By Michael H. Crawford, MD, Editor
Synopsis: An analysis of a very large database of patients with recent-onset atrial fibrillation has shown that whether sex was included in the formulas to predict thromboembolic risk and guide the use of oral anticoagulants probably is not as important as it was decades ago.
Source: Lam SHM, Romiti GF, Corica B, et al. Stroke risk stratifications according to CHA2DS2-VASc vs. CHA2DS2-VA in patients with atrial fibrillation: Insights from the GLORIA-AF registry. Eur Heart J Cardiovasc Pharmacother. 2025; Apr 28. doi/10.1093/ehjcvp/pvaf031. [Online ahead of print].
Recent data suggest that female sex is less significant as a stroke risk predictor in patients with atrial fibrillation (AF) than previously thought. It has been proposed that it be removed from the CHA2DS2-VASc score. Thus, these investigators from Phase III of the Global Registry on Long-term Oral Antithrombotic Treatment in Patients with Atrial Fibrillation (GLORIA-AF) sought to analyze the performance of the CHA2DS2-VASc score compared to the CHA2DS2-VA score in this large global contemporary database of patients with a recent diagnosis of AF.
In Phase III, patients were enrolled between 2014 and 2016 and followed for three years regardless of whether they were on oral anticoagulants (OAC), antiplatelet drugs, or neither. The primary outcome was thromboembolic events (TE; ischemic stroke, transient ischemic attacks, or systemic emboli) at one-year post-baseline. A secondary endpoint was ischemic stroke (the predominant event).
The total population consisted of 21,260 patients with recent AF (mean age 70 years, 45% women). At baseline, the women tended to be older, with lower body mass index, lower systolic blood pressure, less cardiovascular disease, less diabetes, less chronic kidney disease, less chronic obstructive lung disease, and fewer prior bleeding episodes (all P < 0.001). At baseline, 82% were taking OACs (22% warfarin), 11% were taking antiplatelet agents, and 7% were not taking any antithrombotic therapy.
As expected, OAC use increased as the CHA2DS2-VASc increased. Women were less likely to be taking OAC (odds ratio [OR], 0.90; 95% confidence interval [CI], 0.83-0.97), especially those at younger ages (P < 0.001 for the interaction). The risk of TE (hazard ratio [HR], 1.02; 95% CI, 0.82-1.26) and ischemic stroke (HR, 1.14; 95% CI, 0.85-1.53) were similar between men and women. The predictive ability of the two scores was similar for both outcomes: TE (area under the curve [AUC] 0.63 and 0.64 for CHA2DS2-VASc and CHA2DS2-VA, respectively) and stroke (AUC 0.64 for both scores).
A subgroup analysis of the 3,747 patients (45% women) not taking OAC showed similar predictive ability of the CHA2DS2-VASc and CHA2DS2-VA scores for TE (AUC for both scores: 0.64) and stroke (0.66 and 0.65, respectively). An analysis of the risk of TE by sex showed that women had a higher risk of stroke at one year (HR, 1.86; 95% CI, 1.03-3.33), but the risk of all TE was not significant (HR, 1.48; 95% CI, 0.97-2.27). The authors concluded that the CHA2DS2-VASc and CHA2DS2-VA had similar predictive ability for TE. However, the use of OAC in women is lower than in men in GLORIA-AF and may differ by age.
Commentary
The historical data showing an increased risk of TE in women with AF is more than 20 years old and more recent data suggest that the increased risk in women observed in these studies was age-dependent, a variable present in the CHA2DS2 scores. Also, the recent data show that the risk of TE in AF is decreasing, making any sex difference less significant. This prompted the European Society of Cardiology guidelines to recommend the CHA2DS2-VA score with a level of evidence of C. They also noted that this avoided consternation regarding transgender individuals. Recent observational studies, including GLORIA-AF, show that the two scores are of equal predictive value.
The CHA2DS2 scores were designed to identify low-risk patients in whom OAC could be withheld. This is important because the standard of care is OAC unless the patient is at very low risk of TE or at a very high risk of bleeding. Both scores are only moderately good at predicting who will experience a TE event (AUCs < 0.70). Also, in the subgroup not taking OAC, there was a slight increase in the risk of stroke, but not all TE, in women compared to men. This may be the result of the lower rate of OAC use in younger women. Thus, there is concern that using the CHA2DS2-VA score would further decrease OAC use in women. However, at each point in both scores the risk of all TE is similar for men and women in GLORIA-AF.
There are limitations to GLORIA-AF. It is an observational study that included patients taking and not taking OAC who may have had other confounders. Only patients with a CHA2DS2-VASc score of ≥ 1 were included. Follow-up was for only one year. Also, it was not powered to be definitive for the effect of sex or age. Although originally industry-sponsored to assess the use of dabigatran, I do not believe this biased this analysis of the data in this registry study.
The bottom line is that whichever score one uses, it is just one component of the decision of whether to prescribe OAC. Bleeding risk, fall risk, very advanced age, renal function, and weight also are important considerations — and do not forget patient preference.
Michael H. Crawford, MD, is Professor Emeritus of Medicine and Consulting Cardiologist, UCSF Health, San Francisco.
An analysis of a very large database of patients with recent-onset atrial fibrillation has shown that whether sex was included in the formulas to predict thromboembolic risk and guide the use of oral anticoagulants probably is not as important as it was decades ago.
You have reached your article limit for the month. Subscribe now to access this article plus other member-only content.
- Award-winning Medical Content
- Latest Advances & Development in Medicine
- Unbiased Content