Magnetic resonance imaging as a noninvasive means for quantitating the dimensions of articular cartilage in the human knee.код для вставкиСкачать
ARTHRITIS & RHEUMATISM Vol. 50, No. 1, January 2004, pp 5–9 DOI 10.1002/art.11495 © 2004, American College of Rheumatology EDITORIAL Magnetic Resonance Imaging as a Noninvasive Means for Quantitating the Dimensions of Articular Cartilage in the Human Knee Laurie D. Hall The report by Cicuttini and colleagues in this issue of Arthritis & Rheumatism (1) is an important demonstration of the substantial progress investigators have made in developing magnetic resonance imaging (MRI) methods for measuring the dimensions of human articular cartilage, particularly in the knee (2–4). Importantly, the findings reported by Cicuttini et al indicate the utility of MRI in monitoring osteoarthritis (OA)– related loss of knee cartilage during a 2-year period. While the most ergonomically efficient method for deriving the key parameters for assessing disease progression from scans is an open question, the good news is that the studies by Cicuttini and others are attracting the attention of a growing number of research groups, as well as of consortia sponsored by government (5) and individual pharmaceutical companies (6). The diverse needs and interests of those groups will ensure that a wide range of methods will become available, each optimized for a specific purpose. Among clinical and radiographic techniques to evaluate the articular joint, MRI occupies a unique position, as highlighted in Table 1. While a major focus of MRI is the articular cartilage, the same 3-dimensional (3-D) MRI scans that are used to visualize the articular cartilage also provide information about the joint as a “complete organ,” since the scans reveal all of the major soft tissues within the joint capsule (meniscal disks, ligaments) and those surrounding it (tendons, musculature, vasculature); depending on the protocol used, the cortical and trabecular bone can also be assessed. The study by Cicuttini and colleagues (1) and earlier investigations of cartilage (7–11) explore one facet of the data available from this powerful technique. In evaluating the utility of MRI, Cicuttini et al used data from a study of 117 patients with radiographically evident OA, who underwent MRI scanning at study entry and 2 years later. Although all scans were acquired using the sagittal orientation, they were transformed to the coronal plane prior to measurements (see below). The boundaries of the cartilage were determined manually using a defined set of rules; since that approach is laborious and subject to observer error, it is a testament to the training of the investigators that the interobserver coefficients of variation (CVs) were only 2–3% (1). Indeed, a main purpose of their study was to demonstrate that, in the future, substantial time could be saved if measurement of only the volume of the tibial cartilage is used for assessment of disease progression. Thus, they investigated the correlations between cartilage volumes in the separate compartments of the knee. Their finding of a strong correlation between changes in femoral cartilage volume and changes in tibial cartilage volume in the medial and lateral tibiofemoral joints led to their suggestion that measurement of tibial cartilage will suffice for monitoring disease progression. Cicuttini and coworkers discuss several aspects of their study that reflect their original decision to acquire the scans with the sagittal orientation (1). In such scans, however, it can be difficult to define the cartilage boundaries; hence, as indicated above, they reformatted the original data into the coronal orientation prior to volume measurement. Clearly, in addition to the work involved, this reformatting can introduce a source of error which could be eliminated in future studies. For interpretation of the existing data, the authors compared tibial cartilage volumes calculated from the original sagittally oriented scans with those from the reformatted coronal scans. With the latter, the lateral tibial volume was systematically overestimated by 3.5% and the medial volume underestimated by 3.8%. Importantly, the authors argue that, for studies of clinical progression, it Laurie D. Hall, PhD, DSc: University of Cambridge School of Clinical Medicine, Cambridge, UK. Address correspondence and reprint requests to Laurie D. Hall, PhD, DSc, Herchel Smith Laboratory for Medicinal Chemistry, University of Cambridge School of Clinical Medicine, Robinson Way, Cambridge CB2 2PZ, UK. E-mail: email@example.com. Submitted for publication June 23, 2003; accepted in revised form October 13, 2003. 5 6 HALL Table 1. Summary of the strengths and weaknesses of 4 methods for assessing osteoarthritic damage to the human knee Method Strengths/weaknesses Clinical rheumatologic assessment Radiographic measurement of joint space Not suitable for assessing early damage Measures late-stage damage Does not measure cartilage quality Invasive Views joint as a complete organ Visualization of soft tissues (cartilage, meniscal disks, tendons, ligaments) Visualization of bone (cortical and trabecular) Visualization of surroundings (musculature, vasculature) Arthroscopy Magnetic resonance imaging is the rate of cartilage change that matters rather than the absolute volume, and that the rates derived from reformatted scans are similar to those from the sagittal MRI (10). The database used by Cicuttini and colleagues is a valuable resource and, regardless of whether others will wish to use this specific measurement protocol, the take-home message is certainly that further MRI studies of OA progression should be encouraged. In this context, it is important to consider the