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


     

Published online before print January 30, 2008, 10.1148/radiol.2463070298
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
2463070298v1
246/3/845    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 Roberts, C.
Right arrow Articles by Rushton, V. E.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Roberts, C.
Right arrow Articles by Rushton, V. E.
(Radiology 2008;246:845-853.)
© RSNA, 2008

Head and Neck Imaging

Glandular Function in Sjögren Syndrome: Assessment with Dynamic Contrast-enhanced MR Imaging and Tracer Kinetic Modeling—Initial Experience1

Caleb Roberts, BSc, Geoff J. M. Parker, PhD, Chris J. Rose, PhD, Yvonne Watson, DCR, James P. O'Connor, MA, MRCS, Stavros M. Stivaros, FRCR, Alan Jackson, FRCP, FRCR, PhD, and Vivian E. Rushton, MFGDP, PhD

1 From the Department of Imaging Science and Biomedical Engineering (C.R., G.J.M.P., C.J.R., Y.W., J.P.O., S.M.S., A.J.) and the School of Dentistry (V.E.R.), the University of Manchester, Stopford Bldg, Oxford Rd, Manchester M13 9PT, England. From the 2007 RSNA Annual Meeting. Received February 13, 2007; revision requested April 17; revision received May 31; accepted June 27; final version accepted August 27. Supported by the British Sjögren's Syndrome Society and the Sir Halley Stewart Trust. Address correspondence to C.R. (e-mail: caleb.roberts{at}manchester.ac.uk).


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
 IMPLICATIONS FOR PATIENT CARE
 References
 
Purpose: To prospectively use dynamic contrast material–enhanced magnetic resonance (MR) imaging and a tracer kinetic model to compare parotid gland microvascular characteristics in patients who have Sjögren syndrome (SS) with those in healthy volunteers.

Materials and Methods: The local research ethics committee approved the study, and written informed consent was obtained from all participants. Twenty-one patients (19 women, two men; age range, 31–73 years) with a diagnosis of SS and 11 healthy volunteers (10 women, one man; age range, 41–68 years) underwent three-dimensional T1-weighted dynamic contrast-enhanced MR imaging of the parotid gland at 1.5 T. A voxel-wise tracer kinetic model and a model-free analysis were applied to the dynamic MR data. Parameter medians and standard deviations were computed to summarize gland microvascular characteristics and gland heterogeneity, respectively. Differences were investigated by using multivariate analysis of variance, t, or U tests. Further investigation was performed by using linear discriminant and receiver operating characteristic analyses.

Results: Compared with the healthy volunteers, the patients with SS had highly significant elevations (P << .001) in the model-free parameter initial area under the curve and in tracer kinetic model parameters, including transcapillary contrast agent transfer constant (P < .001) and extracellular extravascular volume (P < .001). Gland heterogeneity was significantly greater (P < .001) in the patients with SS. Parameter medians and standard deviations enabled excellent differentiation (areas under receiver operating characteristic curve, 0.96 and 1.00, respectively) between the patients with SS and the healthy volunteers.

Conclusion: Dynamic contrast-enhanced MR imaging has the potential to be used in clinical settings to quantify microvascular function in SS and to differentiate between patients with and those without SS.

© RSNA, 2008


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
 IMPLICATIONS FOR PATIENT CARE
 References
 
Sjögren syndrome (SS) is a clinicopathologic syndrome characterized by dry eyes (keratoconjunctivitis sicca) and dry mouth (xerostomia) that result from autoimmune destruction of the lacrimal and salivary glands, respectively. It can occur as an isolated syndrome (primary SS) or in conjunction with associated autoimmune disorders (secondary SS). Histologic analysis typically reveals lymphocytic and plasma cell infiltration and loss of normal gland architecture. The disease affects predominantly women older than 40 years and causes substantial discomfort, with therapy limited to management of the symptoms.

Conventional diagnostic methods used to detect SS-associated structural glandular changes in the exocrine glands include gland biopsy (1,2), x-ray sialography (3), scintigraphy (4–6), and clinical criteria stipulated by means of international consensus (7,8). However, most of these diagnostic methods are invasive and may cause substantial complications for the patient (5). Noninvasive techniques such as ultrasonography (US) and magnetic resonance (MR) imaging have been used to reliably characterize and diagnose SS and offer promise as replacements for the earlier invasive and potentially harmful techniques (9–14). Current quantitative methods of measuring glandular structure and function include classification of the glandular parenchyma according to the size of nodules, ducts, and cavities (15,16); calculation of the standard deviation (SD) of the mean MR signal intensity in a portion of the gland (17); US texture analysis (11,16); and, more recently, assessment of functional glandular change with diffusion-weighted MR imaging (18) and dynamic MR sialography (13).

The vascular nature of the parotid gland makes it an ideal candidate organ for dynamic contrast material–enhanced MR imaging (ie, dynamic MR imaging) evaluation. Dynamic MR imaging has been used extensively in oncologic studies (19–21), in which the tumor microenvironment was described quantitatively in terms of the vascular plasma volume (vp), transcapillary contrast agent transfer constant (Ktrans), and extracellular extravascular volume (ve), with use of models of intravenously administered contrast agent kinetic parameters (22). These parameters have been investigated as potential biomarkers of disease status and treatment effectiveness in studies of solid tumors in humans, and the technique may yield detailed pathophysiologic information about tissue vascularity in nonmalignant diseases (23,24).

