Fig. 1

Clonal hematopoiesis in hospitalized COVID-19 patients. A Number of patients with CH at a VAF ≥ 2.0% with the number of mutations per patient identified (x-axis). B Percentage of patients with CH in specific genes. C Percentage of patients with mortality within 40 days of COVID-19 hospitalization stratified by the presence (n = 57) or absence (n = 185) of CH, CH with the highest VAF mutation in DNMT3A (n = 35), or CH with the highest VAF mutation in a gene other than DNMT3A (n = 22). D Forest plot displaying odds ratios (ORs) and 95% confidence intervals (CIs) predicting mortality within 40 days of initial COVID-19 hospitalization from overall CH, DNMT3A mutant CH, and non-DNMT3A mutant CH measured at VAF ≥ 2.0%. Overall and gene-specific CH results were derived from the same model. Odds ratios predicting pre-existing cardiovascular and pulmonary disease from overall CH at VAF ≥ 2.0% are also shown. Results from both raw (gray) and covariate adjusted (black) logistic regression models are shown. The dashed vertical line indicates the threshold for no effect from CH. E Percentage of COVID-19 patients with pre-existing CKD stratified by the presence or absence of CH, CH with the highest VAF mutation in DNMT3A, or CH with the highest VAF mutation in a gene other than DNMT3A. F Forest plot displaying odds ratios (ORs) and 95% confidence intervals (CIs) predicting CKD from overall CH, DNMT3A mutant CH, and non-DNMT3A mutant CH measured at VAF ≥ 2.0%. Overall and gene specific CH results were derived from the same model. Results from both raw (gray) and covariate adjusted (black) logistic regression models are shown. The dashed vertical line indicates the threshold for no effect from CH