949 - Stability of Automated Quantitation of Dynamic Contrast Enhanced (DCE) 3T Wrist Magnetic Resonance Imaging (MRI) in Healthy Volunteers; A Year Long Longitudinal Study

Monday, November 7, 2011: 9:00 AM-6:00 PM
Hall F2 - Poster Hall (McCormick Place West)
A. Rastogi, Barts and the London NHS Trust, London, United Kingdom, Olga Kubassova, Image Analysis Ltd., Leeds, United Kingdom, Mikael Boesen, The Parker Institute, Copenhagen University Hospital at Frederiksberg, Copenhagen, Denmark, J.V. Hajnal, Imperial College, London, United Kingdom and Peter Taylor, Kennedy Institute of Rheumatology, London, United Kingdom
Presentation Number: 949

Background/Purpose: A desirable property of new therapeutics is to arrest progression of joint damage at early stages of rheumatoid arthritis (RA). Dynamic Contrast Enhanced (DCE)-MRI has recently been recognized as a promising biomarker of disease modification in early phase inflammatory arthritis as it permits quantitative assessment of perfusion. To recommend DCE-MRI for quantification of treatment effect in patients, thorough validation should be conducted using the data from healthy volunteers where no disease related changes are expected over the course of imaging. This study is concerned with validation of variability inherent in newer automated DCE-MRI methods by testing software performance in healthy cohorts over a year.

Method: Dominant wrists of 7 healthy volunteers (3 males:4 females, age range:24-40 years) were imaged by 3T MR at baseline, week 12, 24 and 52 using T2w TSE, pre and post contrast T1wFFE, DCE-MRI sequences with the following parameters: TR/TE/FA: 3.8 ms/2.1 ms/20°, slices: 127, 40 frames (scan mode: 3D) and T1wFFE Proset. DCE-MRI data were analysed using Dynamika, Image Analysis, UK. Using the software, parameters of maximum enhancement (ME), initial rate of enhancement (IRE) and the number of pixels (e.g. volume of enhancing synovium) with persistent and wash-out pattern of enhancement were measured using pixel-by pixel model based approach in the entire image and within regions of interest (ROI) drawn around the wrist and placed precisely around enhancing synovium.

Result: 

The average ± standard deviation for baseline parameters and change over time is shown in Table 1. In wrist ROI analysis and precise ROI analysis, ME and IRE change are consistently low and less then 0.1. The number of pixels with persistent, plateau and washout pattern of enhancement, reflecting the volume of synovium become more accurate as a % of the ROI size and are low too (normalised persistent voxel change over 12 weeks: 0.2±1.1%(wrist ROI) & -0.2±2.1%(precise ROI), normalised plateau voxel change over 52 weeks: 3.7±3.4%(wrist ROI) & 1.0±10.9% (precise ROI) and normalised washout voxel change over 52 weeks: 1.9±0.9% (wrist ROI) & 5.8±7.4% (precise ROI)).

Table 1: Shows baseline values of parameters and their change over time.

 

                        Baseline

               Wk12-baseline change

             Wk24-baseline change

                   Wk52-baseline change

 

           Average

         Std dev

            Average

            Std dev

             Average

                Std dev

             Average

               Std dev

ROI Wrist

 

 

 

 

 

 

 

 

MEmean

1.266095

0.0786

-0.007619

0.0632

0.007595

0.0940

0.041133

0.0239

MEst. dev.

0.105619

0.0758

-0.000228

0.0576

0.033142

0.1001

0.041866

0.0265

IREmean

0.001952

0.0008

-0.000285

0.0009

9.52381E-05

0.0011

0.0006

0.0009

IREst. dev.

0.000904

0.0007

0.000809

1

0.001476

0.0014

0.002066

0.0015

Normalised Persistent: (%)

0.885665

0.7006

0.202174

1.1978

-0.427921

0.8410

-0.305485

0.3177

Normalised Plateau: (%)

11.283851

10.2615

0.938550

8.9309

0.469898

8.8623

3.790429

3.4510

Normalised Wash-out (%)

1.233881

1.1631

0.874174

0.6031

-0.383588

1.471956

1.951587

0.9091

ROI Precise

 

 

 

 

 

 

 

 

MEmean

1.285428

0.0903

-0.016619

0.0675

0.035595

0.1381

0.048966

0.0428

MEst. dev.

0.106076

0.0798

0.006590

0.0694

0.034757

0.1100

0.041333

0.0382

IREmean

0.001761

0.0008

0.000285

0.0011

0.000809

0.0019

0.0008

0.0009

IREst. dev.

0.000952

0.0004

0.001142

0.0021

0.001809

0.0021

0.0013

0.0009

Normalised Persistent: (%)

1.578668

1.6659

-0.209247

2.1411

-0.540966

2.3994

-0.824105

1.5893

Normalised Plateau: (%)

22.432640

21.2280

-4.270537

13.8230

-3.060829

11.9250

1.078427

10.9669

Normalised Wash-out (%)

1.728650

1.5729

1.475777

1.8877

-0.619386

1.8319

5.819983

7.4759

Conclusion:  This study shows that in healthy subjects, we can expect DCE-MRI parameters remain stable for as long as a year. The absence of variability associated with analysis methodology give confidence that these parameters can be used as imaging biomarkers in patients with inflammatory arthritis where early and sensitive quantification of change due to therapeutic intervention is desirable. Both wrist and precise ROI give similar results in healthy volunteers over a year and thus either could be used alone as part of the analysis.


Keywords: magnetic resonance imaging (MRI)

Disclosure: A. Rastogi, None; O. Kubassova, Image Analysis Ltd., 4 ; M. Boesen, Image Analysis Ltd, 9 ; J. V. Hajnal, None; P. Taylor, None.