-
Views
-
Cite
Cite
Michel D. Crema, Frank W. Roemer, Daichi Hayashi, Ali Guermazi, Comment on: Bone marrow lesions in people with knee osteoarthritis predict progression of disease and joint replacement: a longitudinal study, Rheumatology, Volume 50, Issue 5, May 2011, Pages 996–997, https://doi.org/10.1093/rheumatology/ker006
Close - Share Icon Share
Extract
Sir, We would like to comment on a recently published article investigating the relationship between bone marrow oedema-like lesions (BMLs) and longitudinal change in tibial cartilage volume and subsequent risk of knee joint replacement in subjects with knee OA [1]. BMLs are known to play a role in pain [2] and progression of structural deterioration [3–6] in subjects with knee OA. Thus, selection of adequate MRI pulse sequences that are able to assess the maximum extent of BMLs is crucial for evaluation. Tanamas et al. [1] used a T1-weighted fat-suppressed 3D gradient recall acquisition in the steady state to assess both BMLs and articular cartilage. Gradient recalled echo (GRE)-type sequences, such as the one used by the authors in their study, are insensitive to marrow abnormalities due to trabecular magnetic susceptibility or T2* effects, which may lead to underestimation of BML size [7–9]. Such sequences also demonstrate limited sensitivity in the detection of BMLs when using routine fast spin echo (FSE) sequences as the reference standard [10, 11]. This was summarized and published in a consensus statement by OMERACT and OARSI (OA Research Society International) in 2006 [12]: ‘GRE techniques, even with robust fat suppression or water excitation, are notoriously insensitive to marrow abnormality’. This consensus statement also postulated that bone marrow abnormalities associated with OA are most sensitively demonstrated with fat-suppressed T2-weighted FSE images and short-tau inversion recovery (STIR) images. A co-author of the present study was also an author of the consensus statement on MRI protocols for assessment of knee OA [1, 12]. Either T2-weighted, proton density (PD)-weighted, intermediate-weighted (IW) fat-suppressed FSE or STIR sequences should be applied to assess BMLs in OA research, which appear as ill-defined high-signal intensity abnormalities in the subchondral bone, in contrast to well-defined high-signal intensity abnormalities, which are related to subchondral cyst-like lesions. Although the authors acknowledged the absence of T2-weighted MRI as a limitation, the fact that GRE-type sequences such as spoiled gradient recalled echo (SPGR) might underestimate larger lesions or even miss some lesions was not mentioned (Fig. 1). Thus, it remains unclear how to interpret these results: the lesion sizes may have been inaccurately assessed and some lesions probably escaped detection entirely with the GRE-type sequence used in this study. We strongly recommend the use of appropriate MRI pulse sequences to determine tissue-specific abnormalities, and whenever there is uncertainty about the adequate protocol, collaborative assistance from trained musculoskeletal radiologists should be considered.
Comments