Benefit and risk of oral anticoagulant initiation strategies in patients with atrial fibrillation and cancer: a target trial emulation using the SEER-Medicare database.

Oral anticoagulants (OACs) are recommended for patients with atrial fibrillation (AFib) having CHA2DS2-VASc score ≥ 2. However, the benefits of OAC initiation in patients with AFib and cancer at different levels of CHA2DS2-VASc is unknown. We included patients with new AFib diagnosis and a record of cancer (breast, prostate, or lung) from the 2012-2019 Surveillance, Epidemiology, and End Results (SEER)-Medicare database (n = 39,915). Risks of stroke and bleeding were compared between 5 treatment strategies: (1) initiated OAC when CHA2DS2-VASc ≥ 1 (n = 6008), (2) CHA2DS2-VASc ≥ 2 (n = 8694), (3) CHA2DS2-VASc ≥ 4 (n = 20,286), (4) CHA2DS2-VASc ≥ 6 (n = 30,944), and (5) never initiated OAC (reference group, n = 33,907). Confounders were adjusted using inverse probability weighting through cloning-censoring-weighting approach. Weighted pooled logistic regressions were used to estimate treatment effect [hazard ratios (HRs) and 95% confidence interval (95% CIs)]. We found that only patients who initiated OACs at CHA2DS2-VASc ≥ 6 had lower risk of stroke compared without OAC initiation (HR 0.64, 95% CI 0.54-0.75). All 4 active treatment strategies had reduced risk of bleeding compared to non-initiators, with OAC initiation at CHA2DS2-VASc ≥ 6 being the most beneficial strategy (HR = 0.49, 95% CI 0.44-0.55). In patients with lung cancer or regional/metastatic cancer, OAC initiation at any CHA2DS2-VASc level increased risk of stroke and did not reduce risk of bleeding (except for Regimen 4). In conclusion, among cancer patients with new AFib diagnosis, OAC initiation at higher risk of stroke (CHA2DS2-VASc score ≥ 6) is more beneficial in preventing ischemic stroke and bleeding. Patients with advanced cancer or low life-expectancy may initiate OACs when CHA2DS2-VASc score ≥ 6.

Development and Validation of Machine Learning Algorithms to Predict 1-Year Ischemic Stroke and Bleeding Events in Patients with Atrial Fibrillation and Cancer.

In this study, we leveraged machine learning (ML) approach to develop and validate new assessment tools for predicting stroke and bleeding among patients with atrial fibrillation (AFib) and cancer. We conducted a retrospective cohort study including patients who were newly diagnosed with AFib with a record of cancer from the 2012-2018 Surveillance, Epidemiology, and End Results (SEER)-Medicare database. The ML algorithms were developed and validated separately for each outcome by fitting elastic net, random forest (RF), extreme gradient boosting (XGBoost), support vector machine (SVM), and neural network models with tenfold cross-validation (train:test = 7:3). We obtained area under the curve (AUC), sensitivity, specificity, and F2 score as performance metrics. Model calibration was assessed using Brier score. In sensitivity analysis, we resampled data using Synthetic Minority Oversampling Technique (SMOTE). Among 18,388 patients with AFib and cancer, 523 (2.84%) had ischemic stroke and 221 (1.20%) had major bleeding within one year after AFib diagnosis. In prediction of ischemic stroke, RF significantly outperformed other ML models [AUC (0.916, 95% CI 0.887-0.945), sensitivity 0.868, specificity 0.801, F2 score 0.375, Brier score = 0.035]. However, the performance of ML algorithms in prediction of major bleeding was low with highest AUC achieved by RF (0.623, 95% CI 0.554-0.692). RF models performed better than CHA2DS2-VASc and HAS-BLED scores. SMOTE did not improve the performance of the ML algorithms. Our study demonstrated a promising application of ML in stroke prediction among patients with AFib and cancer. This tool may be leveraged in assisting clinicians to identify patients at high risk of stroke and optimize treatment decisions.

Screening for clinically relevant drug-drug interactions between direct oral anticoagulants and antineoplastic agents: a pharmacovigilance approach.

BACKGROUND: Use of direct oral anticoagulants (DOACs) in patients with cancer remains suboptimal due to the concern regarding potential drug-drug interactions (DDIs) with antineoplastic treatments. However, the clinical relevance of these DDIs is unknown. METHODS: We conducted a pharmacovigilance study of adverse event (AE) reports from the US Food and Drug Administration Adverse Event Reporting System from 1/1/2004 to 12/31/2021. AE reports containing DOACs and antineoplastic agents with CYP3A4/P-gp inhibitory or inducing activity suggested by published pharmacokinetic studies were included (n = 36,066). The outcomes of interest were bleeding or stroke, identified by MedDRA dictionary version 25.0. We used disproportionality analyses (DPA), logistic regression models (LR), and Multi-item Gamma-Poisson Shrinker (MGPS) (Empirical Bayes Geometric Means (EBGM) and 90% credible intervals (90% CIs)) algorithms to identify the safety signal of DDIs. RESULTS: The highest bleeding reporting rates for each drug class were the combination of DOACs with neratinib (39.08%, n = 34), tamoxifen (21.22%, n = 104), irinotecan (20.54%, n = 83), and cyclosporine (19.17%, n = 227). The highest rate of stroke was found for prednisolone (2.43%, n = 113). In the primary analysis, no signal of DDIs by the antineoplastic therapeutic class was detected by MGPS, DPA, and LR approaches. By individual antineoplastic drug, DOACs-neratinib was the only signal detected [EBGM (EB05-EB95) = 2.71 (2.03-3.54)]. CONCLUSION: No signal of DDIs between DOACs and antineoplastic agents was detected, except for DOAC-neratinib. Most DDIs between DOACs and antineoplastic agents may not be clinically relevant. The DDIs between DOACs and neratinib should be further examined in future research.

