Objectives: Cardiac MRI (CMR) measured myocardial T2 and T1 values provide assessment of myocardial tissue abnormalities. There has been a growing interest in using these values to detect myocardial changes over time in various diseases. However, the inherent variability in these measurements is not known. We sought to quantify the expected temporal variability in T2 and T1 pre-contrast measurements over time in healthy volunteers.
Methods: Adult healthy volunteers were prospectively recruited to undergo 3 CMR studies on 1.5T (Avanto fit, Siemens Healthcare) at 3-month intervals. We obtained native T1 mapping using a 5(3)(3) MOLLI technique and T2 mapping using a T2-prepared single-shot SSFP technique at the basal, mid, and apical short axis locations, and SSFP cines for LV function analysis. Measurements of T1/T2/and LVEF were performed by single observer blinded to patient identity and imaging time-point using commercially available algorithms. Temporal variability was calculated as the coefficient of variance (CoV) and as the Standard Error of the Measurement (SEM) using one-way ANOVA. This study was funded by the Canadian Institutes of Health and Research.
Results: Thirty healthy volunteers (60% female, age: 45.5 ± 13.7 years, range: 23 – 80 years) were studied with 90.4 ± 10.9 days between each examination. The SEM for mean T2 was 1.4msec (95% CI 1.2–1.6ms), for T1 was 12.0msec (95% CI 10.4–14.0ms). There were no statistically significant changes in LVEF over time. Slice specific variability, the corresponding COV, and the mean values at each of the 3 time points are summarized in the table1. The apical slices had higher temporal variability than the basal and mid myocardial segments both for T1 and T2 values.
Conclusion: Temporal variability of T1 and T2 values in healthy volunteers was low. Temporal changes in native T1 values ≤14.0 ms and T2≤1.6 ms likely represent expected changes due to measurement variability and are unlikely to represent pathological changes.