ORIGINAL RESEARCH

Assessment of Active and Inactive

Sacroiliitis in Patients With Ankylosing
Spondylitis Using Quantitative Dynamic
Contrast-Enhanced MRI
Mengchao Zhang, MD,1 Le Zhou, MD, PhD,2 Ning Huang, PhD,3
Hong Zeng, MD, PhD,4 Songyan Liu, MD, PhD,5 and Lin Liu, MD, PhD1

Purpose: To investigate the feasibility of using quantitative dynamic contrast enhanced magnetic resonance imaging
(DCE-MRI) to differentiate the active and inactive stage of sacroiliitis and the correlation between quantitative parameters
and disease activity as measured by clinical scores.
Materials and Methods: Forty-two patients with ankylosing spondylitis underwent DCE-MRI on a 3.0T MRI unit. Accord-
ing to the results of the blood sedimentation rate (ESR), C-reactive protein (CRP), and Bath Ankylosing Spondylitis
Disease Activity Index (BASDAI), the patients were grouped into inactive and active groups. Pharmacokinetic models
were used to generate the semiquantitative and quantitative hemodynamic parameters of DCE-MRI. The between-
group differences were analyzed using the Wilcoxon rank sum test, and the correlations between the pharmacokinetic
parameters and BASDAI score were analyzed using Spearmans correlation coefficient. The efficacies of different param-
eters in differentiating the active and inactive phase of sacroiliitis were evaluated and compared using receiver operator
characteristics (ROC) curve analysis.
Results: Ktrans, Kep, Ve, time to peak (TTP), max concentration (MAX Conc), and area under the curve (AUC) of the
active group were significantly higher than those of the inactive stage group (P < 0.05). There were significant correla-
tions between all parameters and BASDAI (P < 0.05). AUC of the receiver operator characteristics curve (AUCR) of
different parameters were not statistically different (P >0.05), except between AUC and MAX Conc (P 5 0.0012).
Conclusion: Quantitative DCE-MRI parameters can differentiate between active and inactive ankylosing spondylitis.
Among those, Ktrans had the highest correlation coefficient with the BASDAI score.
Level of Evidence: 2
J. MAGN. RESON. IMAGING 2016;00:000000

A nkylosing spondylitis (AS) is a spinal arthritis disease

that usually presents before the age of 40, mostly in
males.1 The diagnosis often depends on the clinical symp-
been widely accepted, including different scoring systems
and functional MRI methods, such as diffusion-weighted
imaging (DWI) and dynamic contrast-enhanced (DCE)-
toms and imaging abnormities, with sacroiliitis being the MRI.36 The Spondyloarthritis Research Consortium of
characteristic hallmark.2 Sacroiliitis is an inflammatory dis- Canada (SPARCC) scoring system is often used to evaluate
ease induced by an immune-mediated mechanism and the the sacroiliac joints.7 This system, however, is subjective,
most important manifestation of AS. Conventional radiogra- and limited by its reliance on the radiologists experience in
phy, computed tomography (CT), and particularly magnetic musculoskeletal radiology. DWI and DCE-MRI may be
resonance imaging (MRI) have been applied to assess sacroi- effective methods in quantifying inflammatory changes at
liitis. Using MRI to diagnose and evaluate active AS had involved skeletal sites and useful for assessing treatment

View this article online at wileyonlinelibrary.com. DOI: 10.1002/jmri.25559

Received Jul 29, 2016, Accepted for publication Nov 1, 2016.

*Address reprint requests to: L.L. or S.L., Department of Radiology or Neurology, China-Japan Union Hospital of Jilin University, No. 126 Xiantai Street,
Changchun, Jilin, P.R. China 130033. E-mail: [email protected] or [email protected]

From the 1Department of Radiology, China-Japan Union Hospital of Jilin University, Changchun, Jilin, P.R. China; 2Department of Surgery, China-Japan
Union Hospital of Jilin University, Changchun, Jilin, P.R. China; 3GE Healthcare, Economic and Technological Development Zone, Beijing, P.R. China;
4
Department of Cardiology, China-Japan Union Hospital of Jilin University, Changchun, Jilin, P.R. China; and 5Department of Neurology, China-Japan
Union Hospital of Jilin University, Changchun, Jilin, P.R. China

