BJR
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
First published online December 10, 2007
British Journal of Radiology (2008) 81, 120-128
© 2008 British Institute of Radiology
doi: 10.1259/bjr/98435332
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Services
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 RADJENOVIC, A
Right arrow Articles by SMITH, M A
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by RADJENOVIC, A
Right arrow Articles by SMITH, M A

Full paper

Measurement of pharmacokinetic parameters in histologically graded invasive breast tumours using dynamic contrast-enhanced MRI

A RADJENOVIC, MSc, PhD 1 B J DALL, FRCR 2 J P RIDGWAY, MSc, PhD 3 and M A SMITH, PhD, DSc, FInstP 4

1 Academic Unit of Medical Physics, University of Leeds, Leeds, Departments of 2 Radiology and 3 Medical Physics, Leeds Teaching Hospitals NHS Trust, Leeds and 4 Sheffield Hallam University, Sheffield, UK

Correspondence: Dr Aleksandra Radjenovic, Research Fellow, Academic Unit of Medical Physics, University of Leeds, Level 10, Worsley Building, Clarendon Way, Leeds LS2 9JT, UK. E-mail: sasha{at}medphysics.leeds.ac.uk


   Abstract
Top
 Abstract
Introduction
 Methods
 Results
 Discussion
 References
 
Dynamic contrast-enhanced MRI (DCE-MRI) has demonstrated high sensitivity for detection of breast cancer. Analysis of correlation between quantitative DCE-MRI findings and prognostic factors (such as histological tumour grade) is important for defining the role of this technique in the diagnosis of breast cancer as well as the monitoring of neoadjuvant therapies. This paper presents a practical clinical application of a quantitative pharmacokinetic model to study histologically confirmed and graded invasive human breast tumours. The hypothesis is that, given a documented difference in capillary permeability between benign and malignant breast tumours, a relationship between permeability-related DCE-MRI parameters and tumour aggressiveness persists within invasive breast carcinomas. In addition, it was hypothesized that pharmacokinetic parameters may demonstrate stronger correlation with prognostic factors than the more conventional black-box techniques, so a comparison was undertaken. Significant correlations were found between pharmacokinetic and black-box parameters in 59 invasive breast carcinomas. However, statistically significant variation with tumour grade was demonstrated in only two permeability-related pharmacokinetic parameters: kep (p<0.05) and Ktrans (p<0.05), using one-way analysis of variance. Parameters kep and Ktrans were significantly higher in Grade 3 tumours than in low-grade tumours. None of the measured DCE-MRI parameters varied significantly between Grade 1 and Grade 2 tumours. Measurement of kep and Ktrans might therefore be used to monitor the effectiveness of neoadjuvant treatment of high-grade invasive breast carcinomas, but is unlikely to demonstrate remission in low-grade tumours.


   Introduction
Top
 Abstract
 Introduction
Methods
 Results
 Discussion
 References
 
The blood circulation at the capillary level, or microcirculation, is determined by the metabolic activity of the tissue. In pathological processes (such as tumour genesis), the microcirculation becomes altered. There can be an increase in microvascular density resulting from the growth of new capillary networks (angiogenesis) as well as vasodilatation of existing vessels. With the relatively recent Federal Drug Administration's approval of drugs to target specifically angiogenesis, there is likely to be a requirement to monitor, non-invasively, the levels of angiogenic activity. Compartmental modelling using an MRI contrast agent, gadopentetate dimeglumine (Gd-DTPA), and dynamic MRI acquisition offer the opportunity to investigate non-invasively and quantitatively the associated pharmacokinetics and hence the degree of angiogenic activity.

Gd-DTPA is an extracellular contrast agent that selectively alters the magnetic resonance signal intensity throughout its distribution volume, which consists of plasma and extravascular extracellular fluid. Physiological parameters that determine tissue microcirculation have a direct influence on the resulting local bulk tissue concentration of Gd-DTPA following intravenous administration. It is therefore possible to monitor the pathophysiological status of tissues by measuring the temporal variation of the MR signal, and qualitative information can be obtained from viewing the changes in image contrast. More importantly, it is also possible to obtain quantitative information associated with angiogenesis by mathematical analysis of dynamic contrast-enhanced MRI (DCE-MRI). The investigation of angiogenesis using DCE-MRI techniques can be divided into two fundamentally different groups: the so-called black-box methods and the more complex pharmacokinetic methods.

In black-box methods, the effect of Gd-DTPA is quantified in terms of heuristic, descriptive parameters describing the degree and the time course of enhancement [1–5]. These black-box parameters include maximal enhancement (ME), initial rate of enhancement (IRE), time to peak (TTP) and wash-out slope (WOS). Arguably, this method of analysis does not utilise optimally the available data as information from only selected parts of the dynamic curves are used. Furthermore, it is not possible to correlate findings obtained by different pulse sequences or to compare parameters measured in different centres. In quantifying the extent of Gd-DTPA-induced contrast enhancement, no presumptions are made about the underlying physical or physiological processes. Although these parameters are certainly related to the physiological parameters that govern tissue microcirculation, the form of this relationship is not considered.

