Clinical usefulness of carotid IMT progression: lessons from the improve study

BACKGROUND: The clinical usefulness of serial measurements of carotid artery intima media thickness (C-IMT) remains controversial. Uncertainties include: a) whether it is possible to accurately assess the IMT change in single individuals, b) whether the measurement of IMT change over time actually improves the physician capacity to stratify the individual cardiovascular risk and c) whether change of C-IMT over time can actually help the physician in a better assessment of the individual response to a therapy. To address these issues we have reanalyzed the dataset of the IMPROVE project, a European, multicenter, longitudinal, observational study specifically designed to assess whether the 15-month carotid IMT progression associate with vascular events (VEs) occurring during the subsequent 21.8 months of follow-up. METHODS: A total of 3703 European subjects with at least three vascular risk factors were included into the study. Progression of mean and maximum IMT of the left and right common carotids, bifurcations, internal carotid arteries and of the whole carotid tree as well as the fastest IMT-progression detected in the whole carotid tree, regardless of its location, were included in Cox proportional hazards models as predictors of vascular events and in reclassification analyses. RESULTS: All IMTs significantly progressed during 15 months (all P<0.0001), but only the Fastest-IMTprogr was significantly associated with the risk of VEs (P=0.0025; after adjustment for centre, risk factors, therapies, baseline-IMT and inter-adventitia common carotid diameter). When data were analyzed with reclassification approach, the inclusion of Fastest-IMTprogr into a model based on Framingham risk factors and baseline-IMTs significantly added to the overall risk discrimination (P=0.002). Based on signals and noises, it can be estimated that using the Fastest-IMTprogr 2.0 years of follow up are sufficient to be 95% confident, at individual level, that an IMT change is an anatomical change and not the result of measurement variability. The same figure for the best of the other variables considered (IMTmean-max-prog) was 10.4 years. The Fastest-IMTprog added significant incremental prediction of VEs over and above that of age, gender and the initial atherosclerotic profile, over and above Framingham risk factors and over and above Framingham risk factors and the initial atherosclerotic profile. In the subgroup of subjects treated with statins, the Fastest-IMTprog allows to recognize those individuals which are fast progressors despite the fact that they are treated with statins. Data on incidence of vascular events tell us that these subjects deserve a more aggressive pharmacological treatment than slow progressors. CONCLUSIONS: Admitted that the fastest-IMTprog is used, assessment of change in the individual atherosclerotic profile can be clinically useful even at individual level.