Proposal for a nomenclature for Magnetic Resonance Imaging based measures of articular cartilage in osteoarthritis

OsteoArthritis and Cartilage (2006) 14, 974e983 ª 2006 OsteoArthritis Research Society International. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.joca.2006.03.005

International Cartilage Repair Society

Proposal for a nomenclature for Magnetic Resonance Imaging based measures of articular cartilage in osteoarthritis1 F. Eckstein M.D.y*, G. Ateshian Ph.D.z, R. Burgkart M.D.x, D. Burstein Ph.D.k, F. Cicuttini Ph.D.{, B. Dardzinski Ph.D.#, M. Gray Ph.D.yy, T. M. Link M.D.zz, S. Majumdar Ph.D.xx, T. Mosher M.D.kk, C. Peterfy M.D., Ph.D.{{, S. Totterman M.D., Ph.D.##, J. Waterton Ph.D.yyy, C. S. Winalski M.D.zzz and D. Felson M.D., M.P.H.xxx y Institute of Anatomy & Musculoskeletal Research, Paracelsus Private Medical University, Salzburg, Austria & Chondrometrics GmbH, Ainring, Germany z Columbia University, New York, NY, USA x Department of Orthopedics, Klinikum rechts der Isar, Technische Universita¨t Mu¨nchen, Munich, Germany k Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA, USA { Department of Epidemiology and Preventive Medicine, Alfred Hospital, Melbourne, Victoria, Australia # Cincinnati Children’s Hospital Medical Center, University of Cincinnati, Cincinnati, OH, USA yy Massachusetts Institute of Technology (MIT), MA, USA zz Department of Radiology, University of California, San Francisco, CA, USA xx Musculoskeletal and Quantitative Imaging Research Group (MQIR), Department of Radiology, University of California, San Francisco, CA, USA kk Department of Radiology, Penn State Milton S. Hershey Medical Center, Hershey, PA, USA {{ Synarc, Inc., San Francisco, CA, USA ## University of Rochester & VirtualScopics LLC, Rochester, NY, USA yyy AstraZeneca, Macclesfield, Cheshire, UK zzz Department of Radiology, Harvard Medical School, Boston, MA, USA xxx Clinical Epidemiology Unit, Boston University School of Medicine, Boston, MA, USA Summary Objective: Magnetic resonance imaging (MRI) of articular cartilage has evolved to be an important tool in research on cartilage (patho)physiology and osteoarthritis (OA). MRI provides a wealth of novel and quantitative information, but there exists no commonly accepted terminology for reporting these metrics. The objective of this initiative was to propose a nomenclature for definitions and names to be used in scientific communications and to give recommendations as to which minimal methodological information should be provided when reporting MRI-based measures of articular cartilage in OA. Methods: An international group of experts with direct experience in MRI measurement of cartilage morphology or composition reviewed the existing literature. Through an iterative process that included a meeting with a larger group of scientists and clinicians (December 2nd, 2004, Chicago, IL, USA), they discussed, refined, and proposed a nomenclature for MRI-based measures of articular cartilage in OA. Results: The group proposes a nomenclature that describes: (1) the anatomical location and (2) the structural feature being measured, each name consisting of a metric variable combined with a tissue label. In addition, the group recommends minimal methodological information that should be described. Conclusions: Utilization of this nomenclature should facilitate communication within the scientific community. Further, the uniform adoption of comprehensive nomenclature to describe quantitative MRI- features of articular cartilage should strengthen epidemiological, clinical, and pharmacological studies in OA. ª 2006 OsteoArthritis Research Society International. Published by Elsevier Ltd. All rights reserved. Key words: Nomenclature, Cartilage, Magnetic Resonance Imaging, Osteoarthritis.

1 Funding Source: This initiative has been supported by grants from: Novartis Pharma (Basel, Switzerland); Pfizer Global Research & Development (Ann Arbor, MI, USA); Amgen (Thousand Oaks, CA, USA); Rotta Research Laboratories (Milan, Italy); Sanofi Aventis (Frankfurt, Germany); DePuy Spine, Inc. (Raynham, MA, USA) and AstraZeneca (Macclesfield, UK);NIH AR47785. *Address correspondence and reprint requests to: Prof. Felix Eckstein, Institute of Anatomy & Musculoskeletal Research, PMU, Strubergasse 21, A5020 Salzburg, Austria. Tel: 43-662-44-2002-1240; Fax: 43-662-44-2002-1249; E-mail: [email protected] Received 3 August 2005; revision accepted 11 March 2006.

974

975

Osteoarthritis and Cartilage Vol. 14, No. 10 Magnetic resonance imaging (MRI) of articular cartilage has become an important diagnostic tool in research on cartilage (patho)physiology and osteoarthritis (OA). The technique has proven to be particularly powerful when quantitative metrics that are continuous variables are derived from serial magnetic resonance (MR) images, using image segmentation and analysis algorithms. The use of MRI biomarkers as endpoints shows great promise in the research and management of OA as well as in the development of disease modifying OA drugs1e6 and may also have value with other joint diseases. Currently, however, there exists no commonly accepted definition and terminology for reporting these measures. That makes it difficult to compare different publications and extrapolate results so that one study can inform another. The objective of this initiative was to propose a nomenclature of MRI-based measures of articular cartilage for use in scientific reports, publications and databases, to facilitate communication between researchers and to ensure that comprehensive and high quality information on structural features is reported. Similar efforts have been made in other fields, such as bone histomorphometry, where the promulgation of standard nomenclature, symbols, and units7 led to widespread use of a common language for measurement between scientists, using comparable concepts of evaluation. This, in turn, made it easier for the scientific community to reach agreement upon how to interpret bone pathology findings and spurred an enhanced appreciation for effects of bone agents on bone histomorphometry8.

