Radiology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online before print March 28, 2006, 10.1148/radiol.2392050205
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
2392050205v1
239/2/351    most recent
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Teifke, A.
Right arrow Articles by Lehr, H.-A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Teifke, A.
Right arrow Articles by Lehr, H.-A.
(Radiology 2006;239:351-360.)
© RSNA, 2006

Breast Imaging

Dynamic MR Imaging of Breast Lesions: Correlation with Microvessel Distribution Pattern and Histologic Characteristics of Prognosis1

Andrea Teifke, MD, Oliver Behr, MD, Markus Schmidt, MD, Anja Victor, PhD, Toni W. Vomweg, MD, Manfred Thelen, MD and Hans-Anton Lehr, MD

1 From the Departments of Radiology (A.T., T.W.V., M.T.), Pathology (O.B., H.A.L.) and Gynaecology (M.S.) and the Institute for Medical Biometry, Epidemiology and Informatics (A.V.), Johannes Gutenberg University of Mainz, Langenbeckstr 1, D-55131 Mainz, Germany. Received February 11, 2005; revision requested April 11; revision received May 30; final version accepted July 1. Supported by a grant from the Deutsche Forschungsgemeinschaft (Th 315/7–1). Address correspondence to A.T. (e-mail: teifke{at}radiologie.klinik.uni-mainz.de).


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 References
 
Purpose: To evaluate the association of dynamic enhancement parameters of benign and malignant breast lesions at magnetic resonance (MR) imaging with microvessel distribution and histologic prognostic tumor characteristics.

Materials and Methods: Regional review board approval and informed consent were obtained. Surgical resection specimens of breast lesions (32 benign, 86 malignant) in 118 patients (age range, 28–86 years; mean, 58 years) who had undergone dynamic T1-weighted MR imaging of both breasts were included in the study. Different MR enhancement parameters and microvessel density (MVD) distribution were determined. In malignant lesions, TNM stage, tumor grade, proliferative activity, and hormone receptor expression were determined. Spearman correlation coefficients; Wilcoxon, Fisher exact, Kruskal Wallis, and {chi}2 tests; and logistic regression analysis were used for evaluation.

Results: Malignant lesions exhibited a higher ratio of microvessels in tumor periphery versus tumor center than did benign lesions (P < .0005). High vessel ratios (P = .001) and low central vessel numbers (P = .007) were associated with high tumor grade. In malignant lesions, initial enhancement ratios of periphery to center of lesion correlated with the corresponding microvessel ratios (r = 0.61). Yet, a high peripheral MVD was not associated with strong peripheral enhancement (r = –0.09). High enhancement ratios, washout rates, and early enhancement peaks were associated with unfavorable, albeit not significant, prognostic indicators. Visible rim enhancement was the most accurate prognostic enhancement criterion for estrogen receptor status (P = .007), tumor grade (P = .06), and lymph node status (P = .046). Washout was the best discriminating criterion for proliferative activity.

Conclusion: The different enhancement behaviors of malignant and benign breast lesions cannot be explained by MVD alone; however, a low MVD in the center of carcinoma is reflected quantitatively by a high enhancement ratio and qualitatively by rim enhancement, with an implication of adverse prognosis.

© RSNA, 2006


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 References
 
In contrast to mammography, magnetic resonance (MR) imaging yields important information not only on the morphology of benign and malignant lesions of the breast but also on the functional aspects reflected by the temporal and spatial uptake of contrast medium. This enhancement is influenced by the extent and pattern of vascularization, vessel permeability, cellularity, interstitial pressure, and the fraction of the extracellular space (1). Established characteristics of malignant tumors are a rapid and distinct enhancement, followed by a washout and pronounced enhancement in the tumor periphery, the so-called rim enhancement (1,2). Several authors (3–9) have reported an association between the degree of tumor malignancy and prognosis with microvessel density (MVD) and permeability, oxygen saturation, and proliferative activity. Maximum mitotic activity and MVDs ("hot spots") alike are usually located at the periphery of the tumor (8,10). It is hence reasonable to assume that the different enhancement characteristics may not only indicate whether the respective lesions are benign or malignant but also provide valuable information about their biologic potential and consequently about their prognosis. In this case, MR imaging could eventually help guide therapeutic decisions (antiangiogenic, cytoxic, radiation, etc).

