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


     

Published online before print March 21, 2002, 10.1148/radiol.2232010428

(Radiology 2002;223:558.)

A more recent version of this article appeared on May 1, 2002
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow All Versions of this Article:
2232010428v1
223/2/558    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 Delille, J.-P.
Right arrow Articles by Garrido, L.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Delille, J.-P.
Right arrow Articles by Garrido, L.

Technical Developments

Breast Cancer: Regional Blood Flow and Blood Volume Measured with Magnetic Susceptibility–based MR Imaging—Initial Results1

Jean-Paul Delille, MD, Priscilla J. Slanetz, MD, MPH, Eren D. Yeh, MD, Daniel B. Kopans, MD and Leoncio Garrido, PhD

1 From the Department of Radiology, NMR Center (J.P.D., L.G.) and Breast Imaging Center (P.J.S., E.D.Y., D.B.K.), Massachusetts General Hospital and Harvard Medical School, Boston, Mass. Received February 7, 2001; revision requested April 3; revision received August 10; accepted November 30. Supported by the Massachusetts General Hospital NMR Center. Address correspondence to L.G., Instituto de Ciencia y Tecnología de Polímeros, C.S.I.C., Juan de la Cierva 3, 28006 Madrid, Spain (e-mail: lgarrido@cetef.csic.es).


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 Materials and Methods
 Results
Discussion
 REFERENCES
 
The purpose of this study was to quantify microcirculation in breast neoplasms with magnetic susceptibility-based contrast material–enhanced magnetic resonance imaging. With this imaging method for invasive cancers, the mean values of the ratios of tumor to normal blood flow and blood volume were significantly higher (P < .002) than those for benign or normal tissue. The method allows independent measurement of regional blood flow and blood volume in breast cancers.

© RSNA, 2002

Index terms: Breast neoplasms, diagnosis, 00.32 • Breast neoplasms, MR, 00.121416 • Magnetic resonance (MR), perfusion study, 00.121416, 00.12144


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
Materials and Methods
 Results
Discussion
 REFERENCES
 
An understanding of the microcirculation of tumors is important for their diagnosis and treatment. Tumor vessels are distinct from normal vessels; they tend to be larger in diameter compared with corresponding normal microvessels, have large pores or vacuolar vesicles in the walls, and lack contractile properties because of discontinuities in the layer of epithelium (1,2). These dissimilarities in microvascular attributes lead to microcirculation patterns in tumors that are different from those in normal tissue (3,4). Thus, measurement of hemodynamic parameters of tissues, such as blood flow, blood volume, microvascular permeability, and extravascular extracellular space can help diagnosis and management of cancer.

During the past decade, a substantial effort has been devoted to the development of magnetic resonance (MR) imaging techniques combined with the administration of a contrast medium for the diagnosis of breast cancer (5–7). Furthermore, the characterization of breast tissue hemodynamics has contributed to great improvement in the specificity of MR imaging for the diagnosis of malignancy. This has been accomplished by estimating kinetic parameters on the basis of pharmacologic or physiologic tracer kinetic modeling of concentration-time curves of contrast material in tumors (8–10).

Results at dynamic contrast material–enhanced echo-planar imaging have shown that the values of the extraction-flow product (EF) for breast carcinomas are higher than those for normal tissue and the average benign lesions (11,12). However, EF provides a combined measure of several physiologic parameters, such as flow, microvascular permeability, and surface area, that cannot be deconvolved. Thus, there are lesions with different origins and microvascular properties that behave similarly in dynamic contrast-enhanced MR imaging measurements. Also, interpretation of the dynamic contrast-enhanced MR imaging measurements is difficult when two or more parameters change as a result of therapeutic intervention. Independent measurement of blood flow and blood volume in the breast may allow better discrimination between the various types of breast diseases and could be valuable for assessment of changes in microcirculation of tumors during therapy.

Magnetic susceptibility–based contrast-enhanced methods have the potential to provide that information. A significant decrease in the magnitude of the MR signal is observed when transverse relaxation time (R2*) gets longer. This increase is induced by intravascular variations in magnetic susceptibility during the first pass of a bolus of contrast medium. With magnetic susceptibility–based contrast-enhanced methods, regional cerebral blood volume has been determined in many brain tumors, and the results show that high values of cerebral blood volume are associated with increased tumor metabolism and aggressiveness (13). More recently, a technique to determine cerebral blood flow with bolus tracking of a nondiffusible contrast agent at MR imaging has been described (14).

In the case of magnetic susceptibility–based breast MR imaging, a potential problem is the absence of a natural barrier (ie, the blood-brain barrier) to maintain vascular compartmentalization of the contrast medium. Extravasation of the contrast agent will lead to a decrease in the T2* effect and a competing T1 enhancement from the extravascular spins. The net result would be a substantial or complete elimination of the effect of change in R2* ({Delta}R2*). However, results in several qualitative studies have shown that {Delta}R2* effects are observable in the breast (15–18). The purpose of this study was to quantify microcirculation in breast neoplasms with magnetic susceptibility-based contrast-enhanced MR imaging.