The superficial location of the parotid gland facilitates an excellent signal-to-noise ratio with use of the appropriate surface coil. However, to our knowledge, tracer kinetic modeling of dynamic MR data in patients with SS has not been previously reported. We hypothesized that owing to the increased vascular permeability, chronic inflammation, and microvessel density with SS, there would be differences in dynamic MR parameters between patients with SS and age-matched healthy volunteers. Thus, the purpose of our study was to prospectively use dynamic MR imaging and a tracer kinetic model to compare parotid gland microvascular characteristics in patients who have SS with those in healthy volunteers.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
 IMPLICATIONS FOR PATIENT CARE
 References
 
Participants
Twenty-one patients with SS (19 women, two men; mean age, 55 years; age range, 31–73 years) were recruited from the University of Manchester School of Dentistry. Patients were included if they had received a diagnosis of SS according to the criteria proposed by the American-European Consensus Group (7,8). Eleven age-matched healthy volunteers (10 women, one man; mean age, 56 years; age range, 41–68 years) were recruited and were included on the basis that no known abnormal parotid gland symptoms existed. Successful age matching was verified by establishing that the two subject samples (a) were normally distributed with use of the Kolmogorov-Smirnov test (P = .68, P > .99 for patients with SS and healthy volunteers, respectively), (b) had equal variances with use of the F test (P = .55), and (c) had equal mean values with use of the t test (P = .71). The research ethics committee of the University of Manchester and Central Manchester Hospitals approved the study, and written informed consent was obtained from all participants.

MR Imaging
MR imaging was performed by using a 1.5-T unit (Gyroscan NT/Intera; Philips Medical Systems, Best, the Netherlands). An 8-cm circular receive-only surface coil was placed over the most symptomatic parotid gland (determined by the patient and V.E.R., 18 years experience with SS) in the patients with SS and over the left or right parotid gland (chosen by V.E.R.) in the healthy volunteers to ensure a good signal-to-noise ratio, coverage of the entire gland, and inclusion of the carotid artery in the sensitive volume (Fig 1). After scout images were acquired, data encompassing the entire parotid gland of interest were acquired by using transverse T1-weighted spin-echo sequences with the following parameters: 500/20 (repetition time msec/echo time msec), 140 x 140-mm field of view, 4-mm section thickness, 4.4-mm intersection gap, 15 sections in the volume, 256 x 256 matrix, and two signals acquired. Transverse T2-weighted fast spin-echo images were also acquired, with the following parameters: 2300/90, 180 x 180-mm field of view, 2-mm section thickness, 2-mm intersection gap, 15 sections in the volume, 256 x 256 matrix, echo train length of six, and three signals acquired.


Figure 1A
View larger version (35K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1a: Glandular enhancement after bolus injection of gadodiamide (Gd) (Omniscan; GE Healthcare, Oslo, Norway) in (a) patient with SS and (b) healthy volunteer. Left: Transverse T2-weighted fast spin-echo MR images (2300/90), with region of interest (outlined) defined around parotid gland. Right: Corresponding graphs illustrate contrast agent uptake data (x) from the parotid gland fitted with a compartmental model (dark line) (see Equation in Materials and Methods), from which microvascular parameter values were calculated, and an arterial input function (faint line) obtained by using an automatic extraction method in the external carotid artery, which was also used in the fitting routine.

 

Figure 1B
View larger version (34K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1b: Glandular enhancement after bolus injection of gadodiamide (Gd) (Omniscan; GE Healthcare, Oslo, Norway) in (a) patient with SS and (b) healthy volunteer. Left: Transverse T2-weighted fast spin-echo MR images (2300/90), with region of interest (outlined) defined around parotid gland. Right: Corresponding graphs illustrate contrast agent uptake data (x) from the parotid gland fitted with a compartmental model (dark line) (see Equation in Materials and Methods), from which microvascular parameter values were calculated, and an arterial input function (faint line) obtained by using an automatic extraction method in the external carotid artery, which was also used in the fitting routine.

 
For the subsequent dynamic MR examination, 0.1 mmol gadodiamide per kilogram of body weight was administered manually through a cannula placed in the antecubital vein at a rate of 3 mL/sec and was immediately followed by a saline flush. Transverse dynamic MR imaging data were acquired by using a three-dimensional T1-weighted fast field-echo (spoiled gradient-echo) sequence (25) with the following parameters: 4.2/1.23, 180 x 180-mm field of view, 128 x 128 x 25 matrix, 4-mm section thickness, and 25 sections in the volume. Before the contrast agent administration, baseline longitudinal relaxation time (T10) values (in milliseconds) were calculated from three image acquisitions with different flip angles (2°, 10°, and 35°) and four acquired signals. The dynamic series involved use of the same three-dimensional fast field-echo acquisition as that used for the T10 measurement (with identical repetition time, echo time, field of view, and matrix size), a 35° flip angle, and one signal acquired, repeated over a period of approximately 5 minutes. A total of 50 volumes were acquired. The temporal resolution was 6.2 seconds.

Quantitative Characterization of Glandular Physiologic Characteristics
To help visualize the gland, two authors (V.E.R. and Y.W., 19 and 22 years experience in head and neck MR imaging, respectively) used the DispImage software package (University College London, London, England) (26) to manually outline the volume occupied by the gland on the 10th postcontrast dynamic series volume, with reference to heavily T1-weighted and T2-weighted anatomic images. The contrast agent uptake in the parotid gland was characterized and evaluated by using model-free and model-based approaches. The change in contrast agent concentration over time, Ct(t), was determined in each voxel in the gland, and the compartmental tracer kinetic model (27) described in the following equation was applied to each voxel by using an arterial input function, Cp(t), measured in each individual:

Formula
where t is the time (in minutes), t' is the time (in minutes) as an integration variable, and Cp(t') is the concentration of contrast agent in the blood plasma as a function of time. For each patient and volunteer, the arterial input function was determined by using an automated technique in the external carotid artery, as previously described (28) (Fig 1). Modeling yielded estimates of vp and ve, which are fractional volumes, and of Ktrans (in min–1). The model-free parameter initial area under the curve (IAUC60) (in milllimoles per kilogram per second) was calculated by using trapezoidal integration of the contrast agent concentration with time during the first 60 seconds after arrival of the contrast agent in the enhancing voxels of interest.