To date, only few groups (11–14) have tried to approach this issue, each examining different aspects of the problem. Motivated by different results in the literature, the purpose of our study was to evaluate the association of dynamic enhancement parameters of benign and malignant breast lesions at MR imaging with microvessel distribution and histologic prognostic tumor characteristics.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 References
 
Patients and Lesions
The study protocol was approved in its entirety by our regional review board. All patients gave informed consent to participate in this study. Between January 2003 and December 2003, 365 women underwent breast MR imaging for the following indications: preoperative staging (n = 75), severe scarring after breast-conserving surgery (n = 72), silicone implant after tumor surgery (n = 55), high-risk (family history indicating hereditary breast cancer, prior lobular carcinoma in situ, or atypical hyperplasia) and dense breast tissue (n = 54), problem lesions at conventional imaging (n = 39), search for a primary tumor (n = 29), monitoring of neoadjuvant chemotherapy (n = 2), suspicious microcalcifications (n = 19), and fibroadenomas (n = 20). These examinations were evaluated prospectively.

After MR imaging, tissue was obtained by means of surgical excision in 142 patients. The histopathologic examination, which included assessment of prognostic characteristics of the malignant lesions, was performed during routine diagnostic work-up. As part of the present study protocol, a team of radiologists (A.T., T.W.V., M.T.) and pathologists (M.S., H.A.L.) subsequently compared MR images and histologic slides. Seven patients were excluded because of excessive movement artifacts, and five patients were excluded because the tumor had not been removed in its entirety.

With the exception of four ductal carcinomas in situ, in which generalized enhancement of the entire breast parenchyma prevented identification of the exact tumor extension, and two nonenhancing fibroadenomas missed at MR imaging, all remaining lesions (n = 124) were found by interdisciplinary team assessment to exhibit a conclusive correlation of MR images and histologic findings in terms of localization, size, and outlines. For each of these lesions, one histologic section was selected that best corresponded to the MR image that had been used preoperatively for enhancement calculation. Subsequently, a corresponding recut tissue block was immunostained for the evaluation of microvessel distribution pattern and density. In six lesions, tissue processing and sectioning artifacts precluded an automated microvessel count. In the end, a total of 118 patients (age range, 28–86 years; mean, 58 years) were included in our study. In 12 patients with multicentric carcinoma, the study was confined to the largest malignant lesion.

At MR imaging, most study lesions appeared as enhancing masses (n = 110). In the remaining patients, segmental enhancing regions (n = 2), diffuse unilateral enhancement (n = 1), homogeneous very weak bilateral enhancement of the entire breast tissue (n = 3), and weakly enhancing regions in the former location of a ductal carcinoma treated with neoadjuvant chemotherapy (n = 2) were seen. Indications for MR imaging in the 118 patients included preoperative staging (n = 62), ambiguous findings at conventional imaging (n = 16), fibroadenomas (n = 15), suspicious microcalcifications (n = 5), high risk (n = 2), differentiation of scar tissue from tumor recurrence (n = 8), silicone implants (n = 4), search for primary tumor (n = 4), and monitoring of neoadjuvant chemotherapy (n = 2). Of 118 lesions, 32 (27%) were benign and 86 (73%) were malignant (Table 1). No size measurements are given for fibrocystic diseases. Two of the malignant lesions were recurrences after conservative surgery and radiation therapy.


View this table:
[in this window]
[in a new window]

  		Table 1. Histologic Diagnosis and Size of 118 Lesions

 
Tissue was obtained at mastectomy in 30 patients, breast-conserving surgery in 56 patients, and excision biopsy in 32 patients. Sixty-nine of 86 malignant lesions and 21 of 32 benign lesions had been confirmed preoperatively with core-needle biopsy. Sixty-two of the 118 lesions were palpable. The remaining 56 lesions were excised after needle localization (29 lesions guided with mammography; 23, with ultrasonography [US]; and four, with MR imaging). Sixteen fibroadenomas and one papilloma were excised because of patient anxiety; three papillomas, because of a specific request of the pathologists. The remaining 12 benign lesions were removed to rule out malignancy. The histologic work-up of these 12 lesions revealed radial scars with enhancing spiculated lesions at MR imaging in two patients, fibrocystic disease after neoadjuvant chemotherapy in two patients, and suspicious microcalcifications at mammography in three patients. The remaining five patients had suspicious findings at mammography, US, and MR imaging. Findings of histologic examination showed inflammatory changes. Patients with benign lesions were noticeably younger (median age, 49 years; range, 28–78 years) than those with breast cancer (median age, 62 years; range, 38–86 years).