C 2016 International Society for Magnetic Resonance in Medicine

V 1
Journal of Magnetic Resonance Imaging

FIGURE 1: a,b: Male, 26 years old, diagnosed as ankylosing spondylitis in 2013, with waist and back pain within a week before
imaging; ESR 5 30 mm/1st hour, CRP 5 32.4 lg /ml, BASDAI 5 8. a: Enhanced T1 image: patch enhancements were observed (white
arrows) were observed in right sacral and iliac area. b: Ktrans color map: high Ktrans values (Ktrans 5 1.975) in red (yellow arrows) could
be observed in the right sacroiliac. c,d: Male, 21 years old, diagnosed as ankylosing spondylitis in 2014, no clinical symptoms,
ESR 5 2 mm/1st hour, CRP 5 3.6 lg/ml, BASDAI 5 1. c: No major enhancement was observed in enhanced T1 image. d: Ktrans values
were relatively low in bilateral sacroiliac joint; multiple ROIs were placed on auricular surface and the average Ktrans was 0.388.

efficacy in ankylosing shown by some recent studies.810 architecture, and increased capillary permeability, which is
However, the ADC (apparent diffusion coefficient) value in different from inactive sacroiliitis. Therefore, we hypothesized
DWI is affected by b value and the fitting model, and that there would be significant differences in the quantitative
semiquantitative parameters derived from the signal intensi- DCE-MRI parameters between active and inactive AS, and
tytime curve in DCE-MRI could only reflect the character- correlations should exist between DCE-MRI parameters and
istics of tissue enhancement. disease activity measured by a clinical score such as the Bath
In recent years, quantitative DCE-MRI has been Ankylosing Spondylitis Disease Activity Index (BASDAI).
applied successfully in the brain, body, and the synovium in The aim of this study was thus to test the above
patients with articular disease, including tumor and inflam- hypothesis.
mation.1117 Quantitative permeability parameters were
derived using a pharmacokinetic model by analysis of the sig- Materials and Methods
nal intensitytime curve of the tissue and the arterial input Study Population
function (AIF).18,19 During active sacroiliitis in AS, infiltra- This study was approved by the local Institutional Review Boards,
tion of inflammatory cell and vasogenic edema cause and written informed consent was obtained from all patients.
increased perfusion of serum and blood transported from Patients were diagnosed as having AS using the 1984 modified
capillary to bone marrow cavity, destruction of microvascular New York criteria for ankylosing spondylitis2 by one doctor with 5

2 Volume 00, No. 00

 Zhang et al.: Active and Inactive Sacroiliitis

TABLE 1. Patient Characteristics and Clinical Laboratory Findings

Sex
Group Cases M F Age ESR (mm/h) CPR (lg/ml) BASDAI
Active 22 19 3 22.4 6 1.3 38.6 6 3.6 29.5 6 4.3 8.5 6 2.4
Inactive 20 18 2 20.6 6 1.1 5.4 6 1.5 3.8 6 1.0 2.8 6 0.5
P Value 0.234* 0.134*
*P > 0.05.
ESR, blood sedimentation rate; CPR, C-reactive protein; BASDAI, bath ankylosing spondylitis disease activity index.