   Materials and Methods
TOP
 ABSTRACT
 INTRODUCTION
 Materials and Methods
Results
Discussion
 REFERENCES
 
Thirteen women (mean age, 51.3 years ± 13.7 [SD]; age range, 30–70 years) with abnormal findings on their mammograms participated in the study after they gave informed consent. Some of the mammographic abnormalities were suggestive of malignancy (eg, stellar mass, architectural distortion, and microcalcificacions), and others were nonspecific for cancer (eg, increased and asymmetric breast density). The subjects were chosen randomly and examined prospectively. The institutional review board for research involving human subjects approved the procedures. There were six premenopausal and seven postmenopausal subjects.

MR imaging of the breast was performed with a 1.5-T MR imager (Signa; GE Medical Systems, Milwaukee, Wis) with a dedicated breast receiver coil. An intravenous catheter was placed into an antecubital vein before imaging and was connected to a power injector (MedRad, Pittsburgh, Pa) that contained gadopentetate dimeglumine (Magnevist; Berlex Laboratories, Wayne, NJ). The imaging protocol consisted of eight data acquisition series: (a) localization, (b) proton-density and T2-weighted imaging, (c) precontrast three-dimensional high-spatial-resolution imaging, (d) T1 mapping, (e) dynamic contrast-enhanced echo-planar imaging and first administration of contrast agent, (f) postcontrast three-dimensional high-spatial-resolution imaging, (g) T1 mapping, and (h) dynamic T2*-weighted echo-planar imaging and second administration of contrast agent.

Localization was performed with a gradient-echo sequence (repetition time msec/echo time msec = 18/5), followed by a transverse fast spin-echo acquisition (3,500/17, 165). With the imager in research mode and after manual shimming on the volume of interest, a three-dimensional spoiled gradient-echo sequence (23/6, flip angle of 30°, one signal acquired) was used to obtain high-spatial-resolution images. Fat suppression was performed by using a binomial water-selective excitation pulse. The matrix size was 512 x 256, with a field of view that ranged from 28 to 35 cm. Sixty 2.0–2.7-mm-thick sections were acquired to provide full coverage of the volume of interest, with a voxel size that ranged from 0.5 x 1.0 x 2.0 mm to 0.7 x 1.4 x 2.7 mm. The T1-weighted functional MR imaging protocol included two echo-planar spin-echo acquisitions. Before administration of the contrast agent, the T1 of breast tissue was measured with an inversion-recovery echo-planar sequence (6,000/30/50–1,400 [inversion time msec], 10 150-msec steps, one signal acquired). The matrix size was 128 x 128, with a field of view between 35 and 40 cm. Seventeen to 19 5–7-mm-thick sections were acquired with a 1.5-mm gap. Voxel size ranged from 2.7 x 2.7 x 5.0 mm to 3.1 x 3.1 x 7.0 mm.

Tissue uptake of contrast material was monitored before, during, and after administration with inversion-recovery echo-planar imaging (8,000/30/160). A fixed inversion time of 160 msec was used to minimize the contribution of fat to the total MR signal. The other parameters were the same as those used for T1 mapping. Twenty-six images were acquired at 8-second intervals during a total imaging time of 3 minutes 29 seconds. Five to seven images were acquired before administration of an intravenous bolus of gadopentetate dimeglumine (0.1 mmol per kilogram of body weight at 3.5 cm3/sec). The bolus of contrast agent was followed by a 10-mL saline flush. Immediately after data acquisition with inversion-recovery echo-planar imaging, a second set of postcontrast high-spatial-resolution images was acquired with a fat-suppressed three-dimensional spoiled gradient-echo sequence performed with the same parameters used before injection, followed by a second measurement of T1 with the inversion-recovery echo-planar protocol described previously.

In the breast 10–15 minutes after administration of the first bolus of contrast agent, {Delta}R2* effects associated with the passage of a second bolus were observed. Thus, the T2*-weighted data were collected with a multisection (16–19 sections) asymmetric spin-echo echo-planar sequence developed in-house (2,000/40, echo offset of -20 msec) performed before, during, and after administration of the second bolus injection of gadopentetate dimeglumine (0.1 mmol/kg at 3.5 cm3/sec) followed by a 10-mL saline flush. Imaging time was approximately 2 minutes. The contrast agent was administered 15 seconds after data acquisition was started. By performing the T2*-weighted study during the second injection of contrast medium (the first injection was administered during the dynamic T1-weighted study), the T1 effects due to microvascular leakage of contrast material were minimized. The section thickness was 5 mm with an intersection gap of 1.5 mm. The field of view varied between 35 and 40 cm with an in-plane resolution of 64 x 64 pixels.