In contrast, the pharmacokinetic methods for quantitative analysis of DCE-MRI provide a framework that can be used to link the physics of the MRI signal acquisition and the underlying pathophysiology that governs Gd-DTPA kinetics [6–9]. Pharmacokinetic (or compartmental) modelling of Gd-DTPA kinetics allows quantification of physiologically relevant parameters such as the volume of the extravascular extracellular space and capillary permeability. The development of methods for the quantification of DCE-MRI based on pharmacokinetic modelling has largely centred on cancer applications and the assessment of blood–brain barrier integrity. Within the context of pharmacokinetic modelling it is theoretically possible to separate the influence of physical and physiological parameters on the measured changes of signal intensity in DCE-MRI, thus enabling an assessment of physiological parameters that characterize pathological microcirculation.

Since its introduction into clinical practice by Heywang et al [10] in 1986, DCE-MRI has almost unequivocally demonstrated high sensitivity for detection of breast cancer [11]. The main limitation of DCE-MRI in the investigation of breast lesions lies in its low specificity, and the majority of studies in this field have centred on the design of methods for improving the distinction between malignant and benign breast lesions. The most basic criterion for the differentiation between benign and malignant lesions is the presence or absence of enhancement; this, however, yields a specificity of only 37% [12]. Particularly problematic is the differentiation between benign fibroadenomas, ductal carcinoma in situ (DCIS), and some of the less angiogenesis-dependent types of cancer (such as invasive lobular carcinomas [13]). Improvement in DCE-MRI specificity in breast cancer (to 75–85%) can be achieved by its integration with other diagnostic findings and the formulation of precise inclusion criteria [13, 14].

The first reports of pharmacokinetic analysis of DCE-MRI were published in 1990 and 1991 by three independent European research groups in Copenhagen [7], Heidelberg [6] and London [9]. They applied this technique to the assessment of the breakdown of the blood–brain barrier in multiple sclerosis [7, 9] and brain tumours [6, 7]. The potential of this approach for the assessment of microcirculatory properties of the tissues in a variety of other pathological states was quickly recognized. All subsequent models reported in the literature presented variations of these three principal models without radically changing the underlying methodology. Pharmacokinetic analysis of DCE-MRI was applied to the assessment of breast cancer [15–17], cervical cancer [18], colorectal cancer [19] and heart disease [20–22].

Although the three principal approaches rely on a common set of assumptions, they differ in the way the final formulation of the model-predicted tissue response curve is represented as a function of physiological parameters, and in the way these parameters are labelled and interpreted [23]. The key differences in the practical implementation of these models are in the treatment of the temporal variation of the Gd-DTPA concentration in plasma, the choice of input function (mode of injection) and the measurement of native (pre-contrast) longitudinal relaxation time T1.

Although DCE-MRI was initially applied to the assessment of brain lesions, it has subsequently been used in the evaluation of a variety of tumours, with the research into Gd-DTPA pharmacokinetics in breast tumours being particularly prominent. Pharmacokinetic analysis was applied in several clinical studies of DCE-MRI in breast lesions where the primary aim of the quantitative analysis was the differentiation between benign and malignant tumours. Significantly higher permeability-related quantifiers of DCE-MRI were reported in invasive breast carcinomas than in benign lesions, although a variable degree of overlap between these groups of lesions was also noted in all published studies, regardless of the choice of the analysis method [24–30]. A comparison between black-box and pharmacokinetic analysis of DCE-MRI has been only sporadically reported in the literature and the results of these comparisons are equivocal. Müller-Schimpfle et al [31], for example, found that the application of pharmacokinetic modelling did not result in the improvement in the discrimination between benign and malignant breast lesions when compared with black-box assessment. Hulka [27] and Mussurakis [24], on the other hand, reported that their pharmacokinetic parameters allowed a more specific classification of breast cancer lesions than black-box measurements (such as enhancement ratios, ME and WOS). Whereas Müller-Schimpfle used Brix's model [6] for the extraction of pharmacokinetic parameters, Hulka applied Larsson's method [7]; Mussurakis used both Brix's [6] and Tofts' methods [8, 9] and found them to be equivalent. Temporal resolution of DCE-MRI in the Müller-Schimpfle study was low (1 min) whereas Hulka and Mussurakis used DCE-MRI sets acquired with a markedly higher temporal resolution of 12 s and 6 s, respectively. The different conclusions reached in these studies regarding the comparative utility of pharmacokinetic and black-box methods are at least partly attributable to the differences in the DCE-MRI acquisition protocols.