years experience in rheumatology between May 2015 and March Germany) using a phased array body coil within a week after the
2016. Our exclusion criteria were patients who had a known aller- laboratory tests and clinical assessments. The scanning sequences
gy to the contrast material, had contraindications to MRI, had included axial T1-weighted imaging (T1-WI) and proton density
poor renal function (glomerular filtration rate [GFR] <30 ml/ (PD), oblique coronal T1-WI, PD, and T1 VIBE (volumetric inter-
min), or were pregnant. polated breath-hold examination) fat-suppressed DCE-MRI. The
All patients underwent laboratory tests and clinical assess- DCE-MRI parameters were: five flips angles, including 38, 68, 98,
ments including blood sedimentation rate (ESR), C-reactive pro- 128, and 158. Repetition time / echo time (TR/TE) 5.0/1.77 msec,
tein (CRP), and BASDAI.2024 According to the results of ESR, slice thickness 3 mm, matrix 192 3 154. A total of 35 phases were
CRP, and BASDAI, the patients with AS were grouped into two acquired, with a temporal resolution of 9.6 seconds per phase and
groups with the following inclusion criteria25,26: active group: a total of 336 seconds. Two phases were acquired before the con-
abnormal ESR (male >15 mm/1st hour, female >20 mm/1st trast injection as baseline, and 0.1 mmol/kg Gd-DTPA-BMA
hour), abnormal CRP (>10 lg/ml) and BASDAI >4; inactive (Omniscan, GE, Ireland) was injected through the intravenous
group: normal ESR (male <15 mm/1st hour, female <20 mm/1st catheter (20G, WeiGao Shan Dong, China) placed in the cubital
hour), normal CRP (<10 lg/ml), BASDAI <4. vein, at a flow rate of 2 ml/s using a binocular power injector
(Medrad Avanta, Pittsburgh, PA), followed by a 20-ml saline flush
ESR and CRP Assay at the same rate.
ESR was measured using the Westergren method and the serum
CRP using the CRP turbidimetric immunoassay reagent kit (Elec- Data Analysis
talab, Italy) in the two groups, with normal range <10 mm/h, All DCE-MRI data were transferred to a quantitative DCE-MRI
<6 mg/L, respectively.27 software package (Omni-Kinetics, GE Healthcare, Milwaukee,
WI). The DCE-MRI T1-weighted images were initially converted
MRI Protocol into Gd intensity using the variable flip angle method, and the
All patients underwent MRI exam of the sacroiliac joint with a pharmacokinetics parameters were computed using the two-
3.0T MRI scanner (Magnetom Trio Tim, Siemens, Erlangen, compartment modified Tofts model. One radiologist (M.Z. with 5

TABLE 2. Comparisons of Different DCE-MRI Parameters in Active and Inactive Groups

Active Group (n 5 22) Inactive Group (n 5 20)

Parameter P25 P50 P75 P25 P50 P75 Z Value P Value
trans 21
K (min ) 0.4760 0.7500 1.6058 0.0835 0.1625 0.4848 3.727 0.000*
21
Kep (min ) 0.7785 1.0080 1.5100 0.2118 0.5050 1.1138 2.317 0.021*
Ve 0.3368 0.4595 2.0715 0.0983 0.1790 0.3453 3.639 0.000*
Vp 0.0433 0.1245 0.3530 0.0843 0.1590 0.4133 1.209 0.227
TTP (min) 1.2768 1.7840 3.3145 0.8900 1.3070 1.8498 2.443 0.015*
MAX Conc (mmol) 0.1725 0.2000 0.4805 0.1058 0.1725 0.2698 2.141 0.032*
AUC (mmol
min) 0.5843 0.6900 1.8560 0.2893 0.5065 0.9380 2.921 0.003*
MAX Slope (mmol/min) 0.3908 0.5180 0.6293 0.2940 0.4445 0.5958 1.209 0.227
trans
K , forward volume transfer constant; kep, reverse transvascular transfer rate constant; Ve, extravascular extracellular space volume;
Vp, plasma volume fraction; TTP, time to peak; MAX Conc, max concentration; AUC, area under the curve; MAX Slope, maximum
slope.
*P < 0.05.

Month 2016 3
Journal of Magnetic Resonance Imaging

FIGURE 2: Comparison of Ktrans, Kep, Ve, TTP, Max Conc, and AUC between active and inactive groups of AS, with all pairs show-
ing significant differences.