The inversion-recovery and dynamic contrast-enhanced echo-planar data were analyzed as described elsewhere (11) by using in-house software. Briefly, the inversion-recovery data were converted into T1 maps of the breast. With these T1 maps, the signal intensity changes observed during dynamic imaging were converted to) change in longitudinal relaxation rate equal to reciprocal of relaxation time, where R1 = 1/T1 ({Delta}R1). The change in {Delta}R1 reflects effective concentration of gadopentetate dimeglumine, and a kinetic model was used to estimate EF (milliliters of blood per gram of tissue per minute). Images with high spatial resolution were used to determine the regions of interest (J.D.P., P.J.S., L.G.) for measuring T1 and EF in the corresponding maps.

In six T2*-weighted data sets, one of every five to seven time points was not saved because of commercial software failure. For these data sets, the missing time points in the series were calculated by means of linear interpolation between the two immediate neighbors. This procedure was tested with a data set saved in its entirety by replacing a similar number of time points with their corresponding interpolations. Comparison between the values of the parameters estimated from the noninterpolated and interpolated sets showed nonsignificant differences (P > .21, results not shown). Dynamic T2*-weighted echo-planar data were analyzed as described previously (14). Tissue and arterial time concentration curves of contrast agent were estimated by assuming a linear relationship between concentration C(t) and {Delta}R2* (19–21): where TE is echo time, S(0) and S(t) are the signal intensities at the baseline and time t, respectively.

Postcontrast three-dimensional images with high spatial resolution were used to guide the identification of arterial branches in the imaging plane of interest. Feeding vessels were identified on the echo-planar images as pixels that displayed an early increase in {Delta}R2* after administration of contrast agent. Thus, the shape of the arterial input function (AIF) was obtained by averaging data in two to five pixels. Although we did not attempt to quantify the arterial concentration in absolute units, this method has been found to closely reflect the actual arterial levels of gadopentetate dimeglumine when imaged with echo-planar techniques (22). To reduce errors in the estimation of blood flow due to dispersion and delay of the AIF throughout the breast, an AIF in the plane of interest or one next to it was chosen. A threshold filter equal to 0.4 times the value of the AIF baseline was applied to exclude pixels with low initial signal intensity from the calculations.

To estimate blood flow independently of the vascular structure, an algebraic approach was used, with the assumption of linear variation in signal intensities during equally sampled small time intervals, as well as a nonparametric singular value decomposition deconvolution method (14). Tissue blood flow maps were generated pixel by pixel by using in-house image processing software. Blood volume maps were calculated by means of numeric integration and gamma-variate fitting (23,24). Voxels with baseline values less than six times the magnitude of the SD of background noise were excluded from these calculations.

No attempt was made to normalize the data to the actual concentration of contrast agent to map absolute blood flow and blood volume in breast tissue. Thus, data reported here represent relative values of tumor flows and volumes with respect to those of normal breast parenchyma. To measure the relevant parameter (flow and volume), the location of the regions of interest was guided (L.G., J.P.D.) by using images with high spatial resolution and drawn around the tumor in each of the planes that tumorous tissue appeared. Visible necrotic areas within the tumors were excluded (two of six cases). The size of the regions of interest varied between six and 62 pixels. Regions of interest ranging from six to 65 pixels were placed over "normal" ipsilateral and contralateral breast tissue. After the relative values were determined, the mean values for the entire tumor and normal breast tissue were calculated by averaging the individual measurements in the planes involved. The results were compared with those reported in the literature (25–27).

Tissue samples were collected in 11 of 13 subjects, with core needle biopsy (n = 2), incisional biopsy (n = 6), and fine needle aspiration (n = 3). For two subjects with negative high-spatial-resolution and dynamic contrast-enhanced MR images, follow-up was recommended. Of 13 lesions found at histologic examination, six were invasive ductal carcinomas; one, lobular carcinoma in situ; five, cysts; and one, ductal papilloma. In addition, two tissue samples exhibited fibrocystic changes: atypical ductal hyperplasia in one and focal adenosis with chronic inflammatory changes in the other. The tumor size varied between 0.4 and 2.0 cm. Results of estrogen- and progestin-receptor status were available for six subjects, and the tests for each receptor were positive in four cases. All 13 subjects were included in the study.

Statistical analysis was performed by using a one-way analysis of variance, with the F test applied to pairwise comparisons between groups. Linear regression was used to determine if increasing values of relative tissue blood flow (rTBF) were positively or negatively correlated with increasing values of relative tissue blood volume (rTBV) and EF.


   Results
TOP
 ABSTRACT
 INTRODUCTION
 Materials and Methods
 Results
Discussion
 REFERENCES
 
T1 Measurements
T1 values were determined as described previously (11). Results of T1 measurements in breast tissue before the administration of contrast agent showed variability. The mean values ± standard error of the mean are summarized in Table 1. The changes represent a decrease in the initial value of T1 that ranged from 44% to 63%; the largest changes were measured in tumors. With the repetition time in our protocol, the effect of T1 on {Delta}R2* measurements was minimized.