Only a few studies have attempted to directly correlate DCE-MRI findings with prognostic factors such as tumour grade and nodal status in clinical studies of breast cancer [32–36]. None of these studies included pharmacokinetic analysis of DCE-MRI. Their results appear to be inconclusive and contradictory. Although Mussurakis et al [33] and Bone et al [35] found a significant correlation between DCE-MRI and prognostic factors, Fischer et al [34] and Stomper et al [32] found no correlation between them. Different acquisition and sampling protocols have been employed in each of these studies, as well as different methods for quantitative analysis of DCE-MRI. Furthermore, there was a considerable variation in the number of patients/lesions studied, their histological mix, the method used for grading as well as the choice of prognostic factors that DCE-MRI was compared with (tumour grade, nodal status, DNA S-phase percentage as well as various immunohistochemical prognostic indicators). The temporal resolution of DCE-MRI acquisitions used in these studies ranged from 12 s [33] to 7 min [35], with tissue coverage ranging from four targeted sagittal slices [33] to 64 transverse slices encompassing both breasts [35].

This paper presents a practical clinical application of a quantitative pharmacokinetic model [37] to study histologically confirmed and graded invasive human breast carcinomas and to investigate the capacity of pharmacokinetic measurements of permeability to reflect histological tumour grade and node status. The hypothesis was that, given a documented difference in capillary permeability between benign and malignant breast tumours, a relationship between permeability-related DCE-MRI parameters and tumour aggressiveness persists within invasive breast carcinomas. In addition, it was hypothesized that pharmacokinetic parameters may demonstrate a stronger correlation with prognostic factors than the more conventional black-box techniques; therefore, a comparison was undertaken.


   Methods
Top
 Abstract
 Introduction
 Methods
Results
 Discussion
 References
 
Pharmacokinetic model
After intravenous injection, Gd-DTPA is rapidly distributed throughout the plasma volume and extravasated into the extracellular space. There is evidence that no metabolic trapping of Gd-DTPA occurs within the body and that it is completely eliminated in an unchanged form by renal excretion [37]. Being a highly hydrophilic molecule, Gd-DTPA is unable to cross cellular membranes. In an open two-compartment model of Gd-DTPA kinetics, the extravasation of Gd-DTPA from the central (plasma) compartment is represented by a transfer constant Ktrans. The back flux of Gd-DTPA from the extravascular extracellular compartment into the plasma compartment is represented by a transfer constant kep = Ktrans/ve, where ve denotes the fractional volume of the extracellular extravascular (leakage) space. Fractional elimination rate kel represents the clearance of Gd-DTPA from plasma. Pharmacokinetic parameters Ktrans and kep therefore reflect the process of Gd-DTPA transfer across the capillary wall and are thus related to capillary permeability.

The pharmacokinetic modelling technique used in this paper combines the features of two earlier methods of Brix et al [6] and Tofts [8, 9]. This model [38] describes the temporal variation of contrast agent concentration in the tissue of interest Ct(t), as a function of two pharmacokinetic parameters: ve, and kep, as shown in Equation 1:



Formula 001

where ve is the fractional volume of extravascular, extra cellular fluid (unit-free fraction); kep is the fractional transfer rate (expressed in min); Formula is the fractional elimination rate of 0.058 min–1 quoted by Weinmann et al [37]; a1 and a2 were determined from published data [37] and have the following values: a1 = 3.99 kg l–1, a2 = 4.78 kg l–1 [9]; D is the injected dose of Gd-DTPA per kg body weight (D = 0.1 mmol kg–1); T is the effective duration of the infusion; {tau} = t for t≤T and {tau} = T for t>T.

For a spoiled gradient echo acquisition sequence, with repetition time TR, flip angle {alpha}, the following approximation can be used at low concentrations Ct(t) to represent temporal variation of normalized signal intensity following intravenous injection of Gd-DTPA:


Formula 003

where SI(t) is the signal intensity at time t; SI0 is the pre-injection signal intensity (i.e. t = 0); a = [(exp(–TR/T1))/(1 – exp(–TR/T1))] TR.

Pre-contrast longitudinal relaxation time (T1) can be measured or an assumed fixed value can be used. In our method we used a published value of 876 ms [39].

The pharmacokinetic parameter Ktrans (the transfer constant) is obtained from the product of kep and ve (i.e. kep x ve). The three conventional pharmacokinetic parameters now extracted and presented are ve, kep and Ktrans.

Clinical implementation of the model
There were a number of criteria for the imaging protocol: (i) complete bilateral coverage of both breasts was required because one of the principal clinical objectives was the detection of possible multifocality, (ii) DCE-MRI was required to yield images of diagnostic quality suitable for qualitative assessment by radiologists, and (iii) the duration of the DCE-MRI acquisition was required to be short in order to minimize problems related to gross patient motion and patient discomfort.

All imaging was performed on a 1.5 T MRI scanner (Gyroscan ACS NT, Philips Medical Systems, Best, The Netherlands). The MR signal detection was performed with a standard bilateral breast coil. The selection of the imaging volume was performed following the acquisition of survey scans in three orthogonal directions ensuring complete coverage of both breasts.

A two-dimensional (2D) multislice, T1 weighted spoiled gradient echo sequence was used (TR/TE/{varphi} = 213/4.6/90°, field of view (FOV) = 300 x 300 mm, 25 slices, 4 mm slice thickness, 12 dynamic scans with temporal resolution of {Delta}T = 32.5 second acquisition intervals, 154 x 256 image matrix, reconstructed to 256 x 256 matrix). The acquisition protocol was based on that proposed by Kuhl et al [5] and the total imaging time was 390 s. The patients were positioned prone with both breasts inside the breast coil. The imaging was performed in the transverse plane, with the imaging volume encompassing both breasts in all three dimensions.