years experience in rheumatologic radiology) analyzed the data. In correlation was used to analyze the correlations between the DCE-
both groups the AIF was placed on the center of the iliac artery, MRI parameters and BASDAI score. Linear regression was per-
with a size of 2 mm2. In patients of the active group, one region formed on the DCE-MRI parameter that had the highest correla-
of interest (ROI) of about 2 mm2 was placed on the oblique coro- tion coefficient.
nal slice with the maximum area of bone marrow, then semiquanti- The receiver operating curve (ROC) analyses were performed
tative and quantitative DCE-MRI parameters were measured on using MedCalc (v. 15.6.1, MedCalc Software, Mariakerke, Bel-
the enhancement area. In patients of the inactive group, the ROI gium) to assess the diagnostic efficacy of Ktrans, Kep, Ve, Vp, TTP,
of the same size was placed on the oblique coronal slice with the MAX Conc, AUC, and MAX Slope between the active group and
maximum area of iliac bone on sacroiliac joint, as shown in Fig. 1. inactive group.

Statistical Analysis
The KolmogorovSmirnov test was performed to check whether Results
the data of all variables was in normal distribution. For data in The patient characteristics are shown in Table 1. Among the
normal distribution, it was expressed as mean 6 SD; otherwise it 42 sacroiliitis patients, 22 patients were in the active stage
was expressed as median (P25, P75). Statistical analysis was per- (M 19, F 3), while 20 were in the inactive stage (M 18, F
formed using IBM SPSS (v. 19.0, IBM, Chicago, IL).
2). There were no significant differences between the groups
An independent t-test (for normal distribution) or Wilcoxon
in sex or age.
test (for nonnormal distribution) was used to compare the semi-
quantitative and quantitative parameters of DCE-MRI between the
KolmogorovSmirnov test showed that the quantitative
active and inactive stage, including Ktrans, Kep, Ve, Vp, and semi- and semiquantitative DCE-MRI parameters were not nor-
quantitative parameters of time to peak (TTP), max concentration mally distributed; the results of the 25, 50 (median), and
(MAX Conc), area under the curve (AUC), and MAX Slope. P < 75 percentiles are shown in Table 2. Quantitative parame-
0.05 was considered statistically significant. The Spearman ters of Ktrans, Kep, Ve, and semiquantitative parameters of

4 Volume 00, No. 00

 Zhang et al.: Active and Inactive Sacroiliitis

FIGURE 5: Regression analysis of Ktrans and BASDAI score, with

Ktrans values on the X-axis, BASDAI score on the Y-axis, solid
dots marking each case, and the line marking the best linear
regression fitting.

2.086. The AUCR, and Youden index (sensitivity, specific-

ity) for detecting AS by Ktrans, Kep, Ve, Vp, TTP, MAX
Conc, AUC, and MAX Slope are shown in Table 3. All
parameters had statistical significance in diagnosis except Vp
and Max Slope. AUCR of hemodynamic parameters were
not statistically different, as shown in Table 4, except
between AUC and MAX Conc (P 5 0.0012).

FIGURE 3: Comparison of Vp and Max Slope between active

and inactive groups of AS, with no significant differences.
Discussion
This study validated the feasibility of using pharmacokinetic
TTP, MAX Conc, and AUC were significantly higher in the parameters of DCE-MRI to assess active and inactive sacroi-
active group than in the inactive group (Fig. 2). There were liitis in patients with AS and showed that there are correla-
no significant differences in Vp and Max Slope between the tions between pharmacokinetic parameters and disease
active and inactive groups (Fig. 3). activity measured by clinical score.
Spearmans correlation showed that Ktrans, Kep, Ve, Vp,
TTP, MAX Conc, AUC, and MAX Slope had correlations
with the BASDAI score, and the correlation coefficients
were 0.842 (P 5 0.000), 0.463 (P 5 0.002), 0.777
(P 5 0.000), -0.401 (P 5 0.008), 0.430 (P 5 0.004), 0.558
(P 5 0.000), 0.666 (P 5 0.000), and 0.393 (P 5 0.010),
respectively (Fig. 4). Regression analysis was performed on
Ktrans, as shown in Fig. 5, and the linear relationship
between Ktrans and BASDAI was: BASDAI 5 3.308*Ktrans 1

FIGURE 4: Correlation coefficients between different quantitative FIGURE 6: Receiving operator characteristic analyses of Ktrans,
or semiquantitative parameters with BASDAI score, including Kep, Ve, TTP, MAX Conc, and AUC. Ktrans and Ve had higher
Ktrans, Kep, Ve, Vp, TTP, MAX Conc, AUC, and MAX Slope. sensitivity and specificity on the optimal critical point.

