Heart-type fatty acid-binding protein was measured at rest, at peak stress and two hours after stress in 2,690 patients undergoing adjudicated assessment for functionally relevant coronary artery disease, of whom 34% had the condition. Diagnostic discrimination was low (AUC 0.561), and adding the marker to clinical assessment moved the AUC only from 0.626 to 0.634. Over eight years the marker showed a graded association with death, from 7.7% all-cause mortality in the lowest quartile to 44% in the highest, with a c-index of 0.71, comparable to high-sensitivity troponins.
Why it is interesting: A marker proposed for detecting coronary disease identified who died rather than who had the disease.