| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Department of Anatomical and Cellular Pathology, Prince of Wales Hospital,
Chinese University of Hong Kong, Ngan Shing St., Shatin, Hong Kong, SAR
China.
2 Department of Radiology and Organ Imaging, Prince of Wales Hospital, Chinese
University of Hong Kong, Shatin, Hong Kong SAR, China.
3 Present address: Department of Radiology, Kulliyyah of Medicine, International
Islamic University Malaysia, PO Box 141, 25710 Kuantan, Pahang DM,
Malaysia.
4 Department of Surgery, Prince of Wales Hospital, Shatin, Hong Kong SAR,
China.
5 Department of Clinical Oncology, Prince of Wales Hospital, Shatin, Hong Kong
SAR, China.
Received December 9, 2002;
accepted after revision April 1, 2003.
Address correspondence to G. M. K. Tse
(garytse{at}cuhk.edu.hk).
Abstract
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
SUBJECTS AND METHODS. We prospectively evaluated 19 breast carcinomas, 21 benign breast lesions (including 18 fibroadenomas, one fibrocystic change, one hamartoma, and one papilloma), and six phyllodes tumors (four benign, two of borderline malignancy) using proton MR spectroscopy. All lesions were larger than 1.5 cm. Tumor Ki-67 proliferative index, tumor angiogenesis, and HER2/neu oncogene overexpression were evaluated by immunohistochemistry of the histologic material.
RESULTS. Spectroscopy findings were positive in 17 (89%) of 19 carcinomas but negative for all benign lesions and phyllodes tumors (sensitivity, 89%; specificity, 100%). Significantly higher levels were obtained for all biologic parameters in carcinomas compared with benign lesions and phyllodes tumors. HER2/neu oncogene overexpression was present in 37% of carcinomas but not in other lesions. The two false-negative findings of breast carcinoma showed similar Ki-67 proliferative index and microvessel density compared with the remaining carcinomas, but both cases were negative for HER2/neu overexpression.
CONCLUSION. Proton MR spectroscopy is useful in the in vivo characterization of breast masses when the lesion exceeds 1.5 cm in maximal dimension. Spectroscopy is unable to reveal benign breast lesions and phyllodes tumors of benign and borderline malignancy. We suggest that a false-negative spectroscopic result may be related to an absence of HER2/neu overexpression in carcinoma of the breast.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Although in vivo 1H MR spectroscopy shows promising results in the differentiation of benign and malignant breast lesions, a review of published results so far shows that the technique has a false-negative rate of approximately 4–18% and a false-positive rate of 14–18% [5, 6, 8, 10]. It would be of interest to evaluate the underlying biologic functional parameters of benign and malignant lesions that might affect the outcome of 1H MR spectroscopy.
In this study, we prospectively analyzed 17 benign breast lesions and six phyllodes tumors and compared the 1H MR spectroscopy results with a previous series of 19 breast carcinomas and four benign breast lesions [8]. All samples were categorized as carcinoma, benign breast lesion, or phyllodes tumor. We further analyzed their differences in biologic parameters including proliferative index (Ki-67 proliferative marker expression), tumor angiogenesis (as measured by microvessel density), and overexpression of HER2/neu oncogene for each group of lesions.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
MRI was performed on a 1.5-T whole-body imaging system (Gyroscan ACS-NT, Philips, Best, The Netherlands). A standard receive-only double breast coil covering both breasts was used for both MRI and spectroscopy. Each patient was examined in the prone position with both breasts suspended in the breast coil. The body coil was used as the transmitter to generate a homogeneous radiofrequency induction field (i.e., B1) for all MRIs.