Parameter Estimation, Heterogeneity, and Statistical Analyses
Model fitting was performed by using a simplex algorithm incorporated into the in-house dynamic modeling software (Manchester Dynamic Modeling [MaDyM]). Time of contrast agent arrival in the gland was included as a variable in the fit. Enhancing voxels were identified in the gland volume of interest, and the model was fitted to these voxels' time series; three-dimensional T10 and dynamic MR parameter maps were also generated. Parameter medians were computed to summarize the data for each individual.

To use a noninvasive diagnostic method for patients with SS, Izumi et al (17) assessed the heterogeneity in the parotid gland in terms of the SD of the mean signal intensity on the T1-weighted image (T1-wSD). We applied this method to our heavily T1-weighted image data to determine the heterogeneity of the glands for each group. For each individual, a rectangular region of interest was defined in the same anatomic region of the gland, as previously described (17), and the T1-wSD was calculated. In the same manner and by using the same region of interest, we also calculated the SD of each dynamic MR parameter and of the T10 for each participant to investigate the sensitivity of dynamic MR imaging to gland heterogeneity.

Thus, after imaging and parameter estimation, the data for each gland were summarized by using two sets of variables: medians (IAUC60, Ktrans, ve, vp, and T10) and SDs (T1-wSD and SDs of mean T10, IAUC60, Ktrans, ve, and vp). The two sets of variables were considered separately in each of the following analyses: For each of the two sets of variables, differences between the SS and healthy volunteer groups were investigated by using three types of analysis (by C.J.R.). Multivariate analysis of variance was used to investigate differences with all variables in a set considered jointly. Independent two-tailed t tests—or U tests when distributional assumptions were violated—were used to investigate differences in each variable in turn. The capacity of each set to facilitate discrimination between the patients with SS and the healthy volunteers was investigated by using leave-one-out linear discriminant analysis, the results of which were summarized by using receiver operating characteristic (ROC) analysis (performed with class membership probabilities).

Distributional assumptions were investigated as follows: Normality (t tests) was assessed by using normal plots and Kolmogorov-Smirnov tests; joint normality (multivariate analysis of variance, linear discriminant analysis), by using marginal normal plots, marginal Kolmogorov-Smirnov tests, and scatterplots; equality of variances (t tests), by using F tests; and equality of covariance matrices (multivariate analysis of variance), by using the Box M test. The area under the ROC curve, a summary statistic of discriminant performance, was computed by using numeric integration, and confidence intervals were computed by using a previously described method (29). For all hypothesis testing, {alpha} = .05 defined the significance level. Because the participants were opportunistically recruited, post hoc power analyses were performed by using G*Power, version 3.0.3, software (Heinrich-Heine-Universität, Düsseldorf, Germany) (30). All other statistical analyses were performed by using Matlab R2006b (Mathworks, Natick, Mass).


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 ADVANCES IN KNOWLEDGE
 IMPLICATIONS FOR PATIENT CARE
 References
 
Dynamic MR Parameters
Tests for normality and equality of variances revealed no significant differences between the patients with SS and the healthy volunteers in terms of median parameter values, with the exception of T10, for which the equality of variances was significantly different (P = .03) between the two groups (Table). Box M test results indicated no significant difference in covariance matrices (P = .65). Multivariate analysis of variance revealed a significant joint difference in the mean values of the median kinetic parameters between the two populations (P < .001). Post hoc power analysis results showed this test to have 99% power and that 12 participants would be required to achieve 80% power. Post hoc power analyses of the individual parameters revealed that the power of tests for these parameters was close to or exceeded 80%, with the exception of the test for vp, which would require 60 participants to achieve at least 80% power (Table).


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

  		Results of Statistical Tests Performed on All Parameters

 
The distribution of individual IAUC60, Ktrans, and ve parameter values was significantly different (P < .001) between the two subject groups, with group median parameters elevated in the SS group (Fig 2). Although the SS group median vp was higher than the healthy volunteer value, the distribution of parameter values was not significantly different (Table). Scatterplots of the microvascular parameters and T10 (Fig 3) show the differences between the SS and healthy volunteer group values; note the differences between ve and Ktrans values and between ve and IAUC60 values. ROC analysis (Fig 4) revealed that approximately 70% of individuals with SS could be correctly identified, with no false-positive cases, and that for the median parameters, 100% sensitivity could be achieved at a specificity of 64%. The area under the ROC curve was 0.96 (95% confidence interval: 0.89, 1.00).


Figure 2A
View larger version (8K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2a: Box plots of median (a) T10 (in milliseconds), (b) IAUC60 (in millimoles per kilogram per second), (c) Ktrans (in min–1), (d) ve (fractional volume), and (e) vp (fractional volume) values for patients with SS and healthy volunteers (HV). Each box shows lower quartile, median (center line), and upper quartile of the parameter value; whiskers indicate the range of values. Notches represent estimates of the 95% confidence interval of the median value. Small dots outside box represent estimates of outliers.

