Expanded Genetic Testing Leads to Additional Cardiomyopathy, Arrhythmia Diagnoses

More comprehensive genetic testing of patients suspected of having cardiomyopathy or arrhythmia results in additional diagnoses that would have been missed through disease-specific testing, a new study has found.

There has been tension between the possibility of increased diagnostic yield from broader testing and the greater likelihood of receiving an uncertain result, the researchers, from Northwestern University, noted.

To better understand the results of expanded testing, they conducted a cohort study involving broad, combined cardiomyopathy and arrhythmia genetic testing through a no-charge program sponsored by Invitae. In the study, nearly 5,000 patients with suspected genetic cardiomyopathy or arrhythmia underwent genetic testing for a range of disease-related genes. The researchers then compared the results of that broad testing with what would have been captured if narrower, disease-specific testing had instead been sought, as they reported  this week in JAMA Cardiology.

"Notably, this study found a 10.9 percent gain in genetic diagnoses that would have been missed if testing had been limited to genes associated with a single cardiomyopathy or arrhythmia subtype," senior author Elizabeth McNally, a researcher at Northwestern, and her colleagues wrote in their paper.

During the 12-month study period, clinicians referred 4,782 unrelated patients for testing, which encompassed up to 150 genes. Patients could be included in the study for any level of clinician suspicion of genetic cardiomyopathy or arrhythmia, based on either personal or family history.

Overall, 19.9 percent of the cohort received a molecular diagnosis. Patients for whom there was increased suspicion of genetic cardiomyopathy or arrhythmia were more likely to have a positive result than patients with lower suspicion levels. However, the researchers noted that 11 patients with a low suspicion level also received a positive result.

Diagnostic yield varied across conditions and by patient age. The highest yield, 40.4 percent, was observed among patients between the ages of 19 and 39 who were suspected of having hypertrophic cardiomyopathy. At the same time, the diagnostic yield for patients thought to have long QT syndrome was 26.1 percent.

This combined testing approach, the researchers noted, identified clinically relevant variants for 1 in 5 patients. In a related commentary in JAMA Cardiology, researchers led by the University of Sydney's Christopher Semsarian noted that this rate could be low, as clinicians may have referred patients with low clinical suspicion because a no-cost, sponsored genetic testing program was used in the study.

For about two-thirds of those with positive results, or 1 in 8 of all patients tested, these findings would have supported more intense clinical monitoring or a change in medication or medical device use. Also, for 6 percent of patients, the positive results were in a gene with a targeted molecular therapy.

The researchers additionally compared their broader testing approach to what disease-specific testing would have revealed. More than 10 percent of positive results would have been missed if testing had been based on clinician-provided diagnostic indications. They noted that of the 689 patients who had a positive result and a specific indication, 3.9 percent had an arrhythmia diagnosis but received a positive test result for a cardiomyopathy, or the reverse.

The findings suggested to the researchers that comprehensive genetic testing provides diagnostic and prognostic information that can change clinical management. They added that the benefits of identifying additional cases outweighs the risks posed by detecting variants of uncertain significance, which were found in 51.2 percent of their cohort.

"An important message is to have a management plan for handling findings of variants of uncertain significance, from pretest genetic counseling to expertise in assessing variant pathogenicity, and to have potential links to functional genomics platforms and machine learning algorithms available to resolve as many of these uncertain findings as possible," Semsarian and colleagues added in their commentary.