MR Imaging and Analysis
MR imaging was performed with a 1.0-T imager (Magnetom Impact Expert 42SP/AS; Siemens Medical Systems, Erlangen, Germany) and the manufacturer's double-breast coil, with the patient in a prone position. At first, coronal T2-weighted turbo spin-echo images (repetition time msec/echo time msec, 5432/90; field of view, 175 x 350 mm; resolution, 1.39 x 1.38 x 3 mm) were obtained. These were followed by a dynamic contrast material–enhanced coronal T1-weighted fast low-angle shot three-dimensional sequence (15/7; flip angle, 30°; field of view, 350 x 175 x 119 mm; resolution, 1.82 x 1.37 x 1.86 mm; 64 sections) with an acquisition time of 93 seconds. One measurement was acquired before and five consecutive measurements were acquired after manual bolus injection of gadopentetate dimeglumine (Magnevist; Schering, Berlin, Germany) at a dose of 0.1 mmol per kilogram of body weight. The system software was used to calculate subtraction images from the first and third postcontrast studies and the precontrast dynamic study.

Signal intensity–time curves were calculated by using a workstation (SPARC Station; Sun Microsystems, Mountain View, Calif) and a software package (MRVision; MR Vision, Menlo Park, Calif). In the 110 enhancing masses, one region of interest (2–6 pixels, depending on lesion size) was placed in the peripheral area with the highest enhancement and another one was placed in the geometric center. In case of segmental and regional enhancement, the signal intensity was also measured in the periphery and the center. In the three patients with homogeneous uptake of contrast medium in both breasts, only one measurement was performed and the result was used for further calculations of both peripheral and central enhancement parameters. Likewise, only one measurement was performed in the patient with a generalized homogeneous unilateral enhancement that later turned out to represent a diffusely growing lobular carcinoma. All measurements were performed by two of three radiologists (A.T., M.T., 8 years of experience in breast MR Imaging; T.W.V., 6 years of experience) in consensus. To avoid any bias, calculation of all enhancement parameters (Table 2) was completed and documented before surgery and histologic work-up.


View this table:
[in this window]
[in a new window]

  		Table 2. MR Enhancement Parameters Modified according to Heywang-Köbrunner et al (15)

 
Histologic Analysis
Each resection specimen was obtained with three sutures of various lengths for orientation in three planes. For later orientation of the sections and for determination of resection margins, all specimens were inked with different colors. The tumors were submitted in their entirety in 0.4-cm-thick slices and were embedded in paraffin, and 4-µm whole-mount sections were placed on 5 x 8-cm glass slides. The histopathologic examination, including determination of prognostic features (Table 3), was performed as part of routine diagnostic work-up by board-certified pathologists. All cases that were entered into this study were specifically reevaluated again by one pathologist (H.A.L.) with 12 years experience in breast disease. All pertinent determinants of the TNM classification, estrogen (clone ER 1D5; Immunotech, Marseille, France) and progesterone (clone PGR 636; Dako, Glostrup, Denmark) receptor expression, and the cell cycle–specific marker MIB-1 (clone Ki 67; Dako) were assessed. As part of routine work-up, tumor grade was defined in all invasive nonlobular carcinomas according to the Scarf-Bloom-Richardson protocol as modified by Elston and Ellis (16) by integrating the extent of tubule formation, nuclear polymorphism, and proliferative activity (grade 1 = highly differentiated, grade 2 = moderately differentiated, and grade 3 = poorly differentiated). Ductal carcinomas in situ were graded according to the classification scheme proposed by Silverstein et al (17). Lymph node involvement was not assessed in six patients with ductal carcinoma in situ and in two patients with breast recurrence after prior breast-conserving surgery. Proliferative activity, tumor grade, and receptor status were not available in two ductal carcinomas in situ.


View this table:
[in this window]
[in a new window]

  		Table 3. Prognostic Indicators of Malignant Lesions

 
In addition to these routine standard diagnostic studies, one whole-mount section through the center of the lesion that corresponded well with the MR image that had been used for enhancement calculation before surgery was selected in consensus by two pathologists (H.A.L. and M.S. with 12 and 8 years of experience, respectively, in breast pathology) and two of the three radiologists (A.T., M.T., T.W.V.) who had evaluated the MR imaging examinations before surgery. A recut of the corresponding paraffin block was immunostained with antibodies to CD34 (clone QBEMI 10; Immunotech), making sure that the entire tumor area was well covered with antibodies and detection reagents. Subsequently a grid was placed over the whole-mount section, and five high-power fields (magnification, x40; 0.47 mm2 in area) were selected in the lesion periphery (at 3-, 6-, 9-, and 12-o'clock positions) and in the geometric center of the grid (ie, tumor center). Because of the use of 5 x 8-cm whole-mount slides, all tumors were successfully placed on a slide in their entirety and, hence, all regions of interest could be assessed in each examined lesions. In noncircumscribed lesions without a defined center (fibrocystic changes and diffusely growing lobular carcinoma), the regions were placed so as to represent the geometric center of the slide and the four perimeters. The microvessels were quantified automatically with a computer-assisted image analysis (18).

