Predicting risk of future events in individuals with chronic coronary syndromes

Background

• Chronic coronary syndrome (CCS) is a dynamic process of atherosclerotic plaque accumulation and functional alterations of coronary circulation that can be modified by lifestyle, pharmacological therapies, and revascularization which result in disease stabilization or regression.
• Investigation of suspicion of CCS is often focused on the diagnosis of local anatomical or functional disease. However, this approach is limited in its ability to detect functionally relevant changes and predict systemic risks for future cardiovascular events.
• The easy, accurate, and safe prognostic risk-stratification of CCS patients is an unmet clinical need.

Objective

Evaluate whether a previously validated 27-protein prognostic model for four-year cardiovascular event risk can be used to stratify patients with suspected chronic coronary syndrome (CCS).

Methods

• BASEL VIII is a large prognostic and diagnostic study in participants with suspected CCS designed to evaluate and advance the non-invasive diagnosis of functionally relevant coronary artery disease (fCAD) (as adjudicated by independent cardiologists) using myocardial perfusion scanning and coronary angiography.
• Measured ~5000 proteins in plasma samples (taken at rest) utilizing SomaScan® Platform from 4106 evaluable participants
  • 1688 participants without prior cardiovascular disease (primary population)
  • 2418 participants with known prior cardiovascular disease (secondary population)
• Evaluated 27-protein model of cardiovascular event risk in primary and secondary populations
  • Continuous output of protein risk score was calculated for each population.
  • Low risk bin was evaluated in participants with low risk for an event as a potential rule-out from additional assessments.
  • High risk bin was evaluated for added utility with imaging.
  • Prognostic performance of 27-protein model was compared to final fCAD diagnosis.

Results: Primary Population

Low category for proteomic risk prediction as a rule-out in primary participants

• 41% of primary participants were categorized as low risk and had an observed event rate of 2%.
• Observed event rates were not significantly different between fCAD- and fCAD+ participants.

High categorical risk scores to identify imaging negative participants at high risk of an event

• High risk categorical scores were present in 48% of primary patients negative for fCAD.

Results: Secondary Population

Low category for proteomic risk prediction as a rule-out in secondary participants

• 27% of secondary participants were categorized as low risk and had an observed event rate of 10%.
• Observed event rates were not significantly different between fCAD- and fCAD+ participants.

High categorical risk scores to identify false negatives of imaging negative participants

• High risk categorical scores were present in 35% of secondary patients negative for fCAD.

Conclusions

• The low-risk category of the proteomic model could effectively rule-out 41% and 27% of the primary and secondary populations from further assessment with negative predictive values of 98% and 92%.
• In subjects with a negative fCAD diagnosis, the proteomic model identified subjects with elevated cardiovascular risk who would have otherwise been overlooked by perfusion imaging and angiography with event rates of 15% for the primary and 26% for the secondary cohorts.
• CVD absolute risk score outperformed fCAD diagnosis alone and in combination at predicting 4-year events.
• These findings suggest that the 27-protein prognostic model could be useful in addition to imaging – by ruling out the need for imaging in a substantive fraction and exposing unresolved risks in imaging negative patients.

Authors

Stephen A. Williams¹
Jessica Chadwick¹
David Astling¹
Michael Hinterberg¹
Rachel Ostroff¹
Emma Troth¹
Joan E. Walter²
Christian Mueller²

¹SomaLogic Operating Co., Inc.
²University of Basel