 

Figure 2B
View larger version (6K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2b: Box plots of median (a) T10 (in milliseconds), (b) IAUC60 (in millimoles per kilogram per second), (c) Ktrans (in min–1), (d) ve (fractional volume), and (e) vp (fractional volume) values for patients with SS and healthy volunteers (HV). Each box shows lower quartile, median (center line), and upper quartile of the parameter value; whiskers indicate the range of values. Notches represent estimates of the 95% confidence interval of the median value. Small dots outside box represent estimates of outliers.

 

Figure 2C
View larger version (8K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2c: Box plots of median (a) T10 (in milliseconds), (b) IAUC60 (in millimoles per kilogram per second), (c) Ktrans (in min–1), (d) ve (fractional volume), and (e) vp (fractional volume) values for patients with SS and healthy volunteers (HV). Each box shows lower quartile, median (center line), and upper quartile of the parameter value; whiskers indicate the range of values. Notches represent estimates of the 95% confidence interval of the median value. Small dots outside box represent estimates of outliers.

 

Figure 2D
View larger version (8K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2d: Box plots of median (a) T10 (in milliseconds), (b) IAUC60 (in millimoles per kilogram per second), (c) Ktrans (in min–1), (d) ve (fractional volume), and (e) vp (fractional volume) values for patients with SS and healthy volunteers (HV). Each box shows lower quartile, median (center line), and upper quartile of the parameter value; whiskers indicate the range of values. Notches represent estimates of the 95% confidence interval of the median value. Small dots outside box represent estimates of outliers.

 

Figure 2E
View larger version (8K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2e: Box plots of median (a) T10 (in milliseconds), (b) IAUC60 (in millimoles per kilogram per second), (c) Ktrans (in min–1), (d) ve (fractional volume), and (e) vp (fractional volume) values for patients with SS and healthy volunteers (HV). Each box shows lower quartile, median (center line), and upper quartile of the parameter value; whiskers indicate the range of values. Notches represent estimates of the 95% confidence interval of the median value. Small dots outside box represent estimates of outliers.

 

Figure 3
View larger version (12K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3: Scatterplots show distributions of individual median IAUC60, Ktrans, ve, vp, and T10 values for patients with SS and healthy volunteers (HV).

 

Figure 4
View larger version (15K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4: ROC curves. Area under dark solid curve, representing median dynamic MR parameter estimates, is 0.96 (95% confidence interval: 0.86, 1.00). Area under dashed curve, representing SD of mean dynamic MR parameter estimates, is 1.00, indicating a perfect scenario, where data are definitively characterized as belonging to a particular group (in this case, SS or healthy volunteer group). Thin diagonal line represents worst-case scenario, where data are randomly classified as data for the SS group or those for the healthy volunteer group, and has an area under the curve of 0.50.

 
Heterogeneity
Testing for normality revealed no significant differences in SD parameter values between the patients with SS and the healthy volunteers. Testing for equality of variances revealed highly significant (P << .001) differences in T1-wSD and T10 SD and significant differences in IAUC60 (P = .004) and Ktrans (P = .002) SD values between the SS and healthy volunteer groups (Table). Box M testing revealed a significant (P < .001) difference in the covariance matrices between the two groups, so multivariate analysis of variance was not performed. Post hoc analysis of only the ve and vp SD parameters was performed; nonparametric testing of the other SD statistics was performed owing to violation of t test assumptions. Post hoc power analysis revealed that for both ve SD and vp SD, 14 individuals would be required to achieve a power of 80% or higher (Table). There was strong evidence that the distributions of SDs for all parameters were significantly (P < .001) different between the two groups, with group median SDs elevated in the SS group compared with those in the healthy volunteer group (Fig 5). Scatterplots of the parameter SDs (Fig 6) show the group differences. ROC analysis results (Fig 4) show that 100% sensitivity could be achieved at a specificity of 100%.


Figure 5A
View larger version (7K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 5a: Box plots for SDs of the mean dynamic tracer kinetic parameters (a) Ktrans, (b) ve, (c) vp, and (d) IAUC60; (e) for T10 SD; and for (f) T1-wSD in patients with SS and healthy volunteers (HV). Each box shows lower quartile, median (center line), and upper quartile of the parameter value; whiskers indicate the range of values. Notches represent estimates of the 95% confidence interval of the median value. Small dots outside box represent estimates of outliers.

 

Figure 5B
View larger version (7K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 5b: Box plots for SDs of the mean dynamic tracer kinetic parameters (a) Ktrans, (b) ve, (c) vp, and (d) IAUC60; (e) for T10 SD; and for (f) T1-wSD in patients with SS and healthy volunteers (HV). Each box shows lower quartile, median (center line), and upper quartile of the parameter value; whiskers indicate the range of values. Notches represent estimates of the 95% confidence interval of the median value. Small dots outside box represent estimates of outliers.

 

Figure 5C
View larger version (8K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 5c: Box plots for SDs of the mean dynamic tracer kinetic parameters (a) Ktrans, (b) ve, (c) vp, and (d) IAUC60; (e) for T10 SD; and for (f) T1-wSD in patients with SS and healthy volunteers (HV). Each box shows lower quartile, median (center line), and upper quartile of the parameter value; whiskers indicate the range of values. Notches represent estimates of the 95% confidence interval of the median value. Small dots outside box represent estimates of outliers.

 

Figure 5D
View larger version (7K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 5d: Box plots for SDs of the mean dynamic tracer kinetic parameters (a) Ktrans, (b) ve, (c) vp, and (d) IAUC60; (e) for T10 SD; and for (f) T1-wSD in patients with SS and healthy volunteers (HV). Each box shows lower quartile, median (center line), and upper quartile of the parameter value; whiskers indicate the range of values. Notches represent estimates of the 95% confidence interval of the median value. Small dots outside box represent estimates of outliers.

