Discriminative Power of Different Combinations of Bones and Joints for Assessing Change in Rheumatoid Arthritis with MRI: Results Based on Pooled Data from Four Multisite Clinical Trials.

 Charles G Peterfy, MD, PhD1,Peter Countryman, PhD2,Julie DiCarlo, PhD2,Annarita Gabriele3,Tim M Shaw, BSC4,Andrew Anisfeld5,Wayne H Tsuji, MD6,Ewa Olech, MD7,Norman B Gaylis, MD8. 1Spire Sciences, LLC, Kentfield, CA,2Spire Sciences, LLC, San Francisco, California,3F Hoffmann La-Roche, Ltd,4Medical Science, F Hoffmann La-Roche, Ltd, Welwyn Gladen City Herts,5Genentech, a member of the Roche Group,6Clinical Research/Inflammation, Amgen, Seattle, WA,7Rheumatology, OK Medical Research Foundation, Oklahoma City, OK,8Arthritis & Rheumatic Disease, Aventura, FL

Background: We previously showed that different combinations of bones and joints in the hand and wrist varied in their ability to discriminate significant change in synovitis, osteitis and erosion with MRI in patients wtih RA in a multicenter clinical trial1. In this study, we corroborated these findings using pooled data from 4 multisite clinical trials.

Methods: A total of 522 RA patients from 4 clinical trials were included in the analysis. Three of these studies (459 patients) included 1.5T MRI, and 2 studies (118 patients) included 0.2T MRI. All scans used a hand frame to ensure reproducible positioning, and were read by the same 2 radiologists. Metacarpophalangeal joints (MCP) 1-5 and all bones and joints in the wrist were scored using the RAMRIS scale2 for erosion and osteitis in all 4 trials and synovitis in all but one of the 1.5T trials. Six combinations (RAMRIS, 23 bones and 7 joints in original RAMRIS; RAMRIS+, RAMRIS + MCP1; V-Sharp, MCP1-5 + 6 wrist bones in van der Heijde-Sharp X-ray score; G-Sharp, MCP1-5 + 5 wrist bones in Genant-Sharp X-ray score [V-Sharp – lunate]; MCPs, MCP1-5; Wrist, 15 bones and 3 joints in RAMRIS) were compared for percentages of patients changing (increase or decrease) ≥ smallest detectable change (SDC)3 at week 12 or 24.

Results: Frequency of involvement at baseline varied by location (Fig.1), but showed a similar pattern in all 4 studies and between patients with early (<2y) RA and those with established (>10y) RA (data not shown). Inter-reader agreement, expressed as intraclass correlation coefficient (ICC), also varied among locations (median = 0.84, 0.88 and 0.81, for erosion, osteitis and synovitis, respectively.) Percent of patients changing ≥SDC at 12 or 24 weeks varied among the different combinations (Fig. 2). Adding MCP-1 to RAMRIS (RAMRIS+) increased discrimination for erosion and synovitis. G-Sharp showed the highest discrimination for osteitis and synovitis, and was equivalent to RAMRIS for erosion. Excluding all wrist locations (leaving only MCPs) actually increased discrimination for erosion over that with RAMRIS, but decreased discrimination for osteitis or synovitis.


Conclusions: Filtering out noisy and infrequently involved bones and joints may improve discriminative power of RAMRIS. Determining the combination that optimally balances functional relevance with reading reliability requires further investigation.
1. Peterfy C, et al. EULAR 2010
2. Ostergaard M, et al. Ann Rheum Dis 2005;64(Suppl1):i3–i7
3. Bruynesteyn K et al. Ann Rheum Dis 2005;64:179–82

Keywords: magnetic resonance imaging (MRI), rheumatoid arthritis (RA), clinical trials, RAMRIS

Disclosure:

  • Charles Peterfy:
    Abbott Laboratories: Consulting fees
    Amgen Inc.: Consulting fees
    Biogen Idec: Consulting fees
    Bristol-Myers Squibb: Consulting fees
    Celgene: Consulting fees
    Genzyme Corporation: Consulting fees
    Hoffmann-La Roche, Inc.: Consulting fees
    Lilly USA, LLC.: Consulting fees
    Merck Pharmaceuticals: Consulting fees
    Novartis Pharmaceuticals Corporation: Consulting fees
    Genentech: Consulting fees
    Spire Sciences, LLC: Stock, stock options or bond holdings, Employment (full or part-time), Ownership or partnership
    Regeneron Pharmaceuticals, Inc.: Consulting fees
    Roche: Consulting fees
    Synarc, Inc.: Stock, stock options or bond holdings
    Wyeth Pharmaceuticals: Consulting fees
  • Peter Countryman:
    Spire Sciences, LLC: Employment (full or part-time)
  • Julie DiCarlo:
    Spire Sciences, LLC: Employment (full or part-time)
  • Annarita Gabriele:
    Roche: Employment (full or part-time)
  • Tim Shaw:
    Roche: Employment (full or part-time)
  • Andrew Anisfeld:
    Genentech: Employment (full or part-time)
  • Wayne Tsuji:
    Amgen Inc.: Employment (full or part-time)
  • Ewa Olech:
    Genentech: Research grants, Speakers' bureau, Consulting fees
  • Norman Gaylis:
    Genentech: Research grants, Consulting fees