substantial progress made by the 2 other groups referenced in the report by Cicuttini et al, namely Peterfy and colleagues in the US (2) and Eckstein and colleagues in Germany (3); both of these groups continue to be very active (12,13). In common with Cicuttini and coworkers, those investiga- tors acquired water-only 3-D scans from which they derived measurements of a range of cartilage dimensions, including volume and thickness. While space does not permit a detailed appraisal of their many studies, it is important to note that Eckstein et al have demonstrated that repeated measurements can be made by different independent observers, with remarkably small CVs. Clearly, this reproducibility of measurement results obtained by different operators is extremely important in assessing OA damage and its progression, as well as measuring quantitatively the efficacy of any pharmaceutical treatment. In that context, a recent publication (6) from a collaborative group provides evidence of the precision with which repeated measurements can be made with the aid of Figure 1. Computer visualization and measurement of the femoral articular cartilage, as derived from a 3-dimensional magnetic resonance imaging scan of the knee of a 57-year-old man. Left, Both the external and internal surfaces of the cartilage and the interface with the cortical bone; total volume is calculated to be 13.16 ml. Right, The same scan data following computer-based “punch biopsy” of a specified region (black rectangle) are shown; the calculated volume of the “biopsy specimen” is 0.14 ml. EDITORIAL 7 Figure 2. Left, Two-dimensional magnetic resonance imaging scan of the distal interphalangeal (DIP) joint of a healthy 23-year-old man, measured at 2.4T, slice thickness 1.0 mm, in-plane resolution 0.075 ⫻ 0.15 mm. All of the major structures are clearly seen, including details of the architecture of the trabecular bone. Importantly, the surface cartilage (sc) is clearly defined by a line of low signal intensity between the high intensity from synovial fluid (sf) and the rest of the cartilage. Right, Sagittal and coronal scans of 2 DIP joints of an osteoarthritis (OA) patient, measured using a protocol in which the signal from the lipid in the bone marrow is suppressed (same scanning parameters as used for the image on left). The upper pair of scans is from a joint in which the effect of age is demonstrated by the loss of clear-cut signal from the surface cartilage. The lower pair is from a joint with advanced OA damage, which clearly shows a substantial osteophyte and focal loss of the adjacent articular cartilage. ip ⫽ interphalanx; sms ⫽ synovial membrane support; ac ⫽ articular cartilage; sus ⫽ subungual space; dp ⫽ distal phalanx; fdp ⫽ flexor digitorum profundus; vp ⫽ volar plate. computer software. That study included data from 10 healthy young volunteers plus 8 randomly selected OA patients and is perhaps the most detailed investigation of its kind, clearly demonstrating the overall robustness of the MRI method when used by a carefully trained team. Given the current status of MRI measurements of knee cartilage, it is relevant to question how this methodology might best be used for monitoring clinical efficacy in pharmaceutical trials. In addition to the problems associated with individual measurements, there are two important issues related to such trials. The first is the need for repeated measurements over a period of months, and possibly years, which in turn necessitates methodology by which repeated measurements of the same cartilage regions can be made. The second is the innate variability of cartilage dimensions between individual subjects. Furthermore, it is not obvious how best to incorporate data from age-matched controls. Among other approaches, one solution to all three of these problems may be the use of MRI as a form of “noninvasive punch biopsy”(4). While utilizing the same basic 3-D MRI scan techniques, this approach differs in the method used to assess cartilage dimensions. Thus, software is used to measure the volume of many small regions of cartilage selected by inspection of the first scan from each patient (Figure 1). Some of those cover regions that exhibit clinical damage or are known statistically to be prone to wear and damage; others are intentionally located remote from such areas and are to be used as the “age-matched controls.” In “noninvasive punch biopsy,” the settings of the software defining each of those spatially localized regions can be used to measure the same regions in each of the series of 3-D MRI scans acquired during the overall duration of a trial. This procedure necessitates coregistration of each of these subsequent scans with the original, using technology that is available in a variety of forms. The report by Kshirsagar et al (4) amply demonstrates the technical feasibility of this approach: in 4 repeated measurements of 4 separate regions of cartilage, the CV was 2%. More recent developments by this author in the Herchel Smith Laboratory have further 8 enhanced the ergonomics of this approach such that all parts of the measurements are now automated. That automation has substantially improved the overall precision and reproducibility of the measurements: remarkably, in normal subjects and those with early-stage damage, repeated measurements can be made with CVs of essentially 0. It remains to be seen how well this methodology performs for monitoring damaged regions during clinical trials. Nevertheless, the approach demonstrates the potential for effectively using local regions of undamaged cartilage in the subjects’ damaged knee as “age-matched controls” against which either progression of damage, or repair, can be assessed. What of the future? It is clear