ANATOMICAL LABELS

The current proposal defines anatomical labels of knee joint cartilage, since the knee has been the focus of quantitative MRI measurements of cartilage to date. Rules for naming ‘‘anatomical labels’’ are: upper case digits label anatomically distinct cartilage plates with clear anatomical boundaries (‘‘P’’ ¼ patella, ‘‘F’’ ¼ femur, and ‘‘T’’ ¼ tibia), and modifiers (small case digits preceding the upper case digits) may label anatomical subregions of these plates. Based on these rules, labels for knee joint cartilage plates are listed in Table I and displayed in Figs. 1 and 2. Since the tibia bears two distinct cartilage plates with different degrees of involvement in different types of OA, it is recommended that values for the medial tibia (‘‘MT’’) and lateral tibia (‘‘LT’’) be reported separately (Fig. 1). Because the femur (‘‘F’’) provides the bearing surface of different joint compartments (femoropatellar, medial femorotibial, and lateral femorotibial), the group recommends corresponding anatomical labels to be used (Table I, Fig. 1). Further, because the femoral cartilage plate is continuous throughout these joint compartments, landmarks need to be defined for their separation. ‘‘TrF’’ (trochlea of the femur) may be used for aspects of the femoral cartilage located anterior

Methods Upon invitation of the first and senior author, an international focus group of experts (the authors) with direct experience in quantitative MRI measures of cartilage morphology or composition in OA was formed. After an informal review of the existing literature, an initial draft of the nomenclature was circulated amongst this ‘‘focus’’ group. This focus group then met on December 2nd, 2004 in Chicago (IL) (1) to discuss the proposal with a larger group of invited scientists and clinicians from academia, industry and government agencies (see Acknowledgment section) and (2) working with this larger group, to propose a nomenclature. Using a consensus approach, the group proposed short parameter names based on consistent rules, to keep naming conventions simple and easy to understand. The minutes of the meeting were circulated amongst the focus group and the larger group for further refinement (one iteration in the larger group, and five iterations within the focus group). The final format of the nomenclature and the recommendations was circulated amongst and approved on by the authors, prior to submission. The same applied to the revised version of this manuscript.

TrF

TrF

P

LT

MT

P pLF

TrF

cLF

cMF

cLF

LT

LT

MT

Results The group recommends a modular approach to the nomenclature: the first label identifies the anatomical location of the measurement (anatomical label) and the second label identifies the ‘‘structural feature’’ being measured. Each component should be separated by a full stop/period (‘‘.’’). Another suffix may be added to describe several statistical aspects of the structural feature being measured (‘‘statistical label’’), or the central tendency (e.g. mean) and dispersion (e.g. s.d.) can be described otherwise.

pMF

pLF

pMF

LT

MT

cMF

MT

Fig. 1. Anatomical labels of knee joint cartilage plates. Sagittal images (left row) and coronal images (right row).

976

F. Eckstein et al.: Nomenclature for MRI cartilage measures

to the intercondylar notch, and ‘‘MF’’ and ‘‘LF’’ for the medial and lateral aspects of femoral cartilage posterior to it (Fig. 1). Because the femoropatellar joint displays anatomical features separating a medial and lateral joint compartment9 (patellar ridge and trochlear groove), small letter modifiers may be used to define subregions (Fig. 2): ‘‘mP’’ and ‘‘mTrF’’ for the medial aspects of the patella and trochlea, and ‘‘lP’’ and ‘‘lTrF’’ for the lateral aspects. ‘‘cP’’ and ‘‘cTrF’’ may be used for the central aspect of the cartilage plates (around the ridge and groove e Fig. 2), but authors must report precisely how these regions are defined. The femoral condyles have frequently been separated into two or more components. The first contains the femoral cartilage that is in contact with the tibial cartilage or meniscus (during standing and walking). The group suggests these regions to be named central medial and central lateral femur (‘‘cMF’’ and ‘‘cLF’’). The second component is located posterior to the central area (Fig. 1) and does contact with the tibia only during knee flexion. This is suggested to be named posterior medial and posterior lateral femur (‘‘pMF’’ and ‘‘pLF’’). Note that ‘‘cMF’’ and ‘‘cLF’’ are directly interposed between ‘‘TrF’’ (anteriorly), and ‘‘pMF’’ and ‘‘pLF’’ (posteriorly). The group does not give a specific recommendation to where the separation should be made, since no obvious anatomical landmark exists to separate these regions and since there exist several reasonable proposals in the literature. However, authors are urged to give comprehensive information on how the separation is made.

tAB cAB lP

mP

cAB dAB

cP lTrF

a

b cMF.tAB cAB

AC

AC

cMF.ThCtAB

dAB

ThCcAB

0mm

cMF.AC

c

d

Fig. 2. (a) Anatomical labels of knee joint cartilage plates and (b) cartilage morphology labels for measurements of areas on axial images. This example displays an area of denuded cartilage in the center of the patella; (c) cartilage morphology labels for measurements of areas and (d) for measurements of cartilage thickness in the medial femoral condyle. This example displays an area of denuded cartilage in the external aspect of the medial femoral condyle. Note that the peripheral osteophyte is not included in the measurement of tAB. When reporting values for cartilage thickness, it is important to differentiate whether or not the values include denuded areas (‘‘dAB’’) as areas of 0 mm cartilage thickness. The group recommends the use of ‘‘ThCtAB’’ for measurements including ‘‘dAB’’ (0 mm cartilage thickness) and ‘‘ThCcAB’’ for those only including cartilage.