View this table:
[in this window]
[in a new window]
  		TABLE 1. Results of T1 Measurements before and after Administration of Contrast Material in Breast Tissue

 
Dynamic Contrast-enhanced Echo-planar Measurements
EF was calculated as described elsewhere (11), and the results are summarized in Table 2. The values of EF in invasive cancers (mean, 61.4 mL/min/100 g ± 9.0; range, 35.2–95.7 mL/min/100 g) were higher than those in benign lesions (mean, 18.8 mL/min/100 g ± 3.9; range, 5.4–37.0 mL/min/100 g); these results are in agreement with previous results (11,12). Four benign pathologic findings showed values of EF at or above our threshold value for suspect lesions of 25 mL/min/100 g, which is in accordance with previous results to minimize the number of missed tumors (11,12).


View this table:
[in this window]
[in a new window]
  		TABLE 2. Individual Results of Measurements from Dynamic Contrast-enhanced and Magnetic Susceptibility Perfusion MR Imaging in Breast Tissues

 
One of the highest values in the group of benign lesions was associated with tissue at the periphery of a cyst near a high-grade invasive cancer. For normal breast tissue, the values of EF were low (mean, 4.7 mL/min/100 g ± 0.6; range, 2.0–10.3 mL/min/100 g). Results with analysis of variance showed significant differences between the three groups (P < .001).

Magnetic Susceptibility–based Contrast-enhanced Echo-planar Measurements
Typical AIFs at different locations in the breast are illustrated in Figure 1. The AIFs have a similar profile (dispersion), and only a slight delay (±2 seconds) in the peak of the AIF is noticed between distant planes. Nevertheless, to minimize errors in the estimation of tissue blood flow maps, an AIF in the plane of interest (ie, at the location of the tumor) was chosen whenever possible. In the cases studied, no measurable delays in the arrival of contrast agent were observed between two or three contiguous planes.



View larger version (35K):
[in this window]
[in a new window]
  		Figure 1. Time course of {Delta}R2* in several mammary vessels located in planes 5, 11, and 3 through the breast in a 46-year-old woman with invasive ductal carcinoma. The maximum amplitude of the curves was normalized to 1 for comparison of the AIF shape. The time course attributed to a vein in plane 3 (solid line) shows a delay with respect to that observed in arteries at planes 5 (dashed line) and 11 (dashed and dotted line).

 
Figure 2b shows the AIF and tissue {Delta}R2* time curves around a mammary artery and selected areas that correspond to normal glandular breast tissue and an invasive ductal carcinoma. In glandular breast tissue, a small transient loss of signal is observed after the first pass of the contrast agent. Conversely, the time course of {Delta}R2* in invasive ductal carcinoma shows a large and sustained transient loss of signal after the first pass of the bolus and in many voxels within the region of the tumor. The signal did not return to baseline during the observation time in this experiment (~2 minutes).



View larger version (89K):
[in this window]
[in a new window]
  		Figure 2a. Invasive ductal carcinoma in a 46-year-old subject. (a) High-spatial-resolution MR image of the breast was acquired with a three-dimensional gradient-echo sequence (23/5, 30° flip angle) at 6 minutes after the first administration of contrast agent. FV = feeding vessel, NGT = normal glandular tissue, T = tumor. (b) Time courses of {Delta}R2* around a feeding vessel to the tumor (a, large and rapid change) and selected areas that correspond to invasive ductal carcinoma (b) and normal glandular tissue (c, very small change) at the locations indicated in a.

 


View larger version (25K):
[in this window]
[in a new window]
  		Figure 2b. Invasive ductal carcinoma in a 46-year-old subject. (a) High-spatial-resolution MR image of the breast was acquired with a three-dimensional gradient-echo sequence (23/5, 30° flip angle) at 6 minutes after the first administration of contrast agent. FV = feeding vessel, NGT = normal glandular tissue, T = tumor. (b) Time courses of {Delta}R2* around a feeding vessel to the tumor (a, large and rapid change) and selected areas that correspond to invasive ductal carcinoma (b) and normal glandular tissue (c, very small change) at the locations indicated in a.

 
Figure 3 shows the rTBF and rTBV maps for the same subject as in Figure 2. Increased blood flow and blood volume ratios were observed in the region that corresponds to the area of enhancement on the three-dimensional high-spatial-resolution images. Individual results for rTBF and rTBV are summarized in Table 2 and plotted in Figure 4. For invasive cancers, the mean rTBF was 5.1 ± 1.3, which is significantly higher than that for benign diseases (mean, 1.0 ± 0.2; P < .002). Similarly, rTBV values in tumors (mean, 7.3 ± 1.1) were higher than those in benign diseases (mean, 1.2 ± 0.2). The difference between rTBV values in tumors and in nonmalignant tissues was significant (P < .001).



View larger version (76K):
[in this window]
[in a new window]
  		Figure 3a. High-spatial-resolution three-dimensional MR images (23/5, 30° flip angle) obtained in the same subject as in Figure 2 depict (a) EF, (b) rTBF, and (c) rTBV. High values of rTBF and rTBV are measured in the region that shows enhancement and high values of EF (50-110 mL/min/100 g). In b, feeding vessels to the tumor (fv T) and normal parenchyma (fv N) are visible in the right breast.

