Can machine learning unravel unsuspected, clinically important factors predictive of long-term mortality in complex coronary artery disease? A call for 'big data'

γ-Glutamyl Transferase and Long-Term Survival in the SYNTAXES Trial: Is It Just the Liver?

BACKGROUND

Recently, machine learning algorithms have identified preprocedural γ-glutamyl transferase (GGT) as a significant predictor of long-term mortality after coronary revascularization in the SYNTAX (Synergy Between PCI [Percutaneous Coronary Intervention] With Taxus and Cardiac Surgery) trial. The aim of the present study is to investigate the impact of preprocedural GGT on 10-year all-cause mortality in patients with complex coronary artery disease after revascularization.

METHODS AND RESULTS

The SYNTAX trial was a randomized trial comparing PCI with coronary artery bypass grafting in 1800 patients with complex coronary artery disease. The present report is a post hoc subanalysis of the SYNTAXES (Synergy Between PCI With Taxus and Cardiac Surgery Extended Survival) trial, an investigator-driven extended 10-year follow-up of the SYNTAX trial. The association between preprocedural GGT and 10-year all-cause mortality was investigated. The mean values of GGT for men and women were 43.5 (SD, 48.5) and 36.4 (SD, 46.1) U/L, respectively. In multivariable Cox regression models adjusted by traditional risk factors, GGT was an independent predictor for all-cause death at 10-year follow-up, and each SD increase in log-GGT was associated with a 1.24-fold risk of all cause death at 10-year follow-up (95% CI, 1.10-1.40). According to previously reported sex-related GGT thresholds, patients with higher GGT level had a 1.74-fold risk of all-cause death at 10-year follow-up (95% CI, 1.32-2.29) compared with patients with lower GGT level.

CONCLUSIONS

Preprocedural GGT is an independent predictor of 10-year mortality after coronary revascularization in patients with complex coronary artery disease. In patients with elevated GGT, strong secondary prevention may be required after revascularization and must be studied prospectively.

REGISTRATION

URL: https://clinicaltrials.gov/study/NCT03417050.

Can Machine Learning Aid the Selection of Percutaneous vs Surgical Revascularization?

BACKGROUND

In patients with 3-vessel coronary artery disease (CAD) and/or left main CAD, individual risk prediction plays a key role in deciding between percutaneous coronary intervention (PCI) and coronary artery bypass grafting (CABG).

OBJECTIVES

The aim of this study was to assess whether these individualized revascularization decisions can be improved by applying machine learning (ML) algorithms and integrating clinical, biological, and anatomical factors.

METHODS

In the SYNTAX (Synergy between PCI with Taxus and Cardiac Surgery) study, ML algorithms (Lasso regression, gradient boosting) were used to develop a prognostic index for 5-year death, which was combined, in the second stage, with assigned treatment (PCI or CABG) and prespecified effect-modifiers: disease type (3-vessel or left main CAD) and anatomical SYNTAX score. The model's discriminative ability to predict the risk of 5-year death and treatment benefit between PCI and CABG was cross-validated in the SYNTAX trial (n = 1,800) and externally validated in the CREDO-Kyoto (Coronary REvascularization Demonstrating Outcome Study in Kyoto) registry (n = 7,362), and then compared with the original SYNTAX score II 2020 (SSII-2020).

RESULTS

The hybrid gradient boosting model performed best for predicting 5-year all-cause death with C-indexes of 0.78 (95% CI: 0.75-0.81) in cross-validation and 0.77 (95% CI: 0.76-0.79) in external validation. The ML models discriminated 5-year mortality better than the SSII-2020 in the external validation cohort and identified heterogeneity in the treatment benefit of CABG vs PCI.

CONCLUSIONS

An ML-based approach for identifying individuals who benefit from CABG or PCI is feasible and effective. Implementation of this model in health care systems-trained to collect large numbers of parameters-may harmonize decision making globally. (Synergy Between PCI With TAXUS and Cardiac Surgery: SYNTAX Extended Survival [SYNTAXES]; NCT03417050; SYNTAX Study: TAXUS Drug-Eluting Stent Versus Coronary Artery Bypass Surgery for the Treatment of Narrowed Arteries; NCT00114972).

10 Years of SYNTAX: Closing an Era of Clinical Research After Identifying New Outcome Determinants

The SYNTAX trial randomized patients equally eligible for coronary artery bypass grafting or percutaneous coronary intervention using the Heart Team approach. The SYNTAXES study achieved a follow-up rate of 93.8% and reported the 10-year vital status. Factors associated with increased mortality at 10 years were pharmacologically treated diabetes mellitus, increased waist circumference, reduced left ventricular function, prior cerebrovascular and peripheral vascular disease, western Europe and North American descent, current smoking, chronic obstructive pulmonary disease, elevated C-reactive protein, anemia, and an increase in HbA1c. Procedural factors associated with higher 10 years mortality include periprocedural myocardial infarction, extensive stenting, small stents, ≥1 heavily calcified lesion, ≥1 bifurcation lesion, residual SYNTAX score >8, and staged percutaneous coronary intervention. Optimal medical therapy at 5 years, use of statins, on-pump coronary artery bypass grafting, multiple arterial grafts, and higher physical component score and mental component score were associated with lower mortality at 10 years. Numerous scores and prediction models were developed to help individualize risk assessment. Machine learning has emerged as a novel approach for developing risk models.