MRI was performed in the axial and sagittal planes. The axial scanning was performed by using a T1-weighted spin-echo sequence (TR/TE, 450/12; slice thickness, 4 mm with no intersection gap; field of view, 350 mm; matrix size, 256 x 256; 2 signals acquired; imaging time, 5 min) with spectral presaturation with inversion recovery (SPIR) for fat saturation. Thirty axial images covering the whole breasts were obtained before contrast administration. After the patient was given a bolus IV injection of gadopentate dimeglumine (Magnevist, Schering, Berlin, Germany; 0.2 mmol/kg of body weight), 30 enhanced axial images were acquired in the same manner without change in the patient's position. Image subtraction was then performed to optimally show enhancing lesions on the subtracted images. Enhanced sagittal images were obtained in the affected breasts by using a T2-weighted turbo spin-echo imaging sequence with SPIR (2,000/100; slice thickness, 4 mm with 10% intersection gap; field of view, 350 mm; matrix size, 256 x 256; 3 signals acquired; imaging time, 4 min) to ascertain the correct positioning of the volume of interest and to rule out abscess or cyst. The largest dimensions of the breast lesions were measured on the axial subtraction images by two radiologists.
Three water-suppressed spectra for each volume of interest were acquired using the point-resolved spectroscopic sequence (2,000/38; 2,000/135; 2,000/270) 15–20 min after the administration of contrast agents. The volume of interest was carefully positioned within the enhancing breast lesion as seen on the subtraction image. Automated parameter optimization consisted of frequency and receiver gain adjustments, shimming, and gradient tuning. Water suppression was applied using a selective inversion recovery method, and signal acquisition was performed at the zero crossing of the water signal. Data were acquired at a spectral bandwidth of 1,000 Hz, and 64 signals were averaged for each water-suppressed spectrum to achieve an adequate signal-to-noise ratio (SNR).
All 1H MR spectra were analyzed by one physicist who was unaware of the histopathologic results. A time domain fitting routine variable projection method [12] that was incorporated in a software package (Magnetic Resonance User Interface) developed by A. van den Boogaart, Katholieke Universiteir Leuven, Belgium [13]) was used for the data analysis. Residual water was first removed using the Hankel-Lanczos singular value decomposition [14] method to obtain a reduced free-induction decay that completely lacked any water signal. The resonance frequency and line width of choline were selected manually; these values were used as the "prior knowledge" input in the fitting process. The criteria used to determine whether choline was present in a lesion were that the peak at 3.2 ppm should be clearly identifiable and should have an SNR greater than two and that these conditions should be met in at least two of the three spectra acquired at different TEs.
For all 46 patients, the histologic material (seven percutaneous needle core biopsies, 15 mastectomies, and 24 excisions) and the subsequent immunostaining materials were reviewed by one pathologist who was unaware of the 1H MR spectroscopic results. All the specimens were fixed in formalin and routinely processed with the 4-µm slides that were stained with H and E. For each patient, a most representative section was prepared for immunostaining using the avidin biotin method. The antibodies used were Ki-67 (Immunotech, Marseille, France; 1:10), CD31 (Dako, Glostrup, Denmark; 1:50), HER2/neu (Dako; 1:150). The Ki-67 staining was used to assess proliferative activity, and the percentage of cells showing positive nuclear staining was evaluated. The CD31 (endothelial marker) was used to assess microvessel density: At a scanning power of 40 times magnification, an area of high vessel count was identified and the number of positively staining discreet entities was counted as a microvessel irrespective of whether a lumen was identified, as described in Weidner et al. (hot-spot method) [15]. Five high-power fields were assessed, and an average was taken as the mean microvessel density. To assay HER2/neu expression, the intensity of the membrane staining was graded as mild, moderate, or strong staining, and the percentage of cells showing staining was assessed. The assay was considered positive if moderate to strong complete membrane staining was present in at least 30% of the cells.