Briefly, the digitized images of the five high-power fields were imported into Photoshop (version 4, Adobe Systems; Mountain View, Calif). By using the magic wand tool and the "select similar" command, all CD34-positive chromogen-stained pixels were selected and labeled black. By using the Gaussian blur command in the filter menu, the edges of the labeled blood vessels were smoothed to prevent multiple counts for rough-edged vessels. The numbers of vessels were then counted with a commercially available plug-in (Image Processing Toolkit, version 2.1; Reindeer Games, Charlotte, NC). MVD was assessed as number of microvessels per surface area (0.47 mm2) in each high-power field and was expressed as microvessels per square millimeters. The following three measurements were assessed: mean MVD of the four peripheral fields (hereafter, peripheral MVD), MVD at the central field (hereafter, central MVD), and the ratio of both measurements (hereafter, MVD ratio). To prevent bias, all analyses were performed in a strictly blinded fashion by a research associate (O.B.) who was not aware of the specific end points of the study protocol and who was specifically trained to perform the computer-based image analysis.

Statistical Analysis
Because of skewed distributions, we used median and quartiles for description of continuous data. To present categoric data, absolute and relative frequencies are given. To quantify the correlation between two continuous variables, Spearman correlation coefficients (r values) were calculated. To test whether there exists a difference between two groups in continuous variables, the Wilcoxon test was performed, and for categoric variables, the Fisher exact test was performed. For more than two groups, Kruskal Wallis and {chi}2 tests were used. A multiple test procedure (Bonferroni correction) was used to ensure that the multiple level was .05 for the five main questions of the study (distribution pattern of MVD of benign and malignant lesions, association between tumor grade and proliferative activity with MVD, association of tumor grade and proliferative activity with enhancement). P values of these targets less than .01 were considered to indicate a significant difference. All other P values in the article are investigations of an explorative nature and have to be considered as descriptive values only. To assess the discriminating ability of the different enhancement criteria with respect to the prognostic indices, logistic regression analysis with stepwise forward-likelihood-ratio variable selection was used. All statistical evaluations were performed with statistical software (SPSS for Microsoft Windows, version 11.0; SPSS, Chicago, Ill). If not stated differently throughout the text, descriptions for continuous variables are presented as the median and interquartile range (25th and 75th percentiles).


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 References
 
MVD in Benign and Malignant Lesions
For benign and malignant lesions, the respective medians of the mean peripheral MVD were 49 (range, 30–61) and 41 (range, 32–54) and those of central MVD were 40 (range, 25–60) and 31 (range, 18–40) (Fig 1). The median ratio of peripheral to central MVD was 1.1 (range, 0.9–1.2) for benign lesions and 1.4 (range, 1.1–2.1) for carcinomas; this difference was significant (P < .0005).


Figure 1
View larger version (22K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1: Box plots of MVDs in center (gray) and periphery (white) of malignant and benign lesions show median, interquartile range (25th and 75th percentiles), total range, and outliers ({circ}) (between 1.5 and 3.0 times the interquartile range away from the box). Central MVD of malignant lesions is noticeably lower than that of benign lesions.

 
Association between MVD and Prognostic Factors
There were significant differences (Table 4) between the three tumor grades in terms of the MVD ratio of lesion periphery to lesion center (P = .001). The relationship with proliferative activity (ie, MIB) showed the same trend, but the differences were not significant (P = .08). Also, estrogen receptor–negative tumors exhibited markedly higher MVD ratios than did receptor-positive tumors (P = .006). There was no relationship between MVD and tumor size or lymph node status.


View this table:
[in this window]
[in a new window]

  		Table 4. Association between MVD and Prognostic Indicators for Malignant Lesions

 
Correlation of MVD with Dynamic Enhancement Characteristics at MR Imaging
MVD ratios correlated (Fig 2) with initial enhancement ratios of periphery versus center of malignant lesions (r = 0.61). High MVDs in the center of malignant tumors are associated with high initial central enhancement values (r = 0.51) (Table 5). For benign lesions, this relationship was less distinct (r = 0.33). Data from peripheral MVDs showed only a weak correlation for benign lesions (r = 0.29) and none for malignant lesions (r = –0.09). The respective associations with maximum enhancement parameters were comparable (Table 5).


Figure 2
View larger version (21K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2: Scatterplot depicts positive association of peripheral to central MVD ratio and between initial peripheral to central enhancement ratio for malignant ({blacktriangleup}) lesions (r = 0.61). No such association is seen for benign ({circ}) lesions (r = 0.11).