Month 2016 5
Journal of Magnetic Resonance Imaging

TABLE 3. AUCR for Detecting Ankylosing Spondylitis by Quantitative or Semiquantitative Parameters

AUC SE 95%CI P value Youden index

(sensitivity, specificity)
Ktrans 0.836 0.066 0.707-0.966 0.000* 0.6545 (95.45, 70.00)
Kep 0.709 0.082 0.548-0.870 0.021* 0.4636 (86.36, 60.00)
Ve 0.828 0.065 0.700-0.956 0.000* 0.6136 (86.36, 75.00)
Vp 0.391 0.088 0.219-0.563 0.227
TTP 0.720 0.079 0.566-0.847 0.015* 0.4000 (100.00, 40.00)
MAX Conc 0.693 0.082 0.533-0.853 0.032* 0.3227 (77.27, 55.00)
AUC 0.764 0.074 0.619-0.908 0.003* 0.4182 (81.00, 60.00)
MAX Slope 0.609 0.089 0.435-0.783 0.227
*P < 0.05.
AUCR, area under the receiver operator characteristic curve, Ktrans, forward volume transfer constant; kep, reverse transvascular transfer
rate constant; Ve, extravascular extracellular space volume; Vp, plasma volume fraction; TTP, time to peak; MAX Conc, max concen-
tration; AUC, area under the curve; MAX Slope, maximum slope; SE, standard error.

This study also showed that there were significant dif- Ktrans and Kep. Another possible reason is that inflammatory
ferences in quantitative and semiquantitative parameters of damage to a joint leads to loss of these negatively charged
DCE-MRI (including Ktrans, Kep, Ve, TTP, MAX Conc, and macromolecules. As a result, there is greater affinity for gad-
AUC), between the active and inactive AS patient groups, olinium complexes from paramagnetic contrast agents,
with parameters of the active group significantly higher than which also possess a negative charge, to the extracellular
the inactive group. The explanation for this could be that in matrix.30,31 This is also consistent with a pathological basis
the acute stage inflammatory tissues (fibroblasts, T-cells, and of synovium enhancement of AS.32 For semiquantitative
macrophages)28,29 with increased microcirculation, and parameters, Max Conc and AUC were higher in the active
greater capillary permeability, which would lead to higher group than the inactive group, indicating more perfusion in

TABLE 4. Comparisons of AUCR of Different DCE-MRI Parameters

AD SE 95% CI Z Value P Value

AUC- Kep 0.0545 0.104 20.150-0.259 0.523 0.601
trans
AUC- K 0.0727 0.0652 20.0550-0.200 1.116 0.2646
AUC-MAXConc 0.0705 0.0218 0.0277-0.113 3.228 0.0012*
AUC-TTP 0.0432 0.0948 20.143-0.229 0.455 0.6489
AUC-Ve 0.0648 0.0668 20.0661-0.196 0.970 0.3321
trans
Kep-K 0.127 0.0782 20.0261-0.281 1.627 0.1038
Kep-MAXConc 0.0159 0.111 20.201-0.233 0.144 0.8857
Kep-TTP 0.0114 0.0976 20.180-0.203 0.116 0.9073
Kep-Ve 0.119 0.0882 20.0536-0.292 1.353 0.1762
trans
K -MAXConc 0.143 0.0747 20.00315-0.290 1.918 0.0551
trans
K -TTP 0.116 0.0869 20.0544-0.286 1.334 0.1822
trans
K -Ve 0.00795 0.0459 20.0821-0.0980 0.173 0.8625
MAXConc-TTP 0.0273 0.0989 20.167-0.221 0.276 0.7827
MAXConc-Ve 0.135 0.0760 20.0138-0.284 1.779 0.0753
TTP-Ve 0.108 0.0752 20.0394-0.255 1.436 0.1511
*P < 0.05.
AD, area difference; SE, standard error.