 

Figure 5E
View larger version (8K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 5e: Box plots for SDs of the mean dynamic tracer kinetic parameters (a) Ktrans, (b) ve, (c) vp, and (d) IAUC60; (e) for T10 SD; and for (f) T1-wSD in patients with SS and healthy volunteers (HV). Each box shows lower quartile, median (center line), and upper quartile of the parameter value; whiskers indicate the range of values. Notches represent estimates of the 95% confidence interval of the median value. Small dots outside box represent estimates of outliers.

 

Figure 5F
View larger version (7K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 5f: Box plots for SDs of the mean dynamic tracer kinetic parameters (a) Ktrans, (b) ve, (c) vp, and (d) IAUC60; (e) for T10 SD; and for (f) T1-wSD in patients with SS and healthy volunteers (HV). Each box shows lower quartile, median (center line), and upper quartile of the parameter value; whiskers indicate the range of values. Notches represent estimates of the 95% confidence interval of the median value. Small dots outside box represent estimates of outliers.

 

Figure 6
View larger version (15K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 6: Scatterplots show distributions of individual IAUC60, Ktrans, ve, vp, and T10 SD values and T1-wSD in patients with SS and healthy volunteers (HV).

 
Parameter maps constructed from compartmental models further illustrate the heterogeneity of the gland in terms of its vascular characteristics (Fig 7). In the examples in Figure 7, the parotid gland in the patient with SS is larger and generally more heterogeneous than the gland in the healthy volunteer. In particular, ve, an indication of inflammation, is largely homogeneous across the gland in the healthy volunteer, whereas there is a sporadic distribution of medium to high values across the gland in the patient with SS.


Figure 7A
View larger version (29K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 7a: Dynamic MR kinetic parameter maps overlaid on T10 maps calculated by using three-dimensional T1-weighted fast field-echo sequence (4.2/1.23; flip angles, 2°, 10°, and 35°) in (a, c, e, g) patient with SS and (b, d, f, h) healthy volunteer. Each image, constructed by using free medical image-viewing software (MRIcro for Linux and Windows, version 1.40 build 1; Chris Rorden, http://www.sph.sc.edu/comt/rorden/mricro), is shown in transverse (bottom left), coronal (top left), and sagittal (top right) orthogonal planes. Intersecting lines pinpoint the gland in each plane.

 

Figure 7B
View larger version (28K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 7b: Dynamic MR kinetic parameter maps overlaid on T10 maps calculated by using three-dimensional T1-weighted fast field-echo sequence (4.2/1.23; flip angles, 2°, 10°, and 35°) in (a, c, e, g) patient with SS and (b, d, f, h) healthy volunteer. Each image, constructed by using free medical image-viewing software (MRIcro for Linux and Windows, version 1.40 build 1; Chris Rorden, http://www.sph.sc.edu/comt/rorden/mricro), is shown in transverse (bottom left), coronal (top left), and sagittal (top right) orthogonal planes. Intersecting lines pinpoint the gland in each plane.

 

Figure 7C
View larger version (30K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 7c: Dynamic MR kinetic parameter maps overlaid on T10 maps calculated by using three-dimensional T1-weighted fast field-echo sequence (4.2/1.23; flip angles, 2°, 10°, and 35°) in (a, c, e, g) patient with SS and (b, d, f, h) healthy volunteer. Each image, constructed by using free medical image-viewing software (MRIcro for Linux and Windows, version 1.40 build 1; Chris Rorden, http://www.sph.sc.edu/comt/rorden/mricro), is shown in transverse (bottom left), coronal (top left), and sagittal (top right) orthogonal planes. Intersecting lines pinpoint the gland in each plane.

 

Figure 7D
View larger version (27K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 7d: Dynamic MR kinetic parameter maps overlaid on T10 maps calculated by using three-dimensional T1-weighted fast field-echo sequence (4.2/1.23; flip angles, 2°, 10°, and 35°) in (a, c, e, g) patient with SS and (b, d, f, h) healthy volunteer. Each image, constructed by using free medical image-viewing software (MRIcro for Linux and Windows, version 1.40 build 1; Chris Rorden, http://www.sph.sc.edu/comt/rorden/mricro), is shown in transverse (bottom left), coronal (top left), and sagittal (top right) orthogonal planes. Intersecting lines pinpoint the gland in each plane.

 

Figure 7E
View larger version (28K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 7e: Dynamic MR kinetic parameter maps overlaid on T10 maps calculated by using three-dimensional T1-weighted fast field-echo sequence (4.2/1.23; flip angles, 2°, 10°, and 35°) in (a, c, e, g) patient with SS and (b, d, f, h) healthy volunteer. Each image, constructed by using free medical image-viewing software (MRIcro for Linux and Windows, version 1.40 build 1; Chris Rorden, http://www.sph.sc.edu/comt/rorden/mricro), is shown in transverse (bottom left), coronal (top left), and sagittal (top right) orthogonal planes. Intersecting lines pinpoint the gland in each plane.

 

Figure 7F
View larger version (28K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 7f: Dynamic MR kinetic parameter maps overlaid on T10 maps calculated by using three-dimensional T1-weighted fast field-echo sequence (4.2/1.23; flip angles, 2°, 10°, and 35°) in (a, c, e, g) patient with SS and (b, d, f, h) healthy volunteer. Each image, constructed by using free medical image-viewing software (MRIcro for Linux and Windows, version 1.40 build 1; Chris Rorden, http://www.sph.sc.edu/comt/rorden/mricro), is shown in transverse (bottom left), coronal (top left), and sagittal (top right) orthogonal planes. Intersecting lines pinpoint the gland in each plane.

