Real-world data on cardiovascular disease incidence in adults with acquired immune aplastic anemia.

Abstract Acquired aplastic anemia (AA) is an immune-mediated bone marrow failure with a high propensity for clonal hematopoiesis (CH), wherein mature blood cells arise from a single hematopoietic stem cell. Emerging evidence suggests heightened cardiovascular disease (CVD) risk with CH and in AA, though the mechanisms for these associations remain unclear. The goal of our study was to examine the effects of AA treatment and markers of CH on CVD incidence. We retrospectively reviewed 77 adult AA patients diagnosed between 2003 to present with a median follow up of 34 months. Median age at diagnosis was 49 years; 49% were male. By Camitta criteria, 45% had non-severe AA (NSAA) and 55% had at least severe AA (SAA). Paroxysmal nocturnal hemoglobinuria (PNH) granulocyte clone size at diagnosis (at AA) was available in 60/77; 25 had no PNH clone and the remaining 35 had a median clone size of 2%. Next-generation sequencing (NGS) at AA was performed in 33/77, with 18 patients harboring pathogenic mutations. Cytogenetics at AA were abnormal in 10 patients. Patients had a median of 1 CVD risk factor (former/current smoker, BMI ≥30, hypertension, chronic kidney disease, hyperlipidemia, diabetes, hypo/hyperthyroidism) pre-AA. Treatment exposure included 81% receiving horse (n=60) or rabbit (n=2) anti-thymocyte globulin (ATG), 83% cyclosporine (CSA), 58% eltrombopag (Epag), and 25% hematopoietic stem cell transplant (HSCT). Median times from AA to first ATG, CSA, and Epag were 6, 4, and 11 weeks, respectively. Median duration from initiation to definitive discontinuation of CSA and Epag were 13 and 13.5 months. Median time to HSCT was 13 months. Transfusion requirements varied considerably. The median packed red blood cells (PRBCs) units at our institution since AA diagnosis was 5 (range 0-238) and median platelet units was 7 (range 0-279). At 6-month response evaluation, 27% achieved complete response (CR), 25% partial response (PR), 14% were refractory, and 34% were not evaluable due early death, HSCT, or insufficient exposure. Clonal evolution (CE) to myelodysplastic syndrome (MDS) occurred in 7 patients. Among patients without CE, 5 acquired new mutations, 5 had a first NGS test later in the course showing new pathogenic variants, and 1 developed abnormal cytogenetics (del(20)q). CVD was defined as any history of ischemic strokes (IS), coronary artery disease (CAD), myocardial infarctions (MI), peripheral arterial disease (PAD), heart failure (HF), and arrhythmia, with persistent arrhythmia further defined as requiring treatment or consistently apparent over time. Twelve out of 77 patients had CVD pre-AA while 13/77 developed new CVD at or post AA; 4 of these 13 patients had only transient arrhythmias. Pre-AA CVD cases included 2 IS, 5 CAD, 2 MI, and 4 persistent arrythmias. New CVD cases at or post AA included 1 IS, 2 CAD, 4 MI, 2 PAD, 4 HF, and 14 arrythmias (6 persistent, 8 transient). All new MIs were related to demand ischemia from anemia. New HF cases had high transfusion burdens (median 40.5 units of PRBCs). Median time from AA to new CVD excluding arrythmias was 10 months. Meanwhile, median times from ATG and CSA to any new arrhythmias were 7 and 59 days, respectively. Compared to the overall cohort, the 25/77 with CVD had similar gender distribution, AA severity, median PNH granulocyte clone size (2%), number with pathogenic mutations, median pre-AA CVD risk factors, ATG and CSA exposure, time from AA to ATG, CSA, and Epag, duration of CSA and Epag, and rate of CE to MDS. Recurrent somatic gene mutations seen in the CVD-free group were BCOR/BCORL1 (n=6), U2AF1 (n=2), and CHEK2 (n=2). Patients with CVD were older (median 67 vs 49 years) and had greater exposure to Epag (72% vs 58%), lower rates of HSCT (16% vs 25%), and higher PRBC (median 14 vs 5) and platelet (median 12 vs 7) requirements. Within the CVD group, those with new CVD (n=13) had fewer pre-AA CVD risk factors (median 2 vs 4) and higher median PNH granulocyte clone size (4 vs 0.2%) than those with preexisting CVD. Among 77 AA patients, 17% developed de novo CVD and 32% had CVD overall, most commonly MI, HF, and transient arrythmias. New MI and HF were associated with anemia-related demand ischemia and high transfusion burden. Arrhythmias appeared temporally linked to ATG and CSA exposure. The impact of CH, CE, and individual somatic mutations on IS, CAD, and PAD incidence remains unclear and warrants further studies in larger cohorts.