6 Volume 00, No. 00

 Zhang et al.: Active and Inactive Sacroiliitis

the active group; TTP was also higher due to the delayed parameters and other factors, such as SPARCC score, ESR,
enhancement. Similar results were acquired by other stud- and CPR. Also, in our ongoing study quantitative DCE-
ies.3,4,9,33,34 In addition, a recent study35 showed that the MRI parameters are being used to predict and assess the
perfusion fraction, f, of intravoxel incoherent motion treatment effect of antitumor necrosis factor (anti-TNF) in
(IVIM) had a significant difference between active and AS sacroiliitis patients.
chronic sacroiliitis, which may support the results of our In conclusion, as a biomarker for sacroiliac joint
study. inflammatory activities, quantitative DCE-MRI parameters
In recent years, correlations between MRI and BAS- could provide a potential valuable method for accurate diag-
DAI, ESR, and CRP have been investigated; however, the nosis and assessment of sacroiliitis, and might potentially
results were not consistent. Zhang et al36 found that the help in treatment planning.
BASDAI score showed a statistically significant correlation
with the MRI indices of sacroiliitis activity, while MacKay
et al37 thought that there were weak, nonsignificant correla-
Acknowledgments
tions between MRI disease activity scores and BASDAI, Contract grant sponsor: Natural Science Foundation of Jilin
ESR, and CRP. This study further investigated the correla- Province, China; contract grant number: 20160101074JC
tions between quantitative or semiquantitative DCE-MRI
parameters and disease activity measured by the BASDAI References
1. Jacobson JA, Girish G, Jiang Y, Resnick D. Radiographic evaluation of
score. All parameters showed correlations. Braun et al9 arthritis: inflammatory conditions. Radiology 2008;248:378389.
found that MR enhancement of sacroiliitis patients correlat-
2. van der Linden S, Valkenburg HA, Cats A. Evaluation of diagnostic cri-
ed with the clinical history, back inflammatory pain, and teria for ankylosing spondylitis. A proposal for modification of the
physical exam results. Bollow et al32 showed that MR New York criteria. Arthritis Rheum 1984;27:361368.

enhancement correlated with the invasion degree of sacroili- 3. Algin O, Gokalp G, Baran B, Ocakoglu G, Yazici Z. Evaluation of sac-
roiliitis: contrast-enhanced MRI with subtraction technique. Skeletal
ac joint inflammatory cells, and could help assess the antiin- Radiol 2009;38:983988.
flammatory treatment effects, and demonstrated that
4. Althoff CE, Feist E, Burova E, et al. Magnetic resonance imaging of
different types of contrast enhancement curve on synovium active sacroiliitis: do we really need gadolinium? Eur J Radiol 2009;
could correlate with the patients degree of pain.38 Further- 71:232236.

more, Zhao et al34 demonstrated that there was a significant 5. Braun J, Sieper J, Bollow M. Imaging of sacroiliitis. Clin Rheumatol
2000;19:5157.
correlation between ADC and T1 enhancement rate of the
bone marrow and BASDAI. According to a previous 6. Lacout A, Rousselin B, Pelage JP. CT and MRI of spine and sacroiliac
involvement in spondyloarthropathy. AJR Am J Roentgenol 2008;191:
report,26 synovial enhancement at MRI correlated with dis- 10161023.
ease activity as measured by laboratory inflammatory 7. Landewe RB, Hermann KG, van der Heijde DM, et al. Scoring sacroili-
markers in AS. However, in our study the bone marrow ac joints by magnetic resonance imaging. A multiple-reader reliability
experiment. J Rheumatol 2005;32:20502055.
enhancement region was evaluated by DCE-MRI and the
disease activity was measured by the BASDAI score. None- 8. Bozgeyik Z, Ozgocmen S, Kocakoc E. Role of diffusion-weighted MRI
in the detection of early active sacroiliitis. AJR Am J Roentgenol
theless, results of the above studies were similar to ours, as 2008;191:980986.
pathological changes associated with the disease activity in 9. Braun J, Bollow M, Eggens U, Konig H, Distler A, Sieper J. Use of
synovial and bone marrow are the same. This study found dynamic magnetic resonance imaging with fast imaging in the detec-
tion of early and advanced sacroiliitis in spondylarthropathy patients.
that, in general, quantitative parameters (in particular Ktrans) Arthritis Rheum 1994;37:10391045.
had higher correlation coefficients than semiquantitative
10. Gaspersic N, Sersa I, Jevtic V, Tomsic M, Praprotnik S. Monitoring
ones. Quantitative permeability parameters using the com- ankylosing spondylitis therapy by dynamic contrast-enhanced and
partmental model reflect the exchange of the contrast agent diffusion-weighted magnetic resonance imaging. Skeletal Radiol 2008;
37:123131.
between plasma and EES (extracellular/extravascular space),
11. Heye AK, Culling RD, Valdes Hernandez Mdel C, Thrippleton MJ,
and thus may accurately reflect the tissue microcirculation.39 Wardlaw JM. Assessment of blood-brain barrier disruption using
ROC curve analysis revealed that all parameters were dynamic contrast-enhanced MRI. A systematic review. NeuroImage
able to identify active and inactive AS except MAX Slope Clin 2014;6:262274.

