AI-ECG age gap predicts risk of advanced heart failure events in PKP2-related arrhythmogenic cardiomyopathy
R Garmany, M Castrichini, G Multinu, A Sularz, J M Bos, Z Attia, P Noseworthy, P Friedman, F Lopez-Jimenz, K Siontis, A Killu, M J Ackerman, M Milone, J R GiudicessiAbstract
Background
Arrhythmogenic cardiomyopathy (ACM) is a genetically heterogeneous disorder of cardiac muscle characterized by fibrofatty replacement of the myocardium and an increased propensity for ventricular arrhythmias. Pathogenic mutations in LMNA-encoded lamin A/C cause ACM as well as progeria. Cardiovascular sensescence in LMNA-ACM has been documented as an increased gap between chronological age and artificial intelligence (AI)-ECG predicted biological age. Recent murine studies have demonstrated that PKP2-ACM may also be characterized by premature cardiac aging. However, evidence of accelerated cardiac aging has not been observed clinically in patients with PKP2-ACM.
Purpose
To determine if accelerated electrocardiographic aging, as reflected by an increased AI-ECG age gap, is a clinical feature of PKP2-ACM akin to what has been observed in LMNA-ACM, but not TTN-related cardiomyopathy (TTN-CM).
Methods
The Genetic ACM/DCM registry (n=1,684) was used to identify all patients ≥18 years of age with a pathogenic/likely pathogenic variant in LMNA (n=75), PKP2 (n=117), or TTN (n=254) with a non-paced 12-lead ECG. The ECG age gap was calculated by subtracting the patients age at the time of ECG from the age predicted by the previously validated AI ECG age algorithm. Heart failure events included transplant, mechanical support, or cardiovascular death.
Results
Overall, there was no difference in AI-ECG age gap observed between PKP2-ACM and TTN-CM patients (3.8±10.7 vs. 6.1±10.5 years; p=0.151). In comparison to PKP2-ACM, LMNA-CM had a significantly higher AI-ECG age gap (3.8±10.7 vs. 16.5±13.0 years; p<0.0000001). However, using multivariable Cox regression adjusting for genotype, sex, and age at ECG, AI-ECG age gap was associated with risk of advanced heart failure events (HR: 1.05; 95% CI 1.022–1.088; p= 0.0008) in patients with PKP2-ACM.
Conclusions
Previously, our group demonstrated an increased AI-ECG age gap as a surrogate for premature cardiac aging in LMNA-ACM. Despite mouse studies suggesting that premature cardiac aging is a feature of PKP2-ACM, our study did not find a similar increase in AI-ECG age gap as seen in LMNA-ACM. Notably, an increased AI-ECG age gap was associated with an increased risk of advanced heart failure events in PKP2-ACM and may serve as a useful digital biomarker to monitor structural disease progression.