 

Figure 7G
View larger version (29K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 7g: Dynamic MR kinetic parameter maps overlaid on T10 maps calculated by using three-dimensional T1-weighted fast field-echo sequence (4.2/1.23; flip angles, 2°, 10°, and 35°) in (a, c, e, g) patient with SS and (b, d, f, h) healthy volunteer. Each image, constructed by using free medical image-viewing software (MRIcro for Linux and Windows, version 1.40 build 1; Chris Rorden, http://www.sph.sc.edu/comt/rorden/mricro), is shown in transverse (bottom left), coronal (top left), and sagittal (top right) orthogonal planes. Intersecting lines pinpoint the gland in each plane.

 

Figure 7H
View larger version (28K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 7h: Dynamic MR kinetic parameter maps overlaid on T10 maps calculated by using three-dimensional T1-weighted fast field-echo sequence (4.2/1.23; flip angles, 2°, 10°, and 35°) in (a, c, e, g) patient with SS and (b, d, f, h) healthy volunteer. Each image, constructed by using free medical image-viewing software (MRIcro for Linux and Windows, version 1.40 build 1; Chris Rorden, http://www.sph.sc.edu/comt/rorden/mricro), is shown in transverse (bottom left), coronal (top left), and sagittal (top right) orthogonal planes. Intersecting lines pinpoint the gland in each plane.

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
ADVANCES IN KNOWLEDGE
 IMPLICATIONS FOR PATIENT CARE
 References
 
To our knowledge, our study is the first in which dynamic contrast-enhanced MR imaging and tracer kinetic modeling are used to quantify the microvascular pathophysiologic features of SS; highly significant differences in dynamic MR tracer kinetic parameters were observed between the patients with SS and the healthy volunteers. Our study results demonstrate that dynamic MR tracer kinetic modeling parameters can enable quantification of the parotid gland microvascular characteristics related to the pathophysiologic features seen with SS.

Parameterization of the glandular microvasculature revealed an elevated group measurement for each of the measured parameters in the patients with SS compared with this measurement in the healthy volunteers. In particular, a marked increase in the size of the ve was demonstrated; this is consistent with the inflammation and infiltration that have been observed at histologic analysis (31). Elevated IAUC60, Ktrans, and vp values in the SS group were also observed; however, these increases were less marked. These changes would also be expected in view of changes in the gland vasculature, where interaction of endothelial cells with activated T- and B-cell leukocytes is associated with defective endothelium-dependent vasodilation and increased vascular permeability (1,31).

It is encouraging that our linear discriminant analysis revealed that microvascular characterization, yielding 100% sensitivity at 64% specificity for parameter medians and 100% sensitivity at 100% specificity for parameter SDs, may provide a basis for differentiating patients with SS from healthy volunteers. These findings suggest that dynamic MR imaging may be a useful additional diagnostic tool for patients suspected of having SS and may aid in the identification of phenotypic subtypes based on the severity of the inflammatory process.

Our findings of considerable heterogeneity in microvascular changes in the parotid gland in patients with SS are in agreement with the results of previous morphologic studies to analyze the SDs of mean MR signal intensity (as repeated here) and the changes in US characteristics (11,16,17). We found that statistically, T1-wSD, when used as a diagnostic indicator of SS as in a previous study (17), was no better than Ktrans, ve, vp, IAUC60, or T10 SD values for discriminating between patients with SS and healthy volunteers. The benefit of dynamic MR microvascular parameters in this context is that they may be used to further characterize and classify the parotid glands on the basis of physiologic status rather than simply rely on signal intensity characteristics.

The pathophysiologic characteristics of the parotid gland, such as disrupted vasculature and expansion of the extracellular extravascular space, can be quantified with dynamic MR imaging, and the extent of damage to the gland, or heterogeneity, may be assessed by using SD estimates of these vascular parameters. The greater overlap in T10 SD than in T1-wSD between the two subject groups implies that T1 cannot be the only factor responsible for the group discrimination, and, therefore, proton density must also have an important influence. We found that a combination of Ktrans SD and ve SD measurements would enable discrimination of the healthy and SS subject groups. Furthermore, three-dimensional imaging of the parotid gland provides an opportunity to quantitatively assess the entire gland and isolate areas of extreme inflammation and thus to follow up these areas during a course of therapy to monitor the glandular changes.

The relationship between parotid gland function and heterogeneity in SS has been studied, and a number of imaging techniques have been suggested (13,17,18,32). The use of dynamic MR imaging, diffusion-weighted MR imaging (18), and functional sialography (13) has demonstrated that recent advances in MR technology have made it possible to use complex imaging strategies to examine the structural and functional changes in the parotid gland. Both dynamic MR sialography (13) and diffusion-weighted MR imaging (18) have demonstrated sensitivity to physiologic changes in the gland. The decreased baseline stimulatory changes in salivary flow seen in the Morimoto et al study (13) correspond to the elevated microvascular characteristics, such as vessel permeability (Ktrans), seen in our study. Quantifiable noninvasive techniques such as these are clearly beneficial, and our study results show that dynamic MR imaging offers the ability to noninvasively detect particular pathophysiologic changes in the parotid gland and therefore has the potential to offer broader insight into SS. However, for assessment of the treatment effectiveness and disease progression associated with SS, we recommend that the described dynamic MR technique be validated by means of reproducibility assessment (33).