 

View this table:
[in this window]
[in a new window]

  		Table 5. Spearman Correlation Coefficients between Enhancement Parameters and MVD in Periphery and Center

 
Difference in Enhancement Behavior in Benign and Malignant Lesions
The enhancement parameters assessed in the lesion periphery were more helpful to distinguish between malignant and benign breast diseases than were those assessed in the center (P values in Table 6). Also, the enhancement ratio of periphery to center showed significant differences (P < .0005). Twenty-nine percent (25 of 86) of carcinomas exhibited a distinct peripheral rim enhancement. These lesions showed much higher initial enhancement ratios and lower central MVDs than did lesions without rim enhancement (Fig 3). Among 32 benign lesions, only one had a rim enhancement (P = .002). This lesion turned out to be an abscess.


View this table:
[in this window]
[in a new window]

  		Table 6. Comparison of Enhancement Values for Benign and Malignant Breast Lesions

 

Figure 3
View larger version (17K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3: Box plots of association between visible rim enhancement with initial enhancement ratio (gray) and MVD ratio (white) of periphery to center show median, interquartile range (25th and 75th percentiles), total range, outliers ({circ}) (between 1.5 and 3 times the interquartile range away from the box), and extreme outliers (*) (> three times interquartile range away from the box). Two outliers (vessel density at 38 and initial enhancement at 44.5) are not shown so as not to distort the graph. Lesions without visible rim enhancement show identical MVD and initial enhancement in lesion periphery as in the center, in contrast to lesions with rim enhancement.

 
Association of Enhancement Parameters with Prognostic Factors
Malignant tumors with histomorphologic indications of a poor prognosis showed higher ratios of peripheral to central enhancement, higher washout rates, and earlier enhancement peaks than those with favorable prognostic indicators (Table 7). However, the association of enhancement ratios with tumor grade and proliferative activity was not significant (P = .192 and P = .394, respectively). The affinity of unfavorable histopathologic features with a visible rim enhancement is demonstrated in Table 8 and Figure 4. The logistic regression revealed that rim enhancement was the most accurate prognostic enhancement criterion for estrogen receptor status, tumor grade, and lymph node status (Table 9). For the proliferative activity, washout was the best discriminating enhancement criterion.


View this table:
[in this window]
[in a new window]

  		Table 7. Association between Enhancement Parameters and Prognostic Indicators for Malignant Lesions

 

View this table:
[in this window]
[in a new window]

  		Table 8. Association between Rim Enhancement and Prognostic Indicators for Malignant Lesions

 

Figure 4
View larger version (29K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4: Bar graphs show relationship between peripheral rim enhancement and prognostic features of malignant lesions. Tumors with adverse prognostic features have a higher incidence of rim enhancement. N = lymph node, T = TNM stage.

 

View this table:
[in this window]
[in a new window]

  		Table 9. Results of Logistic Regression Analysis for Malignant Lesions at Periphery

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
References
 
MVD in Benign and Malignant Lesions
Angiogenesis constitutes a prerequisite for the growth of malignant tumors beyond a certain size. The higher the MVD the better the nutritional tumor situation, which in turn facilitates tumor growth and poor prognosis (3,4,8,19). In light of this theoretical background, we were rather surprised that in our present study malignant lesions did not show increased MVDs when compared with benign lesions. Rather, MVDs were lower in the center of malignant tumors than of benign tumors. A careful analysis of the literature on this subject shows that virtually all studies on tumor MVD are limited to the examination of malignant lesions. Only rarely were malignant lesions compared with benign lesions. Da Silva and co-workers (20) found that malignant lesions have a denser microvessel network than do fibroadenomas. We concede that the discrepancy with our data may—at least in part—be due to a selection bias. Benign lesions with low MVD may show no or only weak enhancement. Such nonenhancing lesions remain frequently occult at MR imaging. Also, they are usually considered to be benign and are hence not excised. For these reasons, they could be underrepresented in our present study.

It is well established that benign nonneoplastic breast lesions such as fibroadenomas, papillomas, and hyperplasias and inflammatory changes may exhibit avid angiogenesis (21). Indeed, most of the benign lesions investigated in our present study did belong to these histologic groups. This interpretation is furthermore corroborated by similar observations in a comparative study (22) of fibroadenomas and invasive carcinomas, in which MR imaging was likewise used for selection of cases.

To date, most studies have rather focused on the assessment of MVDs in the most densely perfused areas of the tumor, the so-called hot spots (8). It is generally assumed that these hot spots are located in the tumor periphery, but a systematic analysis has been performed in only a small number of studies and almost exclusively in malignant lesions (12,22–26). In agreement with findings of Jitsuiki et al (24) and Hasebe et al (25), who examined ductal carcinomas, we found a distinct increase of MVD from the center toward the periphery of malignant tumors. We also analyzed benign lesions and observed only small differences in microvessel distribution from center to periphery.

