Chiswell, Karen

Abstract Background Direct oral anticoagulation (DOACs) prevent stroke in patients with atrial fibrillation (AF) and have a superior safety profile compared with vitamin K antagonists (VKA). Yet, better definition of incidence and risk factors for major bleeding associated with DOACs in clinical practice may be important given emerging stroke prevention technologies, both pharmacologic and nonpharmacologic. Purpose To describe the incidence of and risk factors for major bleeding in individuals with AF on DOAC therapy. Methods We reviewed electronic health record data for two patient cohorts with AF prescribed DOACs: (1) Duke University Health System (DUHS) (2010–2018) and (2) Sites within the Patient-Centered Clinical Research Network (PCORnet) (2015–2019) which had ≥6 years assimilated data from both inpatient and outpatient encounters (7 sites). In each cohort, we assessed the 5-year incidence of major bleeding events defined as hospitalization for intracranial hemorrhage, or hospitalization for gastro-intestinal bleeding or procedure to control bleeding accompanied by transfusion within ±7 days or death within 30 days. Multivariable Fine-Gray proportional hazards modeling in each cohort was performed to evaluate independent risk factors for major bleeding on DOAC therapy. Results The cohorts included 10,625 patients (DUHS) and 58,321 patients (PCORnet) with AF. Major bleeding events occurred within 5 years of diagnosis in 639 (7.9%) of DUHS patients and 2568 (6.6%) of PCORnet patients (Table 1). The DUHS model predicted time to first major bleeding event with a C-index of 0.756 (95% CI 0.737, 0.775) and the PCORNet model had a c-index of 0.745 (0.736, 0.755) (Table 2). Independent factors associated with major bleeding consistent across both models (p<0.001 in PCORnet for all unless noted) were higher CHA2DS2-VASc scores, lower eGFR, anemia (HR per 1-point increase in hemoglobin up to 12 g/dL 0.79 [0.76, 0.82]), prior major bleeding (HR 2.70 [2.22, 3.30]), cancer (HR 1.23 [1.12, 1.36]), recent cardiac surgery (HR 0.70 [0.51, 0.97]; p=0.030), alcohol use (HR 1.56 [1.29, 1.88]), aspirin use (HR 1.44 [1.32, 1.57]), and selective serotonin reuptake inhibitor use (HR 1.30 [1.19, 1.42]). Conclusions Across a large and geographically diverse contemporary population, risk of bleeding on DOAC for stroke prevention in AF remains a frequent and important clinical problem. There is an unmet need for stroke prevention therapies with improved safety profiles. We identified risk factors for major bleeding events on DOAC therapy, some of which are not represented in traditional risk scores, which may inform shared decision making for stroke prevention. Funding Acknowledgement Type of funding sources: Private company. Main funding source(s): Bayer Pharmaceuticals

Introduction: Elevated lipoprotein(a) [Lp(a)] is associated with atherosclerotic cardiovascular disease (ASCVD), yet little is known about Lp(a) testing patterns in real-world practice. We hypothesized that Lp(a) testing occurs in a different population than calculated low density lipoprotein (LDL-C) testing alone, and that elevated Lp(a) is associated with greater lipid-lowering therapy (LLT) initiation and cardiovascular (CV) events. Methods: We used data from 11 health systems participating in the National Patient-Centered Clinical Research Network. We created two cohorts based on lab tests from Jan 1, 2015 to Dec 31, 2019: 1) the Lp(a) cohort, of adults with an Lp(a) test and 2) the LDL-C cohort, of 4:1 date- and site-matched adults with an LDL-C test, but no Lp(a) test. We used logistic regression to assess the relationship between Lp(a) levels (< 50, 50-100, and > 100mg/dL) and initiation of LLT within 3 months; and cox proportional hazards to evaluate Lp(a) levels and time to all-cause hospitalization and the composite of MI, revascularization and ischemic stroke hospitalization. Results: Overall, 20,551 patients (0.06% per year) had Lp(a) tests, with median Lp(a) value 17.5mg/dL (Q1, Q3: 7.0, 56.0). Compared with the LDL-C cohort (N=82,204), the Lp(a) cohort more frequently had ASCVD (24.3% vs. 8.5%) or multiple prior CV events (8.6% vs. 2.6%). Elevated Lp(a) was associated with greater odds of LLT initiation (Figure 1). Elevated Lp(a) was not significantly associated with all-cause hospitalization; however, Lp(a) > 100mg/dL was correlated with composite CV hospitalization compared with Lp(a) < 50mg/dL [adj HR (95% CI): 1.24 (1.06-1.44), p=0.006]. Conclusions: Lp(a) testing occurs more often in adults with ASCVD and multiple CV events, but testing is infrequent. Elevated Lp(a) was associated with LLT initiation, including niacin, despite lack of evidence of CV benefit. There was an association between elevated Lp(a) and CV hospitalization outcomes.