and Vp. Ktrans and Ve had higher sensitivity and specificity 12. Huang W, Li X, Chen Y, et al. Variations of dynamic contrast-
enhanced magnetic resonance imaging in evaluation of breast cancer
than other parameters (Table 3). Therefore, these two therapy response: a multicenter data analysis challenge. Transl Oncol
parameters are recommended in clinical practice. 2014;7:153166.
While this study was primarily focused on the estab- 13. Ledsam JR, Hodgson R, Moots RJ, Sourbron SP. Modeling DCE-MRI
lishment of a new method for differentiation of active and at low temporal resolution: a case study on rheumatoid arthritis.
J Magn Reson Imaging 2013;38:15541563.
inactive sacroiliitis in patients with AS, there are some limi-
14. Nielsen T, Wittenborn T, Horsman MR. Dynamic contrast-enhanced
tations. First, the sample size is relatively small. Second, we magnetic resonance imaging (DCE-MRI) in preclinical studies of anti-
did not study the correlation between quantitative vascular treatments. Pharmaceutics 2012;4:563589.

Month 2016 7
Journal of Magnetic Resonance Imaging

15. Rastogi A, Kubassova O, Krasnosselskaia LV, et al. Evaluating auto- 28. Francois RJ, Gardner DL, Degrave EJ, Bywaters EG. Histopathologic
mated dynamic contrast enhanced wrist 3T MRI in healthy volunteers: evidence that sacroiliitis in ankylosing spondylitis is not merely enthe-
one-year longitudinal observational study. Eur J Radiol 2013;82:1286 sitis. Arthritis Rheum 2000;43:20112024.
1291.
29. Maksymowych WP, Inman RD, Salonen D, et al. Spondyloarthritis
16. Sourbron S, Sommer WH, Reiser MF, Zech CJ. Combined quantifica- Research Consortium of Canada magnetic resonance imaging index
tion of liver perfusion and function with dynamic gadoxetic acid- for assessment of spinal inflammation in ankylosing spondylitis. Arthri-
enhanced MR imaging. Radiology 2012;263:874883. tis Rheum 2005;53:502509.
17. Verma S, Turkbey B, Muradyan N, et al. Overview of dynamic 30. Hermann KG, Bollow M. Magnetic resonance imaging of sacroiliitis in
contrast-enhanced MRI in prostate cancer diagnosis and manage- patients with spondyloarthritis: correlation with anatomy and histolo-
ment. AJR Am J Roentgenol 2012;198:12771288. gy. Rofo 2014;186:230237.
18. Sourbron SP, Buckley DL. Classic models for dynamic contrast- 31. Taupitz M, Stolzenburg N, Ebert M, et al. Gadolinium-containing
enhanced MRI. NMR Biomed 2013;26:10041027. magnetic resonance contrast media: investigation on the possible
19. Tofts PS. Modeling tracer kinetics in dynamic Gd-DTPA MR imaging. transchelation of Gd(3)(1) to the glycosaminoglycan heparin. Contrast
J Magn Reson Imaging 1997;7:91101. Media Mol Imaging 2013;8:108116.