 


View larger version (82K):
[in this window]
[in a new window]
  		Figure 3b. High-spatial-resolution three-dimensional MR images (23/5, 30° flip angle) obtained in the same subject as in Figure 2 depict (a) EF, (b) rTBF, and (c) rTBV. High values of rTBF and rTBV are measured in the region that shows enhancement and high values of EF (50-110 mL/min/100 g). In b, feeding vessels to the tumor (fv T) and normal parenchyma (fv N) are visible in the right breast.

 


View larger version (72K):
[in this window]
[in a new window]
  		Figure 3c. High-spatial-resolution three-dimensional MR images (23/5, 30° flip angle) obtained in the same subject as in Figure 2 depict (a) EF, (b) rTBF, and (c) rTBV. High values of rTBF and rTBV are measured in the region that shows enhancement and high values of EF (50-110 mL/min/100 g). In b, feeding vessels to the tumor (fv T) and normal parenchyma (fv N) are visible in the right breast.

 


View larger version (21K):
[in this window]
[in a new window]
  		Figure 4. Mean rTBV plotted against mean rTBF in tumors and benign breast diseases. There is a clear separation between malignant and other breast lesions.

 
The wide range of rTBF values (from 2.3 to 10.5) and rTBV values (from 3.9 to 11.0) in tumors reflects their heterogeneity (28). All cancers in this study were invasive ductal carcinomas grade 1-3/III associated with ductal carcinoma in situ, according to the modified Bloom-Richardson classification.

Mean values of rTBF and rTBV for all benign diseases were not widely dispersed. In one case, the value of rTBV in the lesion was twice the magnitude of the mean in this group. This value was associated with tissue at the periphery of a cyst near an invasive ductal carcinoma grade 2/III with high-grade ductal carcinoma in situ. Also, it exhibited a high value of EF. The mean values of left to right flow and volume ratios for normal breast parenchyma were low, 1.2 ± 0.1 and 1.1 ± 0.1, respectively.

The comparison of EF with rTBF measurements is illustrated in Figure 5. For all lesions, increasing values of EF and rTBF were correlated (P < .05).



View larger version (18K):
[in this window]
[in a new window]
  		Figure 5. EF plotted against mean rTBF for all breast diseases (r2 = 0.28, P < .05). The correlation between EF and rTBF for benign lesions ({circ}) is positive (r2 = 0.66, P < .008). For breast tumors ({bullet}), EF and rTBF are not correlated (r2 = 0.08, P = .58), which indicates that contrast agent flux in breast tumors is limited by flow and vascular permeability and surface area or by only vascular permeability and surface area.

 

   Discussion
TOP
 ABSTRACT
 INTRODUCTION
 Materials and Methods
 Results
Discussion
REFERENCES
 
We have confirmed findings in previous reports that indicated T2* effects associated with the first pass of a contrast agent in the breast are detectable when the bolus is administered 10–15 minutes after the first bolus is administered (15–17). A decrease in signal intensity was observed approximately 15 seconds after injection. In those studies, a gradient-echo sequence was used to acquire data through a single section in the breast. We used an asymmetric spin-echo echo-planar sequence (29) for two reasons. First, echo-planar imaging allows data collection over the entire breast with sufficient temporal resolution for adequate sampling of the AIF and tissue residue function (30).

Second, fluctuations associated with large vessels and susceptibility-induced distortions in the breast are reduced, while magnetic-susceptibility contrast enhancement in the breast parenchyma is retained. In addition, by performing the T2*-weighted study during the second administration of contrast agent with a repetition time of 2 seconds, the signal enhancement due to the shortening of T1 associated with the leakage of contrast agent to the extravascular space is minimized. This T1 effect could be substantial in breast tissue and, in fact, the T2* effect is not observed when the contrast agent is administered once.

Our results show that it is feasible to characterize the microcirculation of breast tumors by using magnetic susceptibility–based contrast-enhanced echo-planar imaging. Absolute quantitation of tissue blood flow would require accurate knowledge of the relationship between susceptibility contrast enhancement around the vessel of interest and the intravascular concentration of contrast medium with the hematocrit in the microvasculature. This may be difficult to accomplish in a clinical setting. Nevertheless, we measured rTBF rates and found that the overall ratio of tumor to normal tissue blood flow in the breast was 5.1. This finding is in good agreement with ratios obtained with positron emission tomography (PET), which ranged from 4.7 to 5.5 (25,26).

Accurate determination of the AIF is important to minimize errors in the estimation of tissue blood flow maps. Although the approach used in this study offered the advantage of being independent of the underlying microvascular structure of tissues, it is possible to underestimate blood flow if there is a wide dispersion and delay in the AIF before the contrast agent reaches the tissue in the region of interest. Thus, we attempted to evaluate the AIF as closely as possible to the region of interest in the breast. However, the vascular anatomy of the breast varies greatly among subjects (31), which poses some difficulties for identification and characterization of a suitable AIF. Results for this task could be improved by increasing the spatial resolution of the T2*-weighted images, coregistering the functional data sets with high-spatial-resolution images, and automating vessel detection.

