William Grant Dunn

Summary: Haematological malignancies are rare in primary care and are hard to predict. Existing prediction models often require a specific working diagnosis and are based on extensive clinical information, access to which may be limited at initial work‐up in primary care. We aimed to develop a tool for estimating overall risk of haematological malignancy in primary care patients using simple data inputs. Including all full blood cell counts (FBCs) measured from 2000 to 2016 in 856 403 adult primary care patients in Eastern Denmark, we used machine learning to develop CBC‐HEMA, a risk assessment tool using FBCs, age, sex and C‐reactive protein measures to stratify ‘any haematological malignancy’ risk within 6 months, 1 year and 3 years with area under the curve (AUC) of 0.81, 0.78 and 0.71 respectively. Predictions of chronic lymphoid leukaemia and myeloproliferative disease were excellent, while CBC‐HEMA was unable to accurately predict non‐Hodgkin lymphoma and plasma cell dyscrasia. External validation was conducted in 123 015 primary care patients from Western Denmark with similar results and in 418 410 individuals from the background population cohort UK Biobank with inferior performance showing unfitness in screening healthy populations. The CBC‐HEMA https://shiny.sund.ku.dk/CopLab/blood‐cancer‐risk‐prediction/ may support risk assessment in primary care but cannot replace a clinical evaluation.

The terms clonal monocytosis of undetermined significance (CMUS) and clonal cytopenia and monocytosis of undetermined significance (CCMUS) were introduced by the International Consensus Classification of Myeloid Neoplasms to describe cases of clonal hematopoiesis (CH) and concurrent monocytosis that did not meet the diagnostic criteria of chronic myelomonocytic leukemia. To date, their practical relevance as clinicopathological entities at a population level has not been assessed. Here, we assess the prevalence, significance, and natural history of CMUS and CCMUS among 431 531 UK Biobank participants through analysis of clinical, genomic, and health outcome data. We find that CMUS with an absolute monocytosis and CCMUS are high-risk entities strongly associated with incident myeloid neoplasia (MN), cardiovascular disease, and renal disease. Noting the overall higher monocyte counts in men and the low rate of progression of DNMT3A-CMUS, we reveal that amending the definition of CMUS/CCMUS to incorporate sex-specific monocyte thresholds and the exclusion of isolated DNMT3A mutations from the definition significantly strengthens the association with incident MN. Finally, given their association with poor outcomes, we develop MoSAIC, a machine-learning classifier, to infer the presence of SRSF2 mutations (associated with high MN risk) among individuals with monocytosis, based on complete blood count indices alone. We corroborate our findings in an independent cohort of 625 328 Danish primary care patients. Our findings underscore the clinical relevance of CMUS and CCMUS as distinct high-risk states within the spectrum of CH and establish an evidence base to refine their diagnostic definition.

Clonal hematopoiesis (CH), the clonal expansion of a hematopoietic stem cell and its progeny driven by somatic mutations, has been associated with inferior survival outcomes among recipients of autologous stem cell transplants (ASCT). Leukocyte telomere length (LTL) has a complex but well-documented interaction with CH, but the impact of this interaction on stem cell transplantation has not been adequately examined. We measured LTL in graft cell DNA from 452 patients undergoing ASCT for myeloma, for whom targeted DNA sequencing for CH driver gene mutations was available. We interrogated clinical and longitudinal large-scale laboratory data for these patients to understand the impact of graft LTL on progression-free survival (PFS) and overall survival after transplantation, as well as blood count indices and their trajectories. In multivariate analyses, longer LTL was associated with increased PFS among patients without CH. However, this protective association was not seen in patients with CH. We also report that among patients with CH, longer LTL was associated with an increased red cell distribution width before myeloablative chemotherapy and after ASCT. Collectively, these data reveal hitherto undescribed interactions between LTL, CH, and ASCT outcomes.