The regions of interest on the femoral condyles are different from the ICRS cartilage injury evaluation package (http:// www.cartilage.org/files/ICRS_evaluation.pdf) and other publications10. An explanation on why a different anatomical separation is proposed is contained in a footnote of Table I.

CARTILAGE MORPHOLOGY LABELS

A number of studies have investigated the accuracy of morphological variables as derived from MRI in healthy and OA subjects11e26 and their ability to characterize change longitudinally27e32. The group recommends the following rules for naming ‘‘morphology labels’’: Lower case modifiers should come first (t ¼ total, c ¼ covered by cartilage, and d ¼ denuded of cartilage). These should be followed by an upper case ‘‘geometric label’’ (A ¼ area, V ¼ volume, and Th ¼ thickness) and then by an upper case identifier of the tissue being measured (B ¼ subchondral bone and C ¼ cartilage). These rules permit researchers easily to adapt the nomenclature for new modifiers, metric variables, tissues (e.g. synovial fluid, menisci, osteophytes), or joints. Based on these rules, the following morphological cartilage labels were proposed (Table II). The actual interface between cartilage and the subchondral bone is the cartilage-covered area of subchondral bone (‘‘cAB’’). The unit recommended for human studies is cm2, and that for small animal studies is mm2. Areas of (original, premorbid) subchondral bone with full thickness cartilage defects or central osteophyte growth are termed the denuded (eroded) area of subchondral bone (‘‘dAB [unit ¼ cm2 or mm2]). The sum of ‘‘cAB’’ and ‘‘dAB’’ is the total area of subchondral bone (‘‘tAB’’ [unit ¼ cm2 or mm2]). Although remodeling of the subchondral bone occurs in OA, the size of ‘‘tAB’’ can be interpreted as an estimate of the original, premorbid area of subchondral bone and a useful proxy for bone size, which may be used to normalize cartilage volume to joint size36. For this reason, peripheral osteophytes are not included in measurements of ‘‘cAB’’, ‘‘dAB’’ and ‘‘tAB’’, since otherwise the premorbid bone area may be overestimated (Fig. 2). Because central osteophytes represent areas of original, premorbid subchondral bone that is denuded of original cartilage, they are regarded as dAB, and segmentation of tAB may be performed by connecting the cAB surrounding the osteophyte through its base (at the location of the original subchondral bone). Note that ‘‘cAB’’ and ‘‘dAB’’ can be expressed as a percentage of ‘‘tAB’’ (‘‘cABp’’ or ‘‘cAB%’’, ‘‘dABp’’ or ‘‘dAB%’’). In a healthy, premorbid joint, ‘‘cAB’’ is identical to ‘‘tAB’’, ‘‘cABp’’/’’cAB%’’ is 100%, and ‘‘dABp’’/’’dAB%’’ is 0%. ‘‘dAB’’ may consist of one or several areas. ‘‘NdAB’’ may be used to additionally report the number of (topographically separate) denuded bone areas. The area of the cartilage surface (‘‘AC’’ [unit ¼ cm2 or mm2]) may be used to describe the size of cartilaginous joint surface that provides the bearing surface of joints (Fig. 2). ‘‘AC’’ is similar but not identical in size to ‘‘cAB’’ and does not include areas of full thickness defects (‘‘dAB’’). The volume of the cartilage (‘‘VC’’ [unit ¼ mm3 or ml]) represents all articular cartilage, with exception of the cartilage cover of osteophytes. Since ‘‘VC’’ scales with bone size17,33e37, ‘‘VC’’ may be reported additionally as divided by ‘‘tAB’’ (’’VCtAB’’ [unit ¼ mm3/mm2 or ml/mm2 ¼ mm]). This is important, because ’’VCtAB’’ has been shown to perform better than ‘‘VC’’ in differentiating subjects with and without OA37,38.

977

Osteoarthritis and Cartilage Vol. 14, No. 10 Table I Anatomical labels of the proposed nomenclature (knee joint) Label

Explanations

Guidelines/recommendations

Knee K P T MT LT F TrF MF* LF*

Knee, total (P þ MT þ LT þ F) Patella Tibia (MT þ LT) Medial tibia Lateral tibia Femur, total (TrF þ MF þ LF) Femoral trochlea (ant. to intercondylar notch) Medial femoral condyle (post. to intercond. notch) Lateral femoral condyle (post. to intercond. notch)

Give data for cMF and pMF separately Give data for cMF and pMF separately

Patella cP mP lP

Subregions Central aspect of patella (around main ridge) Medial aspect of patella (medial facet) Lateral aspect of patella (lateral facet)

Borders need to be defined Medial to main ridge Lateral to main ridge

Trochlea cTrF mTrF lTrF

Subregions Central aspect of trochlea (around groove) Medial aspect of trochlea (medial facet) Lateral aspect of trochlea (lateral facet)