20. Ruof J, Stucki G. Validity aspects of erythrocyte sedimentation rate 32. Bollow M, Fischer T, Reisshauer H, et al. Quantitative analyses of
and C-reactive protein in ankylosing spondylitis: a literature review. sacroiliac biopsies in spondyloarthropathies: T cells and macrophages
J Rheumatol 1999;26:966970. predominate in early and active sacroiliitis- cellularity correlates with
the degree of enhancement detected by magnetic resonance imag-
21. Sieper J, Rudwaleit M, Baraliakos X, et al. The Assessment of Spondy- ing. Ann Rheum Dis 2000;59:135140.
loArthritis International Society (ASAS) handbook: a guide to assess
spondyloarthritis. Ann Rheum Dis 2009;68(Suppl 2):ii144. 33. Pagnini I, Savelli S, Matucci-Cerinic M, Fonda C, Cimaz R, Simonini G.
Early predictors of juvenile sacroiliitis in enthesitis-related arthritis.
22. Soliman E, Labib W, el-Tantawi G, Hamimy A, Alhadidy A, Aldawoudy J Rheumatol 2010;37:23952401.
A. Role of matrix metalloproteinase-3 (MMP-3) and magnetic reso-
nance imaging of sacroiliitis in assessing disease activity in ankylosing 34. Zhao Y, Han X, Hu S, The relevance of ADC value, T(1)intensive rate
spondylitis. Rheumatol Int 2012;32:17111720. and the clinical activities in ankylosing spondylitis sacroiliitis. Zhong-
hua Yi Xue Za Zhi 2014;94:27502753.
23. van der Heijde D, Bellamy N, Calin A, Dougados M, Khan MA, van
der Linden S. Preliminary core sets for endpoints in ankylosing spon- 35. Zhao YH, Li SL, Liu ZY, et al. Detection of active sacroiliitis with anky-
dylitis. Assessments in Ankylosing Spondylitis Working Group. losing spondylitis through intravoxel incoherent motion diffusion-
J Rheumatol 1997;24:22252229. weighted MR imaging. Eur Radiol 2015;25:27542763.

24. Zochling J, van der Heijde D, Burgos-Vargas R, et al. ASAS/EULAR 36. Zhang P, Yu K, Guo R, et al. Ankylosing spondylitis: correlations
recommendations for the management of ankylosing spondylitis. Ann between clinical and MRI indices of sacroiliitis activity. Clin Radiol
Rheum Dis 2006;65:442452. 2015;70:6266.

25. Garrett S, Jenkinson T, Kennedy LG, Whitelock H, Gaisford P, Calin 37. MacKay JW, Aboelmagd S, Gaffney JK. Correlation between clinical
A. A new approach to defining disease status in ankylosing spondyli- and MRI disease activity scores in axial spondyloarthritis. Clin Rheu-
tis: the Bath Ankylosing Spondylitis Disease Activity Index. matol 2015;34:16331638.
J Rheumatol 1994;21:22862291.
38. Bollow M, Braun J, Hamm B, et al. Early sacroiliitis in patients with
26. Jee WH, McCauley TR, Lee SH, Kim SH, Im SA, Ha KY. Sacroiliitis in spondyloarthropathy: evaluation with dynamic gadolinium-enhanced
patients with ankylosing spondylitis: association of MR findings with MR imaging. Radiology 1995;194:529536.
disease activity. Magn Reson Imaging 2004;22:245250.
39. OConnor JP, Jackson A, Parker GJ, Jayson GC. DCE-MRI biomarkers
27. Lewis SM, Bain BJ, Bates I, Dacie JV, Dacie JV. Dacie and Lewis prac- in the clinical evaluation of antiangiogenic and vascular disrupting
tical haematology. Philadelphia: Churchill Livingstone/Elsevier; 2006. agents. Br J Cancer 2007;96:189195.

8 Volume 00, No. 00