The mechanisms through which mutations in splicing factor genes drive clonal hematopoiesis (CH) and myeloid malignancies, and their close association with advanced age, remain poorly understood. Here we show that telomere maintenance plays an important role in this phenomenon. First, by studying 454,098 UK Biobank participants, we find that, unlike most CH subtypes, splicing-factor-mutant CH is more common in those with shorter genetically predicted telomeres, as is CH with mutations in PPM1D and the TERT gene promoter. We go on to show that telomere attrition becomes an instrument for clonal selection in advanced age, with splicing factor mutations ‘rescuing’ HSCs from critical telomere shortening. Our findings expose the lifelong influence of telomere maintenance on hematopoiesis and identify a potential shared mechanism through which different splicing factor mutations drive leukemogenesis. Understanding the mechanistic basis of these observations can open new therapeutic avenues against splicing-factor-mutant CH and hematological or other cancers.

Background: Clonal haematopoiesis (CH), the disproportionate expansion of a haematopoietic stem cell and its progeny, driven by somatic DNA mutations, is a common age-related phenomenon that engenders an increased risk of developing myeloid neoplasms (MN). At present, CH is identified by targeted sequencing of peripheral blood DNA, which is impractical to apply at population scale. The complete blood count (CBC) is an inexpensive, widely used clinical test. Here, we explore whether machine learning (ML) approaches applied to CBC data could predict individuals likely to harbour CH and prioritise them for DNA sequencing. Methods: The UK Biobank was filtered to identify 431,531 participants with paired CBC and whole exome sequencing (WES). Somatic mutations were previously identified from blood WES using Mutect2 to classify individuals with CH driver mutations. Using 18 CBC indices/features and basic demographics (age and sex), we trained a range of tree-based ML classifiers to infer as binary output, the presence/ absence of CH. Findings: Using Random Forest (RF) classifiers, we predicted the presence/absence of CH driven by mutations in one of five genes known to confer a high-risk of incident MN ( JAK2, CALR, SF3B1, SRSF2 and U2AF1 ). We subsequently developed a unified, optimised RF classifier for high-risk CH driven by any of these genes and assessed its performance (median AUC 0.85). However, the low prevalence of high-risk CH implies that our model cannot be generalised to population scale without compromising its sensitivity (20.1% using stringent cutoff probability score). Interpretation: We showcase a proof-of-concept that the presence of high-risk CH can be inferred from CBC perturbations using RF classifiers. The future integration of raw blood cell analyser data can help improve the performance of our model and facilitate its application at scale. Funding: Cancer Research UK. Research in context: Evidence before this study: We searched PubMed for articles published, in English, between database inception and 5 ^th of June 2024, using the terms “clonal hematopoiesis” AND (“machine learning” OR “artificial intelligence”). We additionally searched for the terms “clonal hematopoesis” AND “complete blood count”. We found 18 research articles: one article used ML approaches (XGBoost classifiers) to differentiate clonal haematopoiesis “driver” mutations from “passenger” mutations, but none linked machine learning frameworks to complete blood count data for predicting the presence of clonal haematopoiesis. Progression from clonal haematopoiesis to myeloid neoplasia is known to be associated with several blood count parameters; two recent publications developed clonal haematopoiesis risk stratification tools that incorporated blood count indices in their final risk prediction models (Gu et al. Nature Genetics, Weeks et al. NEJM Evidence). However, we found no study assessing whether blood count indices could be used to infer the presence of clonal haematopoiesis. Added value of this study: Here we show that CH driven by mutations in genes associated with high risk of progression to myeloid neoplasia can be reliably differentiated using ML approaches applied on peripheral blood indices; however, low-risk forms of CH (driven by mutations in the DNMT3A or TET2 genes) cannot be reliably inferred from CBC indices. While optimising the model we identified challenges in upscaling its applicability; we propose that the integration of single-cell resolution “raw” blood analyser data might overcome these issues. Previous efforts to enhance the scalability of CH screening focused on reducing DNA sequencing costs. Here, we provide a proof-of-concept that an extensively used clinical test, the CBC, can, using machine learning approaches, predict individuals more likely to harbour high-risk CH, who should be prioritised for genetic testing. Implications of all the available evidence: Our study proposes a model for predicting high-risk CH mutations by applying a Random Forest classifier on CBC indices; this represents an important step towards scalable screening for identifying individuals at high risk of developing myeloid neoplasia in the future. This is an attractive approach, as it relies solely on a routine, inexpensive test. Despite good sensitivity, the low prevalence of high-risk CH leads to a low positive predictive value that precludes the use of the predictive model as a population-wide pre-screening tool. To overcome this, we propose the future integration of raw blood analyser data into models like ours to improve the performance and scalability of this approach.