Statistical analysis was carried out to compare the results of the 1H MR spectroscopy, histologic diagnosis, the expression of Ki-67, mean microvessel density, and HER2/neu oncogene overexpression, using the Student's t test. The statistical significance threshold was set at a p value of 0.05.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
|
|
The 1H MR spectroscopy result was positive for choline-containing compounds in 17 of 19 patients with breast carcinoma. The two remaining patients had a medullary carcinoma and an infiltrating ductal carcinoma NOS type. All 21 benign breast lesions and the six phyllodes tumors were negative for choline-containing compounds on 1H MR spectroscopy (Fig. 3). For Ki-67 nuclear expression, the percentage in the carcinomas ranged from 0% to 45% (mean, 17.5%); in the benign breast lesions, it ranged from 0% to 10% (mean, 1.6%), and in the phyllodes tumors it ranged from 0% to 3% (mean, 0.6%). The mean microvessel density count for carcinomas ranged from eight to 20 vessels (mean, 13 vessels) per high-power field; for the benign group, the microvessel density count ranged from 3.8 to 10.8 vessels (mean, 7.4 vessels) per high-power field; for the phyllodes tumors, the microvessel density count ranged from four to 15 vessels (mean, 7.6 vessels) per high-power field. Twelve of the 19 carcinomas were negative and seven were positive for HER2/neu expression (Figs. 2B, 2C). All the benign breast lesions and phyllodes tumors had negative results for HER2/neu expression (Figs. 1B and 1C). The carcinomas had a higher proportion of Ki-67 expression, HER2/neu overexpression, and high mean microvessel density. The difference for all parameters was statistically significant (p < 0.001).
|
|
|
|
|
For the two false-negative carcinomas (medullary carcinoma and infiltrating ductal carcinoma), the Ki-67 proliferative indexes were 30% and 0%, respectively, and the microvessel densities were 13.4 and 11.8 vessels per high-powered field, respectively. All values were similar to the mean for the carcinoma group. Both cases were negative for overexpression of the HER2/neu oncogene. This observation was different from the generally high expression of this oncogene in the malignant lesions.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Detection of choline-containing compounds in benign breast lesions has also been reported in three series, and the reported false-positive rate was low, ranging from 14% to 18% [4, 5, 10]. Two reports in the literature also described the total absence of detectable choline-containing compounds in vivo in healthy, clinically normal breasts [5, 6]. In our series, we examined 27 noncarcinomatous breast lesions in vivo using 1H MR spectroscopy, and we found that none of these lesions showed any detectable choline-containing compounds. The group of 27 noncarcinomatous lesions—the largest series reported, to our knowledge—was heterogeneous and consisted of 18 fibroadenomas, four benign phyllodes tumors, two phyllodes tumors of borderline malignancy, one case of fibrocystic changes of breast, one hamartoma, and one papilloma. We found a relatively high proportion of phyllodes tumors because of the practice in our institution of routine fine-needle aspiration of breast masses at initial consultation, which resulted in early diagnosis of fibroadenomas, some of which were not excised. Other fibroadenomas smaller than 1.5 cm were excluded from this study because they were judged too small for 1H MR spectroscopy to detect.
The underlying mechanism for the elevated level of choline-containing compounds in malignant breast lesions has been the subject of intense investigation. Many in vitro studies have been conducted on perchloric acid extract of tumors using high-field-strength magnets to perform 1H MR spectroscopy for tissue characterization. The choline signal found in the spectra of in vivo studies resonating at 3.2 ppm is the summation of three choline-containing metabolites: phosphocholine, glycerophosphocholine, and free choline, the chemical shifts of which are too close to be adequately resolved in vivo at 1.5-T magnetic field strength. In vitro studies, with much higher field strength magnets, however, have shown that phosphocholine is the principal component that is elevated in breast cancer [16–20] and breast cancer cell line MCF7 [21].
The relationship between the levels of these metabolites and the known prognostic markers of breast carcinoma has remained elusive. It has been reported that the degree of elevation of choline-containing compounds is related to the grade of the tumor, with higher levels being found in higher grade lesions [22]. In a mouse mammary carcinoma model, the level of phosphocholine correlated with the proliferative index of the tumor cells [23], including BrdU labeling and S-phase fraction. In another study, tumor cell mitotic count and Ki-67 were shown not to be related to the level of choline compounds, even though higher levels were observed in higher grade lesions [24]. Another group of researchers found that phosphocholine levels were different in two breast cancer cell lines with similar doubling times [20].
HER2/neu is an oncogene that is commonly overexpressed in breast cancer and is associated with poor prognosis because it contributes to tumor progression. In one study, forced overexpression of HER2/neu in a tumor cell line resulted in a dramatic increase in levels of choline compounds [20].