Assessment of novel oral anticoagulant use within a community teaching hospital.

Background: Novel oral anticoagulants (NOACs) are considered to be at least as effective and safe as warfarin with several advantages such as predictable pharmacokinetics, allowing for standardized dosing without monitoring, a lack of food interactions and fewer drug interactions; however, their misuse could potentially result in patient harm. Objective: To evaluate the appropriate use of the NOACs within a community teaching hospital. SETTING: A community teaching hospital in the United States. Method: A retrospective chart review of patients that were prescribed dabigatran, rivaroxaban, or apixaban at our institution from October 2012 through November 2014 was conducted. Main outcome measure: The primary objective was to determine the percentage of patients that were appropriately prescribed NOACs. Secondary objectives were to determine the number of patients who were inappropriately transitioned from warfarin or parenteral anticoagulants to a NOAC or vice versa, the number of incidents when a NOAC was held or discontinued inappropriately before a procedure and the number of bleeding or thrombotic events while taking a NOAC. Results: Of the 113 patients receiving therapy with an NOAC, appropriate prescribing was observed in 79.7%. Dabigatran, rivaroxaban, and apixaban were appropriately prescribed in 73.8%, 88.3%, and 85.8% of patients respectively. Lack of renal dose-adjustment in patients with reduced renal function was the most common reason for inappropriate use (8.8%). Ten out of 38 patients (26%) were inappropriately transitioned from/to other anticoagulants. Two out of six patients underwent a procedure without holding NOACs as recommended prior to surgery. Of all patients receiving NOACs, a total of 3 bleeding incidents were observed, one with each NOAC. Conclusion: The NOACs were appropriately prescribed for the majority of patients within our institution. Future efforts however should focus on ensuring appropriate dose adjustments for renal impairment, procedures for transitioning between NOACs and parenteral anticoagulants, and adequate withholding times for NOACs prior to surgery in order to optimize the management of NOACs usage within our institution.

Acquired Thrombophilia.

Acquired thrombophilia is associated with an increased risk of venous thromboembolism (VTE). Antiphospholipid syndrome (APS) is the most prevalent acquired thrombophilia and is associated with both venous and arterial thromboses. Human immunodeficiency virus (HIV) is another form of acquired thrombophilia. Risk factors associated with VTE in this population include those related to the disease itself, host factors, and the pharmacotherapy for HIV. A significant proportion of VTE events occur in patients with malignancies. There is an increase in mortality associated with patients having cancer who experience VTE when compared to patients having cancer without VTE. Combination oral contraceptive (COC) use infers risk of thromboembolic events. The risk is dependent upon the presence of an underlying inherited thrombophilia, the estrogen dose, and generation of progestin. Patients at highest risk of VTE include those receiving high-dose estrogen and fourth-generation, progesterone-containing contraceptives. With the exception of APS, thrombophilia status does not alter the acute treatment of an initial VTE in nonpregnant patients.

Pregnancy-Related Venous Thromboembolism.

Pregnancy is associated with an increased risk of venous thromboembolism (VTE), with a reported incidence ranging from 0.49 to 2 events per 1000 deliveries. Risk factors include advanced maternal age, obesity, smoking, and cesarian section. Women with a history of previous VTE are at a 4-fold higher risk of recurrent thromboembolic events during subsequent pregnancies. Additionally, the presence of concomitant thrombophilia, particularly factor V Leiden (homozygosity), prothrombin gene mutation (homozygosity), or antiphospholipid syndrome (APS), increases the risk of pregnancy-related VTE. Low-molecular-weight heparin (LMWH) and unfractionated heparin (UFH) are the drugs of choice for anticoagulation during pregnancy. LMWH is preferred due to ease of use and lower rates of adverse events. Women with high thromboembolic risk particularly those with a family history of VTE should receive antepartum thromboprophylaxis. Women with low thromboembolic risk or previous VTE caused by a transient risk factor (ie, provoked), who have no family history of VTE, may undergo antepartum surveillance. Postpartum anticoagulation can be considered in women with both high and low thromboembolic risk.

Thrombophilia Screening.

Although controversial, screening for thrombophilia has become common. Testing for antiphospholipid antibodies is indicated in order to guide treatment decisions if there is clinical suspicion for antiphospholipid syndrome. The utility of identifying other thrombophilias in symptomatic venous thromboembolism (VTE) is questionable, as the risk of recurrence does not appear to be increased by an appreciable degree with the most common disorders (heterozygosity for factor V Leiden or prothrombin mutation). Although recurrence appears to be increased in those with homozygous or multiple abnormalities and potentially deficiencies in natural anticoagulants, screening to detect these conditions is difficult to justify based on their rarity. The American College of Chest Physicians' current guidelines note the increased risk of recurrence with idiopathic, proximal events regardless of thrombophilia status. They suggest duration of anticoagulation therapy be based on location and provoking factors rather than whether or not the individual has a thrombophilia. Because routine prophylaxis in asymptomatic individuals with thrombophilia is not recommended, screening of asymptomatic family members is difficult to justify. Screening prior to prescribing combination oral contraceptives is not cost effective, may result in unwanted pregnancies, and may have little effect on the overall rate of VTE.