A standard dose of 0.1 mmol per kilogram body weight of Gd-DTPA (Magnevist®, Schering, Berlin, Germany) was used. The Gd-DTPA injection was followed by a 10 ml saline flush. The duration of the contrast administration was Tinf = ({Delta}T)/2 and effective duration of the infusion (T), which was used for the modelling, was approximated by 2xTinf. This approximation was based on the data reported by Fritz-Hansen et al [20] and Andersen et al [40] for short peripheral injections. Therefore, the effective injection duration T was 32.5 s.

Figure 1Go represents a pair of pre- and post-contrast images and a resulting subtraction image of the transverse slice cutting through the centre of the lesion. In routine clinical practice, the lesion is evaluated by placing a region of interest (ROI) on a subtraction image and displaying a SI/time curve on an MR console. The subtraction method is effective in delineating the extent of the lesion. However, these images are less suitable for the analysis of the internal architecture of the lesion and its relationship to the surrounding parenchyma.


Figure 1
View larger version (49K):
[in this window]
[in a new window]

  		Figure 1. (a) Pre-contrast, (b) post-contrast and (c) subtraction images derived from a DCE-MRI dataset.

 
To improve the visualization of the lesions, parametric maps of the black-box parameters ME, IRE and WOS were computed on a voxel-by-voxel basis and displayed superimposed on greyscale anatomical images (Figure 2Go). These parameters were extracted from dynamic curves derived from each voxel in the three-dimensional (3D) array using algorithms implemented in C programming language. A three-point moving window algorithm encompassing temporal segments of 65 s was used for measurement of ME and IRE. Parameter WOS was computed on a voxel-by-voxel basis by measuring the slope of the least-squares straight line through the fixed five-point window encompassing the last 130 s of the dynamic curves. Three resulting colour-coded images were interrogated simultaneously. No segmentation or motion correction was applied and a uniform colour-coding scheme was used in all studies. The computation of colour-coded parametric maps effectively condensed the information contained in the original DCE-MRI datasets. Following the visual inspection of parametric images, ROI selection was performed using an image processing package AnalyzeTM (Biomedical Imaging Resource; Mayo Foundation, Rochester, MN).


Figure 2
View larger version (56K):
[in this window]
[in a new window]

  		Figure 2. Parametric maps of the variables(a) ME, (b) IRE and (c) WOS corresponding to the images presented in Figure 1.

 
The three pharmacokinetic parameters (Ktrans, ve and kep) were calculated for each dynamic curve derived from a user-selected ROI. All processing was performed using a computer program for non-linear least-squares fitting employing the Levenberg–Marquardt algorithm adapted from Press et al [41]. The program was written in C programming language and run on a standard PC. The processing was performed in a single batch operation as no user input was required.

Patients
MRI examination of the breasts was performed in patients with breast lesions where conventional triple assessment (X-ray mammography, ultrasound and clinical examination) did not provide conclusive diagnosis and where further information about the extent of a known lesion and/or possible multifocality was being sought. The study was approved by the regional ethics review board and written informed consent was obtained from every patient. From the total of 255 consecutive patients who underwent the MRI examination, surgery was subsequently carried out in 66 cases. A full pathology report, including tumour grade and lymph node status, was available for 53 patients (60 lesions). Tumour grading was performed using the Nottingham Prognostic Index for primary breast cancer [42]. In one examination, quantitative analysis was not possible because of excessive patient motion.

Full DCE-MRI analysis was undertaken retrospectively in 59 lesions (in 52 patients). All patients were female with a median age of 55 years (ranging from 32 to 80 years). The lesions were classified according to their histological grade into three groups. 12 lesions were found to be Grade 1 tumours, 29 were Grade 2 and 18 were Grade 3 tumours. 30 lesions had negative node status and 29 were node positive. 44 lesions were classified as invasive ductal carcinomas not otherwise specified (NOS), 11 were invasive lobular carcinomas, 2 were invasive tubular carcinomas and 2 were invasive mucinous carcinomas. 34 out of 59 lesions had a significant in situ (DCIS) component. Table 1Go presents a summary of the pathology grading and lymph node status for the set of 59 evaluated lesions.


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

  		Table 1. Summary of histological status of breast cancer lesions

 
Following the inspection of parametric maps, the most representative (usually central) cross-section was identified by a trained radiologist and a single circular 16 voxel ROI was placed close to the lesion rim and away from the necrotic, central areas, if present (Figure 3Go). Figure 4Go illustrates dynamic curves extracted from two different lesions and the superimposed least-squares lines obtained after non-linear fitting of the experimental data to the pharmacokinetic model. The corresponding pharmacokinetic and black-box parameters are listed in Table 2Go.