After the administration of a contrast agent, a rapid decrease in signal intensity distinguishes image pixels in the vicinity of arteries from those near veins, which exhibit a slower and usually delayed response. Also, veins are easily identified in the high-spatial-resolution three-dimensional data sets because of their size and high contrast enhancement; therefore, they can be avoided when AIFs are determined. These differences may be exploited for the design of an operator-independent method to identify arteries in the breast. In addition, the estimation of blood flow in the breast may be improved by using vascular models that take flow heterogeneity into account (32–34).

In some cases, physiologic motion is another factor that contributes to a reduction in the {Delta}R2* effect observed in arteries (blurring) and, hence, in the profile of the AIF.

The ratio of tumor to normal tissue blood volume was 7.3, which was higher than the 1.3 obtained with PET (25). In our measurements, the vascular fraction in normal breast tissue may be underestimated because the {Delta}R2* effect is reduced by the leakage of contrast agent from the intravascular to the interstitial space. Extravasation of the contrast medium contributes to the reduction of {Delta}R2* by decreasing the concentration gradient between the two compartments and by further shortening the value of T1 in regions with low accumulation of contrast material during the time since the first injection.

We observed a T1 effect in some areas of normal parenchyma that was manifested by an increase in signal intensity above the baseline after the second injection of contrast agent. As a result, the values of flow and volume in normal parenchyma are generally low with large fluctuations. In tumors, however, high leakage of contrast material into the interstitium induces additional signal loss due to the effect of transverse relaxation rate equal to reciprocal of relaxation time (R2 = 1/T2), which leads to a slow recovery to baseline values. If R2 = R20 + K(c), where R20 is the native tissue relaxation, K is a constant, and c is the concentration of contrast agent, then the effect of dipole-dipole interactions on T2 is small at low concentrations. However, this is not the case in tumors after the first administration of contrast agent.

On the basis of changes in T1, the concentration of gadolinium-based contrast agent in tumors is approximately 0.35 mmol/L and is likely to increase after the second injection. Thus, if we consider only T2 effects and assume that the concentration of gadolinium-based contrast agent in tissue doubles during the first 2 minutes after the second injection and that the native T2 (1/R20) of the breast tissue is 40 msec, then the change in MR signal due to a direct effect on T2 would be approximately 20%. In those regions, the area under {Delta}R2*(t) curve was estimated from the gamma-fitted curves to minimize the overestimation of blood volume due to leakage and recirculation. Nevertheless, if we assume a blood volume fraction for pectoral muscle of 4% (35), then the normalized vascular fraction for normal breast tissue is 1.9%, which is in agreement with previous measurements of blood volume fraction in the breast (36). The calculated tumor vascular fraction is 10.6%, which is in agreement with blood volume estimates in mammary and other solid tumors (3,35,36).

Our results show that increasing rTBF is strongly correlated with increasing rTBV in breast diseases (P < .001). The combination of blood flow and blood volume measurements clearly distinguishes the group of malignant lesions from nonmalignant lesions and normal breast tissue.

EF represents a measure of microvascular flow, permeability, and surface area. In only those cases where the transport of contrast agent between the intravascular and extravascular-extracellular space is flow limited, EF will be equal to flow. Our results show that EF is positively correlated with rTBF in breast disease (P < .05). Analysis of the results for benign lesions also shows a positive correlation (P < .008), but for the group of tumors, no correlation was observed (P = .58). Therefore, these results suggest that the transport of contrast agent exhibited an intermediate regime controlled by flow and vascular permeability and surface area or by only vascular permeability and surface area. Studies of tumor models might help elucidate the mechanisms that govern the relationship between these hemodynamic parameters.

The scope of this study was limited by the few subjects (n = 13) and, therefore, the small number of cancers investigated. Thus, appropriate validation of the approach outlined here is required. Nevertheless, the preliminary results show promise toward achieving the goal of quantification of microcirculation in breast tumors in a noninvasive manner without ionizing radiation. The independent measurement of blood flow and blood volume might prove useful to assess the hemodynamic changes that occur in the breast during and after radiation and chemotherapeutic interventions.

In summary, magnetic susceptibility-based contrast-enhanced echo-planar methods can be used to assess regional blood flow and blood volume in the entire breast. Preliminary results show high values of rTBF and rTBV in tumors with respect to glandular breast tissue. The ratio of blood flows between malignant and normal breast tissue is in good agreement with values in the literature that were obtained with other methods.


   ACKNOWLEDGMENTS
 
The authors thank Mary T. Foley for assistance with data acquisition, James M. Vevea for assistance with computer programming, and Dr Timothy Reese and Ms Ona Wu for helpful discussions related to echo-planar data acquisition and processing.