Borders need to be defined Medial to main ridge Lateral to main ridge

Femoral condyle* cMF pMF cLF pLF

Subregions

Always give data for MT and LT separately Give data for TrF, MF, and LF separately

Central medial femoral condyle (anterior aspect of condyle) Posterior medial femoral condyle (posterior aspect of condyle) Central lateral femoral condyle (anterior aspect of condyle) Posterior lateral femoral condyle (posterior aspect of condyle)

Border Border Border Border

between between between between

cMF cMF cMF cMF

and and and and

pMF pMF pMF pMF

needs needs needs needs

to to to to

be be be be

defined defined defined defined

*The proposal to separate the femoral condyles into two anatomical regions of interest (c and p, respectively) is in contrast to the ICRS cartilage injury evaluation package (http://www.cartilage.org/files/ICRS_evaluation.pdf) and other publications in the literature10. These have divided the femoral condyles into three regions (anterior, central and posterior). In the current nomenclature, the ‘‘anterior’’ region of the femoral condyle (between the trochlea and the region of the femoral condyle that is in contact with the tibia during standing) was discarded, because this region would be very small, difficult to define (dependence on knee flexion angle) and thus less suitable for quantitative measurements. If authors would like to measure this part of the femoral condyle (between the trochlea and tibial contact zone) separately or discard it from measurements of cMF or cLF (i.e. because chemical shift from the patellar fat pad may introduce error to measurements of T1 or T2) this should be clearly stated in the manuscript.

When reporting values for cartilage thickness (‘‘ThC’’) [unit ¼ mm or mm], it is important to clearly differentiate whether the values include denuded areas (‘‘dAB’’) as areas of 0 mm cartilage thickness (Fig. 2). The group recommends the use of ‘‘ThC’’ as either ‘‘ThCtAB’’for measurements including ‘‘dAB’’ (0 mm cartilage thickness) or ‘‘ThCcAB’’ for measurement of only the regions with cartilage coverage (Fig. 2). Note that ‘‘ThCcAB’’ may or may not decrease in OA, because areas of thin articular cartilage that progress to full thickness lesions will no longer be included, whereas the remaining cartilage may become thinner, remain unaltered or even swell. To report ‘‘ThCtAB’’, 0 mm cartilage thickness values need to be added for ‘‘dAB’’ and weighted in proportion to values from ‘‘cAB’’. ‘‘ThCtAB’’ and ’’VCtAB’’ should be very similar in the same cartilage plate, since they represent two methods of computing the same entity (thickness). However, values are usually not identical because different methodologies are applied. It is therefore recommended that both variables be reported independently (Table III). When reporting metric variables of cartilage thickness, authors will want to report the mean, median, maximum, minimum, s.d., coefficient of variation, or other descriptive statistical aspects of the thickness values measured within one cartilage plate. The authors may want to note that they report means or s.d.’s, etc., or they may add, a suffix that includes abbreviations of these statistical measures.

CARTILAGE COMPOSITION LABELS

Various methods for compositional imaging have been described, but only delayed gadolinium enhanced magnetic resonance imaging (dGEMRIC)39e42 and T2 mapping of cartilage43e50 (Fig. 3) were felt to be sufficiently explored to date to be included in a nomenclature. However, the nomenclature presented here is easily adaptable to other compositional variables. As compositional labels, ‘‘DG’’ (dGEMRIC index value [unit ms]) and ‘‘T2’’ (T2 relaxation time [unit ms]) are proposed in conjunction with a ‘‘C’’ for cartilage (Table II). ‘‘VlDGC’’ is to be used to express the volume [unit mm3 or ml] of low dGEMRIC values (voxels) below a threshold value [in ms] as a measure of size of a compositional cartilage lesion. The threshold that is used to define a dGEMRIC lesion should be stated (e.g. <350 ms). To express how large ‘‘VlDGC’’ is in relation to all cartilage (or of the cartilage within a specific zone) ‘‘VlDGCp’’ or ‘‘VlDGC%’’ [VlDGC/VC
100%, unit %] may be reported. However, this may provide an insufficient relative measure if a large proportion of the cartilage has been lost. ‘‘VlDGCtAB’’ [VlDGC divided by tAB; unit ¼ mm3/mm2) may then be preferable. If values for the ‘‘DGC’’ are not derived volumetrically, but from a single slice, it is recommended that the term ‘‘AlDGC’’ [area, unit ¼ mm2) be used and ‘‘AlDGC%’’ or ‘‘AlDGCp’’ and ‘‘AlDGCtLB’’ [tLB ¼ total length of bone within the slice in mm].