There were limitations in our study. First, limited histologic information prevented our evaluation of disease grade with dynamic MR parameters. Second, we optimized the data quality in this study by using a surface coil in close proximity to the parotid gland to maximize the signal-to-noise ratio. This enabled us to achieve a high spatial resolution while maintaining a high temporal resolution. However, some errors due to missampling of the arterial input function and the tissue contrast agent uptake curve may have occurred (34).

In conclusion, our findings represent preliminary results regarding the potential use of microvascular biomarkers of SS, and future work must include a comparison of dynamic MR kinetic parameters with histologic findings to more accurately evaluate the sensitivity of dynamic MR imaging to specific pathologic processes. There may also be a particular benefit in using further statistical techniques to investigate the heterogeneity of the parotid gland as assessed by using dynamic MR parameters.


   ADVANCES IN KNOWLEDGE
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
IMPLICATIONS FOR PATIENT CARE
 References
 


   IMPLICATIONS FOR PATIENT CARE
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
 IMPLICATIONS FOR PATIENT CARE
References
 


   ACKNOWLEDGMENTS
 
The authors acknowledge the Wellcome Trust Clinical Research Facility for access to the 1.5-T MR imaging unit at the University of Manchester; Sue Cheung, MSc, Barry Whitnall, DCR(R), and David Pollitt, DCR(R), for assistance with data acquisition and analysis; and David L. Buckley, PhD, for invaluable input.


   FOOTNOTES
 

Abbreviations: IAUC60 = initial area under the curve • Ktrans = transcapillary contrast agent transfer constant • ROC = receiver operating characteristic • SD = standard deviation • SS = Sjögren syndrome • T10 = baseline longitudinal relaxation time • T1-wSD = SD of mean signal intensity on T1-weighted image • ve = extracellular extravascular volume • vp = vascular plasma volume

Author contributions: Guarantors of integrity of entire study, C.R., G.J.M.P., A.J., V.E.R.; 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, C.R., C.J.R., J.P.O., V.E.R.; clinical studies, C.R., Y.W., S.M.S., A.J., V.E.R.; statistical analysis, C.R., G.J.M.P., C.J.R., Y.W., V.E.R.; and manuscript editing, C.R., G.J.M.P., C.J.R., J.P.O., A.J., V.E.R.

Authors stated no financial relationship to disclose.


   References
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
 IMPLICATIONS FOR PATIENT CARE
 References
 

