Gene Fitness as a Predictor of Pathophysiological Outcomes in Clonal Haematopoiesis of Indeterminate Potential

Clonal haematopoiesis of indeterminate potential (CHIP) is defined as the clonal expansion of haematopoietic stem and progenitor cells (HSPCs) in healthy aged individuals [1–3]. Although mostly inconsequential, the constant rate of acquisition of mutations in HSPCs (17 mutations/year [4]) leads to an increasing probability, with respect to age, of a somatic variant occurring that can destabilise the tightly regulated homeostasis of haematopoiesis. In healthy individuals, differentiated blood cells are the net progeny of an approximately balanced HSPC pool and together produce a spectrum of differentiated cells without any single mutations reaching high variant allele frequencies (VAFs). Clonal haematopoiesis, however, is marked by the population of blood cells showing increasing oligoclonality through selection - becoming increasing dominated by single (or multiple) large genetic clones that are genotypically identical [5]. The growth rate – or fitness – of HSPC clones is defined as the proliferative advantage over cells carrying no or only neutral mutations. If mutations in different genes lead to distinct fitness advantages, this could enable patient stratification [6].
Clonal haematopoiesis is the consequences of the outgrowth of high fitness clones that are generally driven by somatic mutations in a small set of functionally diverse genes [1]. Although captured under this nomenclature due to their capacity to permit progenitor and stem cell expansion in the haematopoietic niche, their diverse roles lead to a variety of differing mechanisms that permit this expansion with corresponding divergencies in their aetiology, pathophysiological outcomes and associated disease risk [7]. 
CHIP increases rapidly in prevalence beyond age 60 and an important focus of recent research has been uncovering its links with myeloproliferative disease as a result of the enrichment of somatic mutations in many key driver genes [3,8]. Recent focus has also surrounded the association to many non-haematological diseases that include many distal pathologies of ageing; including cardiovascular disease and ischaemic stroke [2]. Other studies have highlighted the increased risk of all-cause mortality in CHIP carriers, and found that the number of mutations, the mutational burden and the size of the primary driver clone (maximum VAF) were associated with this risk [9].
The study of CHIP in large cross-sectional cohorts has provided a wealth of perspective on the genetic drivers, prevalence and associations with numerous clinical features [2,3,10]. However, cross-sectional studies – providing a single snapshot in time across a population – leave numerous questions regarding how CHIP develops, the dynamics of clone growth and how it might interact with ageing. In 2022, several groups (including ourselves) generated the first large-scale longitudinal studies of CHIP and effectively characterised the spectrum of fitness effects of mutations in a range of driver mutations and highlighted the common dynamics of these variants across several independent cohorts [6,11,12]. In Robertson and Latorre-Crespo et al. (Nature Medicine, 2022), we described the potential of using gene fitness as a novel proxy to predict survival (amongst other clinically relevant features) showing improved effect sizes over measurements of maximum clone size alone (maximum VAF), while noting that our study was currently underpowered to adequately predict such associations (n=82) [6].
We would like to continue this analysis by combining the three existing longitudinal CHIP cohorts (Robertson and Latorre-Crespo et al. (2022), n = 82; Fabre et al. (2022), n = 385; Uddin et al. (2022), n = 182 and van Zeventer et al. (2023) n= 3,359 [NL Lifelines]) to call gene fitness using our computational pipelines and improve our statistical power to make meaningful pathophysiological associations [6,11,12]. By understanding the links between gene fitness and relevant clinical outcomes, we hope that we can better inform clinicians and improve the prospects for effective patient monitoring in tandem.