Longitudinal ultrasound evaluation of carotid atherosclerosis in one, two and three dimensions.

The aim of this study was to compare the sensitivity of three ultrasound phenotypes of carotid atherosclerosis in a longitudinal study of patients with diabetic nephropathy. B-mode ultrasound-derived intima-media thickness (IMT), total plaque area (TPA) as well as three-dimensional ultrasound (3DUS) vessel wall volume (VWV) of the common carotid artery (CCA) (VWV(CCA)) and internal carotid artery (ICA). (VWV(CCA+ICA)) were all evaluated in subjects enrolled in a randomized placebo-controlled double blind study of vitamin B therapy. Of 106 subjects randomized, 77 subjects were scanned at baseline and 2.3 +/- 1 y later (range: 0.5 to 4.5 y); of these subjects, 71 had images of sufficient quality for complete analysis of all three measurements. Subjects were analyzed according to the two treatment groups (A and B) and the analysis was performed blinded to treatment group description to prevent any potential for bias in future analyses. There were differences in sensitivity to longitudinal changes observed in all the ultrasound measurements. Specifically, there was no difference in IMT change between treatment groups (0.02 +/- 0.07 mm/y and 0.02 +/- 0.1 mm/y p = 0.15, group A and B, respectively, rates not different from zero [p > 0.05]) or TPA rate between treatment groups (0.09 +/- 0.2 cm(2)/y, significantly different from 0, p = 0.013 and -0.02 +/- 0.3 cm(2)/y in group A and B, respectively). However, the VWV(CCA+ICA) rate of change was significantly greater than 0 for group B (53 +/- 110 mm(3)/y) (p = 0.008), which was significantly (p = 0.034) higher than the rate of change of VWV(CCA+ICA) (nonsignificant, p = 0.6) for group A (-12 +/- 137 mm(3)/y). The relationship between DeltaVWV and DeltaIMT was significant, such that in group A, DeltaVWV(CCA) was positively associated with DeltaIMT (r = 0.44, p < 0.05), and in group B, DeltaVWV(CCA) was negatively correlated with DeltaIMT (r = -0.44, p < 0.01). These results suggest that 3DUS-derived VWV provides necessary and sufficient sensitivity and specificity to measure longitudinal changes in small numbers of carotid atherosclerosis patients at risk of disease progression and over short periods of time.