Background Quantitative analysis of tissue microvascular function using dynamic contrast enhanced MRI (DCE-MRI) shows promise for improved understanding of synovial pathophysiology in rheumatoid arthritis (RA). Most studies to date used 2D measures, and do not use physiologically-interpretable measures such as ktrans.
Objectives Develop a 3D method for DCE-MRI., using validated quantitative parameters. Provide precise and consistent regions of interest (ROI) that are unaffected by hand position in the coil using active appearance models (AAMs)
Methods MR images of the hand were acquired pre- and post-contrast (Gd) in 27 patients with established RA who had recently commenced the same biological therapy. Subjects were imaged at 0,3,6 months. MRI protocol included pre- and post-Gd high-resolution 3D FLASH acquisitions. DCE-MRI scan protocol included 3 FLASH images for baseline T1 measurement. 15 sequential volumes were acquired over 5 minutes (flip angle =35°), during which a bolus of gadolinium (0.1 mmol/kg) was administered at 2 ml/s at the beginning of the 5th volume measurement. Concentration of contrast agent was determined via T1 measurement. The Extended Kety model was applied to each voxel concentration-time series within ROIs, allowing voxelwise estimates of the capillary transfer coefficient of contrast agent (ktrans). Pre-contrast T1 images were searched using 3D AAMs to identify bones and capsular structures and generate 3D ROIs. Images from the dynamic series were registered to the high resolution pre-contrast images, providing standardised 3D ROIs for each of the RAMRIS joints. 3D visualisation was performed by dividing voxels into 3 ktrans bins: (High >0.2 min-1, Medium 0.1-0.2 min-1, Low <0.1 min-1), displayed over AAM-identified bone surfaces. RAMRIS scoring was performed by a single experienced reader blinded to time point.
Results Change in ktrans from baseline (95% CI) at 3 months was -0.007 min-1(0.001,-0.015), p=0.06, at 6 months -0.012 (-0.004,-0.020), p=0.005 (Fig 1). This indicates a reduction in joint perfusion and capillary permeability associated with reduced inflammation. RAMRIS scoring showed no significant change at either 3 or 6 months. 3D visualisation of individual participants revealed extra details of the response to the biological treatment. For example, the total amount of synovial tissue with a ktrans value >0 decreased by 40% in 6 months in one participant (Figure 2); DCE-MRI revealed that most of the remaining “synovitis” at 6 months was of Low permeability class with almost all of the High permeability class of synovitis disappearing after drug treatment.
Conclusions Novel 3D DCE-MRI measures are practical in RA MR imaging trials, and offer sensitivity to differential tissue response not visible with other methods. Full hand coverage maximises the potential to monitor disease, and allows the pharmacokinetic modelling of drugs used in RA.
Disclosure of Interest C. Roberts Employee of: Bioxydyn, G. Guillard Shareholder of: Imorphics, Employee of: Imorphics, M. Bowes Shareholder of: Imorphics, Employee of: Imorphics, R. O'Connor Employee of: Imorphics, N. Maguire Employee of: Bioxydyn, A. Morgan Employee of: Bioxydyn, G. Parker Shareholder of: Bioxydyn, Employee of: Bioxydyn, R. Hodgson: None declared, E. Vital: None declared, P. Bird: None declared, G. Vincent Shareholder of: Imorphics, Employee of: Imorphics, P. Emery: None declared, P. Conaghan: None declared
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.