Figure 3
View larger version (146K):
[in this window]
[in a new window]

  		Figure 3. The ROI illustrated is superimposed onto a colour map of variable IRE. ROI selection was based on simultaneous inspection of all three parametric maps in Figure 2.

 

Figure 4
View larger version (10K):
[in this window]
[in a new window]

  		Figure 4. Examples of dynamic curves from two ROIs derived from a(a) Grade 1 lesion and a (b) Grade 3 lesion. DCE-MRI parameters are listed in Table 2.

 

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

  		Table 2. Pharmacokinetic and"black-box" parameters measured in two different lesions (with DCE-MRI curves presented in Figure 4Go)

 
Statistical analysis
SPSS statistical software package (Version 13.0, SPSS, Chicago, IL) was used for statistical analysis. All statistical tests were performed at {alpha} = 0.05 confidence level.


   Results
Top
 Abstract
 Introduction
 Methods
 Results
Discussion
 References
 
The technique worked robustly in this group of patients, adding only 5 min on to the investigation time. Of all the tumours and patients studied the technique was unsuccessful on only one occasion (owing to excessive patient motion).

A summary of the black-box and pharmacokinetic parameters is presented in GoTables 3 and 4Go, respectively. The mean values of measured parameters and their standard deviations are listed for each of the three subgroups. A pattern can be seen with a number of parameters (IRE, WOS, Ktrans, kep) with the parameter value changing in a consistent manner when compared with tumour grade. However, when a comparison is made of the pharmacokinetic and black-box parameters for the three different tumour groups using one-way analysis of variance, statistically significant variation with tumour grade was detected only in Ktrans (p<0.005) and kep (p<0.05), although the parameter WOS approaches significance (p = 0.054).


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

  		Table 3. Summary of"black-box" variables

 

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

  		Table 4. Summary of pharmacokinetic variables

 
The summary of results of the post hoc analysis of the differences between individual groups of measurements is presented in Table 5Go. A least significant difference correction for multiple comparisons was used. Although there were no significant differences between Grade 1 and Grade 2 tumours, Grade 3 tumours were significantly different from Grade 1 and Grade 2 tumours with respect to kep and Ktrans.


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

  		Table 5. Significance of the difference of pharmacokinetic variablesKtrans and kep between tumour grades

 
The correlation of the DCE-MRI parameters with tumour grade and with each other is listed in Table 6Go. There are significant correlations between tumour grade and WOS, kep and Ktrans (p<0.01) and IRE (p<0.05). Ktrans exhibited the highest degree of correlation with tumour grade (Spearman's {rho} = 0.473 p<0.0005 ). The data for Ktrans for each tumour group are plotted in Figure 5Go.


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

  		Table 6. Correlations between DCE-MRI parameters and tumour grade (all correlation coefficients are Pearson's {rho}, apart from those related to tumour grade, where Spearman's {rho} is listed instead)

 

Figure 5
View larger version (9K):
[in this window]
[in a new window]

  		Figure 5. Distribution ofKtrans values in three histological groups.

 
There was no significant association between DCE-MRI parameters and nodal status (Student's t-test, p>0.05). Furthermore, groups with and without a significant DCIS component also did not vary significantly (Student's t-test, p>0.05).


   Discussion
Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
References
 
In our study, the pharmacokinetic parameter Ktrans demonstrated a stronger relationship with tumour grade than the conventional black-box parameters, suggesting greater sensitivity to differences in microcirculation between different tumour grades.

The measurements obtained in this study are in good agreement with ve and Ktrans values in invasive breast carcinomas reported by Tofts et al [23] (Ktrans of 0.1–1.2 min–1 and ve of 0.3–0.8) [16], and den Boer et al [28] (Ktrans of 1.05±0.75 min–1 and ve of 0.47±0.20). Whereas Tofts did not measure T10, den Boer included a pre-contrast measurement of T10 in the pharmacokinetic analysis. Our measurements of Ktrans are somewhat higher than those obtained by Ikeda [30] (0.52±0.22 min–1) and Hulka et al [26, 27] (0.45±0.22 min–1) possibly as a result of different Cp(t) models. Both Ikeda et al [30] and Hulka et al [26, 27] have modelled Cp(t) as a three-exponential function. None of these studies, however, included measurements of ve and Ktrans in subgroups of invasive cancers defined by histological grade or nodal status. Furthermore, the proportion of high-grade tumours and tumours of different histological type will have influenced the mean values of Ktrans and ve measured in all these studies.

Prior to undertaking this study a comparable measurement of permeability in different histological grades of human breast cancer had not, to our knowledge, been reported in the literature. Our measurements are in broad agreement with permeability-related measurements in invasive breast carcinomas in humans reported elsewhere in studies involving an unspecified mix of histological grades and nodal involvement [16, 28]. However Furman-Haran et al [43] have recently demonstrated the capacity of high-resolution DCE-MRI to detect the differences in perfusion-related pharmacokinetic parameters between low-grade and high-grade invasive breast carcinomas [43].

