my personal edition > diagnostic radiology > news


  E-Mail this DGDispatch to a colleague

DGDispatch

Elastography for Breast Cancer Detection Appears to Get Reproducible Results in Multicentre Trial: Preseneted at RSNA

By Ed Susman

CHICAGO, IL -- December 3, 2007 -- Radiologists at six different sites were able to determine with high degree of accuracy the difference between malignant and benign breast tumours using ultrasound-based elasticity imaging (EI).

"We achieved 100% sensitivity at five of the sites and 97% sensitivity at the sixth site, and had specificity that ranged from 67% to 95%," said investigator Richard Barr, MD, a private practice radiologist, Radiology Consultants Inc., Youngstown, Ohio, United States.

"Elasticity generates images based on the stiffness or hardness of a lesion and correlate to a physical examination -- but with imaging," he said in his oral presentation on November 30 here at the 93rd Scientific Assembly and Annual Meeting of the Radiological Society of North America (RSNA).

The imaging system produces side-by-side simultaneous images of the breast in both B-mode and the elastogram in real time. In addition to differences in shading compared with other tissues, which denotes stiffness or hardness, lesions appear to have changes in size.

"Malignant lesions appear to be larger on the elastogram while benign lesions appear to be smaller," Dr. Barr explained.

A year ago, he reported that in a single institution study of more than 100 specimens, elasticity imaging had a sensitivity of 100% and a specificity of 98%. In the new study, his team aimed to determine whether they could reproduce the results from the one-centre study with other sites among radiologists who were recently trained in the technique, he said.

The researchers recruited patients who were scheduled for biopsy for suspected breast lesions. The 635 women in the study underwent a usual workup plus the elastogram, which added about 1 or 2 minutes to the examination, he said.

The radiologist at each site measured the lesions and determined if it should be classified as benign or malignant before the results of the pathologist were known. They also measured the longest length of the mass on B-mode image and on EI. Ductal carcinoma in situ and hyperplasia were classified as malignancies.

There were 413 biopsy-proven lesions among the women studied, Dr. Barr said, and 222 were biopsy-proven malignancies -- mostly invasive ductal carcinomas; 10% were ductal carcinoma in situ.

The relationship of the length of the mass from the elastogram to the B-mode EI ranged from 0.2 to 1.5 among the benign lesions. The malignant lesions relationship ranged from 0.9 to 3.1. The difference was significant P <.001, Dr. Barr reported.

"Additional work is needed to incorporate higher levels of interpretation to decrease the number of false positives and negatives," he said. "We believe that with this technique we will be able to select lesions which need to be biopsied and help decrease the number of biopsies that need to be done."

Funding for this study was provided by Siemens Ultrasound.


[Presentation title: Elasticity Imaging of the Breast: Initial Results of a Multicenter Trial. Abstract SST01-07]

E-Mail this DGDispatch to a colleague

To print, use this version