  1. Greenspan JS, Daniels TE, Talal N, Sylvester RA. The histopathology of Sjogren's syndrome in labial salivary gland biopsies. Oral Surg Oral Med Oral Pathol 1974;37:217–229.[CrossRef][Medline]
  2. Lee M, Rutka JA, Slomovic AR, McComb J, Bailey DJ, Bookman AA. Establishing guidelines for the role of minor salivary gland biopsy in clinical practice for Sjogren's syndrome. J Rheumatol 1998;25:247–253.[Medline]
  3. Ohbayashi N, Yamada I, Yoshino N, Sasaki T. Sjogren syndrome: comparison of assessments with MR sialography and conventional sialography. Radiology 1998;209:683–688.[Abstract/Free Full Text]
  4. Aung W, Yamada I, Umehara I, Ohbayashi N, Yoshino N, Shibuya H. Sjogren's syndrome: comparison of assessments with quantitative salivary gland scintigraphy and contrast sialography. J Nucl Med 2000;41:257–262.[Abstract/Free Full Text]
  5. Markusse HM, Pillay M, Breedveld FC. The diagnostic value of salivary gland scintigraphy in patients suspected of primary Sjogren's syndrome. Br J Rheumatol 1993;32:231–235.[Abstract/Free Full Text]
  6. Lindvall AM, Jonsson R. The salivary gland component of Sjogren's syndrome: an evaluation of diagnostic methods. Oral Surg Oral Med Oral Pathol 1986;62:32–42.[CrossRef][Medline]
  7. Vitali C. Classification criteria for Sjogren's syndrome. Ann Rheum Dis 2003;62:94–95.[Free Full Text]
  8. Vitali C, Bombardieri S, Jonsson R, et al. Classification criteria for Sjogren's syndrome: a revised version of the European criteria proposed by the American-European Consensus Group. Ann Rheum Dis 2002;61:554–558.[Abstract/Free Full Text]
  9. Spath M, Kruger K, Dresel S, Grevers G, Vogl T, Schattenkirchner M. Magnetic resonance imaging of the parotid gland in patients with Sjogren's syndrome. J Rheumatol 1991;18:1372–1378.[Medline]
  10. Chikui T, Okamura K, Tokumori K, et al. Quantitative analyses of sonographic images of the parotid gland in patients with Sjogren's syndrome. Ultrasound Med Biol 2006;32:617–622.[CrossRef][Medline]
  11. Ariji Y, Ohki M, Eguchi K, et al. Texture analysis of sonographic features of the parotid gland in Sjogren's syndrome. AJR Am J Roentgenol 1996;166:935–941.[Abstract/Free Full Text]
  12. Makula E, Pokorny G, Kiss M, et al. The place of magnetic resonance and ultrasonographic examinations of the parotid gland in the diagnosis and follow-up of primary Sjogren's syndrome. Rheumatology (Oxford) 2000;39:97–104.[CrossRef][Medline]
  13. Morimoto Y, Habu M, Tomoyose T, et al. Dynamic magnetic resonance sialography as a new diagnostic technique for patients with Sjogren's syndrome. Oral Dis 2006;12:408–414.[CrossRef][Medline]
  14. Niemela RK, Takalo R, Paakko E, et al. Ultrasonography of salivary glands in primary Sjogren's syndrome: a comparison with magnetic resonance imaging and magnetic resonance sialography of parotid glands. Rheumatology (Oxford) 2004;43:875–879.[CrossRef][Medline]
  15. Niemela RK, Paakko E, Suramo I, Takalo R, Hakala M. Magnetic resonance imaging and magnetic resonance sialography of parotid glands in primary Sjogren's syndrome. Arthritis Rheum 2001;45:512–518.[CrossRef][Medline]
  16. El Miedany YM, Ahmed I, Mourad HG, et al. Quantitative ultrasonography and magnetic resonance imaging of the parotid gland: can they replace the histopathologic studies in patients with Sjogren's syndrome? Joint Bone Spine 2004;71:29–38.[CrossRef][Medline]
  17. Izumi M, Eguchi K, Ohki M, et al. MR imaging of the parotid gland in Sjogren's syndrome: a proposal for new diagnostic criteria. AJR Am J Roentgenol 1996;166:1483–1487.[Abstract/Free Full Text]
  18. Thoeny HC, De Keyzer F, Claus FG, Sunaert S, Hermans R. Gustatory stimulation changes the apparent diffusion coefficient of salivary glands: initial experience. Radiology 2005;235:629–634.[Abstract/Free Full Text]
  19. Jackson A, Buckley D, Parker G, eds. Dynamic contrast-enhanced magnetic resonance imaging in oncology. New York, NY: Springer, 2005.
  20. Jayson GC, Parker GJ, Mullamitha S, et al. Blockade of platelet-derived growth factor receptor-beta by CDP860, a humanized, PEGylated di-Fab', leads to fluid accumulation and is associated with increased tumor vascularized volume. J Clin Oncol 2005;23:973–981.[Abstract/Free Full Text]
  21. O'Connor JP, Jackson A, Parker GJ, Jayson GC. DCE-MRI biomarkers in the clinical evaluation of antiangiogenic and vascular disrupting agents. Br J Cancer 2007;96:189–195.[CrossRef][Medline]
  22. Tofts PS, Brix G, Buckley DL, et al. Estimating kinetic parameters from dynamic contrast-enhanced T(1)-weighted MRI of a diffusable tracer: standardized quantities and symbols. J Magn Reson Imaging 1999;10:223–232.[CrossRef][Medline]
  23. Leach MO, Brindle KM, Evelhoch JL, et al. Assessment of antiangiogenic and antivascular therapeutics using MRI: recommendations for appropriate methodology for clinical trials. Br J Radiol 2003;76(spec no 1):S87–S91.[Free Full Text]
  24. Evelhoch JL. Key factors in the acquisition of contrast kinetic data for oncology. J Magn Reson Imaging 1999;10:254–259.[CrossRef][Medline]
  25. Li KL, Zhu XP, Waterton J, Jackson A. Improved 3D quantitative mapping of blood volume and endothelial permeability in brain tumors. J Magn Reson Imaging 2000;12:347–357.[CrossRef][Medline]
  26. Plummer D. DispImage: a display and analysis tool for medical images. Rev Neuroradiol 1992;5:489–495.
  27. Tofts PS. Modeling tracer kinetics in dynamic Gd-DTPA MR imaging. J Magn Reson Imaging 1997;7:91–101.[Medline]
  28. Parker GJ, Roberts C, Macdonald A, et al. Experimentally-derived functional form for a population-averaged high-temporal-resolution arterial input function for dynamic contrast-enhanced MRI. Magn Reson Med 2006;56:993–1000.[CrossRef][Medline]
  29. Hanley JA, McNeil BJ. The meaning and use of the area under a receiver operating characteristic (ROC) curve. Radiology 1982;143:29–36.[Abstract/Free Full Text]
  30. Faul F, Erdfelder E, Lang AG, Buchner A. G*Power 3: a flexible statistical power analysis for the social, behavioral, and biomedical sciences. Behav Res Methods 2007;39:175–191.[Medline]
  31. Fox RI. Sjogren's syndrome. Lancet 2005;366:321–331.[CrossRef][Medline]
  32. Izumi M, Eguchi K, Nakamura H, Nagataki S, Nakamura T. Premature fat deposition in the salivary glands associated with Sjogren syndrome: MR and CT evidence. AJNR Am J Neuroradiol 1997;18:951–958.[Abstract]
  33. Roberts C, Issa B, Stone A, Jackson A, Waterton JC, Parker GJ. Comparative study into the robustness of compartmental modeling and model-free analysis in DCE-MRI studies. J Magn Reson Imaging 2006;23:554–563.[CrossRef][Medline]
  34. Roberts C, Buckley DL, Parker GJ. Comparison of errors associated with single- and multiple-bolus injection protocols in low-temporal-resolution dynamic contrast-enhanced tracer kinetic analysis. Magn Reson Med 2006;56:611–619.[CrossRef][Medline]



This article has been cited by other articles:


Home page
 Journal of the American Dental AssociationHome page
P. C. Fox, S. J. Bowman, B. Segal, F. B. Vivino, N. Murukutla, K. Choueiri, S. Ogale, and L. McLean
Oral involvement in primary Sjogren syndrome
J Am Dent Assoc, December 1, 2008; 139(12): 1592 - 1601.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
2463070298v1
246/3/845    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 Roberts, C.
Right arrow Articles by Rushton, V. E.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Roberts, C.
Right arrow Articles by Rushton, V. E.

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