Because of the poor resolution of the spectra, it is not possible in the in vivo setting to make quantitative measurements of the choline levels or the resolution of the choline-containing compounds into phosphocholine, glycerophosphocholine, and free choline in comparison with the degree or grade of malignancy of the breast lesion. The following factors may play a role in influencing the result of the in vivo study: a benign lesion with high proliferative activity may give a false-positive result, whereas an HER2/neu-negative, low-grade, and low–proliferative rate carcinoma may give a false-negative result. The relationship of high proliferative activity and presence of choline-containing compounds in the breasts has been well illustrated in an in vivo study that revealed positive 1H MR spectroscopy results in most patients (5/7, 71%) with lactating breasts [5].
Ki-67 is an antibody that recognizes an antigen expressed by cells in G1, S, G2, and M (but not G0) phases of the cell cycle, thus identifying cells that are actively proliferating. In breast cancer, the higher Ki-67 level correlates with higher mitotic index, tumor grade, and tumor necrosis [25]. It has also been shown that a high Ki-67 level has a predictive value for cancer recurrence [26]. In our series, carcinomas had a significantly higher level (17.5%) of Ki-67 index compared with that in the benign breast lesions (1.6%) and the phyllodes tumors (0.6%). For the two false-negative carcinomas, the level of Ki-67 expression was negative in one and high (30%) in the other.
Tumor angiogenesis, as measured by microvessel density, has been established as a prognostic factor in breast cancer, and it has been shown to be much higher in intratumoral areas than in the adjacent benign breast tissue. Furthermore, microvessel density did not correlate with tumor cell Ki-67 expression or the mitotic index, suggesting that angiogenesis is independent of Ki-67 expression as a prognostic factor [27]. Other researchers have shown that angiogenesis also possesses predictive power for long-term survival and tumor recurrence [28]. In our series, the mean microvessel density was significantly different between the two groups, with the cancer group showing higher mean microvessel density (12.5 vessels per high-powered field) compared with the benign and the phyllodes tumor groups (7.4 and 7.6 vessels per high-powered field, respectively). Interestingly, the two false-negative carcinomas seen on 1H MR spectroscopy showed a mean microvessel density that was similar to the mean value for the carcinoma group, suggesting that the microvessel density value was potentially unrelated to the MR spectroscopic finding.
HER2/neu oncogene overexpression was found in 37% (7/19) of patients with breast carcinomas. The remaining 12 patients with carcinomas (including the two carcinomas with false-negative results on 1H MR spectroscopy), all 21 patients with benign breast lesions, and the six patients with phyllodes tumors had negative findings for this oncogene overexpression. Although the number of cancerous lesions with false-negative spectroscopy results was small, we considered this a potentially significant observation because it suggests a possible relationship between 1H MR spectroscopic results and HER2/neu oncogene overexpression. In carcinogenesis, the amplification of the HER2/neu oncogene, which encodes for epidermal growth factor receptor, results in an abnormal persistent activation of tyrosine kinase activity and thus triggers mitosis even in the absence of the growth factor [29]. Our observation suggests that overexpression of HER2/neu oncogene may be one of the factors contributing to positivity for choline-containing compounds at 1H MR spectroscopy. False-negative in vivo 1H MR spectroscopic results may occur if the tumor does not overexpress the oncogene. This hypothesis would be in agreement with the observation by Aboagye and Bhujwalla [20], who found that forced overexpression of HER2/neu in a tumor cell line resulted in a dramatic increase in levels of choline-containing compounds. However, the number of such false-negative findings observed in this series is too small to allow any inference to be drawn. Because the sensitivity of 1H MR spectroscopy in detecting breast carcinoma is high, further work with much larger series may be needed to elucidate the factors contributing to the accuracy and sensitivity of in vivo 1H MR spectroscopy, particularly with respect to HER2/neu overexpression.
We have reported on the in vivo 1H MR spectroscopic findings of six phyllodes tumors. To our knowledge, no previous report on the in vivo 1H MR spectroscopic findings in phyllodes tumors has appeared in the literature. Two phyllodes tumors have been reported in a large series of in vitro 1H MR spectroscopy, but they were excluded from the final statistical analysis, and the histologic nature (whether benign, borderline malignant, or malignant) was not documented [1].