Although it is not possible to trace all possible sources of discrepancy between the results presented in this study and other clinical studies where the relationship between tumour grade and black-box quantifiers of DCE-MRI was investigated, one probable source of variability lies in the different acquisition sensitivity to underlying T1 changes. The most T1-sensitive acquisition sequence was used by Stomper et al [32, 36]. However, their studies included only a small number of subjects, and the imaging volume encompassed only five contiguous slices. Fischer et al [34] conducted a large study but employed a suboptimal acquisition protocol, with respect to both temporal resolution (1.5 min) and T1 sensitivity. In two studies where simple enhancement ratios displayed significant association with tumour grade [33, 35] and nodal status [33], T1 sensitivity was somewhat higher than that achieved by our acquisition protocol. Their superior T1 sensitivity, however, was associated with the concomitant loss of spatial coverage [33] and temporal resolution [35]. The present study provided a compromise between the conflicting requirements for high temporal and spatial resolution, tissue coverage and T1 sensitivity, all of which are important for determining the utility of breast cancer DCE-MRI examinations.

Pharmacokinetic analysis of DCE-MRI was put forward as a tool for non-invasive monitoring of the effects of neoadjuvant chemotherapy in breast cancer and the reduction in Ktrans was associated with positive response to therapy [44, 45]. However, the reports presented in the literature to date are contradictory. Whereas Manton et al [46] report that pharmacokinetic parameters had no prognostic value, Padhani et al [47] found that the change in Ktrans was an accurate predictor of response.

In the present study, all lesions were evaluated by MRI before surgical excision without the administration of pre-surgical (neoadjuvant) chemotherapy. Successful neoadjuvant chemotherapy could be viewed as an effective downgrading of the tumour (e.g. from Grade 3 to Grade 2, or from Grade 2 to Grade 1). Therefore, our measurements of DCE-MRI pharmacokinetic parameters in graded primary breast carcinomas may offer an insight into the mechanisms involved in the monitoring of the effects of neoadjuvant chemotherapy by pharmacokinetic analysis of DCE-MRI.

There was a high degree of correlation between black-box and pharmacokinetic factors (Table 6Go). However, pharmacokinetic parameters Ktrans and kep exhibited the highest degree of correlation with tumour grade.

Furthermore, our results indicate that permeability-related pharmacokinetic parameters Ktrans and kep vary significantly between Grade 3 and Grade 2 tumours, whereas there is no significant difference between Grade 2 and Grade 1 tumours (Table 2Go). This suggests that pharmacological downgrading of Grade 3 tumours can be detected by measuring the changes in Ktrans and kep, and that further remission (from Grade 2 to Grade 1) will not result in significant change in Ktrans and kep.


   Acknowledgments
 
We would like to thank The Wellcome Trust for its financial support. We are grateful to Daina Dambitis and Sarah Bacon for their assistance with data collection.

Received for publication January 26, 2007. Revision received May 14, 2007. Accepted for publication May 16, 2007.


