Standardization of Compositional MRI of Knee Cartilage: Why and How

HomeRadiologyVol. 301, No. 2 PreviousNext Reviews and CommentaryFree AccessEditorialStandardization of Compositional MRI of Knee Cartilage: Why and HowRichard Kijowski Richard Kijowski Author AffiliationsFrom the Department of Radiology, New York University Grossman School of Medicine, 660 First Ave, 3rd Floor, New York, NY 10016.Address correspondence to the author (e-mail: [email protected]).Richard Kijowski Published Online:Sep 7 2021https://doi.org/10.1148/radiol.2021211957MoreSectionsPDF ToolsImage ViewerAdd to favoritesCiteTrack CitationsPermissionsReprints ShareShare onFacebookTwitterLinked In See also the article by Chalian et al in this issue.Richard Kijowski, MD, is a professor of radiology at New York University Grossman School of Medicine (New York, NY). He has a special clinical and research interest in osteoarthritis and cartilage imaging and serves as an associate editor of the journal Osteoarthritis and Cartilage.Download as PowerPointOpen in Image Viewer Compositional MRI can depict changes in cartilage macromolecular and water content and collagen fiber ultrastructure associated with cartilage degeneration before morphologic cartilage loss. As a result, compositional MRI is superior to radiographs and conventional MRI using fast-spin-echo and gradient-echo-based sequences in detecting early cartilage degeneration in clinical practice and monitoring disease-related and treatment-related changes in cartilage in osteoarthritis (OA) research studies. Whereas various compositional cartilage MRI techniques have been described, T2 and T1ρ are the most widely used methods with the largest body of literature (1). Two recent meta-analyses summarizing the results of more than 50 published studies concluded that knee cartilage T2 and T1ρ measurements can reliably distinguish between individuals with and without knee OA although greater standardization of MRI methods is required for biomarker validation (2,3).In this issue of Radiology, the Musculoskeletal Biomarker Committee of the Radiological Society of North America (RSNA) published a Quantitative Imaging Biomarkers Alliance (QIBA) Profile for compositional cartilage MRI to provide recommendations for obtaining reproducible measurements of cartilage T2 and T1ρ of the knee joint. The QIBA Profile provided two complementary claims on the reproducibility of compositional cartilage MRI biomarkers from a comprehensive review of the literature. First, knee cartilage T2 and T1ρ can be measured with high scan-rescan precision with a within-subject coefficient of variation between 4% and 5%. Second, the minimum detectable change with 95% confidence of knee cartilage T2 and T1ρ in a single individual over time is 14% given this scan-rescan precision. The minimum detectable change decreases to 12% if only an increase in cartilage T2 and T1ρ is expected because of the progression of cartilage degeneration. However, these claims are only valid if compositional cartilage MRI is performed on the same 3.0-T scanner model from the same manufacturer by using the proposed imaging protocol in individuals with limited existing knee cartilage degeneration, defined as knees without joint space loss on radiographs (4).Whereas the claims provided in the QIBA Profile appear to be technical in nature and primarily focused on OA research and clinical drug trials, the information provided could have high potential impact on routine clinical care. The lack of reproducible compositional cartilage MRI biomarkers has mainly limited OA research studies to small patient samples evaluated on the same MRI scanner at a single institution. Standardization of compositional cartilage MRI biomarkers across different scanner models from different manufacturers would allow for large multicenter OA research studies and a better comparison of quantitative imaging data acquired at different institutions. Longitudinal cartilage T2 measurements acquired in middle-aged and older adults in the Osteoarthritis Initiative have led to a wealth of important information on knee OA (5). Imagine the potential knowledge if measurements of knee cartilage T2 and T1ρ could be standardized and analyzed from larger and more diverse patient populations as part of routine clinical care at hundreds of institutions worldwide. Larger patient samples evaluated during clinical knee MRI would facilitate cost-effective research for a better understanding of disease mechanisms and a more rapid investigation of the efficacy of surgical and pharmaceutical therapies.In addition, standardization of cartilage MRI biomarkers across different scanner models from different manufacturers would allow for the development of normative data of cartilage T2 and T1ρ measurements on different articular surfaces for individuals of different ages, sex, and ethnicity. The feasibility of creating normative cartilage T2 measurements has been previously described in the Osteoarthritis Initiative (6). MRI vendors are devoting resources to developing rapid cartilage