that the pioneering work of several investigative groups has, at long last, now started to attract many other researchers from a diverse range of backgrounds, some of whom are forming consortia to tackle a complex area that requires integration of the disparate skills of interdisciplinary teams. It is also clear that it is important to create a substantial database of MRI scans, which could facilitate understanding of the etiologic relationships between clinical assessments and radiologic findings in patients with OA. Indeed, such comparisons alone will provide muchneeded information on the clinical diagnosis of OA. Furthermore, measurements from scans repeated over a period of years will provide unique insight into the progression of OA; at present, no other method can provide such knowledge. In considering the future of MRI studies, it is important to note the large commitment to this endeavor made by the National Institutes of Health (NIH) (5). The NIH will support the acquisition of 4 new MRI scanners that will be used in identical protocols over a period of 5 years to create a huge database of knee MRI scans. Importantly, each subject in this study will also be assessed clinically and by standard radiographic measurements. Given the present knowledge about MRI scanning protocols, it is likely that the quality of the scans will be state-of-the-art. It is important that future improvements in image measurement software be accommodated within the program and that the scan data be made widely available to other researchers. Although the precise scanning and measurement protocols have not yet been determined, the overall objectives of the program were discussed at 2 meetings organized by the NIH and are described on the NIH Web site (5). The use of MRI in studies of the etiology and clinical course of OA will likely expand, especially as HALL new technology is developed and refined, and applied even to small joints. Indeed, as shown in studies in this laboratory (14) and others, MRI can now provide sufficiently high spatial resolution to visualize cartilage damage in finger joints (Figure 2), specifically the distal interphalangeal joint. Furthermore, with MRI protocols, it is possible to quantitatively assess structural properties of articular cartilage with precision and to distinguish OA-damaged cartilage from that of clinically normal “age-matched controls,” using 2 parameters referred to as the “spin-spin relaxation time” (T2 value) and the “magnetization transfer rate.” These parameters reflect the interactions of water molecules with collagen and proteoglycan and enable quantitation of the local concentrations of all 3 species. Although much further work is needed before widespread clinical application of MRI can be achieved, the future of this technology appears very bright, given its potential to provide an unparalleled picture of the joint in OA. Hopefully this picture, which will be richer and more detailed than anything before, will lead to both new understanding of the disease and new approaches to treatment. REFERENCES 1. Cicuttini FM, Wluka AE, Wang Y, Stuckey SL. Longitudinal study of changes in tibial and femoral cartilage in knee osteoarthritis. Arthritis Rheum 2004;50:94–7. 2. Peterfy CG, van Dijke CF, Janzen EL, Gluer CC, Namba R, Majumdar S, et al. Quantification of articular cartilage in the knee with pulsed saturation transfer subtraction and fat-suppressed MR imaging: optimization and validation. Radiology 1994;192:485–91. 3. Eckstein F, Schnier M, Haubner M, Priebsch J, Glaser C, Englmeier KH, et al. Accuracy of cartilage volume and thickness measurements with magnetic resonance imaging. Clin Orthop 1998;352:137–48. 4. Kshirsagar AA, Watson PJ, Tyler JA, Hall LD. Measurement of localised cartilage volume and thickness of human knee joints by computer analysis of 3D images. Invest Radiol 1998;289:299–33. 5. National Institute of Arthritis and Musculoskeletal and Skin Diseases. Osteoarthritis Initiative. URL: www.niams.nih.gov/ne/oi/ imaging.htm. 6. Raynauld J-P, Kauffmann C, Beaudoin G, Berthiaume M-J, de Guise JA, Bloch DA, et al. Reliability of a quantification imaging system using magnetic resonance images to measure cartilage thickness and volume in human normal and osteoarthritic knees. Osteoarthritis Cartilage 2003;11:351–60. 7. Cicuttini F, Forbes A, Morris K, Darling S, Bailey M, Stuckey S. Gender differences in knee cartilage volume as measured by magnetic resonance imaging. Osteoarthritis Cartilage 1999;7: 265–71. 8. Jones G, Glisson M, Hynes K, Cicuttini F. Sex and site differences in cartilage development: a possible explanation for variations in knee osteoarthritis in later life. Arthritis Rheum 2000;43:2543–9. 9. Cicuttini FM, Wluka AE, Stuckey SL. Tibial and femoral cartilage changes in knee osteoarthritis. Ann Rheum Dis 2001;60:977–80. EDITORIAL 10. Wluka AE, Stuckey S, Snaddon J, Cicuttini FM. The determinants of change in tibial cartilage volume in osteoarthritic knees. Arthritis Rheum 2002;46:2065–72. 11. Cicuttini FM, Wluka A, Forbes A, Wolfe R. Comparison of tibial cartilage volume and radiologic grade of the tibiofemoral joint. Arthritis Rheum 2003;48:682–8. 12. Peterfy CG. Imaging of the disease process. Curr Opin Rheumatol 2002;14:590–6. 9 13. Eckstein F, Muller S, Faber SC, Englmeier KH, Reiser M, Putz R. Side differences of knee joint cartilage volume, thickness, and surface area, and correlation with lower limb dominance: an MRI-based study. Osteoarthritis Cartilage 2002;10:914–21. 14. Hodgson RJ, Barry MA, Carpenter TA, Hall LD, Hazleman BL, Tyler JA. Magnetic resonance imaging protocol optimisation for evaluation of hyaline cartilage in the distal interphalangeal joint of fingers. Invest Radiol 1995;30:522–31.