   References
Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 

  1. Heywang SH, Wolf A, Pruss E, Hilbertz T, Eiermann W, Permanetter W. MR imaging of the breast with Gd-DTPA: use and limitations. Radiology 1989;171:95–103.[Abstract/Free Full Text]
  2. Kaiser WA, Zeitler E. MR imaging of the breast – fast imaging sequences with and without Gd-DTPA – preliminary observations. Radiology 1989;170:681–6.[Abstract/Free Full Text]
  3. Gilles R, Guinebretiere JM, Lucidarme O, Cluzel P, Janaud G, Finet JF, et al. Nonpalpable breast tumors – diagnosis with contrast-enhanced subtraction dynamic MR imaging. Radiology 1994;191:625–31.[Abstract/Free Full Text]
  4. Boetes C, Barentsz JO, Mus RD, van der Sluis RF, van Erning L, Hendriks J, et al. MR characterization of suspicious breast lesions with a gadolinium enhanced Turboflash subtraction technique. Radiology 1994;193:777–81.[Abstract/Free Full Text]
  5. Kuhl CK, Mielcareck P, Klaschik S, Leutner C, Wardelmann E, Gieseke J, et al. Dynamic breast MR imaging: are signal intensity time course data useful for differential diagnosis of enhancing lesions? Radiology 1999;211:101–10.[Abstract/Free Full Text]
  6. Brix G, Semmler W, Port R, Schad LR, Layer G, Lorenz WJ. Pharmacokinetic parameters in CNS Gd-DTPA enhanced MR imaging. J Comput Assist Tomogr 1991;15:621–8.[Medline]
  7. Larsson HBW, Stubgaard M, Frederiksen JL, Jensen M, Henriksen O, Paulson OB. Quantitation of blood-brain barrier defect by magnetic resonance imaging and Gadolinium-DTPA in patients with multiple sclerosis and brain tumors. Magn Reson Med 1990;16:117–31.[Medline]
  8. Tofts PS. Modeling tracer kinetics in dynamic Gd-DTPA MR imaging. J Magn Reson Imaging 1997;7:91–101.[Medline]
  9. Tofts PS, Kermode AG. Measurement of the blood-brain barrier permeability and leakage space using dynamic MR imaging. 1. Fundamental concepts. Magn Reson Med 1991;17:357–67.[Medline]
  10. Heywang SH, Hahn D, Schmidt H, Krischke I, Eiermann W, Bassermann R, et al. MR imaging of the breast using gadolinium-DTPA. J Comput Assist Tomogr 1986;10:199–204.[Medline]
  11. Kuhl CK, Schild HH. Dynamic image interpretation of MRI of the breast. J Magn Reson Imaging 2000;12:965–74.[CrossRef][Medline]
  12. Harms SE, Flamig DP, Hesley KL, Meiches MD, Jensen RA, Evans WP, et al. MR imaging of the breast with rotating delivery of excitation off resonance – clinical experience with pathological correlation. Radiology 1993;187:493–501.[Abstract/Free Full Text]
  13. Kuhl CK. MRI of breast tumors. Eur Radiol 2000;10:46–58.[CrossRef][Medline]
  14. Heywang Kobrunner SH, Viehweg P, Heinig A, Kuchler C. Contrast-enhanced MRI of the breast: accuracy, value, controversies, solutions. Eur J Radiol 1997;24:94–108.[CrossRef][Medline]
  15. Hoffmann U, Brix G, Knopp MV, Hess T, Lorenz WJ. Pharmacokinetic mapping of the breast – a new method for dynamic MR mammography. Magn Reson Med 1995;33:506–14.[Medline]
  16. Tofts PS, Berkowitz B, Schnall MD. Quantitative analysis of dynamic Gd-DTPA enhancement in breast tumors using a permeability model. Magn Reson Med 1995;33:564–8.[Medline]
  17. 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;10:233–41.[CrossRef][Medline]
  18. Hawighorst H, Knapstein PG, Weikel W, Knopp MV, Zuna I, Knof A, et al. Angiogenesis of uterine cervical carcinoma: characterization by pharmacokinetic magnetic resonance parameters and histological microvessel density with correlation to lymphatic involvement. Cancer Res 1997;57:4777–86.[Abstract/Free Full Text]
  19. Müller-Schimpfle M, Brix G, Schlag P, Engenhart R, Frohmuller S, Hess T, et al. Recurrent rectal cancer – diagnosis with dynamic MR imaging. Radiology 1993;189:881–9.[Abstract/Free Full Text]
  20. Fritz-Hansen T, Rostrup E, Larsson HBW, Sondergaard L, Ring P, Henriksen O. Measurement of the arterial concentration of Gd-DTPA using MRI: a step toward quantitative perfusion imaging. Magn Reson Med 1996;36:225–31.[Medline]
  21. Larsson HBW, Stubgaard M, Sondergaard L, Henriksen O. In-vivo quantification of the unidirectional influx constant for Gd-DTPA diffusion across the myocardial capillaries with MR imaging. J Magn Reson Imaging 1994;4:433–40.[Medline]
  22. Larsson HBW, Fritz-Hansen T, Rostrup E, Sondergaard L, Ring P, Henriksen O. Myocardial perfusion modeling using MRI. Magn Reson Med 1996;35:716–26.[Medline]
  23. Tofts PS, Brix G, Buckley DL, Evelhoch JL, Henderson E, Knopp M, et al. Estimating kinetic parameters from dynamic contrast-enhanced T1-weighted MRI of a diffusable tracer: standardized quantities and symbols. J Magn Reson Imaging 1999;10:223–32.[CrossRef][Medline]
  24. Mussurakis S, Buckley DL, Drew PJ, Fox JN, Carleton PJ, Turnbull LW, et al. Dynamic MR imaging of the breast combined with analysis of contrast agent kinetics in the differentiation of primary breast tumours. Clin Radiol 1997;52:516–26.[CrossRef][Medline]
  25. Knopp MV, Weiss E, Sinn HP, Mattern J, Junkermann H, Radeleff J, et al. Pathophysiologic basis of contrast enhancement in breast tumors. J Magn Reson Imaging 1999;10:260–6.[CrossRef][Medline]