   FOOTNOTES
 
Abbreviations: AIF = arterial input function, EF = extraction-flow product, rTBF = relative tissue blood flow, rTBV = relative tissue blood volume, R1 = transverse relaxation rate equal to reciprocal of relaxation time (R1 = 1/T1), {Delta}R1 = change in R1, R2 = longitudinal relaxation rate equal to reciprocal of relaxation time (R2 = 1/T2), R2* = transverse relaxation rate equal to reciprocal of magnetic susceptibility–weighted relaxation time (R2* = 1/T2*), {Delta}R2* = change in R2*

Author contributions: Guarantors of integrity of entire study, P.J.S., D.B.K., L.G.; study concepts, L.G.; study design, P.J.S., L.G.; literature research, J.P.D., L.G.; clinical studies, P.J.S., E.D.Y.; data acquisition, J.P.D.; data analysis/interpretation, J.P.D., L.G.; statistical analysis, J.P.D., L.G.; manuscript preparation, J.P.D., L.G.; manuscript definition of intellectual content, L.G.; manuscript editing, P.J.S., L.G.; manuscript revision/review and final version approval, all authors.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 Materials and Methods
 Results
Discussion
 REFERENCES
 

  1. Dvorak HF, Nagy JA, Dvorak JT, Dvorak AM. Identification and characterization of the blood vessels of solid tumors that are leaky to circulating macromolecules. Am J Pathol 1988; 133:95-109.
  2. Dvorak HF, Brown LF, Detmar M, Dvorak AM. Vascular permeability factor/vascular endothelial growth factor, microvascular hyperpermeability, and angiogenesis. Am J Pathol 1995; 146:1029-1039.
  3. Jain RK. Determinants of tumor blood flow: a review. Cancer Res 1988; 48:2641-2658.
  4. Jain RK. Delivery of molecules, particles and cells to solid tumors. Ann Biomed Eng 1996; 24:457-473.
  5. Heywang S, Hahn D, Schmidt H, et al. Magnetic resonance imaging of the breast using Gd-DTPA. J Comput Assist Tomogr 1986; 10:199-204.
  6. Kaiser WA, Zeiler E. MR imaging of the breast: fast imaging sequences with and without Gd-DTPA. Radiology 1989; 170:681-686.
  7. Pierce WB, Harms SE, Flamig DP, Griffey RH, Evans WP, Hagans JE. Three-dimensional gadolinium-enhanced MR imaging of the breast: pulse sequence with fat suppression and magnetization transfer contrast. Radiology 1991; 181:757-763.
  8. Hoffman U, Brix G, Knopp MV, He T, Lorenz WJ. Pharmacokinetic mapping of the breast: a new method for dynamic MR mammography. Magn Reson Med 1995; 33:506-514.
  9. Tofts PS, Berkowitz B, Schnall MD. Quantitative analysis of dynamic Gd-DTPA enhancement in breast tumors using a permeability model. Magn Reson Med 1995; 33:564-568.
  10. Port RE, Knopp MV, Hoffmann U, Milker-Zabel S, Brix G. Multicompartment analysis of gadolinium chelate kinetics: blood-tissue exchange in mammary tumors as monitored by dynamic MR imaging. J Magn Reson Imaging 1999; 10:233-241.
  11. Hulka CA, Smith BL, Sgroi DC, et al. Benign and malignant breast lesions: differentiation with echo-planar MR imaging. Radiology 1995; 197:33-38.
  12. 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.
  13. Rosen BR, Aronen HJ, Cohen MS, et al. Diffusion and perfusion fast scanning in brain tumors. Neuroimaging Clin North Am 1993; 3:631-648.
  14. Ostergaard L, Weisskoff RM, Chesler DA, Gyldensted C, Rosen BR. High resolution measurement of cerebral blood flow using intravascular tracer bolus passages. I. Mathematical approach and statistical analysis. Magn Reson Med 1996; 36:715-725.
  15. Kuhl CK, Bieling H, Gieseke J, et al. Breast neoplasms: T2* susceptibility-contrast first-pass perfusion MR imaging. Radiology 1997; 202:87-95.
  16. Kvistad KA, Lundgren S, Fjosne HE, Smenes E, Smethurst HB, Haraldseth O. Differentiating between benign and malignant breast lesions with T2*-weighted first pass perfusion imaging. Acta Radiol 1999; 40:45-51.
  17. Kvistad KA, Rydland J, Vainia J, et al. Breast lesions: evaluation with dynamic contrast-enhanced T1-weighted MR imaging and with T2*-weighted first-pass perfusion MR imaging. Radiology 2000; 216:545-553.
  18. Zhu X, Hawnaur JM, Stringfellow J, et al. Quantification of relative blood volume and endothelium permeability of breast neoplasm using dynamic MR imaging (abstr) In: Proceedings of the Seventh Meeting of the International Society for Magnetic Resonance in Medicine. Berkeley, Calif: International Society for Magnetic Resonance in Medicine, 1999; 1076.
  19. Villringer A, Rosen RB, Belliveau JW, et al. Dynamic imaging with lanthanide chelates in normal brain: contrast due to magnetic susceptibility effects. Magn Reson Med 1998; 6:164-174.