Phyllodes tumors are uncommon fibroepithelial tumors and are classified as benign, borderline malignant, or frankly malignant on the basis of a combination of morphologic criteria including stromal cellularity, cellular pleomorphism and mitotic rate, margin morphology (whether rounded or infiltrative), and stromal overgrowth. Malignant phyllodes tumors tend to possess high stromal cellularity and stromal cell pleomorphism, high mitotic count, an infiltrative margin, and stromal overgrowth. Benign phyllodes tumors tend to show the reverse. A borderline malignant category is defined in those lesions having some but not all malignant histologic parameters. In general, benign phyllodes tumors can recur locally but do not metastasize, but malignant phyllodes tumors can metastasize, and borderline malignant phyllodes tumors show intermediate behavior [30].
Many studies have shown that the degree of malignancy in phyllodes tumors forms a continuum. Progressively increasing expression of p53 tumor suppressor gene protein product, Ki-67 proliferative marker, and microvessel density are associated with increasing degrees of malignancy from benign to borderline to malignant [31–33]. In our series, all six patients with phyllodes tumors (four benign and two with borderline malignancy) were negative for choline compounds on 1H MR spectroscopy. Given that the stromal cell component is considered the principal neoplastic element in phyllodes tumors, we suggest that in vivo 1H MR spectroscopy may be of limited value in the detection and characterization of stromal lesions or the differentiation of phyllodes tumors from fibroadenomas.
We have shown that in vivo 1H MR spectroscopy is useful as a noninvasive adjunct in the radiologic workup of breast lesions larger than 1.5 cm. In this series, the sensitivity and specificity were 89% and 100%, respectively. Our preliminary results suggest that positivity of 1H MR spectroscopy may be related to HER2/neu oncogene overexpression, but not to the Ki-67 proliferative index and tumor angiogenesis, and experience with more cases is needed. We also showed that phyllodes tumors, including both benign and borderline malignant varieties, are negative on 1H MR spectroscopy, suggesting that this noninvasive technique may be useful in the detection of epithelial (carcinoma) but not nonepithelial (stromal) malignancy.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  L. Bartella, S. B. Thakur, E. A. Morris, D. D. Dershaw, W. Huang, E. Chough, M. C. Cruz, and L. Liberman Enhancing Nonmass Lesions in the Breast: Evaluation with Proton (1H) MR Spectroscopy Radiology, October 1, 2007; 245(1): 80 - 87. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Bartella and W. Huang Proton (1H) MR Spectroscopy of the Breast RadioGraphics, October 1, 2007; 27(suppl_1): S241 - S252. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. Stanwell and C. Mountford In Vivo Proton MR Spectroscopy of the Breast RadioGraphics, October 1, 2007; 27(suppl_1): S253 - S266. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Bartella, E. A. Morris, D. D. Dershaw, L. Liberman, S. B. Thakur, C. Moskowitz, J. Guido, and W. Huang Proton MR Spectroscopy with Choline Peak as Malignancy Marker Improves Positive Predictive Value for Breast Cancer Diagnosis: Preliminary Study Radiology, June 1, 2006; 239(3): 686 - 692. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R K Jacklin, P F Ridgway, P Ziprin, V Healy, D Hadjiminas, and A Darzi Optimising preoperative diagnosis in phyllodes tumour of the breast J. Clin. Pathol., May 1, 2006; 59(5): 454 - 459. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Meisamy, P. J. Bolan, E. H. Baker, M. G. Pollema, C. T. Le, F. Kelcz, M. C. Lechner, B. A. Luikens, R. A. Carlson, K. R. Brandt, et al. Adding in Vivo Quantitative 1H MR Spectroscopy to Improve Diagnostic Accuracy of Breast MR Imaging: Preliminary Results of Observer Performance Study at 4.0 T Radiology, August 1, 2005; 236(2): 465 - 475. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. D. Yeh Characterization of Breast Lesions with Proton MR Spectroscopy Am. J. Roentgenol., November 1, 2003; 181(5): 1273 - 1274. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
