Challenges in the Diagnosis and Management of Hypertrophic Cardiomyopathy and the Promise of Artificial Intelligence

Hypertrophic cardiomyopathy (HCM) is the most common genetic cardiomyopathy marked by left ventricular (LV) hypertrophy (LVH) not caused by secondary factors. It often manifests with asymmetric wall thickening, particularly of the interventricular septum, and can lead to LV outflow tract obstruction in the majority of patients. Key pathological features include myocyte hypertrophy and disarray, as well as interstitial fibrosis, often resulting in diastolic dysfunction with a preserved or elevated ejection fraction. Patients with HCM have a highly variable clinical course. HCM is a leading cause of sudden cardiac death (SCD) in young individuals. Atrial fibrillation is also common and poorly tolerated in patients with HCM 1, 2, 3

There are unique challenges in diagnosing HCM. In a retrospective observational study conducted from 2009 to 2019 involving 3,888 patients with HCM, ∼60% of participants experienced a delay of ≤2 years before receiving a diagnosis. Patients were given an average of 4.0 misdiagnoses before the correct diagnosis of HCM was made 4 HCM exhibits incomplete penetrance and variable expressivity, resulting in diverse clinical presentations that obscure the diagnosis further 5

Guidelines published by multiple cardiac imaging societies have aimed to improve the appropriate understanding of the utilization of multimodality imaging in the diagnosis and longitudinal care of patients with HCM 6 Electrocardiography (ECG) and transthoracic echocardiography (TTE) are the first-line diagnostic modalities utilized to evaluate patients with suspected HCM. Advanced imaging techniques aid in distinguishing HCM from other conditions that mimic its presentation, such as hypertensive heart disease and infiltrative or storage diseases. Cardiac Magnetic Resonance (CMR) is emphasized for its ability to characterize myocardial tissue, particularly in identifying fibrosis via late gadolinium enhancement (LGE) and parametric mapping, with prognostic implications and helping distinguish between HCM and its phenocopies. Advanced imaging techniques, in conjunction with clinical parameters, can aid with SCD risk stratification and inform decisions regarding implantable cardioverter-defibrillator (ICD) therapy 6 According to 2024 American College of Cardiology (ACC)/American Heart Association (AHA) guidelines, CMR imaging can be helpful for SCD risk-stratification, or when the decision to proceed with ICD placement remains uncertain after comprehensive clinical evaluation. CMR provides additional risk stratification by evaluating LV wall thickness, ejection fraction, the presence of an LV apical aneurysm, and the extent of myocardial fibrosis as detected by LGE 2

Screening and early detection of HCM are crucial aspects of care in the community, as HCM is the most common inherited cardiac disease and can be associated with increased cardiovascular mortality and morbidity. Since HCM is a hereditary trait, an accurate diagnosis not only benefits the affected individuals but may also allow for genetic or clinical screening of asymptomatic family members. Over the past decade, there has been a significant increase in studies evaluating the potential of AI methods for routine diagnostic tasks in cardiovascular medicine. AI is increasingly being integrated into the diagnosis and management of HCM. In the HCM field, the majority of research has focused on the use of AI-ECG algorithms for screening and diagnosis of HCM, therapeutic monitoring, and prediction of genotype, myocardial fibrosis, and clinical complications in HCM. Research studies have also evaluated the potential role of AI in biomarker-based prediction of HCM. Additionally, AI algorithms have been used to aid in the analysis and interpretation of echocardiographic and CMR images.

This comprehensive review critically examines current data on the clinical implications of AI-driven techniques in screening, diagnosis, prognosis, monitoring, and management of patients with HCM (Central Illustration).

Comments (0)

No login
gif