T2 and T1ρ mapping sequences and fully automated methods for cartilage segmentation and cartilage T2 and T1ρ analysis for commercial use. This will make it possible in the near future to obtain regional measurements of cartilage T2 and T1ρ in all patients undergoing routine knee MRI. The development of normative data and threshold values of MRI cartilage biomarkers could be used to define abnormal cartilage compositional values and values associated with different disease burdens and different risks of disease onset and progression. Early disease diagnosis and preventive strategies that target individuals at high risk for knee OA without imaging findings of joint degeneration could maximize the effectiveness of conservative interventions (7). Furthermore, identifying patients with early knee OA at high risk for disease progression could help triage referrals for more expensive and invasive treatment options, which would ultimately improve clinical outcomes while reducing health care costs.Measuring knee cartilage T2 and T1ρ is a complex process consisting of acquiring images by using compositional and high spatial resolution morphologic sequences. Cartilage is segmented on the morphologic images to create three-dimensional cartilage masks. The cartilage masks are divided into six regions, defining each articular surface. Each region is occasionally divided into additional compartments for subregional analysis and into superficial and deep halves for laminar analysis. The cartilage masks are then superimposed over the compositional images to measure average T2 and T1ρ in each region. Errors occurring at any step in the process can lead to variations in measured cartilage T2 and T1ρ when the same in individual is scanned at different points. Coefficients of variation for scan-rescan precision for the six articular surfaces range between 2.3% and 6.5% for T2 measurements and between 4.2% and 7.4% for T1ρ measurements with much higher coefficients of variance for subregional and laminar analysis (2,3). These inherent variations in cartilage T2 and T1ρ may serve as confounding factors that make it challenging to detect changes in compositional cartilage MRI biomarkers because of cartilage degeneration.The QIBA Profile provides a detailed description of the steps needed to maximize the repeatability of compositional cartilage MRI biomarkers when the same 3.0-T scanner model from the same manufacturer is used. Installation of the pulse sequences and coils should be performed by using manufacturer-defined protocols. The same 3.0-T scanner model from the same manufacturer should be used with the same minimum eight-channel phased-array knee coil. A magnetization-prepared angle-modulated partitioned k-space spoiled gradient-echo snapshots (MAPPS) sequence should be used to measure cartilage T2 and T1ρ with parametric maps reconstructed with pixel-by-pixel nonlinear fitting. Semiautomated or fully automated software should be used for cartilage segmentation on morphologic images acquired by using a high-spatial-resolution fat-suppressed three-dimensional gradient-echo–based sequence such as fast acquisition double echo (known as FADE), dual-echo steady-state (known as DESS), or multiecho in steady-state acquisition (known as MENSA). A quality assurance program should also be in place to confirm adequate repeatability. This quality assurance program should ensure monthly assessments of scan-rescan precision by using calibration phantoms. Periodic conformity assessments by using phantoms and volunteers should also be performed when different coils are used or when vendor hardware and software are updated.The QIBA Profile has limitations: Its claims are only valid if compositional cartilage MRI is performed by using the same 3.0-T scanner model from the same manufacturer with the proposed imaging protocol. The proposed protocol requires the use of a MAPPS sequence for measuring cartilage T2 and T1ρ. The MAPPS sequence is currently not commercially available on any MRI vendor platform although it can be obtained on most MRI vendor platforms as a research prototype. Furthermore, additional work will be needed to standardize cartilage MRI biomarkers across different scanner models from different manufacturers, which will be a challenging task. A recent study has shown that the coefficients of variance for measuring knee cartilage T2 and T1ρ by using the same MAPPS sequence performed on three different 3.0-T scanner models from three different manufacturers were 10% for T2 and 8% for T1ρ (8). Coefficients of variance are even higher when different MRI sequences are used to measure cartilage T2 and T1ρ on different scanner models from different manufacturers (8–10). Future standardization of compositional cartilage MRI biomarkers will require all MRI vendors to agree on a single sequence for measuring cartilage T2 and T1ρ and then work together to limit variability in factors such as MRI system and coil design, B0 homogeneity profiles, shimming algorithms, and reconstruction and postprocessing methods to better