978

F. Eckstein et al.: Nomenclature for MRI cartilage measures Table II Morphological and compositional labels of the proposed nomenclature

Label Morphology tAB cAB cAB% [cABp] dAB dAB% [dABp] AC VC VCtAB NdAB ThCtAB ThCcAB

Composition DGC VlDGC VlDGC% [VlDGCp] VlDGCtAB T2C

VhT2C VhT2C% VhT2CtAB VlT2C VlT2C% VlT2CtAB

Explanations

Guidelines/recommendations

Total area of subchondral bone (dAB þ cAB) Area of subchondral bone covered with cartilage (tAB covered by AC) Percent of subchondral bone covered with cartilage (cAB/tAB
100%) Area of subchondral bone, denuded, eroded, full thickness defect (tAB not covered by AC) Percent of subchondral bone area that is denuded (dAB/tAB
100%) Area of cartilage surface Volume of cartilage Volume of cartilage divided by total area of subchondral bone (VC/tAB) Number of (separate) cartilage denuded areas Cartilage thickness over total subchondral bone area; denuded areas counting as 0 mm thickness Cartilage thickness over cartilaginous area of subchondral bone; denuded areas not included dGEMRIC index value across cartilage Volume of low dGEMRIC index values below certain threshold Percent cartilage volume of dGEMRIC index value below certain threshold (VlDGC/VC
100%) VlDGC divided by tAB (VlDGC/tBA) T2 value across entire cartilage

dAB not included Specify technical implementation (see text) Accounts for differences in bone size Specify technical implementation (see text); Add statistical label for database purposes Specify technical implementation (see text); Add statistical label for database purposes

Threshold needs to be specified Threshold needs to be specified

Volume of high T2 value above certain threshold Percent cartilage volume of high T2 above certain threshold (VhT2C/VC
100%) Volume of high T2 value above certain threshold divided by tAB (VhT2C/tAB) Volume of T2 low value below certain threshold Percent cartilage volume of low T2 below certain threshold (VlT2C/VC
100%) Volume of low T2 value below certain threshold divided by tAB (VlT2C/tAB)

Similar considerations apply to the analysis of T2 relaxation times of articular cartilage: the T2 value may be given throughout the entire cartilage (‘‘T2C’’ [unit ¼ ms]). However, because ‘‘T2’’ may vary as a function of distance from ‘‘AC’’ to ‘‘cAB’’5,43,49, values may be reported for a superficial zone (e.g. ‘‘szT2C’’); ‘‘mz’’ and ‘‘dz’’ may be used as modifiers preceding the compositional label for the middle and deep zones, respectively. Whether cartilage is separated in two (sz and dz) or three zones (sz, mz and dz), the extension of the layers should always be clearly defined, preferably as percent values of the distance from the cartilage surface (as used to compute ‘‘AC’’, 0%) to the bone surface (as used to compute ‘‘cAB’’, 100%). In contrast to ‘‘DGC’’ both low (e.g. xx ms) may indicate a lesion (Fig. 3). ‘‘VlT2C’’ may be

Represents premorbid subchondral bone area, peripheral osteophytes to be excluded, base of central osteophytes to be included

Accounts for differences in bone size Should be given for different layers/zones (sz ¼ superficial zone, mz ¼ middle zone, dz ¼ deep zone); Add statistical label for database purposes Threshold needs to be specified Threshold needs to be specified Accounts for differences in bone size Threshold needs to be specified Threshold needs to be specified Accounts for differences in bone size

used to express the volume [unit mm3 or ml] of voxels with low T2 values and ‘‘VhT2C’’ that with high T2 values. The same relative measures may be used for lesion size as have been described for ‘‘DGC’’ previously (Table II). Since T2 values can depend on the measurement method, echo spacing and other parameters should be explicitly reported (see below). Both for dGEMRIC and T2, describing means or other measures of central tendency and measures of dispersion such as s.d. (by suffix or otherwise) are recommended. GUIDELINES FOR MINIMAL METHODOLOGICAL INFORMATION

The group recommends some minimal methodological information that should be provided when reporting the above metric variables. Please note that general requirements for reporting of methods of clinical and preclinical studies have been formulated previously and should serve as primary

979

Osteoarthritis and Cartilage Vol. 14, No. 10

 Loading conditions of the knee during or prior to imag-

Table III Examples of nomenclature Example variable

ing (period of rest or unloading, traction)

 Type of analysis (inter- or intra-subject comparisons,

Explanation

Cartilage morphology P.tAB Total subchondral bone area in the patella MT.VCtAB Cartilage volume divided by total subchondral bone area of the medial tibia mTrF.CThtAB Cartilage thickness (over total subchondral bone area) of the medial femoral trochlea

cross sectional calculation

or

longitudinal

analysis),

power

guidelines. The following specific information was deemed particularly important in the context of quantitative MRI studies on articular cartilage:

b) Sample information/animal models and cartilage samples  Number of animals/samples, species and specific strain, gender  Age (range and s.d.); level of skeletal maturity; developmental stage  Joint(s) examined (right, left, treated, contralateral)  Type of intervention (spontaneous model; surgical e what type? enzymatic e what agent? what dosage?)  Anesthesia used for MR scanning, and any physiological parameters monitored  Details of joint positioning joint in coil (flexion angle)  Time between intervention and MR examination, level of physical activity in between  Inter- or intra-subject comparisons (cross sectional or longitudinal analysis)  Temperature of specimens/samples during imaging  Media in which samples have been equilibrated (e.g. protease inhibitor vs saline)

a) Sample information/studies on human subjects  Number of patients or healthy volunteers, gender (distribution), age (range and s.d.), ethnicity  Method(s) for subject recruitment/selection  Joint(s) examined (right, left, dominant, non-dominant, signal, non signal)  Grade of osteoarthritis (symptomatic, asymptomatic, Kellgren Lawrence Grade)  Axial alignment, if measured, and type of measurement (weightbearing ap knee radiographs or full limb hip-toankle images)

c) MRI acquisition conditions  Magnetic field strength, MR scanner and coil type  Quality assurance (QA) or control (QC) procedures in scanner maintenance and potential drift in scanner gradient calibration  Scan orientation (sagittal, coronal, axial, or double oblique orientated perpendicular/parallel to.)  Were parallel imaging methods used?  Angle between B0 and the cartilage surface under investigation (to assist interpretation of magic angle effects)