Association of MVD with Prognostic Indicators
The association of maximum MVD and prognostic indicators such as tumor grade and proliferative activity has been addressed by several authors (10,27–30), but the results are not conclusive. In our present study, we found no significant association of MVD in the tumor periphery (which were higher than those in the tumor center) and prognostic features. The divergent results may or may not be secondary to differences in the techniques of microvessel quantification. Our idea to assess MVD also in the geometric center of the tumor was derived from the phenomenon of rim enhancement empirically observed almost exclusively in malignant lesions. To our knowledge, the MVD ratio of the periphery to center that we obtained in this manner has previously not been studied in detail. In our present study, we found that this gradient correlates significantly with tumor grade (P < .001) and with other prognostic indicators. Of particular interest was the observation that this association was due not to a denser microvessel network in the periphery of less differentiated tumors but rather to a gradual loss of central microvessels with decreasing tumor differentiation. Poor central vascularization can be explained by the development of a central scar and central necrosis.

Correlation of MVD with Dynamic MR Enhancement Characteristics
Some authors (13,31–34) have described a correlation of initial enhancement with MVD, but authors of more recent studies (14,35,36) found no such association. To our knowledge, the association of the ratio of peripheral to central enhancement with the respective MVD has not been examined before. We found that differences in enhancement between tumor periphery and center in malignant tumors are well reflected by the corresponding differences in MVD (r = 0.61 for initial enhancement ratio) and that no such correlation was seen for benign lesions in which enhancement and microvessel values were virtually identical in the periphery and center of the lesions. While in the tumor center a dense microvessel network was associated with high initial enhancement (r = 0.51), no such interrelation was seen for the tumor periphery (r = –0.09). Matsubayashi and co-workers (26) observed higher vascular endothelial growth factor concentrations in the periphery of those breast carcinomas that exhibited early rim enhancement. Complementary to this finding, our study results showed that rim enhancement is associated with high enhancement and microvessel ratio from tumor periphery versus center.

Association of Enhancement Parameters with Prognostic Indicators
The postulate that MVD leads to strong enhancement, which results in poor prognosis, is not supported by the results of our present study. Nevertheless, we found an association between the degree of "typical malignant" MR imaging enhancement parameters and unfavorable biologic markers of prognosis. These were initial enhancement, saturation, and washout on the one hand and tumor grade, proliferative activity, estrogen receptor expression, and nodal status on the other. The correlation of rim enhancement with poor tumor differentiation (P = .003) and negative hormone receptor status (P = .006) was significant.

These findings are consistent with observations made by others. Mussurakis et al (11,37) found a significant association between the degree of enhancement and lymph node status, as well as tumor grade. Szabo et al (14) found that the lack of hormone receptor expression and high tumor grade were linked with a rapid peak enhancement and also found an association between tumor grade and washout. Likewise Stomper et al (38) reported an association of peripheral enhancement with a high S-phase percentage of tumor cells as an indicator of tumor proliferation.

In conclusion, our study findings show that the characteristic enhancement in the periphery of breast carcinomas at MR imaging is not caused by an elevated MVD in the tumor periphery but rather by a lower MVD in the tumor center. A higher ratio of microvessels in the tumor periphery to tumor center in malignant lesions compared with that in benign lesions correlates with a higher enhancement ratio and visible rim enhancement, both associated with unfavorable histomorphologic prognostic tumor features.


   FOOTNOTES
 

Abbreviations: MVD = microvessel density

Author contributions: Guarantors of integrity of entire study, A.T., H.A.L.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; manuscript final version approval, all authors; literature research, A.T., O.B.; clinical studies, A.T., O.B., M.S., T.W.V., M.T., H.A.L.; statistical analysis, A.V.; and manuscript editing, A.T., T.W.V., H.A.L.

Authors stated no financial relationship to disclose.


