Improving the yield of genetic testing in dilated cardiomyopathy using late gadolinium enhancement

Abstract

Abstract Background Genetic testing has become an important tool in the work-up of dilated cardiomyopathy (DCM), aiding in family screening and providing information about prognosis and arrhythmic risk. Nonetheless, a significant proportion of patients with DCM present negative genetic test results. The "Madrid Score" was created to evaluate the likelihood of positive genetic tests and increase their yield in DCM patients. Purpose To determine if late gadolinium enhancement (LGE) assessment can improve the Madrid Score accuracy and increase the yield of genetic tests for DCM. Methods A single-centre retrospective study of patients diagnosed with DCM who underwent cardiac magnetic resonance (CMR) and genetic test between January 2017 and October 2023. The CMR imaging included the assessment of LGE. Multivariate logistic regression was performed and the model’s change in performance was evaluated when adding LGE. Results A total of 90 patients with the diagnosis of DCM underwent CMR and genetic test during the study period. Patient’s mean age was 51.0±14.6 years, and 57.1% were male. The genetic test was positive (G+) in 42 (46.7%), and negative (G-) in 48 (53.3%) patients. A variant of uncertain significance was present in 31 (34.4%), and these were included in the G- group. Most of the known independent predictors of a G+ given by the Madrid Score were significant in our cohort: Family history of DCM (OR: 4.82; CI: 1.73-11.60; p < 0.001), absence of hypertension (OR: 0.19; CI: 0.06-0.58; p < 0.001), absence of left bundle branch block (OR: 0.22; CI: 0.05-0.98; p = 0.047), and low electrocardiogram voltage in peripheral leads (OR: 3.13; CI: 0.87-11.26; p = 0.081). When analyzing the effect of LGE in the regression model (R2 = 0.622, p < 0.001), we found that the number of LGE segments (OR: 2.54; CI: 1.10-5.90; p = 0.031), and the presence of LGE in the anterior wall (OR: 3.01; CI: 1.42-12.04; p = 0.041), were associated with a higher rate of positive genetic tests. When comparing both regression models, the incorporation of LGE increased the prediction power of the standard Madrid Score, correctly predicting 82.0% of the tests in our sample, with a probability of a G+ result ranging from 2% when all predictors were absent to 84% when ≥6 predictors, including LGE, were present. Conclusion Our study suggests that LGE assessment may enhance the Madrid Score performance and consequently improve genetic testing decisions in DCM, potentially leading to better use of the limited resources in our health care system.