| BackgroundBreast cancer, which seriously threatens female’s health, is one of the most common form of cancer in women. The global prevalence of breast cancer has been increasing in recent years. It has been the leading cause of cancer-related mortality for women. Ultrasonography has become an indispensable tool in breast imaging and has been commonly used in the diagnosis and follow-up of breast disease. Ultrasonography allows confident characterization of benign cysts and better evaluation of dense breast disease. The differentiation of benign and malignant breast lesions with conventional ultrasonography is on the basis of the shape, margin, boundary, orientation, echo pattern, posterior acoustic features, calcification, vascularity of the lesions and the changes of the surrounding tissue. Unfortunately, the stiffness of the lesions can’t be assessed by conventional ultrasonography. Recent studies have shown that ultrasound elastography (UE) is a promising technique in the differentiation of benign and malignant breast lesions in the clinical practice by evaluating the tissue elasticity (or stiffness). Nowadays, two different techniques are currently available in breast elastography: compression elastography (CE) and acoustic radiation force impulse imaging (ARFI). They possess different physical performance properties and basic principles, so both have their limitations and suitability for the clinical applications in breast disease. CE is based on the application of a compressive force to the tissue and on the measurement of the shape-deforming effect. The calculation of strain ratio (SR) is a semi-quantitative assessment of CE. CE is intuitive and easy to performed, but still with high operator dependency and obvious interobserver variability. In virtual touch tissue quantification (VTQ) of ARFI technique, an independent mechanical ultrasound beam is transmitted to generate perpendicular shear waves localized peripherally to the positioned region of interest (ROI). Tissue elasticity (or stiffness) could be assessed by measurement of shear wave velocity (SWV) within ROI. VTQ is with a high repeatability and less operator dependent. Unfortunately, there is the potential to cause artifacts in the estimated SWV, especially in the case of high stiffness contrast. Previous studies show that both of these two techniques are helpful in the differentiation of benign and malignant lesions. They may improve the diagnostic performance of ultrasonography in breast disease. However, the most appropriate cutoff point of strain or velocity and the elasticity (or stiffness) of different histopathology breast lesions is still not determined. What’s more, few studies have been conducted to compare the diagnostic performances of the two different techniques in the same study population.Objective1. To investigate the elasticity (or stiffness) of breast lesions of different histopathological types.2. To compare the diagnostic performances of CE and VTQ in the differentiation of benign and malignant breast lesions.3. To investigate the diagnostic performance of the combination of CE and VTQ in the differentiation of benign and malignant breast lesions. Materials and methods1. Patient sample49female patients with breast lesions underwent surgery in Qilu hospital of Shandong University from October2013to February2014were enrolled in this study. Totally52breast lesions were studied. Of the52lesions,18were invasive ductal carcinomas,3invasive lobular carcinomas,3ductal carcinomas in situ,1micropapillary carcinoma,10fibroadenomas,7intraductal papillomas,5breastadenosis,2cystic hyperplasia,2lipomyomas and1benign phyllodes tumor. The mean age of the patient sample were (46.7±11.8) years, ranging from22to81years old. The mean diameter of the lesions were (17.7±10.9) mm, ranging from4.0to41.0mm.2. Ultrasound scanning2.1Conventional ultrasonography and CEPatients were scanned on a Philips iU22sonography system with a LI2-5linear array probe. After conventional imaging of the lesions, CE cine loops were captured by applying slight manual pressure perpendicular to the skin under the Elasto mode. The SR ratio of the lesions to the adjacent healthy glandular tissue was calculated with the elastography quantification software.2.2VTQPatients were scanned on a Siemens S2000sonography system with a9L4linear array probe. The maximum cross section of the lesions were selected and the sampling frame was placed in the center of the lesions. VTQ were performed and the SWV (m/s) of the lesions were measured.3. Statistical analysisStatistical analyses were performed using SPSS17.0software. All measurement data were expressed as x±s. Student’s t tests, one-way ANOVA tests and χ2tests were used for comparisons. Receiver operating characteristic (ROC) curves were used to analyze the diagnostic performances of these two elastography techniques in the differentiation of benign and malignant breast lesions. Z tests were used to compare the area under the curve (AUC). A P value<0.05was considered statistically significant. Results1. The comparison of SR ratio and SWV between malignant group and benign group. SR ratio was (0.894±0.304) in benign group and (0.435±0.175) in malignant group. There was significant difference in SR ratio between the two groups (P<0.001). SWV was (1.78±0.71) m/s in benign group and (4.19±2.51) m/s in malignant group. There was significant difference in SWV between the two groups (P<0.001).2. The comparison of SR ratio and SWV between different histopathology groups. The SR ratio of invasive ductal carcinomas, fibroadenomas, intraductal papillomas and breastadenosis was (0.455±0.193),(0.8180.376),(1.020±0.260),(0.796±0.177) respectively. The SWV was (4.12±2.68) m/s,(2.13±0.65) m/s,(2.08±0.79) m/s,(1.71±0.77) m/s respectively. There were significant differences in SR ratio and SWV between the four groups (P<0.001, P=0.012).3. The diagnostic performance of conventional ultrasonography in the differentiation of benign and malignant breast lesions. The sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) and accuracy was88.3%,65.4%,71.0%,85.7%,76.9%, respectively.4. ROC analysis of the diagnostic performances of CE and VTQ in the differentiation of benign and malignant breast lesions. The AUC of SR ratio was0.919. The optimal cutoff value of SR ratio was0.499to differentiate benign and malignant breast lesions, with a sensitivity of93.6%, specificity of72.0%, PPV of78.8%, NPV of94.7%and accuracy of84.6%. The AUC of SWV was0.796. The optimal cutoff value of SWV was2.90m/s to differentiate benign and malignant breast lesions, with a sensitivity of68.0%, specificity of92.6%, PPV of89.5%, NPV of75.8%and accuracy of80.8%. Both of the two techniques could improve the specificity and accuracy of ultrasound to differentiate benign and malignant breast lesions.5. The comparison of the diagnostic performances of CE and VTQ in the differentiation of benign and malignant breast lesions. There was no significant difference in AUC of the two techniques (Z=1.580, P=0.057). They have the similar diagnostic performance to differentiate benign and malignant breast lesions. 6. The diagnostic performance of the combination of CE and VTQ in the differentiation of benign and malignant breast lesions. The sensitivity, specificity, PPV, NPV and accuracy was84.0%.88.9%.87.5%,85.7%.86.5%respectively. The combination of CE and VTQ could significantly improve the diagnostic performance of ultrasound to differentiate benign and malignant breast lesions.ConclusionsCE and VTQ are helpful in the differentiation of benign and malignant breast lesions and they have the similar diagnostic performance. Both have the potential to improve the specificity, PPV and accuracy of ultrasonography to characterize breast lesions. The combination of CE and VTQ could improve the ability of ultrasonography to detect and diagnose breast cancer. Moreover, ultrasound elastography may be a promising method to predict the histopathological types of breast lesions. This work provides useful information for the medical management of breast disease in the clinical practice. |
PDF Full Text Request