Variability in interventional trials in clonal haematopoiesis

Abstract Background Clonal haematopoiesis (CH) is an important cause of age-related diseases, characterised by the clonal outgrowth of hematopoietic stem cells without haematologic neoplasms. It is found in 10-30% of people over 70 and is a silent driver of mortality and life-limiting complications, most commonly cardiovascular disease.(Jaiswal, 2014). Treatment for CH remains unestablished and is being explored through various clinical trials. However, there are unique considerations to developing interventional trials in CH. The inherent heterogeneity requires the selection of individuals who are most likely to benefit from intervention. Additionally, there are limitations to applying commonly used outcome measures due to the long latency period before complications develop.(Haque, 2025) Methods This study assessed all published or registered interventional trials in CH up to 27/06/2025. The databases searched were PubMed, CENTRAL, and Clinicaltrials.gov. The search was performed using the terms “clonal hematopoiesis” or “clonal cytopenia”, then filtered to interventional trials only. Results A scoping search identified 4893 studies on PubMed, 127 on CENTRAL, and 137 on ClinicalTrials.gov. Non-interventional studies were then excluded, leaving four interventional trials on PubMed, 17 on CENTRAL, and 15 on ClinicalTrials.gov. One preprint and eighteen duplicate entries were excluded, leaving thirteen registered and four published trials: NCT03418038, NCT03682029, CANTOS (exploratory analysis), and ASPREE (sub-study). There are four trials (23.5%) in Clonal haematopoiesis of indeterminate potential (CHIP), including one specifically investigating TET2, and another specifically investigating TET2 and DNMT3A. There are 13 trials (76.5%) in Clonal cytopenia of undetermined significance (CCUS) with significant variability in the inclusion criteria. Seven different cytopenia thresholds are employed. Three thresholds, including the World Health Organisation fifth edition criteria, are each applied in three trials, while the remaining four are each applied in a single trial. These range between haemoglobin (Hb) 9-13 g/dL, absolute neutrophil count (ANC) 0.75-1.8 × 109/L, and platelet (plt) count 30-150 × 109/L. A lower limit was defined in one trial (ANC > 0.5 × 109/L, plt > 50 × 109/L). Six trials require persistent cytopenia for at least six months, while four specify a shorter duration between one and four months. One trial defines the inclusion of high-risk cases based on the CH Risk Score, and another two based on variants involved and their allelic frequencies (VAF) (e.g., TP53 > 5%). The therapeutics evaluated are targeting the inflammatory pathway (n=5), epigenetics (n=3), and IDH-1/2 (n=4). The treatment duration varies between 3-36 months/cycle, most commonly 12 (n=7), followed by 18 (n=4), and 3 (n=3). Cytopenia responses are most commonly measured (n=11), including four trials employing the 2018 International Working Group (IWG) response criteria and two trials employing the 2006 IWG response criteria, respectively. VAF changes are measured in ten trials, with varying endpoints (continuous or categorical) and timepoints of measurement. Twenty-three different inflammatory biomarkers are measured in eight trials, with IL-6 being the most commonly measured (n = 5), followed by hsCRP (n = 3). Survival is measured in six trials, most widely overall (n=3), followed by disease-free (n=2). Five trials are evaluating patient-reported outcomes using four different measures, most commonly with the EORTC QLQ-C30 (n=3) and SF-36 (n=2). Other exploratory outcomes include changes in methylation (n=4) and gut microbiota (n=2). These findings reflect the challenges in designing interventional trials in CH, likely due to its inherent heterogeneity and the early but evolving research landscape. The implications of the variability need to be considered. Harmonising the eligibility criteria may benefit implementation of therapies in future.(Haque, 2025) In contrast, variability allowing comparison between distinct subgroups may benefit the development of personalised therapy.(Patel, 2025) Establishing a core outcome set may reduce variability in the outcomes measured. Conclusion There is considerable variability across interventional trials in CH, including participant selection and outcome measures. The implications, including the impact on patient representation and inter-trial comparability, need to be carefully considered.