  20. Weisskoff RM, Zuo CS, Boxerman JL, Rosen BR. Microscopic susceptibility variation and transverse relaxation: theory and experiment. Magn Reson Med 1994; 31:601-610.
  21. Fisel CR, Ackerman JL, Buxton RB, et al. MR contrast due to microscopically heterogeneous magnetic susceptibility: numerical simulations and applications to cerebral physiology. Magn Reson Med 1991; 17:348-356.
  22. Porkka L, Neuder M, Hunter G, Weisskoff RM, Belliveau JW, Rosen BR. Arterial input function measurements with MRI (abstr) In: Book of abstracts: Society of Magnetic Resonance in Medicine 1991. Berkeley, Calif: Society of Magnetic Resonance in Medicine, 1991; 120.
  23. Rosen BR, Belliveau JW, Vevea JM, Brady TJ. Perfusion imaging with NMR contrast agents. Magn Reson Med 1990; 14:249-265.
  24. Rosen BR, Belliveau JW, Buchbinder BR, et al. Contrast agent and cerebral hemodynamics. Magn Reson Med 1991; 19:285-292.
  25. Beaney RP, Lammertsma AA, Jones T, McKenzie CG, Halnan KE. Positron emission tomography for in-vivo measurement of regional blood flow, oxygen utilization, and blood volume in patients with breast carcinoma. Lancet 1984; 1:131-134.
  26. Wilson CB, Snook DE, Dhokia B, et al. Quantitative measurement of monoclonal antibody distribution and blood flow using positron emission tomography and 124iodine in patients with breast cancer. Int J Cancer 1991; 47:344-347.
  27. Wilson CBJH, Lammertsma AA, McKenzie CG, Sikora K, Jones T. Measurements of blood flow and exchanging water space in breast tumors using positron emission tomography: a rapid and noninvasive dynamic method. Cancer Res 1992; 52:1592-1597.
  28. Vaupel P. Vascularization, blood flow, oxygenation, tissue pH, and bioenergetic status of human breast cancer. Adv Exp Med Biol 1997; 411:243-254.
  29. Baker JR, Hoppel BE, Stern CE, Kwong KK, Weisskoff RW, Rosen BR. Dynamic functional imaging of the complete human cortex using gradient-echo and asymmetric spin-echo echo-planar magnetic resonance imaging (abstr) In: Proceedings of the 12th Annual Scientific Meeting of the Society of Magnetic Resonance in Medicine. Berkeley, Calif: Society of Magnetic Resonance in Medicine, 1993; 1400.
  30. Henderson E, Rutt BK, Lee TY. Temporal sampling requirements for the tracer kinetics modeling of breast tissue. Magn Reson Imaging 1998; 16:1057-1073.
  31. Cunningham L. The anatomy of the arteries and veins of the breast. J Surg Oncol 1977; 9:71-85.
  32. King RB, Deussen A, Raymond GM, Bassingthwaighte JB. A vascular transport operator. Am J Physiol 1993; 265:H2196-H2208.
  33. King RB, Raymond GM, Bassingthwaighte JB. Modeling blood flow heterogeneity. Ann Biomed Eng 1996; 24:352-372.
  34. Kroll K, Wilke N, Jerosch HM, et al. Modeling regional myocardial flows from residue functions of an intravascular indicator. Am J Physiol 1996; 271:H1643-H1655.
  35. Brix G, Bahner ML, Hoffman U, Horvath A, Schreiber W. Regional blood flow, capillary permeability, and compartmental volumes: measurement with dynamic CT—initial experience. Radiology 1999; 210:269-276.
  36. Edmister WB. Improved quantification of breast cancer physiology using MRI. Thesis Cambridge: Massachusetts Institute of Technology, 1999.



This article has been cited by other articles:


Home page
 Am. J. Roentgenol.Home page
C. El Khoury, V. Servois, F. Thibault, A. Tardivon, L. Ollivier, M. Meunier, C. Allonier, and S. Neuenschwander
MR Quantification of the Washout Changes in Breast Tumors Under Preoperative Chemotherapy: Feasibility and Preliminary Results
Am. J. Roentgenol., May 1, 2005; 184(5): 1499 - 1504.
[Abstract] [Full Text] [PDF]

Home page
 RadiologyHome page
H. E. Daldrup-Link, J. Rydland, T. H. Helbich, A. Bjornerud, K. Turetschek, K. A. Kvistad, E. Kaindl, T. M. Link, K. Staudacher, D. Shames, et al.
Quantification of Breast Tumor Microvascular Permeability with Feruglose-enhanced MR Imaging: Initial Phase II Multicenter Trial
Radiology, December 1, 2003; 229(3): 885 - 892.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow All Versions of this Article:
2232010428v1
223/2/558    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 Delille, J.-P.
Right arrow Articles by Garrido, L.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Delille, J.-P.
Right arrow Articles by Garrido, L.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH
RADIOLOGY RADIOGRAPHICS RSNA JOURNALS ONLINE