzing a base-case scenario and assuming CHIP stability for 2.5 years on both sides of enrollment.^16,17 Sensitivity models were performed in an attempt to further mitigate this limitation, considering extreme left scenarios (clone stability 5 years after CHIP ascertainment), extreme right scenarios (clone stability 5 years prior to CHIP ascertainment), and considering CHIP as a time-dependent covariate, which showed overall consistent results for primary and secondary outcomes. Given the limitations of this asynchronous timing, we agree with this analytic approach; however, this leaves space for future work with longitudinal samples to assess the acquisition and stability of CHIP clones in the OHT population.

In 2022, Scolari et al. performed a retrospective study of CHIP in 127 OHT recipients in Toronto, Canada (median age of CHIP carriers, 54 years), and found a high prevalence (20%) of CH that was associated with higher prevalence of CAV and mortality (mean follow-up, 3.2 years).¹⁰ However, and importantly, CAV rates were low (3%) in this population, with no CAV cases in subjects without CHIP. Conversely, in 2023, in a retrospective study of 479 OHT recipients at Vanderbilt University and Columbia University (median age CHIP carriers 59—60 years, prevalence 15%), Amancherla et al. found no association with CAV or mortality (median follow-up 18.1 years Vanderbilt; 5.1 years Columbia), but those with CHIP were more likely to have an ischemic etiology of HF.¹¹ More broadly, this year Silver et al. published work across 2 large biobanks, All of Us and UK Biobank, which included 2610 heart, kidney, liver, and lung allografts, and they found an enrichment for TET2 CHIP in this solid-organ transplant population, without clear risk of CHIP and transplant outcomes.¹³

The importance of CHIP is less ambiguous in populations after hematopoietic stem cell transplants, in whom adverse outcomes, including therapy-related myeloid neoplasms, have been reported for both autologous and allogeneic stem cell transplant when CHIP is present.^18,19 In solid-organ transplantation, although there is recipient exposure to donor-derived blood at the time of transplantation and in preclinical models, donor-derived macrophages persist in transplanted hearts for up to 14 days, so CHIP can be considered largely a recipient-related factor.²⁰ There was hope for CHIP to be a novel, noninvasive transplant-outcome biomarker, but the present analyses by Simitsis and coauthors move the needle away from this promise. Our understanding, to date, of the pathogenesis of CAV is thought to be driven by T-cell activation, with coronary branch vessel pruning that is distinct from traditional atherosclerosis pathobiology.²¹ One explanation for the lack of association between CHIP and CAV is that CHIP exists in the myeloid compartment, where monocytes and macrophages may play a more specific role in atherosclerotic plaque formation than in CAV development. The lower prevalence of ischemic HF in the OHT population studied here may further weaken this signal. Ultimately, it seems most likely that immunosuppressive therapies required in the post-OHT setting dampen the inflammatory effects of CHIP on long-term OHT outcomes. Regardless, the findings of potential protective outcomes for graft failure warrant further prospective study to understand the complex relationships between specific CHIP mutations (ie, TET2) and immunosuppressive therapies. This should include comparisons across populations after solid-organ transplantations, which require varying degrees of immunosuppression, with detailed assessment of treatment regimens, including any interruptions or intermittent discontinuations of immunosuppression.

Despite the neutral findings of this work, the door for studying CHIP in HF and OHT should not be closed. With the reduced cost of next-generation sequencing, CHIP is being more commonly detected in clinical settings, necessitating specialized, multidisciplinary CHIP clinics to assess and follow patients for long-term hematological and cardiovascular outcomes. There is still great potential for CHIP as a marker of incident atherosclerotic and HF risk, and targeted therapeutics may be impactful in preventing progression to end-stage HF. Deeper phenotyping of CHIP, including metabolic traits, which are also highly prevalent in post-OHT patients, is needed to understand complex inflammatory and metabolic mechanisms of risk. Multi-omic approaches to CHIP biology to understand epigenetic, transcriptional and inflammatory proteomic processes will help us hone the ways in which we apply CHIP to the precision-medicine approaches to cardiovascular care. We applaud the current work, which provides another data point on the scorecard of the association between CHIP and OHT outcomes (Fig. 1), and we look forward to results from larger, multicenter, prospective studies that include ascertainment of CHIP at the time of transplantation, as well as long-term and detailed patient follow-up which, it is hoped, will break the tie in this important match.

Fig. 1.

Scorecard for CHIP and OHT outcomes. To date, 3 studies have evaluated the prognostic value of CHIP in OHT. The present work by Simitis et al. aims to serve as a tiebreaker, with neutral effects on long-term OHT outcomes, with the longest duration of follow-up. X indicates lack of significant association. CAV, cardiac allograft vasculopathy; CHIP, clonal hematopoiesis of indeterminate potential; OHT, orthotopic heart transplantation; yo, years old. (Created with Biorender.com.)

Biography