Regression of left ventricular hypertrophy with echocardiography: some lessons from the LIVE study

Objective Large trials on left ventricular hypertrophy (LVH) regression are indispensable to allow accurate evaluation of the different classes of treatments. Such a trial, the LIVE study, gave us the opportunity to examine the influence of the mode of reading of the echo tracings on results and to underline the value of two initial recordings performed at entry. Methods The LIVE study was designed to compare 1.5 mg indapamide SR with 20 mg enalapril on the regression of LVH evaluated echocardiographically. M-mode tracings of the left ventricle were performed at selection and after a 2 week placebo run-in. Recordings were read by investigators and by three experts all through the study, and finally read again by the three experts blind to treatment and sequence. Results The two initial examinations produced an estimate of the reproducibility of the measurement of the left ventricular mass (LVM) that can be proposed as an indicator of overall quality provided no patients are excluded. The standard deviation of the differences between these two estimates of LVM was 52 g. These two examinations may also help quantify regression to the mean, which was not significant for the whole group. Knowledge of the sequence of examinations overestimates the change in LVM with treatment, as shown by comparison of measurements carried out during quality control with the measurements made blind (−19 ± 52 versus −6 ± 53 g, P < 0.01). Conclusions The comparison of two initial echoes at entry allow one to define a quality criteria for such trials and to quantify a possible regression to the mean. The reading of recordings with knowledge of sequence significantly overestimates LVM variations on treatment.