Rohan Khera, MD, MS @rohan_khera @YaleMed @YaleCardiology @cards_lab #ASSIST #ChestPain #Cardiology #Research A pheno...

2 years 131 Views

Rohan Khera, MD, MS, Assistant Professor at Yale School of Medicine speaks about A phenomapping-derived tool to personalize the selection of anatomical vs. functional testing in evaluating chest pain (ASSIST).

Link to Lab's ASSIST:

Link to Article:



Coronary artery disease is often identified after anatomical or clinical examination is used to evaluate persistent chest pain. Personalized testing could be possible with a more detailed knowledge of patient phenotypes that benefit from either approach.

Methods and outcomes

We generated a topological representation of the sample population based on 57 pre-randomization variables using participant-level results from 9572 patients undergoing anatomical (n = 4734) vs. functional (n = 4838) research in the PROMISE (PROspective Multicenter Imaging Study for Evaluation of Chest Pain) trial. Cox regression models presented individual patient-centered hazard ratios for significant adverse cardiovascular outcomes within each patient's 5% topological neighbourhood and showed marked heterogeneity around the phenomap [median 1.11 (10th to 90th percentile: 0.52–2.61]], suggesting distinct phenotypic neighbourhoods favoring anatomical or functional research. We used an intense gradient boosting algorithm in 80% of the PROMISE population to estimate the personalized value of anatomical vs. functional tests using 12 model-derived, regularly collected variables and generated a decision support method called ASSIST based on this vulnerability phenomap (Anatomical vs. Stress teSting decIsion Support Tool). The testing strategy recommended by ASSIST was associated with a significantly lower incidence of each study's primary endpoint in both the remaining 20% of PROMISE and an external validation set consisting of patients from SCOT-HEART (Scottish COmputed Tomography of the HEART Trial) undergoing anatomical-first vs. functional-first assessment (P = 0.0024 and P = 0.0321 for interaction, respectively) in both the remaining 20% of PROMISE and an external validation set consist

Final thoughts

We suggest a new phenomapping-based decision support method to standardize the use of structural vs. functional research in the treatment of controlled chest pain, which has been tested in two broad and geographically disparate clinical trial populations.