  26. Hulka CA, Edmister WB, Smith BL, Tan LJ, Sgroi DC, Campbell T, et al. Dynamic echo-planar imaging of the breast: experience in diagnosing breast carcinoma and correlation with tumor angiogenesis. Radiology 1997;205:837–42.[Abstract/Free Full Text]
  27. Hulka CA, Smith BL, Sgroi DC, Tan LJ, Edmister WB, Semple JP, et al. Benign and malignant breast lesions – differentiation with echo-planar MR imaging. Radiology 1995;197:33–8.[Abstract/Free Full Text]
  28. den Boer JA, Hoenderop RK, Smink J, Dornseiffen G, Koch PW, Mulder JH, et al. Pharmacokinetic analysis of Gd-DTPA enhancement in dynamic three-dimensional MRI of breast lesions. J Magn Reson Imaging 1997;7:702–15.[Medline]
  29. Daniel BL, Yen YF, Glover GH, Ikeda DM, Birdwell RL, Sawyer Glover AM, et al. Breast disease: dynamic spiral MR imaging. Radiology 1998;209:499–509.[Abstract/Free Full Text]
  30. Ikeda O, Yamashita Y, Takahashi M. Gd-enhanced dynamic magnetic resonance imaging of breast masses. Top Magn Reson Imaging 1999;10:143–51.[Medline]
  31. Müller-Schimpfle M, Ohmenhauser K, Sand J, Stoll P, Claussen CD. Dynamic 3D-MR mammography: is there a benefit of sophisticated evaluation for enhancement curves for clinical routine? J Magn Reson Imaging 1997;7:236–40.[Medline]
  32. Stomper PC, Penetrante RB, Edge SB, Arredondo MA, Blumenson LE, Stewart CC. Cellular proliferative activity of mammographic normal dense and fatty tissue determined by DNA S phase percentage. Breast Cancer Res Treat 1996;37:229–36.[CrossRef][Medline]
  33. 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;70:446–51.[Abstract]
  34. Fischer U, Kopka L, Brinck U, Korabiowska M, Schauer A, Grabbe E. Prognostic value of contrast-enhanced MR mammography in patients with breast cancer. Eur Radiol 1997;7:1002–5.[CrossRef][Medline]
  35. Bone B, Aspelin P, Bronge L, Veress B. Contrast-enhanced MR imaging as a prognostic indicator of breast cancer. Acta Radiol 1998;39:279–84.[Medline]
  36. Stomper PC, Herman S, Klippenstein DL, Winston JS, Edge SB, Arredondo MA, et al. Suspect breast lesions – findings at dynamic gadolinium enhanced MR imaging correlated with mammographic and pathological features. Radiology 1995;197:387–95.[Abstract/Free Full Text]
  37. Weinmann HJ, Laniado M, Mutzel W. Pharmacokinetics of Gd-DTPA dimeglumine after intravenous-injection into healthy volunteers. Physiol Chem Phys Med NMR 1984;16:167–72.[Medline]
  38. Radjenovic A, Ridgway JP, Smith MA. A method for pharmacokinetic modelling of dynamic contrast enhanced MRI studies of rapidly enhancing lesions acquired in a clinical setting. Phys Med Biol 2006;51:N187–N197.[CrossRef][Medline]
  39. Merchant TE, Thelissen GRP, Degraaf PW, Nieuwenhuizen C, Kievit HCE, Denotter W. Application of a mixed imaging sequence for MR imaging characterization of human breast disease. Acta Radiol 1993;34:356–61.[Medline]
  40. Andersen C, Taagehoj JF, Muhler A, Rehling M. Approximation of arterial input curve data in MRI estimation of cerebral blood-tumor-barrier leakage: comparison between Gd-DTPA and Tc-99m-DTPA input curves. Magn Reson Imaging 1996;14:235–41.[CrossRef][Medline]
  41. Press WH, Teukolsky SA, Vetterling WT, Flannery BP. Numerical recipes in C, 2nd edn. Cambridge, USA: Cambridge University Press; 1994.
  42. Galea MH, Blamey RW, Elston CE, Ellis IO. The Nottingham Prognostic Index in primary breast cancer. Breast Cancer Res Treat 1992;22:207–19.[CrossRef][Medline]
  43. Furman-Haran E, Schechtman E, Kelcz F, Kirshenbaum K, Degani H. Magnetic resonance imaging reveals functional diversity of the vasculature in benign and malignant breast lesions. Cancer 2005;104:708–18.[CrossRef][Medline]
  44. Knopp MV, von Tengg-Kobligk H, Choyke PL. Functional magnetic resonance imaging in oncology for diagnosis and therapy monitoring. Mol Cancer Ther 2003;2:419–26.[Abstract/Free Full Text]
  45. Padhani AR. Dynamic contrast-enhanced MRI in clinical oncology: current status and future directions. J Magn Reson Imaging 2002;16:407–22.[CrossRef][Medline]
  46. Manton DJ, Chaturvedi A, Hubbard A, Lind MJ, Lowry M, Maraveyas A, et al. Neoadjuvant chemotherapy in breast cancer: early response prediction with quantitative MR imaging and spectroscopy. Br J Cancer 2006;94:427–35.[CrossRef][Medline]
  47. Padhani AR, Hayes C, Assersohn L, Powles T, Makris A, Suckling J, et al. Prediction of clinicopathologic response of breast cancer to primary chemotherapy at contrast-enhanced MR imaging: Initial clinical results. Radiology 2006;239:361–74.[Abstract/Free Full Text]



This article has been cited by other articles:


Home page
 RadiologyHome page
M. Oostendorp, M. J. Post, and W. H. Backes
Vessel Growth and Function: Depiction with Contrast-enhanced MR Imaging
Radiology, May 1, 2009; 251(2): 317 - 335.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Services
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 RADJENOVIC, A
Right arrow Articles by SMITH, M A
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by RADJENOVIC, A
Right arrow Articles by SMITH, M A

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
BJR DMFR IMAGING  ALL BIR JOURNALS