Arthritis & Rheumatism, Volume 62,
November 2010 Abstract Supplement

Abstracts of the American College of
Rheumatology/Association of Rheumatology Health Professionals
Annual Scientific Meeting
Atlanta, Georgia November 6-11, 2010.

TNF Blocking Agents Promote Resolution of Vertebral Erosions in Patients with Spondyloarthritis.

Maksymowych2,  Walter P., Chiowchanwisawakit1,  Praveena, Lambert2,  Robert

Mahidol University, Bangkok, Thailand
University of Alberta, Edmonton, AB, Canada


While the impact of anti-TNF agents on new bone formation remains unclear their impact on other structural lesions such as erosions has not yet been studied in SpA because these lesions are not commonly observed in the spine on radiography. Bone erosion is readily discerned on T1-weighted (T1W) MRI at vertebral corners (VC) and affecting the vertebral endplate. MRI can reliably detect these lesions and systematic study has demonstrated erosions in the majority of patients. It is possible that such lesions could be a secondary cause of symptoms in SpA. We aimed to assess the impact of both TNF blocking agents and standard treatment on the evolution of bone erosions using MRI.


MRI scans were performed at baseline and 2 years in 61 AS patients of whom 28 received TNF blocking agents in open label follow up of clinical trials while 33 received either TNF blocking agents (n = 16) or standard therapy (n = 17) in an observational cohort. A bone erosion on MRI was defined as full-thickness loss of the dark appearance of cortical bone at its anticipated location and loss of the normal bright appearance of adjacent bone marrow on T1W MRI. Via televideoconference, reference images were developed in which erosions were assigned by consensus amongst an international MRI working group. Lesions were independently recorded dichotomously (present/absent) from lower C2 to the upper sacrum of the spine at both anterior and posterior vertebral corners and at each vertebral endplate. Anonymized MRI scans were assessed independently by 2 readers who were blinded to treatment and time point. The primary analysis compared the resolution of lesions according to treatment and was based on concordant data indicating agreement on resolution of erosions. Improvement or complete resolution was defined on the basis of >= 50% and complete restoration, respectively, towards a normal vertebral contour. Proportions were compared by Fisher's exact test.


Analysis was based on concordant reads for 2801 VC and vertebral endplate sites. Complete resolution of VC erosions was recorded significantly more frequently in the TNF blocker treatment group (11 of 13 erosions (84.6%) whereas this was not observed in a single erosion in patients on standard therapy (p = 0.0002). Similar differences between treatments were noted for individual readers. Vertebral endplate erosions were more frequent than corner erosions but showed minimal resolution after 2 years regardless of treatment.

Table. Number (percentage) of vertebral erosions resolving.

 VC ErosionVertebral Endplate Erosion
TreatmentResolvedImprovedNo changeResolvedImprovedNo change
        Concordant11 (84.6)02 (15.4)2 (4.8)2 (4.8)40 (90.4)
        Reader 126 (74.3)2 (5.7)7 (20)8 (9.4)23 (27.1)54 (63.5)
        Reader 219 (59)011 (41)17 (15.7)3 (2.8)88 (81.5)
        Concordant008 (100)0017 (100)
        Reader 11 (7.7)4 (30.8)8 (61.5)06 (22.2)21 (77.8)
        Reader 27 (31.8)4 (18.2)11 (50)4 (11.4)4 (11.4)27 (77.2)


Our preliminary data supports a beneficial effect of TNF blocker treatment on the resolution of vertebral erosions. The lack of impact on vertebral endplate erosions may reflect greater time dependence due to larger size and/or differences in etiology.

To cite this abstract, please use the following information:
Maksymowych, Walter P., Chiowchanwisawakit, Praveena, Lambert, Robert; TNF Blocking Agents Promote Resolution of Vertebral Erosions in Patients with Spondyloarthritis. [abstract]. Arthritis Rheum 2010;62 Suppl 10 :1959
DOI: 10.1002/art.29724

Abstract Supplement

Meeting Menu