Choice of pulse sequences for magnetic resonance imaging–based semiquantitative assessment of cartilage defects in osteoarthritis research: Comment on the article by Doré et al

Abstract

We would like to comment on magnetic resonance imaging (MRI) assessment of focal cartilage defects in osteoarthritis (OA) research from a radiologic point of view. Doré et al recently reported the association between baseline tibial bone area and tibial subchondral bone mineral density (BMD) with tibial cartilage defect development and cartilage volume loss (1). They concluded that bone area predicted medial and lateral cartilage defect development and medial cartilage volume loss, while subchondral BMD predicted medial defect development, but not cartilage loss. Tibial cartilage defects were graded on a scale of 0-4 using T1-weighted fat suppressed 3-dimensional (3-D) gradient-recalled acquisition in the steady-state sequence, which is one of several gradientrecalled echo (GRE) type of sequences. GRE-type sequences such as 3-D spoiled gradient-recalled acquisition in the steadystate (SPGR), fast low-angle shot, 3-point Dixon, and doubleecho steady-state are well suited for depicting the cartilage volume, and thus they are the sequence of choice for quantitative (volumetric) analysis of cartilage (2,3).However, for depicting subtle cartilage abnormalities, GRE-type sequences are less suited than fluid-sensitive sequences (such as fat-suppressed intermediate [IM]-weighted, proton density [PD]-weighted, or fast spin-echo [FSE] T2-weighted image sequences [4,5]). With fat-suppressed IM-, PD-, and T2-weighted FSE sequences, normal hyaline cartilage has intermediate signal intensity, and intraarticular fluid is bright and displays good contrast due to an "arthrographic" effect that identifies surface abnormalities as well as abnormalities in cartilage matrix (6) ( Figures 1A and B). Moreover, GRE-type sequences are very susceptible to artifacts (motion artifacts from long imaging time and susceptibility artifacts), leading to difficulty differentiating between true cartilage defect and signal changes due to artifact (Figures 1C and D).Recent studies applying MRI semiquantitative assessment of focal cartilage defects use fluid-sensitive sequences (7-9) rather than GRE-type sequences because of these difficulties. Doré and colleagues' choice of MR sequence for assessment of focal cartilage defects is therefore not appropriate, according to the latest reports that advocate using these techniques. The authors should have at least acknowledged these facts in their report.As a consequence of the inadequate choice of pulse sequences in their study, it is unclear what the results presented by this group pertaining to cartilage defects represent. Detection of lesions and the estimations of their size will be inaccurate, so adequately interpreting the results and drawing conclusions based upon them seems impossible. We would strongly encourage others to use appropriate MRI pulse sequences to evaluate focal cartilage abnormalities, and if there is uncertainty, to seek advice from adequately trained musculoskeletal radiologists when developing studies, in order to achieve highest possible image quality with currently available t...

Assessment of Cartilage Changes Over Time in Knee Osteoarthritis Disease‐Modifying Osteoarthritis Drug Trials Using Semiquantitative and Quantitative Methods: Pros and Cons

“…Interestingly, although the score obtained with the IW‐FSE sequence is higher than with the GRE sequence and the precision of grading is allegedly better in the GRE sequence (22, 24), neither sequence demonstrated a superior sensitivity to detect cartilage defect change over 2 years, since the SRM values were similar between the 2 sequences. These present data do not support the finding that for the scoring of cartilage defects, the water‐sensitive FSE sequences are superior to the high‐resolution GRE sequences (20, 38). Results could have possibly been different if IW‐FSE sequences were performed with a higher spatial resolution (i.e., with the same voxel size as for the GRE sequences).…”

“…Clefts in the cartilage surface, for example, seem to be better depicted using water‐sensitive intermediate‐weighted fast spin‐echo (IW‐FSE) sequences with fat suppression (FS), which favor contrast resolution, i.e., differentiation of tissue with different water content (17–19). On the other hand, the spatial resolution is superior with GRE sequences, which allow for less partial volume effects and a more precise delineation of the cartilage (12, 20–22), as reflected by a better specificity than FSE when using arthroscopic grading as a reference (23, 24). Peterfy et al (3) suggested the use of 2 sequences simultaneously for the scoring (WORMS): an FS–3‐dimensional (3‐D) spoiled gradient‐recalled (SPGR) sequence and an FS–T2‐weighted FSE sequence, unfortunately without specifically indicating how to merge the findings.…”

Higher dynamic medial knee load predicts greater cartilage loss over 12 months in medial knee osteoarthritis

“…This is somewhat surprising given that we have previously reported cross-sectional relationships between these parameters,17 18 and in the current study we observed a relationship between loading and loss of cartilage volume. It could be argued that the MRI sequence we used is less suited than fluid-sensitive sequences to depict focal cartilage defects as it is susceptible to artefacts 43. This may have led to a degree of misclassification and reduced our ability to detect a relationship.…”

Why radiography should no longer be considered a surrogate outcome measure for longitudinal assessment of cartilage in knee osteoarthritis

“…In contrast, GRE sequences such as 3D spoiled gradient echo at a steady state (SPGR), fast low angle shot (FLASH), 3-point Dixon, and double echo steady state (DESS) are well suited for quantitative (volumetric) analysis of cartilage [55,56]. Use of inappropriate pulse sequences will inevitably invalidate the results of a study [57]. Expert musculoskeletal radiologists who can ensure the use of adequate imaging protocol may be available in only a limited number of institutions.…”