Free-hand ultrasound strain elastography in evaluation of soft tissue tumors.

Objective: The purpose of this study is to evaluate elastography in a wide spectrum of soft tissue superficial lesions by correlating the elastographic characteristics of these lesions with the elastographic score (ES) system established by Asteria.

Methods: Forty patients with different superficial lesions of the soft tissues were studied, including lipomas, schwannomas, neuromas, epidermal inclusion cysts, "in transit" melanoma metastasis, arterio-venous malformation, and giant-cell tumor. An ultrasound examination was performed combined with color-Doppler and elastographic module. The B-mode criteria were echogenicity, margins, and structural homogeneity of the lesion. The color-Doppler criterion was irregular and mainly intra-nodular vascularization. ES 1-4 was attributed, in relation with the increasing tissue stiffness, according to the classification of Asteria adapted for soft tissues. Subsequently, we added to each single B-mode and color-Doppler criterion the ES 3 and 4, thus crossing two parameters of malignancy. All the presumptive diagnoses formulated were confirmed with the clinical data or with the histopathological result.

Results: The hypoechoic appearance had the best diagnostic performance. Sensitivity was 87%, specificity 71%, positive predictive value (PPV) 80%, negative predictive value (NPV) 80%, and diagnostic accuracy 80%. There was a good correlation with the clinical and biopsy data, the irregularity of margins the worst performance, the inhomogeneity an intermediate. Color-Doppler had sensitivity 74%, specificity 82%, PPV 85%, NPV 70% and diagnostic accuracy 77.5%. Elastography had sensitivity 87%, specificity 94%, PPV 95%, NPV 84%, and diagnostic accuracy 90%. The combination hypoechoic appearance + ES3/ES4 showed sensitivity 83%, specificity 100%, PPV 100%, NPV 81%,and diagnostic accuracy of 90%. The combination of irregularity of margins + ES3/ES4 showed sensitivity 43%, specificity 100%, PPV 100%, NPV 59%, and diagnostic accuracy of 67.5%. The combination of inhomogeneity of the lesion + ES3/ES4 showed sensitivity 65%, specificity 94%, PPV 94%, NPV 68%, and diagnostic accuracy of 78%. The combination of the color-Doppler with the ES3/ES4 showed sensitivity 69.5%, specificity 100%, PPV 100%, NPV 71%, and diagnostic accuracy of 82.5%.In the combined evaluation, there was a significant increase in specificity, allowing healthy subjects to be categorized as correctly negative, with a reduction in false positives which also translates into an increase in PPV.

Conclusions: Elastography alone is not sufficient for a correct diagnostic classification and must be considered as an additional parameter in the study of soft-tissue lesions. Although there was a good agreement between B-mode malignancy criteria and ES3/ES4, there is no significant improvement in sensitivity. Ultrasound assessment, especially of superficial lesions, cannot be separated from an integrated approach that foresees the additional and routine use of the elastographic examination.