Abstract
Background Meta-analysis can identify biological factors that moderate cardiac magnetic resonance myocardial tissue markers such as native T(1) (longitudinal magnetization relaxation time constant) and T(2) (transverse magnetization relaxation time constant) in cohorts recovering from COVID-19 infection. Methods and Results Cardiac magnetic resonance studies of patients with COVID-19 using myocardial T(1), T(2) mapping, extracellular volume, and late gadolinium enhancement were identified by database searches. Pooled effect sizes and interstudy heterogeneity (I(2)) were estimated with random effects models. Moderators of interstudy heterogeneity were analyzed by meta-regression of the percent difference of native T(1) and T(2) between COVID-19 and control groups (%ΔT(1) [percent difference of the study-level means of myocardial T(1) in patients with COVID-19 and controls] and %ΔT(2) [percent difference of the study-level means of myocardial T(2) in patients with COVID-19 and controls]), extracellular volume, and the proportion of late gadolinium enhancement. Interstudy heterogeneities of %ΔT(1) (I(2)=76%) and %ΔT(2) (I(2)=88%) were significantly lower than for native T(1) and T(2), respectively, independent of field strength, with pooled effect sizes of %ΔT(1)=1.24% (95% CI, 0.54%-1.9%) and %ΔT(2)=3.77% (95% CI, 1.79%-5.79%). %ΔT(1) was lower for studies in children (median age: 12.7 years) and athletes (median age: 21 years), compared with older adults (median age: 48 years). Duration of recovery from COVID-19, cardiac troponins, C-reactive protein, and age were significant moderators for %ΔT(1) and/or %ΔT(2). Extracellular volume, adjusted by age, was moderated by recovery duration. Age, diabetes, and hypertension were significant moderators of the proportion of late gadolinium enhancement in adults. Conclusions T(1) and T(2) are dynamic markers of cardiac involvement in COVID-19 that reflect the regression of cardiomyocyte injury and myocardial inflammation during recovery. Late gadolinium enhancement and to a lesser extent extracellular volume, are more static biomarkers moderated by preexisting risk factors linked to adverse myocardial tissue remodeling.