Electrocardiography and cardiac magnetic resonance imaging in the detection of left ventricular hypertrophy: the impact of indexing methods

Background: Discrepancies between increased left ventricular mass (LVM) and electrocardiographic (ECG) criteria for the diagnosis of left ventricular hypertrophy (LVH) are described in the literature. Aims: This study aimed to evaluate the usefulness of ECG criteria in the diagnosis of LVH, as determined by cardiac magnetic resonance (CMR) imaging, using various LVM indexing methods. Methods: We included 53 patients who underwent CMR imaging and had electrocardiograms of appropriate quality available in their medical records. The majority of the study patients had cardiovascular diseases. We defined CMR‑LVH as increased LVM, also assessed after LVM indexing to body surface area (LVM/BSA), height1.7, height2.7, or as the percentage of predicted LVM (%pLVM). To determine ECG‑LVH, 10 different ECG-LVH criteria were used. Results: The prevalence of CMR‑LVH ranged from 11% (for %pLVM) to 72% (for LVM/BSA). At the same time, for a single criterion, the prevalence of ECG‑LVH ranged between 1.9% (for R wave amplitude in lead V5 / V6 greater than 2.6 mV, Sokolow–Lyon product, and Gubner–Ungerleider criterion) and 45.3% (for Peguero–Lo Presti criterion), showing high sensitivity, from 55.3% (95% CI, 38.3–71.4) to 100% (95% CI, 54.1–100). The sensitivity of ECG‑LVH criteria when all criteria were applied together ranged from 57.9% (95% CI, 40.8–73.7) to 100% (95% CI, 63.1–100). The best performance regarding the endpoint of CMR‑LVH diagnosis after LVM indexing was achieved by the Peguero–Lo Presti and Cornell criteria (area under the curve, 0.621–0.876; P, 0.001–0.17). Conclusions: Thediagnosis of LVH strongly depends on ECG- and CMR‑based definitions. ThePeguero–Lo Presti criterion and the Cornell criteria, which are sex‑specific, may provide the highest level of diagnostic accuracy and should be considered when screening patients with cardiovascular diseases for LVH.