Cartilage composition LT.VlDGC Volume of low dGEMRIC values (below certain threshold) in the lateral tibia lP.szT2C T2 in the superficial zone (to be defined) of the lateral patellar facet MF.VhT2CtAB Volume of high T2 values (above specified threshold) of the medial femoral condyle, divided by the total subchondral bone area to account for differences in bone size

800

Example of high dGEMRIC index values

255 670

550

430

Example of low dGEMRIC index values

80 320

200

DG index T1Gd (ms)

0

T2 (ms)

Fig. 3. Examples of dGEMRIC. The figure displays a person with high dGEMRIC values (top left) and one with low dGEMRIC values (bottom left). Because chemical shift from the patellar fat pad may introduce error to measurements, the areas of cartilage adjacent to it should be discarded. Examples of T2 mapping of the knee, axial orientation (top right) and sagittal orientation (bottom right). The arrow (top right) in the enlarged portion of the patella cartilage highlights a T2 lesion. Note the magic angle effects at 55( on the T2 examples.

980

F. Eckstein et al.: Nomenclature for MRI cartilage measures

 Imaging sequence and measure of validity (accuracy)   

      



and precision (reproducibility) Number of slices acquired Repetition time (TR), echo time (TE), bandwidth Flip angle (FA) and, in the case of cartilage composition measures, variability in flip angle across the joint either due to mis-set pulses, or due to transmitter radiofrequency inhomogeneity Number of acquisition averages (NEX) Slice- or phase oversampling if >0%, elliptical scanning (if on), asymmetric echo (if allowed) Field of view (FOV), matrix, in-plane resolution (with or without interpolation), slice thickness, phase resolution (if less than 100%), slice resolution (if less than 100%) Image acquisition time for each sequence Fat saturation: prepulse or water excitation Non-fat-suppressed images: size (in mm) and direction of chemical shift artifact dGEMRIC: dose of intravenous Gd(DTPA) injection; type and time of joint motion between injection and image acquisition; time between injection and image acquisition, cut off value (
d) Image analysis  Definition of anatomical regions/subregions, in particular borders not corresponding to clear anatomical landmarks (e.g. cMF vs pMF)  Data analyzed by one reader or by several readers, level of training of readers, measure of reliability (accuracy, repeatability) in view of reader intervention  Time sequence of image analysis, e.g. longitudinal data (baseline, follow-up) read in pairs, unblinded to sequence of acquisition; longitudinal data read in pairs, blinded to sequence of acquisition; longitudinal data read randomly, blinded to subject ID  2D analysis (within one slice or several slices) vs 3D analysis (across slices, independent of slice orientation)  Detailed methodological description how quantitative parameters were calculated  Level of interpolation of data (in-plane, thickness) prior to segmentation Specific to cartilage morphology Volume measurements: numerical integration of segmented voxels or surface reconstruction model B Surface areas computation: 2D (length of contours
slice thickness) or 3D (triangulation or other)/degree of smoothing, if applied B Thickness computations: 2D (within slice) or 3D (independent of slice orientation); Euclidean distance transformation (EDT), vector method, other; direction from cAB to AC or from AC to cAB B

Specific to cartilage composition Registration used to align images for computing T1 or T2, 2D or 3D/translation and/or rotation/rigid or non-linear B Method of dGEMRIC index computation: map vs univoxel method B Processing algorithm for determining T2 (2 vs 3 parameter fit), B

B

B B

Number of echoes and inter-echo spacing of the echo train used to calculate T2 Spatial variation of T2 characterization Cut off value (< or >xx ms) that defines a lesion

Discussion The objective of this initiative was to propose a nomenclature of metric variable names to be used in scientific communications of quantitative MRI-based measures of articular cartilage in OA, and to give recommendations as to which minimal methodological information should be provided. Please note that this proposal on nomenclature should neither be construed as an endorsement of particular metric variables that might be most useful in characterizing OA (pathophysiological validity), nor an endorsement about the accuracy and reproducibility of any particular metrics (technical validity). Also, we do not endorse specific techniques or methodologies on how these measures should be obtained. No previous publication has promulgated a comprehensive and widely accepted set of guidelines and nomenclature to include anatomical, morphological and compositional labels of knee joint cartilage. By doing this we hope to promote uniform reporting of these measurements. A strength of the modular nomenclature proposed here is that it is systematic, consistent, and easily adaptable to technical innovation and to new anatomical, morphological, compositional and tissue labels. The parameter names deliberately do not enforce the use of a specific methodology, in order to gain wide acceptance throughout the scientific community. A limitation of the current nomenclature is that some of the recommendations (as for instance the subdivision in the patellar and femoral cartilage, and some of the morphological and compositional labels) are primarily ‘‘opinionbased’’. The current proposal should thus be considered as work in progress and may need to be revised from time to time, based on further scientific evidence emerging. Also, this nomenclature is designed to describe cartilage changes occurring in OA and is less suitable for cartilage undergoing surgical manipulation and repair. MRI-based technologies may be developed further for the latter purpose, and the nomenclature may therefore need to be extended in this respect in the future. Since quantitative parameters on non-cartilage tissues are of increasing interest in the realm of OA, the nomenclature may also have to be expanded in the future to include other tissues. However, during the meeting it was decided that further exploration was necessary before a nomenclature of non-cartilaginous tissues can be established. Widespread utilization of this nomenclature should facilitate communication and avoid misunderstandings amongst statisticians, data managers, investigators, drug developers and regulators. The nomenclature should aid the scientific community in agreeing upon how to interpret pathological findings in OA and treatment effects on cartilage. The nomenclature may consolidate and disseminate the expertise of scientists in the field by informing others which measurement parameters are central. This, in turn, should substantially strengthen the use of quantitative MRI-based measures of articular cartilage in epidemiological, clinical, and pharmacological studies in OA research. Conflict of interest statement: The authors acknowledge the financial support of companies with a commercial interest in the field in the organization of the meeting at Chicago on December 2nd 2004, as outlined in the Acknowledgment