   References
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 References
 

  1. Kuhl CK, Schild HH. Dynamic image interpretation of MRI of the breast. J Magn Reson Imaging 2000;12:965–974.[CrossRef][Medline]
  2. Nunes LW, Schnall MD, Siegelman ES, et al. Diagnostic performance characteristics of architectural features revealed by high spatial-resolution MR imaging of the breast. AJR Am J Roentgenol 1997;169:409–415.[Abstract/Free Full Text]
  3. Weidner N, Semple JP, Welch WR, Folkman J. Tumor angiogenesis and metastasis: correlation in invasive breast carcinoma. N Engl J Med 1991;324:1–8.[Abstract]
  4. Bosari S, Lee AK, DeLellis RA, et al. Microvessel quantitation and prognosis in invasive breast carcinoma. Hum Pathol 1992;23:755–761.[CrossRef][Medline]
  5. Hall NR, Fish DE, Hunt N, et al. Is the relationship between angiogenesis and metastasis in breast cancer real? Surg Oncol 1992;1:223–229.
  6. Horak ER, Leek R, Klenk N, et al. Angiogenesis, assessed by platelet/endothelial cell adhesion molecule antibodies, as indicator of node metastases and survival in breast cancer. Lancet 1992;340:110–1124.
  7. Magno WB, Hirschfield L, Bhuiya T, Harrison G, Mir R. Correlation of proliferative index (PCNA reactivity and Ki-67 reactivity) in primary breast carcinoma with hormone status, lymph node status, and disease-free survival. Conn Med 1992;56:667–669.[Medline]
  8. Weidner N, Folkman J, Pozza F, et al. Tumor angiogenesis: a new significant and independent prognostic indicator in early-stage breast carcinoma. J Natl Cancer Inst 1992;84:1875–1887.[Abstract/Free Full Text]
  9. Haerslev T, Jacobsen GK. Proliferating cell nuclear antigen in breast carcinomas: an immunohistochemical study with correlation to histopathological features and prognostic factors. Virchows Arch 1994;424:39–46.[Medline]
  10. Belien JA, van Diest PJ, Baak JP. Relationships between vascularization and proliferation in invasive breast cancer. J Pathol 1999;189:309–318.[CrossRef][Medline]
  11. Mussurakis S, Buckley DL, Horsman A. Prediction of axillary lymph node status in invasive breast cancer with dynamic contrast-enhanced MR imaging. Radiology 1997;203:317–321.[Abstract/Free Full Text]
  12. Stomper PC, Winston JS, Herman S, et al. Angiogenesis and dynamic MR imaging gadolinium enhancement of malignant and benign breast lesions. Breast Cancer Res Treat 1997;45:39–46.[CrossRef][Medline]
  13. Bone B, Aspelin P, Bronge L, Veress B. Contrast-enhanced MR imaging as a prognostic indicator of breast cancer. Acta Radiol 1998;39:279–284.[Medline]
  14. Szabo BK, Aspelin P, Kristoffersen Wiberg M, Tot T, Bone B. Invasive breast cancer: correlation of dynamic MR features with prognostic factors. Eur Radiol 2003;13:2425–2435.[CrossRef][Medline]
  15. Heywang-Kobrunner SH, Bick U, Bradley WG Jr, et al. International investigation of breast MRI: results of a multicentre study (11 sites) concerning diagnostic parameters for contrast-enhanced MRI based on 519 histopathologically correlated lesions. Eur Radiol 2001;11:531–546.[CrossRef][Medline]
  16. Elston CW, Ellis IO. Pathological prognostic factors in breast cancer. I. The value of histological grade in breast cancer: experience from a large study with long-term follow-up. Histopathology 1991;19:403–410.
  17. Silverstein MJ, Poller DN, Waisman JR, et al. Prognostic classification of breast ductal carcinoma-in-situ. Lancet 1995;345:1154–1157.[CrossRef][Medline]
  18. Lehr HA, Mankoff DA, Corwin D, Santeusanio G, Gown AM. Application of photoshop-based image analysis to quantification of hormone receptor expression in breast cancer. J Histochem Cytochem 1997;45:1559–1565.[Abstract/Free Full Text]
  19. Folkman J, Klagsbrun M. Angiogenic factors. Science 1987;235:442–447.[Abstract/Free Full Text]
  20. Da Silva ID, De Lima GR, Gebrim LH, et al. Invasive ductal carcinoma, fibroadenoma and adjacent breast stroma angiogenesis: an immunohistochemical and morphometrical study. Bull Assoc Anat (Nancy) 1996;80:17–19.
  21. Brem SS, Jensen HM, Gullino PM. Angiogenesis as a marker of preneoplastic lesions of the human breast. Cancer 1978;41:239–244.[CrossRef][Medline]
  22. Weind KL, Maier CF, Rutt BK, Moussa M. Invasive carcinomas and fibroadenomas of the breast: comparison of microvessel distributions—implications for imaging modalities. Radiology 1998;208:477–483.[Abstract/Free Full Text]