standardize quantitative measurements (8).The QIBA Profile serves as an important first step toward standardizing compositional cartilage MRI biomarkers. The claims provided in the QIBA Profile will allow standardization of cartilage T2 and T1ρ measurements for multicenter cross-sectional and longitudinal studies performed on the same 3.0-T scanner model from the same manufacturer by using the proposed imaging protocol. However, additional future collaborative efforts between the Musculoskeletal Biomarker Committee and MRI vendors will be needed to accomplish the ultimate goal of better standardization of compositional cartilage MRI biomarkers across different scanner models from different manufacturers at different institutions.Disclosures of Conflicts of Interest: R.K. disclosed consulting fees from Boston Imaging Core Lab.References1. Guermazi A, Alizai H, Crema MD, Trattnig S, Regatte RR, Roemer FW. Compositional MRI techniques for evaluation of cartilage degeneration in osteoarthritis. Osteoarthritis Cartilage 2015;23(10):1639–1653. Crossref, Medline, Google Scholar2. MacKay JW, Low SBL, Smith TO, Toms AP, McCaskie AW, Gilbert FJ. Systematic review and meta-analysis of the reliability and discriminative validity of cartilage compositional MRI in knee osteoarthritis. Osteoarthritis Cartilage 2018;26(9):1140–1152. Crossref, Medline, Google Scholar3. Atkinson HF, Birmingham TB, Moyer RF, et al. MRI T2 and T1ρ relaxation in patients at risk for knee osteoarthritis: a systematic review and meta-analysis. BMC Musculoskelet Disord 2019;20(1):182. Crossref, Medline, Google Scholar4. Chalian M, Li X, Guermazi A, et al. The QIBA Profile for MRI-based Compositional Imaging of Knee Cartilage. Radiology 2021.https://doi.org/10.1148/radiol.2021204587. Published online September 7, 2021. Link, Google Scholar5. Eckstein F, Kwoh CK, Link TM;. OAI investigators. Imaging research results from the osteoarthritis initiative (OAI): a review and lessons learned 10 years after start of enrolment. Ann Rheum Dis 2014;73(7):1289–1300. Crossref, Medline, Google Scholar6. Joseph GB, McCulloch CE, Nevitt MC, et al. A reference database of cartilage 3 T MRI T2 values in knees without diagnostic evidence of cartilage degeneration: data from the osteoarthritis initiative. Osteoarthritis Cartilage 2015;23(6):897–905. Crossref, Medline, Google Scholar7. Felson DT, Hodgson R. Identifying and treating preclinical and early osteoarthritis. Rheum Dis Clin North Am 2014;40(4):699–710. Crossref, Medline, Google Scholar8. Kim J, Mamoto K, Lartey R, et al. Multi-vendor multi-site T1ρ and T2 quantification of knee cartilage. Osteoarthritis Cartilage 2020;28(12):1539–1550. Crossref, Medline, Google Scholar9. Li X, Pedoia V, Kumar D, et al. Cartilage T1ρ and T2 relaxation times: longitudinal reproducibility and variations using different coils, MR systems and sites. Osteoarthritis Cartilage 2015;23(12):2214–2223. Crossref, Medline, Google Scholar10. Balamoody S, Williams TG, Wolstenholme C, et al. Magnetic resonance transverse relaxation time T2 of knee cartilage in osteoarthritis at 3-T: a cross-sectional multicentre, multivendor reproducibility study. Skeletal Radiol 2013;42(4):511–520. Crossref, Medline, Google ScholarArticle HistoryReceived: Aug 2 2021Revision requested: Aug 10 2021Revision received: Aug 17 2021Accepted: Aug 20 2021Published online: Sept 07 2021Published in print: Nov 2021 FiguresReferencesRelatedDetailsAccompanying This ArticleThe QIBA Profile for MRI-based Compositional Imaging of Knee CartilageSep 7 2021RadiologyRecommended Articles Association of Mucoid Degeneration of the Anterior Cruciate Ligament at MR Imaging with Medial Tibiofemoral Osteoarthritis Progression at Radiography: Data from the Osteoarthritis InitiativeRadiology2018Volume: 287Issue: 3pp. 912-921The QIBA Profile for MRI-based Compositional Imaging of Knee CartilageRadiology2021Volume: 301Issue: 2pp. 423-432Deep Learning Approach for Evaluating Knee MR Images: Achieving High Diagnostic Performance for Cartilage Lesion DetectionRadiology2018Volume: 289Issue: 1pp. 160-169State of the Art: Imaging of Osteoarthritis—Revisited 2020Radiology2020Volume: 296Issue: 1pp. 5-21Can Signal Abnormalities Detected with MR Imaging in Knee Articular Cartilage Be Used to Predict Development of Morphologic Cartilage Defects? 48-Month Data from the Osteoarthritis InitiativeRadiology2016Volume: 281Issue: 1pp. 158-167See More RSNA Education Exhibits Don't Be Afraid About T2 Mapping of the Knee Articular Cartilage: A Simple Way to Decrease Reports Ambiguity and Improve the Communication with ArthroscopistsDigital Posters2022Articular Cartilage Injury of the Knee Reporting and Data System (ACIK-RADS): A New Proposal for the Report of Cartilage Injury in the Knee by MRIDigital Posters2019Multimodality Imaging Of Synovitis In OsteoarthritisDigital Posters2021 RSNA Case Collection Osteochondritis dissecansRSNA Case Collection2020Osteoarthritis of L hipRSNA Case Collection2020Chondroblastoma of the glenoidRSNA Case Collection2021 Vol. 301, No. 2 Metrics Altmetric Score PDF download