981

Osteoarthritis and Cartilage Vol. 14, No. 10 section. The authors would like to state that this support did not influence the choice of nomenclature. The authors would like to stress further that the nomenclature only means to provide a common language amongst researchers, that it has been formulated based on various proposals and finally on consensus within the group of authors and participants, and that it does not ratify any particular group’s approach or methodology. Therefore, and because no organization gains or loses financially from adoption of the nomenclature recommended here, the authors feel there is no competing interest with regard to publication of this proposal. In supplementing the journal’s Disclosure of Interest forms, the author wish to disclose here information about their involvement in companies with a commercial interest in the field, and financial support and consulting activities that might be interpreted as constituting a possible conflict of interest: Felix Eckstein is Chief Executive Officer (CEO) of Chondrometrics GmbH, Ainring, Germany, an international company providing consulting and centralized imaging services for research studies and clinical trials to other research groups and pharmaceutical companies. Felix Eckstein provides consulting services to Pfizer Inc., Glaxo-Smith-Kline Inc., and Virtualscopics Inc. Gerard Ateshian has provided consulting services to Arthrovision Inc. Rainer Burgkart None. Deborah Burstein provides consulting services to Pfizer, Inc. and VirtualScopics, Inc. Flavia Cicuttini None. Bernard Dardzinski None. Martha Gray provides consulting services to Pfizer, Inc. Thomas Link None. Sharmila Majumdar receives research grants from Pfizer Inc. Tim Mosher provides consulting services for GlaxoSmithKline and has received honorarium from Pfizer, Inc. Charles Peterfy is Executive Vice President and Chief Medical Officer for Synarc Inc., San Francisco, CA, USA, an international company providing centralized imaging and molecular marker services for clinical trials to the pharmaceutical, biotechnology and medical devices industries. John Waterton is Director of Imaging at Global Sciences & Information of AstraZeneca PLC and holds stock and stock options in AstraZeneca PLC, Macclesfield, Cheshire, UK. Carl S. Winalski has grant support from Genzyme Biosurgery, Inc. and provides consulting services to DePuy Spine, Inc. Saara Totterman is Chief Medical Officer of VirtualScopics Inc., Rochester, NY, USA, an international company providing centralized imaging services for clinical trials to the pharmaceutical, biotechnology and medical devices industries. David Felson None.

Acknowledgment We would like to thank the following participants of the December 2nd, 2004 meeting in Chicago (IL) for their invaluable contributions: John D. Bradley, MD (Eli Lilly & Company, Indianapolis, IN, USA); Robert Buck, PhD (Pfizer Global Research and Development, Ann Arbor, MI, USA); September Cahue, PhD (Northwestern University, Chicago, IL, USA), Jeffrey L. Evelhoch, PhD (Amgen Inc., Thousand

Oaks, CA, USA); Klaus Flechsenhar, MD (GELITA, AG, Eberbach, Germany); Christian Glaser, MD (Institut fu¨r Klinische Radiologie der LMU Mu¨nchen, Munich, Germany); Garry E. Gold, MD (Department of Radiology, Stanford University, CA, USA); Heiko Graichen, MD (Orthopaedic Department, University of Frankfurt, Frankfurt, Germany); Marie-Pierre Hellio Le Graverand-Gastineau, MD, DSc, PhD (Pfizer Global Research and Development, Ann Arbor, MI, USA); David Hunter, MD (Boston University School of Medicine, Boston, MA, USA); Mark Hurtig, DVM (Ontario Veterinary College, University of Guelph, ON, Canada); Dean Inglis, PhD (McMaster University, Hamilton, ON, Canada); Sakeba Issa, MD (Northwestern University, Chicago, IL, USA); Sudha Kadiyala, PhD (DePuy Biologics, Raynham, MA, USA); John J. Kotyk, PhD (Pfizer Global Research and Development, St. Louis, MO, USA); Gayle E. Lester, PhD (Clinical Osteoarthritis & Diagnostics Imaging NIAMS/NIH/DHHS, Bethesda, MD, USA); Rose A. Maciewicz, PhD (AstraZeneca, Macclesfield, UK); Lachy McLean, MD, PhD (Merck and Company, Rahway, NJ, USA); Emily McWalter, MASc (University of British Columbia, Vancouver, Canada); Carol Muehleman, PhD (Rush Medical College, Chicago, IL, USA); Michael C. Nevitt, PhD (University of California, San Francisco, CA, USA); Ann E. Remmers, PhD (Pfizer Global Research and Development, Ann Arbor, MI, USA); Erika Schneider, PhD (SciTrials, LLC, Westwood, MA, USA); Michael S. Vincent, MD, PhD (Amgen Inc., Thousand Oaks, CA, USA); Lydia Wachsmuth, PhD (Institute of Medical Physics, Friedrich-AlexanderUniversity, Erlangen, Germany); Kathryn Wildy, MD, Donald S. Williams, PhD (Merck & Co., Inc., West Point, PA, USA); and Thasia Woodworth, MD (Novartis Pharma, Basel, Switzerland). We would also like to thank the following sponsors for generously supporting this initiative: Novartis Pharma (Basel, Switzerland); Pfizer Global Research & Development (Ann Arbor, MI, USA); Amgen (Thousand Oaks, CA, USA); Rotta Research Laboratories (Milan, Italy); Sanofi Aventis (Frankfurt, Germany); DePuy Spine, Inc. (Raynham, MA, USA) and AstraZeneca (Macclesfield, UK). This support did not influence the outcome of the meeting and the sponsors who provided support were not asked to approve the nomenclature.

