ST-T/T-Wave Changes Most Common ECG Abnormalities in SLE; Age, Disease Duration and Activity Are Risk Factors for Abnormalities

The most common abnormalities seen on electrocardiogram (ECG) in patients with systemic lupus erythematosus were found to be ST-segments, T-waves, and QT intervals, according to study results published in Arthritis Care & Research. Age, disease duration and activity, and hypertension were observed to be influential factors for these ECG abnormalities.

A team of researchers in China conducted a cross-sectional analysis to determine the prevalence of ECG abnormalities in patients with SLE, and the factors associated with these abnormalities factor, using machine learning approaches.

Researchers collected ECGs with abnormalities and categorized them based on the presence of tachycardias, atrioventricular or bundle branch blocks, nonspecific ST-segment changes, T-wave abnormalities, prolonged QT intervals, axis deviations, and ventricular hypertrophies. Logistic regression and 4 other machine learning approaches were compared to evaluate the ECG abnormalities associated with these factors. Machine learning and methods included random forest, support vector machine, least absolute shrinkage and selection operator (LASSO), and multivariate adaptive regression spline.

Of the 299 patients who were enrolled in the study (mean age, 37.6 years; 89.9% women), 128 patients (42.8%) had significant ECG abnormalities. Some ECGs without clinical significance included sinus bradycardia, sinus tachycardia, and sinus irregularity. Among the entire cohort, the most common findings were abnormal ventricular repolarization (39.5%), arrythmia (5.0%), and conduction abnormalities (2.7%).

Among the 128 patients with ECG abnormalities, changes in T-waves occurred in 67 (52.3%). Nonspecific ST-T changes (n=34; 26.6%) and prolonged QT intervals (n=11; 8.6%) were also common among these patients. Of note, elevations in U-waves and pathologic Q-waves were noted in 4 (3.1%) and 2 (1.6%) cases, respectively.

Of the models constructed using the 4 machine learning algorithms, LASSO had the highest area under the curve (AUC) for predicting nonspecific ST-T changes and prolonged QT intervals, while random forest model had the highest AUC for predicting T-wave changes.

Other factors that were associated with nonspecific ST-T changes, longer QT intervals, and T-wave changes included age, disease duration, disease activity (based on SLE Disease Activity Index 2000), and antinuclear antibody titers. In addition, factors that influenced ST-T changes, longer QT-intervals, and T-wave changes included hypertension, positive anti-SSA, and secondary Sjögren syndrome.

“Further longitudinal, prospective studies assessing the role of potential risk factors will help clarify the mechanism of ECG abnormalities among patients [with SLE],” the researchers noted.

Reference

Hu Z, Wu L, Lin Z, Liu X, Zhao C, Wu Z. Prevalence and associated factors of Electrocardiogram abnormalities in patients with systemic lupus erythematosus: a machine learning study. Arthritis Care Res. Published online March 30, 2021. doi:10.1002/acr.24612