Clinical risk stratification in spontaneous coronary artery dissection: development and incremental validation of a prognostic model

Background

Spontaneous coronary artery dissection (SCAD) is an increasingly recognised cause of acute coronary syndromes, predominantly affecting women, yet prognostic assessment remains poorly defined and no dedicated risk models are available.

Purpose

To develop and internally validate a clinical risk model for adverse outcomes in SCAD and to evaluate the incremental prognostic value of treatment-related variables.

Methods

A single-centre observational study was conducted including 64 patients diagnosed with SCAD between November 2009 and August 2025 and followed for a median of 69 months (IQR 34–114). The primary endpoint was major adverse cardiac events (MACE), defined as a composite of cardiovascular death, myocardial infarction, stroke, or hospitalisation for heart failure. Two multivariable logistic regression models were developed. Model A included baseline clinical and angiographic variables (age, TIMI risk score, dissection type, and culprit vessel) selected based on biological plausibility and availability at presentation. Model B incrementally added treatment-related variables (percutaneous coronary intervention, beta-blocker therapy, and statin therapy) to assess their additional prognostic value. Model performance was assessed using discrimination (area under the receiver operating characteristic curve [AUC]) and internal validation with bootstrap resampling.

Results

The study population was predominantly female (93.8%), with a median age of 51.5 years (IQR 44.0–64.0). Conservative management was the most frequent strategy, while 15.9% (10/63) experienced the composite endpoint during follow-up.

Model A demonstrated good apparent discrimination (AUC 0.79), reflecting the prognostic contribution of baseline clinical and angiographic characteristics. Model B showed a higher apparent AUC (0.93), largely driven by treatment variables reflecting initial disease severity; however, this improvement was not sustained after internal validation, indicating limited incremental prognostic value beyond baseline characteristics.

Conclusion

In patients with SCAD, a clinical and angiographic risk model provides meaningful prognostic discrimination for adverse outcomes. The addition of treatment-related variables does not confer robust incremental prognostic value, suggesting that baseline disease characteristics primarily drive long-term risk. This pragmatic model may support early risk stratification and highlights the need for external validation in larger SCAD cohorts.ROCFor image description, please refer to the figure legend and surrounding text.ModelFor image description, please refer to the figure legend and surrounding text.