Precision Medicine in Atrial Fibrillation: Biomarker-Based ABC-AF Risk Scores Fail to Outperform Standard Care

Highlights

No Superiority Over Standard Care

The multidimensional treatment strategy guided by ABC-AF risk scores did not show a statistically significant reduction in the composite primary outcome of stroke or death compared to conventional clinical management.

Early Termination and Safety Concerns

The trial was terminated prematurely due to safety signals, specifically a trend toward higher mortality in the ABC-AF strategy arm among patients with higher baseline clinical risk scores.

High Baseline Performance of Standard Care

Both groups achieved high rates of oral anticoagulant (OAC) therapy, suggesting that contemporary guideline-based care provides a robust baseline that is difficult to improve upon even with advanced biomarker tools.

Background: The Evolution of Risk Stratification in Atrial Fibrillation

Atrial fibrillation (AF) remains a leading cause of stroke, heart failure, and mortality worldwide. For decades, clinical risk scores such as CHA2DS2-VASc and HAS-BLED have served as the cornerstone of decision-making for thromboprophylaxis. However, these scores rely on clinical variables that may not fully capture the dynamic biological state of the patient.

In recent years, the Age, Biomarkers (N-terminal pro–B-type natriuretic peptide [NT-proBNP] and high-sensitivity cardiac troponin [hs-cTn]), and Clinical history (ABC) risk scores emerged as promising tools. Validated in several retrospective cohorts, the ABC-AF stroke and bleeding scores appeared to offer superior predictive accuracy compared to traditional clinical-only models. Despite this, prospective evidence demonstrating that using these scores actually improves patient outcomes remained sparse. The ABC-AF study was designed to bridge this gap by testing whether a precision medicine approach, integrating laboratory biomarkers into routine care, could translate into better clinical results.

Study Design: A Registry-Based Randomized Approach

The ABC-AF study was a multicenter, registry-based, randomized, controlled, open-label trial. It enrolled adults with AF across various clinical settings. The study utilized a unique design where investigators in the intervention arm were provided with the participant’s specific ABC-AF risks for stroke and bleeding. These scores served as a decision-support tool to tailor treatment recommendations, including the selection of specific direct oral anticoagulants (DOACs) or other management strategies.

In contrast, the standard of care (SoC) arm managed patients at the discretion of the investigator, following existing clinical guidelines. The primary endpoint was a composite of stroke or death. Secondary endpoints included individual components of the primary outcome, major bleeding events, and a broader composite of stroke, death, or major bleeding. The median follow-up period was 2.6 years, providing a substantial window to observe clinical events in a high-risk population.

Key Findings: Analyzing the Data

The intention-to-treat population included 3,933 patients with a median age of 73.7 years. The baseline characteristics reflected a typical AF population: roughly one-third were women, and over 11% had a history of prior stroke or transient ischemic attack.

Primary and Secondary Outcomes

Over the follow-up period, 175 primary events occurred in the ABC-AF strategy arm (3.18 per 100 patient-years) compared to 148 events in the standard of care arm (2.67 per 100 patient-years). The resulting hazard ratio (HR) was 1.19 (95% CI, 0.96-1.48; P=0.12), indicating no significant benefit for the biomarker-guided strategy.

Table 3. Primary and Secondary Outcomes

Figure 2.

Secondary outcomes mirrored this lack of differentiation:
– Major Bleeding: HR 1.08 (95% CI, 0.86-1.36; P=0.50)
– Stroke: HR 1.18 (95% CI, 0.78-1.79; P=0.44)
– All-Cause Mortality: HR 1.21 (95% CI, 0.94-1.55; P=0.13)
– Composite of Stroke, Death, or Major Bleeding: HR 1.14 (95% CI, 0.96-1.36; P=0.13)

Subgroup Consistency

The results were consistent across various subgroups. Specifically, when analyzing patients categorized by their ABC-AF scores, there was no significant interaction (P=0.98), suggesting that the intervention failed to provide added value regardless of the patient’s baseline biomarker risk profile.

The Safety Signal and Early Termination

One of the most critical aspects of the trial was its premature termination. The data monitoring committee identified a concerning trend: patients in the ABC-AF strategy arm who had a CHA2DS2-VASc score of 3 or higher exhibited a trend toward higher mortality. While this did not reach the threshold for definitive harm, the safety concerns, combined with the low likelihood of demonstrating the primary benefit, led to the cessation of the study. This termination meant the trial was ultimately underpowered to reach its original statistical objectives, though the trends observed strongly suggest that the intervention was not on track to show superiority.

Expert Commentary: Interpreting the Neutral Result

Why did a strategy that appeared so promising in retrospective validation fail in a prospective trial? Several factors may explain these results. First, the standard of care in the participating centers was exceptionally high. At the time of randomization, over 85% of patients were already on OACs, and this rose to over 92% in the SoC arm during the study. When the baseline care is already optimized according to evidence-based guidelines, the room for incremental improvement by adding biomarkers is significantly narrowed.

Second, the ‘human factor’ in decision support cannot be ignored. Providing a risk score to a clinician is only effective if it leads to a meaningful and superior change in management. In this trial, the ABC-AF strategy arm did see a slightly higher rate of OAC use (97.8% vs 92.6%), but this 5% difference may not have been enough to overcome the inherent risks of the population, or perhaps the ‘tailored’ changes in DOAC types did not offer a biological advantage over the standard choices already being made by experienced clinicians.

Finally, the safety signal in higher-risk patients suggests that biomarker-guided intensification of care might inadvertently lead to adverse outcomes, perhaps through over-treatment or shifts in management that disrupt a stable clinical balance. This underscores a vital lesson in medical science: predictive accuracy does not always translate to therapeutic utility.

Conclusion: The Path Forward for Precision Medicine

The ABC-AF trial serves as a sobering reminder that the implementation of precision medicine tools requires rigorous prospective validation before they are integrated into routine clinical practice. While biomarkers like NT-proBNP and high-sensitivity troponin remain invaluable for prognosis, their role in actively directing AF therapy remains unproven.

For clinicians, the takeaway is clear: current guideline-based care, focused on the CHA2DS2-VASc score and clinical judgment, remains highly effective for the majority of patients. Future research should focus on identifying specific niches where biomarkers might truly change management, such as in patients with borderline clinical risk scores or those at exceptionally high risk of bleeding where the benefit-risk ratio is narrowest.

Funding and Registration

This study was a registry-based, multicenter trial. Detailed funding information and trial protocols are available at ClinicalTrials.gov.

Registration: URL: https://www.clinicaltrials.gov; Unique identifier: NCT03753490.

References

Oldgren J, Hijazi Z, Arheden H, et al. Biomarker-Based ABC-AF Risk Scores for Personalized Treatment to Reduce Stroke or Death in Atrial Fibrillation: A Registry-Based, Multicenter, Randomized, Controlled Study. Circulation. 2025 Nov 25;152(21):1457-1469. doi: 10.1161/CIRCULATIONAHA.125.076725 IF: 38.6 Q1 B1.