Researchers developed a propensity model to assess the risk of developing either wild-type or hereditary transthyretin amyloidosis (ATTR)-associated cardiomyopathy, according to a study published in the Permanente Journal.

ATTR is underrecognized and often misdiagnosed; ATTR-associated cardiomyopathy is likely even more so. Epidemiological studies suggest that ATTR-associated cardiomyopathy is becoming more common as life expectancy increases. 

Like regular cardiac disease, the patient prognosis improves if ATTR-associated cardiomyopathy is diagnosed and treated early. A number of imaging tests excel at detecting ATTR-associated cardiomyopathy among vulnerable individuals. The challenge is to arrive at this point; symptoms tend to be nonspecific and awareness of this condition is low. 

A machine learning model previously developed successfully predicted patients with wild-type ATTR-associated cardiomyopathy with 87% sensitivity and 88% positive predictive value. The authors of the present study chose to build upon this work by incorporating additional data to develop a model capable of predicting the probability of ATTR-associated cardiomyopathy among real-world patients with heart failure. 

Read more about hereditary ATTR diagnosis 

The research team recruited patients with a confirmed diagnosis of wild-type/hereditary ATTR-associated cardiomyopathy (observation group) and patients with heart failure but without ATTR-associated cardiomyopathy (control group). A propensity model was created based on a wide range of demographic and medical data. The researchers identified samples of 50 control patients with the highest and lowest propensity scores and fed the information into a data set. 

A cardiologist with prior experience diagnosing and treating ATTR-associated cardiomyopathy was blinded to the disease status of the participants and was asked to independently adjudicate whether further diagnostic tests were warranted for each participant. 

The study included 7651 participants (31 in the observation group and 7620 in the control group). The research team successfully identified a number of risk factors for ATTR-associated cardiomyopathy and developed a propensity score that was able to distinguish between control patients with high and low probabilities of developing this disorder. In addition, researchers found that the model had an excellent record of distinguishing patients who warranted further workup for this condition. 

“Future studies should expand upon this methodology to incorporate different health systems and larger samples of patients with confirmed ATTR-associated cardiomyopathy to develop more refined predictive tools for targeted ATTR-associated cardiomyopathy testing,” the authors concluded. 


Suh AA, Shaw PB, Jeong MY, Olson KL, Delate T. Transthyretin amyloid cardiomyopathy risk evaluation in a cohort of patients with heart failurePerm J. Published online March 27, 2023. doi:10.7812/TPP/22.135