Meeting News

AI-powered blood test predicts PAD, need for revascularization in patients with diabetes

MUNICH — An artificial intelligence-powered blood test consisting of one clinical variable and six biomarkers predicted with high accuracy the presence of peripheral artery disease and need for revascularization in patients with and without diabetes, researchers reported at the European Society of Cardiology Congress.

The HART PAD test (Prevencio) comprises one clinical variable — history of hypertension — plus concentrations of six biomarkers: midkine, kidney injury molecule-1, interleukin-23, follicle-stimulating hormone, angiopoietin-1 and eotaxin-1. The predictors of PAD included in the panel were identified via proteomics and machine learning.
Researchers analyzed the accuracy of the HART PAD test in a prospective cohort of 354 patients in the CASABLANCA study who were referred for diagnostic peripheral angiography and/or coronary angiography. Ninety-four patients had diabetes.

According to the findings, the HART PAD panel showed excellent performance for predicting peripheral stenosis greater than 50% in patients with diabetes. The area under the receiver operating characteristic curve was 0.85 for obstructive PAD. The higher the score, the greater the severity of angiographic stenosis.

For detection of PAD in patients with diabetes, the model performed with a sensitivity of 84%, specificity of 88%, positive predictive value of 76% and negative predictive value of 81%. Using a 5-point score, a score of 1 had a 100% negative predictive value and a score of 5 had a 95% positive predictive value, according to a press release.

In other findings, the HART PAD panel was highly accurate in predicting the need for revascularization in patients with PAD. A higher score was associated with shorter time to revascularization during 4.3 years of follow-up, according to the data reported here.

Results were comparable to those in patients without diabetes.

The researchers noted that a limitation of the study is the lack of biomarkers measured at a single point in time; thus, biomarker concentrations may not reflect levels and future time periods.

“Clinically, use of a tool such as this could act as a gatekeeper prior to imaging or invasive testing, thus reducing cost and exposures to intravenous contrast and/or ionizing radiation. The score may also be used to evaluate at-risk patients for risk of vascular complications; as such, a role in clinical trials to enrich for PAD-related events or to identify patients at risk for adverse effects of drug therapies is plausible,” McCarthy and colleagues wrote in the posted abstract. - by Katie Kalvaitis

References:

McCarthy C, et al. P732. Presented at: European Society of Cardiology Congress; Aug. 25-29, 2018; Munich.

McCarthy C, et al. Eur Heart J. 2018;doi:10.1093/eurheartj/ehy564.P732.

Disclosures: McCarthy reports no relevant financial disclosures. Please see the full study for a list of the other authors’ relevant financial disclosures.

MUNICH — An artificial intelligence-powered blood test consisting of one clinical variable and six biomarkers predicted with high accuracy the presence of peripheral artery disease and need for revascularization in patients with and without diabetes, researchers reported at the European Society of Cardiology Congress.

The HART PAD test (Prevencio) comprises one clinical variable — history of hypertension — plus concentrations of six biomarkers: midkine, kidney injury molecule-1, interleukin-23, follicle-stimulating hormone, angiopoietin-1 and eotaxin-1. The predictors of PAD included in the panel were identified via proteomics and machine learning.
Researchers analyzed the accuracy of the HART PAD test in a prospective cohort of 354 patients in the CASABLANCA study who were referred for diagnostic peripheral angiography and/or coronary angiography. Ninety-four patients had diabetes.

According to the findings, the HART PAD panel showed excellent performance for predicting peripheral stenosis greater than 50% in patients with diabetes. The area under the receiver operating characteristic curve was 0.85 for obstructive PAD. The higher the score, the greater the severity of angiographic stenosis.

For detection of PAD in patients with diabetes, the model performed with a sensitivity of 84%, specificity of 88%, positive predictive value of 76% and negative predictive value of 81%. Using a 5-point score, a score of 1 had a 100% negative predictive value and a score of 5 had a 95% positive predictive value, according to a press release.

In other findings, the HART PAD panel was highly accurate in predicting the need for revascularization in patients with PAD. A higher score was associated with shorter time to revascularization during 4.3 years of follow-up, according to the data reported here.

Results were comparable to those in patients without diabetes.

The researchers noted that a limitation of the study is the lack of biomarkers measured at a single point in time; thus, biomarker concentrations may not reflect levels and future time periods.

“Clinically, use of a tool such as this could act as a gatekeeper prior to imaging or invasive testing, thus reducing cost and exposures to intravenous contrast and/or ionizing radiation. The score may also be used to evaluate at-risk patients for risk of vascular complications; as such, a role in clinical trials to enrich for PAD-related events or to identify patients at risk for adverse effects of drug therapies is plausible,” McCarthy and colleagues wrote in the posted abstract. - by Katie Kalvaitis

References:

McCarthy C, et al. P732. Presented at: European Society of Cardiology Congress; Aug. 25-29, 2018; Munich.

McCarthy C, et al. Eur Heart J. 2018;doi:10.1093/eurheartj/ehy564.P732.

Disclosures: McCarthy reports no relevant financial disclosures. Please see the full study for a list of the other authors’ relevant financial disclosures.

    See more from European Society of Cardiology Congress