Aims

Coronary artery calcium (CAC) scoring is an established tool for cardiovascular risk stratification. However, the lack of widespread availability and concerns about radiation exposure have limited the universal clinical utilization of CAC. In this study, we sought to explore whether machine learning (ML) approaches can aid cardiovascular risk stratification by predicting guideline recommended CAC score categories from clinical features and surface electrocardiograms.

Methods and results

In this substudy of a prospective, multicentre trial, a total of 534 subjects referred for CAC scores and electrocardiographic data were split into 80% training and 20% testing sets. Two binary outcome ML logistic regression models were developed for prediction of CAC scores equal to 0 and ≥400. Both CAC = 0 and CAC ≥400 models yielded values for the area under the curve, sensitivity, specificity, and accuracy of 84%, 92%, 70%, and 75%, and 87%, 91%, 75%, and 81%, respectively. We further tested the CAC ≥400 model to risk stratify a cohort of 87 subjects referred for invasive coronary angiography. Using an intermediate or higher pretest probability (≥15%) to predict CAC ≥400, the model predicted the presence of significant coronary artery stenosis (P = 0.025), the need for revascularization (P < 0.001), notably bypass surgery (P = 0.021), and major adverse cardiovascular events (P = 0.023) during a median follow-up period of 2 years.

Conclusion

ML techniques can extract information from electrocardiographic data and clinical variables to predict CAC score categories and similarly risk-stratify patients with suspected coronary artery disease.

Background

Deficiencies of iron-sulfur (Fe-S) clusters, metal complexes that control redox state and mitochondrial metabolism, have been linked to pulmonary hypertension (PH), a deadly vascular disease with poorly defined molecular origins. BOLA3 (BolA Family Member 3) regulates Fe-S biogenesis, and mutations in BOLA3 result in multiple mitochondrial dysfunction syndrome, a fatal disorder associated with PH. The mechanistic role of BOLA3 in PH remains undefined.

Methods

In vitro assessment of BOLA3 regulation and gain- and loss-of-function assays were performed in human pulmonary artery endothelial cells using siRNA and lentiviral vectors expressing the mitochondrial isoform of BOLA3. Polymeric nanoparticle 7C1 was used for lung endothelium-specific delivery of BOLA3 siRNA oligonucleotides in mice. Overexpression of pulmonary vascular BOLA3 was performed by orotracheal transgene delivery of adeno-associated virus in mouse models of PH.

Results

In cultured hypoxic pulmonary artery endothelial cells, lung from human patients with Group 1 and 3 PH, and multiple rodent models of PH, endothelial BOLA3 expression was downregulated, which involved hypoxia inducible factor-2α-dependent transcriptional repression via histone deacetylase 1-mediated histone deacetylation. In vitro gain- and loss-of-function studies demonstrated that BOLA3 regulated Fe-S integrity, thus modulating lipoate-containing 2-oxoacid dehydrogenases with consequent control over glycolysis and mitochondrial respiration. In contexts of siRNA knockdown and naturally occurring human genetic mutation, cellular BOLA3 deficiency downregulated the glycine cleavage system protein H, thus bolstering intracellular glycine content. In the setting of these alterations of oxidative metabolism and glycine levels, BOLA3 deficiency increased endothelial proliferation, survival, and vasoconstriction while decreasing angiogenic potential. In vivo, pharmacological knockdown of endothelial BOLA3 and targeted overexpression of BOLA3 in mice demonstrated that BOLA3 deficiency promotes histological and hemodynamic manifestations of PH. Notably, the therapeutic effects of BOLA3 expression were reversed by exogenous glycine supplementation.

Conclusions

BOLA3 acts as a crucial lynchpin connecting Fe-S-dependent oxidative respiration and glycine homeostasis with endothelial metabolic reprogramming critical to PH pathogenesis. These results provide a molecular explanation for the clinical associations linking PH with hyperglycinemic syndromes and mitochondrial disorders. These findings also identify novel metabolic targets, including those involved in epigenetics, Fe-S biogenesis, and glycine biology, for diagnostic and therapeutic development.