| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Commentary |
1 Department of Radiology, Division of Breast Imaging, ACC 219, Massachusetts General Hospital and Harvard Medical School, 15 Parkman St., Boston, MA 02114.
Received May 15, 2003;
accepted after revision May 15, 2003.
This article is a commentary on the preceding article by Tse et al.
Contrast-enhanced MRI of the breast has been shown in a number of studies to detect breast cancers with a high degree of sensitivity ranging from 95% to 100% [2]. One of the major limitations is that false-positive enhancement occurs in benign breast lesions and causes a relatively low specificity ranging from 37% to 97%. The combination of morphologic and dynamic enhancement curves may increase the sensitivity and specificity. In addition, further characterization of lesions by MR spectroscopy may be useful in improving the specificity.
MR spectroscopy of the breast is a recently described noninvasive method that has shown promising results in differentiating benign from malignant lesions. High levels of composite choline are likely to be found in malignant lesions on proton MR spectroscopy. Low choline levels are expected in benign lesions and normal breast tissue. The sensitivity in differentiating breast cancer from benign tumors of the breast has been reported to be 70–92% with a specificity ranging from 82% to 87% [3].
In their article, Tse et al. [1] address three issues: They compare the proton MR spectroscopy results in 19 breast carcinomas from a series they reported previously with 27 cases of noncarcinomatous breast lesions. They report on the spectroscopy results of six phyllodes tumors, which is the largest series of phyllodes tumors for which in vivo proton MR spectroscopy data have been reported to date. They correlate the results with biologic functional parameters, which has not previously been described.
To date, at least six studies that use composite choline levels to differentiate benign from malignant breast lesions have been published. The data in the study by Tse et al. [1] confirm the previous results. Their current article deals with 19 breast carcinomas (18 infiltrating ductal carcinomas not otherwise specified and one medullary carcinoma) and 21 benign breast lesions (18 fibroadenomas, one fibrocystic change, one papilloma, and one hamartoma). In addition, they analyzed six phyllodes tumors, four of which were benign and two of which were of borderline malignancy. The proton MR spectroscopy findings were positive for choline-containing compounds in 17 of the 19 cases of breast cancer; of the two remaining cases, one was medullary carcinoma and one was infiltrating ductal carcinoma. None of the 21 benign breast lesions nor six phyllodes tumors showed detectable choline-containing compounds. Therefore, the sensitivity was 89% and the specificity was 100% in this series, similar to results previously reported.
The authors address the issue of whether biologic functional parameters of benign and malignant lesions may affect the findings of proton MR spectroscopy. Ki-67 antigen identifies actively proliferating cells and correlates with higher tumor grades and worse prognoses. In the carcinomas in this study by Tse et al. [1], the percentage of Ki-67 nuclear expression ranged from 0% to 45% with a mean of 17.5%, which was significantly higher than that in the benign lesions (range, 0–10%; mean, 1.6%) and in the phyllodes tumors (range, 0–3%; mean, 0.6%). In the two false-negative carcinomas, the Ki-67 proliferative index was 30% for the medullary carcinoma and 0% for the infiltrating ductal carcinoma [1].
Tumors require increased vascularity for growth. Tumor angiogenesis as measured by intratumoral microvessel density correlates with prognosis in breast cancer. In this series, the mean microvessel density count for the carcinomas ranged from 8 to 20 (mean, 13) vessels per high-power field and was significantly higher than that of the benign lesions (range, 3.8–10.8; mean, 7.4) and for the phyllodes tumors (range, 4–15; mean, 7.6). The two false-negative findings of carcinoma had a similar mean microvessel density (13.4 and 11.8 vessels per high-power field) to the mean of the carcinoma group despite one having a high and one having a negative Ki-67 proliferative index. This suggests that intratumoral microvessel density and intratumoral endothelial cell proliferation are independent of each other, as has been shown in a previous study [4]. In addition, the tumor angiogenesis may be unrelated to the proton MR spectroscopy results because both false-negative findings of carcinoma had a mean microvessel density similar to that of the carcinoma group. However, further studies with larger samples are needed because the number of subjects here is too small to draw any conclusions.
The breast cancers that show HER2/neu overexpression are usually high-grade and are associated with unfavorable prognoses. The authors found HER2/neu overexpression in seven (36%) of the 19 patients with carcinoma but not in patients with the other lesion types including the two false-negative carcinomas. The authors suggest that false-negative proton MR spectroscopy results may arise if the tumor does not overexpress the oncogene. However, as Tse et al. [1] point out, the number of patients is too small to confirm this theory, and further work needs to be done to draw any conclusions.
One of the limitations of proton MR spectroscopy is that it has a higher sensitivity for detection of larger malignant lesions than for small ones. All lesions in this series were 1.5 cm or greater in the largest linear dimension. In other series of proton MR spectroscopy that have been reported, the mean largest dimension of the malignant tumors was 2.0–4.9 cm [2]. It is relatively easy to differentiate benign from malignant conditions in large lesions using breast MRI interpretation of morphologic and dynamic information. An adjunct method for differentiating malignancy would be helpful in smaller lesions, particularly those less than 5 mm.
The sensitivity of proton MR spectroscopy of the breast may be limited by technical factors. Most of the studies that have been performed to date have used 1.5-T magnets. Improvements in the signal-to-noise ratio using higher field strength magnets, such as 3–4 T, may increase the detection of composite choline compounds and therefore may increase the sensitivity in detection of smaller breast cancers. An additional technical factor that may cause incorrect sampling of the lesions and lead to false-negative results is patient motion or MRI misrepresentation of the unenhanced and enhanced images. Image misrepresentation would be particularly troublesome for smaller lesions and could be alleviated by decreasing scanning times so that patients would be less likely to move—or, ideally—by using methods to register the images.
The data that Tse et al. [1] present regarding phyllodes tumors suggest that proton MR spectroscopy is not useful in differentiating phyllodes tumors anywhere along the spectrum from benign to borderline to malignant. Phyllodes tumor should therefore be considered in the differential diagnosis of a lesion that has a negative composite choline signal and is therefore characterized as benign on MR spectroscopy.
Currently, proton MR spectroscopy may be useful as an adjunct to contrast-enhanced breast MRI in lesions greater than 1.5 cm. Further research is needed to increase the sensitivity and specificity of proton MR spectroscopy, particularly in differentiating small (< 5 mm) benign from malignant lesions. Until technical innovations or further correlative data with biologic markers can accomplish this, another potential use for proton MR spectroscopy is in monitoring locally advanced breast cancer response to treatment with neoadjuvant chemotherapy [5]. A comparison of the composite choline level before treatment with subsequent choline reduction or disappearance after treatment may be a useful noninvasive indicator of patient response to treatment.
 Top References |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
