(Radiographics. 2002;22:280-281.)
© RSNA, 2002
Invited Commentary
Robyn Birdwell, MD
1 Department of Radiology, Stanford University Medical Center, Stanford, California
 |
Commentary
|
|---|
Two challenging topics related to US are addressed in the preceding article: the recognition of patterns associated with DCIS, and efforts to extend the use of breast US beyond the evaluation of palpable and mammographically detected masses. Mammography is far superior to clinical examination in the detection of DCIS, with microcalcifications being the most common mam-mographic finding. However, unlike the reported US depiction of masses associated with microcalcification clusters (1), mammographic findings of calcification alone are unreliable for predicting associated invasion. As emphasized by Moon et al
, lesions that are visible at US not only allow the use of this modality for biopsy guidance but may also allow visualization of frank masses. The use of US to target these masses may decrease underestimation of lesion microcalcifications, which has been reported at percutaneous stereotactic core biopsy performed with 11- and 14-gauge needles (2).
Although no lexicon yet exists for the US characterization of breast masses, guidelines from Stavros et al (3) are regularly quoted and practiced. At least one addition to the original descriptors is included in the study by Moon et al, namely, that of a thick echogenic rim surrounding the (often markedly) hypoechoic mass. This is a very important observation and should be critically differentiated from the thin echogenic rim described as a benign characteristic. This peripherally thickened echogenic rim of tissue (or perhaps "rind") is abnormal, albeit not necessarily tumor filled, and suggests invasive carcinoma rather than DCIS. A possible correlative finding might be that of the rim-enhancing lesions described in some cases of invasive breast carcinomas evaluated with gadolinium-enhanced breast magnetic resonance imaging. Also, at US-guided percutaneous biopsy, even before the needle is seen to enter the hypoechoic portion of a mass, resistance may be met at the level of this thick echogenic rim. Indeed, our experience has been that some of our better biopsy samples were obtained using this "tactile plus visual" approach.
Moon et al
could have highlighted a second characteristic associated with solid breast masses: posterior acoustic enhancement. This feature is very important because it is so closely allied with cysts, a common benign breast lesion, and may be misinterpreted by the breast imager as a benign characteristic. Posterior acoustic enhancement is not a benign finding in and of itself and may be seen in benign lesions (eg, fibroadenomas), high-grade malignant lesions, and DCIS.
The importance of careful assessment of margin status is well known to all practicing mammographers and is no different in the US analysis of a mass. Margin detail may be helpful in predicting the presence of DCIS associated with invasive tumors. Moon et al carefully describe the features of the masses associated with DCIS and include illustrations of microlobulated margins with ductal extension. Stavros (4) states that 70%–80% of all invasive but not otherwise specified ductal carcinomas contain DCIS and that, in most of these cases, the DCIS is located predominantly in the tumor periphery, often thereby determining the margin characteristics.
Breast US has many limitations, including those related to lesion size, lesion location, the attenuating nature of the breast tissue, differentiation of masses that are isoechoic relative to fat, and detection of microcalcifications. Specific limitations related to identifying DCIS include difficulty in (a) determining lesion size when the ducts are not distended, (b) establishing the differential diagnosis of distended hypoechoic ducts or terminal duct lobular units (including almost the entire spectrum of fibrocystic change and benign proliferative conditions [4]), and (c) demonstrating calcifications. Most breast microcalcifica-tions are too small to cause posterior acoustic shadowing; with high-frequency transducers, however, tiny echogenic foci can be identified. A phantom experiment performed with a 7.5-MHz transducer identified 100-mm-diameter glass beads when they were in a hypoechoic area (5). In a US study published in 2000, Moon et al
(1) visualized more calcifications in malignant than in benign breast masses associated with mammographically detected microcalcifications. This may be because more masses were found in cases of malignant cancer (both DCIS and invasive cancer) than in clustered microcalcifications associated with benign pathologic findings, and these isoechoic or hypoechoic masses allowed better visualization of the calcifications.
How much of a radiologist’s time and money should be dedicated to breast US? Although Moon et al
clearly state the relationship between US examination quality and the examiner’s skill, they do not say whether their US examinations were performed by physicians or technologists or how much time was required per examination. Also, they make a strong statement regarding the linear-array transducers that should be used, stating that maximum frequencies of 10–13 MHz are "required to depict DCIS in the breast." Clearly, investment of more physician time and use of more advanced equipment should improve lesion detection and characterization, and recent studies by Kolb et al (6) (in which screening US was performed in women with negative mammograms and clinical examinations) and Berg and Gilbreath (7) (in which the ipsilateral extent of disease was evaluated in women with a known index lesion) help confirm this. However, the cost of such a practice often exceeds that which can be afforded by breast imaging facilities. Hopefully, articles such as the preceding one by Moon et al
that provide excellent images and descriptions of what can be discerned with high-quality equipment will encourage vendors to focus on the production of affordable dedicated breast US machines that are capable of superior b
Top
Commentary
References
