TITLE:
Reliability and sensitivity to change of IW-TSE versus DESS magnetic resonance imaging sequences in the assessment of bone marrow lesions in knee osteoarthritis patients: Longitudinal data from the Osteoarthritis Initiative (OAI) cohort
AUTHORS:
Jean-Pierre Raynauld, Lukas Martin Wildi, François Abram, Thomas Moser, Jean-Pierre Pelletier, Johanne Martel-Pelletier
KEYWORDS:
Osteoarthritis; Magnetic Resonance Imaging; Bone Marrow Lesions
JOURNAL NAME:
Journal of Biomedical Science and Engineering,
Vol.6 No.3A,
March
29,
2013
ABSTRACT: Background: Bone marrow lesions (BMLs) are associated
with osteoarthritis (OA). We assessed the performance of two commonly used
MRI sequences, IW-TSE and DESS, for reliability in the detection of BMLs and
sensitivity to estimate change over time. We suggested that the IW-TSE would
demonstrate higher sensitivity to change than DESS in the assessment of BML
prevalence and change over time. This study was performed using a subset of the
Osteoarthritis Initiative (OAI) cohort. Methods: A sub-group of 144 patients
was selected from the OAI progression cohort who all had IW-TSE and DESS MRI
acquisitions at baseline and 24 months. BMLs were assessed using a
semi-quantitative scale in the global knee, medial and lateral compartments,
and subregions. Intra-reader reliability was assessed on a subset of 51
patients. Results: Intra-reader reliability was substantial for the global knee
≥ 0.64, medial ≥ 0.70, and lateral ≥ 0.63 compartments for IW-TSE and DESS. The
prevalence of BML detected at baseline was only slightly greater for IW-TSE
compared to DESS. The mean BML score at baseline was significantly higher (p ≤
0.006) for the IW-TSE than the DESS. However, mean change at 24 months was
similar for both sequences for all regions except the medial compartment (p = 0.034)
and medial femur (p = 0.015) where they were significantly higher for DESS than
IW-TSE. Moreover, the prevalence of BML change at 24 months was similar in all
regions except the global knee (p = 0.047)
and the lateral tibial plateau (p = 0.031). Conclusion: This study does
not suggest superior sensitivity to change of one sequence over the other for
almost all the regions. The only difference is a higher BML mean change over
time detected by the DESS sequence in the medial compartment and femur. These
data bring into perspective that both sequences seem equivalent regarding
their use for the assessment of BML in clinical trials.