The Role of Sonoelastography in the Diagnosis of Supraspinatus Tendinosis

Abstract

Purpose: The aim of the study was to evaluate the sensitivity and specificity of compressive elastosonography (“Strain Elastography”), integrated with the B-mode ultrasound examination, in supraspinatus tendinosis, using magnetic resonance imaging as the gold standard for diagnosis. We also verified the intra-observer reproducibility of this elastosonographic method. Materials and Methods: The supraspinatus tendons of 30 patients with shoulder pain and positive magnetic resonance imaging for tendinosis were examined with B-Mode ultrasound and compressive elastosonography in the absence of full-thickness tendon ruptures. At the same time, 30 control individuals, asymptomatic for shoulder pain and with a negative history of previous diseases affecting the rotator cuff, were examined with B-Mode ultrasound and compressive elastosonography. The MRI exam was used as a gold standard reference exam for the diagnosis of tendinosis. At the “Strain” elastosonography the images were evaluated qualitatively with the color elastogram and quantitatively with the percentages of deformation of the supraspinatus tendon (Strain T), of the deltoid muscle used as reference tissue (Strain R) and of the SR/ST ratio (Ratio). Results and Conclusions: The comparison between the group of individuals with tendinosis and the control group shows a statistically significant difference (p<0.05) for the Ratio values (SR/ST), respectively: 4.7 ± 0.14 vs 5.3 ± 0.19. The different Ratio values indicate that the supraspinatus tendon, when compared with the deltoid muscle, is softer in patients with tendinosis than in the control group. As indicated by the statistical analysis of the ROC curve, the Ratio value of 4.8 represents the best sensitivity and specificity data (70% and 61% respectively). Elastosonography can therefore provide added value, compared to the conventional ultrasound examination only, to confirm the diagnosis of supraspinatus tendinosis. The analysis showed a low variance of results, due to its excellent intra-observer reproducibility (intraclass correlation coefficient = 0.86). Future studies could aim to describe any changes in Stain T, Strain R and Ratio values in a cohort of individuals stratified by gender, age and BMI.