Germ line variants in the DDX41 gene have been linked to myelodysplastic syndromes (MDS) and acute myeloid leukemia (AML) development. However, the risks associated with different variants remain unknown, as do the basis of their leukemogenic properties, impact on steady-state hematopoiesis, and links to other cancers. Here, we investigate the frequency and significance of DDX41 variants in 454 792 United Kingdom Biobank (UKB) participants and identify 452 unique nonsynonymous DNA variants in 3538 (1/129) individuals. Many were novel, and the prevalence of most varied markedly by ancestry. Among the 1059 individuals with germ line pathogenic variants (DDX41-GPV) 34 developed MDS/AML (odds ratio, 12.3 vs noncarriers). Of these, 7 of 218 had start-lost, 22 of 584 had truncating, and 5 of 257 had missense (odds ratios: 12.9, 15.1, and 7.5, respectively). Using multivariate logistic regression, we found significant associations of DDX41-GPV with MDS, AML, and family history of leukemia but not lymphoma, myeloproliferative neoplasms, or other cancers. We also report that DDX41-GPV carriers do not have an increased prevalence of clonal hematopoiesis (CH). In fact, CH was significantly more common before sporadic vs DDX41-mutant MDS/AML, revealing distinct evolutionary paths. Furthermore, somatic mutation rates did not differ between sporadic and DDX41-mutant AML genomes, ruling out genomic instability as a driver of the latter. Finally, we found that higher mean red cell volume (MCV) and somatic DDX41 mutations in blood DNA identify DDX41-GPV carriers at increased MDS/AML risk. Collectively, our findings give new insights into the prevalence and cognate risks associated with DDX41 variants, as well as the clonal evolution and early detection of DDX41-mutant MDS/AML.

Many hematological diseases are characterized by altered abundance and morphology of blood cells and their progenitors. Myelodysplastic syndromes (MDS), for example, are a group of blood cancers characterised by cytopenias, dysplasia of hematopoietic cells and blast expansion. Examination of peripheral blood slides (PBS) in MDS often reveals changes such as abnormal granulocyte lobulation or granularity and altered red blood cell (RBC) morphology; however, some of these features are shared with conditions such as haematinic deficiency anemias. Definitive diagnosis of MDS requires expert cytomorphology analysis of bone marrow smears and complementary information such as blood counts, karyotype and molecular genetics testing. Here, we present Haemorasis, a computational method that detects and characterizes white blood cells (WBC) and RBC in PBS. Applied to over 300 individuals with different conditions (SF3B1-mutant andSF3B1-wildtype MDS, megaloblastic anemia, and iron deficiency anemia), Haemorasis detected over half a million WBC and millions of RBC and characterized their morphology. These large sets of cell morphologies can be used in diagnosis and disease subtyping, while identifying novel associations between computational morphotypes and disease. We find that hypolobulated neutrophils and large RBC are characteristic ofSF3B1-mutant MDS. Additionally, while prevalent in both iron deficiency and megaloblastic anemia, hyperlobulated neutrophils are larger in the latter. By integrating cytomorphological features using machine learning, Haemorasis was able to distinguish SF3B1-mutant MDS from other MDS using cytomorphology and blood counts alone, with high predictive performance. We validate our findings externally, showing that they generalize to other centers and scanners. Collectively, our work reveals the potential for the large-scale incorporation of automated cytomorphology into routine diagnostic workflows.