P2713A genetic risk score improves discrimination of hemodynamically obstructive coronary artery disease (CAD) beyond the CAD Consortium scores in patients at low-to-intermediate risk of CAD

Abstract

Abstract Background Genetic risk scores (GRSs), based on variants identified in genome-wide association studies (GWAS), have been shown to predict risk of coronary artery disease (CAD). However, the clinical potential remains unknown. Purpose To investigate whether a GRS improves discrimination of hemodynamically obstructive CAD beyond the CAD Consortium scores and coronary artery calcium score (CACS) in patients referred for coronary computed tomography angiography (CTA). Methods We consecutively included and genotyped 1645 patients undergoing CACS scoring and coronary CTA on a suspicion of CAD. Using LDPred, a recently validated GRS was calculated as the weighted sum of the number of CAD risk variants identified from the CARDIoGRAMplusC4D GWAS meta-analysis. Patients with a ≥50% stenosis on CTA further underwent invasive coronary angiography (ICA) with fractional flow reserve (FFR). Hemodynamically obstructive CAD was defined as a visual ICA stenosis >90%, FFR <0.80, or a quantitative coronary analysis stenosis >50% if FFR was not feasible. Discrimination was evaluated by receiver-operating characteristics. Results Median age was 57 (interquartile range 50–64) years and 799 (49%) were males. Hemodynamically obstructive CAD was present in 14 (4%) with a low GRS (<20th percentile), 91 (9%) with an intermediate GRS (20th–80th percentile) and 53 (16%) with a high GRS (>80th percentile) (p<0.0001). Adding the GRS improved the area under the receiver-operating curve (AUC) on top of the CAD Consortium basic score (from 0.67 to 0.72, p=0.0052), and the CAD Consortium clinical score (0.70 to 0.74, p=0.0084), but not on top of the CAD Consortium clinical score + CACS (0.85 to 0.86, p=0.30). Improvement in discrimination on top of the CAD Consortium scores was predominantly driven by females ≤57 years (CAD Consortium basic score ± GRS: 0.60 to 0.78, p=0.0004; CAD Consortium clinical score ± GRS: 0.63 to 0.78, p=0.0007). The GRS did not improve discrimination in any subgroups including CACS (CAD Consortium clinical score + CACS ± GRS: all p-values >0.05). Conclusion A GRS improves discrimination of hemodynamically obstructive CAD beyond CAD consortium scores, particularly in young women. However, the additive discriminative value is attenuated in models including CACS.