  23. Buadu LD, Murakami J, Murayama S, et al. Patterns of peripheral enhancement in breast masses: correlation of findings on contrast medium enhanced MRI with histologic features and tumor angiogenesis. J Comput Assist Tomogr 1997;21:421–430.[CrossRef][Medline]
  24. Jitsuiki Y, Hasebe T, Tsuda H, et al. Optimizing microvessel counts according to tumor zone in invasive ductal carcinoma of the breast. Mod Pathol 1999;12:492–498.[Medline]
  25. Hasebe T, Mukai K, Tsuda H, et al. New prognostic histological parameter of invasive ductal carcinoma of the breast: clinicopathological significance of fibrotic focus. Pathol Int 2000;50:263–272.[CrossRef][Medline]
  26. Matsubayashi R, Matsuo Y, Edakuni G, et al. Breast masses with peripheral rim enhancement on dynamic contrast-enhanced MR images: correlation of MR findings with histologic features and expression of growth factors. Radiology 2000;217:841–848.[Abstract/Free Full Text]
  27. Kato T, Kimura T, Miyakawa R, et al. Clinicopathologic study of angiogenesis in Japanese patients with breast cancer. World J Surg 1997;21:49–56.[CrossRef][Medline]
  28. Fridman V, Humblet C, Bonjean K, Boniver J. Assessment of tumor angiogenesis in invasive breast carcinomas: absence of correlation with prognosis and pathological factors. Virchows Arch 2000;437:611–617.[CrossRef][Medline]
  29. Valkovic T, Dobrila F, Melato M, et al. Correlation between vascular endothelial growth factor, angiogenesis, and tumor-associated macrophages in invasive ductal breast carcinoma. Virchows Arch 2002;440:583–588.[CrossRef][Medline]
  30. Vermeulen PB, Fox SB, Dirix LY, Colpaert C, Van Marck EA. Heterogeneity of vascularisation in invasive breast carcinoma. Eur J Cancer 2002;38:1951–1952.[CrossRef][Medline]
  31. Frouge C, Guinebretiere JM, Contesso G, Di Paola R, Blery M. Correlation between contrast enhancement in dynamic magnetic resonance imaging of the breast and tumor angiogenesis. Invest Radiol 1994;29:1043–1049.[Medline]
  32. Buadu LD, Murakami J, Murayama S, et al. Breast lesions: correlation of contrast medium enhancement patterns on MR images with histopathologic findings and tumor angiogenesis. Radiology 1996;200:639–649.[Abstract/Free Full Text]
  33. Buckley DL, Drew PJ, Mussurakis S, Monson JR, Horsman A. Microvessel density of invasive breast cancer assessed by dynamic Gd-DTPA enhanced MRI. J Magn Reson Imaging 1997;7:461–464.[Medline]
  34. Esserman L, Hylton N, George T, Weidner N. Contrast-enhanced magnetic resonance imaging to assess tumor histopathology and angiogenesis in breast carcinoma. Breast J 1999;5:13–21.[CrossRef][Medline]
  35. Hulka CA, Edmister WB, Smith BL, et al. Dynamic echo-planar imaging of the breast: experience in diagnosing breast carcinoma and correlation with tumor angiogenesis. Radiology 1997;205:837–842.[Abstract/Free Full Text]
  36. Su MY, Cheung YC, Fruehauf JP, et al. Correlation of dynamic contrast enhancement MRI parameters with microvessel density and VEGF for assessment of angiogenesis in breast cancer. J Magn Reson Imaging 2003;18:467–477.[CrossRef][Medline]
  37. Mussurakis S, Buckley DL, Horsman A. Dynamic MR imaging of invasive breast cancer: correlation with tumour grade and other histological factors. Br J Radiol 1997;70:446–451.[Abstract]
  38. Stomper PC, Herman S, Klippenstein DL, et al. Invasive breast carcinoma: analysis of dynamic magnetic resonance imaging enhancement features and cell proliferative activity determined by DNA S-phase percentage. Cancer 1996;77:1844–1849.[CrossRef][Medline]



This article has been cited by other articles:


Home page
 RadiologyHome page
S. Makkat, R. Luypaert, T. Stadnik, C. Bourgain, S. Sourbron, M. Dujardin, J. De Greve, and J. De Mey
Deconvolution-based Dynamic Contrast-enhanced MR Imaging of Breast Tumors: Correlation of Tumor Blood Flow with Human Epidermal Growth Factor Receptor 2 Status and Clinicopathologic Findings--Preliminary Results
Radiology, September 9, 2008; (2008) 2492071147.
[Abstract] [Full Text]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
2392050205v1
239/2/351    most recent
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Teifke, A.
Right arrow Articles by Lehr, H.-A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Teifke, A.
Right arrow Articles by Lehr, H.-A.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
RADIOLOGY RADIOGRAPHICS RSNA JOURNALS ONLINE