References 1. Peterfy CG. Imaging of the disease process. Curr Opin Rheumatol 2002;14:590e6. 2. Burstein D, Gray M. New MRI techniques for imaging cartilage. J Bone Joint Surg Am 2003;85-A(Suppl 2): 70e7. 3. Gray ML, Eckstein F, Peterfy C, Dahlberg L, Kim YJ, Sorensen AG. Toward imaging biomarkers for osteoarthritis. Clin Orthop 2004;S175e81. 4. Eckstein F, Glaser C. Measuring cartilage morphology with quantitative magnetic resonance imaging. Semin Musculoskelet Radiol 2004;8:329e53. 5. Mosher TJ, Dardzinski BJ. Cartilage MRI T2 relaxation time mapping: overview and applications. Semin Musculoskelet Radiol 2004;8:355e68. 6. Raynauld JP. Quantitative magnetic resonance imaging of articular cartilage in knee osteoarthritis. Curr Opin Rheumatol 2003;15:647e50. 7. Parfitt AM, Drezner MK, Glorieux FH, Kanis JA, Malluche H, Meunier PJ, et al. Bone histomorphometry: standardization of nomenclature, symbols, and

982

8. 9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

F. Eckstein et al.: Nomenclature for MRI cartilage measures units. Report of the ASBMR Histomorphometry Nomenclature Committee. J Bone Miner Res 1987;2: 595e610. Delmas PD. Treatment of postmenopausal osteoporosis. Lancet 2002;359:2018e26. Kwak SD, Colman WW, Ateshian GA, Grelsamer RP, Henry JH, Mow VC. Anatomy of the human patellofemoral joint articular cartilage: surface curvature analysis. J Orthop Res 1997;15:468e72. Biswal S, Hastie T, Andriacchi TP, Bergman GA, Dillingham MF, Lang P. Risk factors for progressive cartilage loss in the knee: a longitudinal magnetic resonance imaging study in forty-three patients. Arthritis Rheum 2002;46:2884e92. Peterfy CG, van Dijke CF, Janzen DL, 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;192:485e91. Eckstein F, Sittek H, Milz S, Putz R, Reiser M. The morphology of articular cartilage assessed by magnetic resonance imaging (MRI). Reproducibility and anatomical correlation. Surg Radiol Anat 1994;16: 429e38. Eckstein F, Gavazzeni A, Sittek H, Haubner M, Losch A, Milz S, et al. Determination of knee joint cartilage thickness using three-dimensional magnetic resonance chondro-crassometry (3D MR-CCM). Magn Reson Med 1996;36:256e65. Eckstein F, Adam C, Sittek H, Becker C, Milz S, Schulte E, et al. Non-invasive determination of cartilage thickness throughout joint surfaces using magnetic resonance imaging. J Biomech 1997;30:285e9. 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;137e48. Cohen ZA, McCarthy DM, Kwak SD, Legrand P, Fogarasi F, Ciaccio EJ, et al. Knee cartilage topography, thickness, and contact areas from MRI: in-vitro calibration and in-vivo measurements. Osteoarthritis Cartilage 1999;7:95e109. 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;7:265e71. Eckstein F, Stammberger T, Priebsch J, Englmeier KH, Reiser M. Effect of gradient and section orientation on quantitative analysis of knee joint cartilage. J Magn Reson Imaging 2000;11:161e7. Burgkart R, Glaser C, Hyhlik-Durr A, Englmeier KH, Reiser M, Eckstein F. Magnetic resonance imagingbased assessment of cartilage loss in severe osteoarthritis: accuracy, precision, and diagnostic value. Arthritis Rheum 2001;44:2072e7. Graichen H, Springer V, Flaman T, Stammberger T, Glaser C, Englmeier KH, et al. Validation of highresolution water-excitation magnetic resonance imaging for quantitative assessment of thin cartilage layers. Osteoarthritis Cartilage 2000;8:106e14. Graichen H, Jakob J, Eisenhart-Rothe R, Englmeier KH, Reiser M, Eckstein F. Validation of cartilage volume and thickness measurements in the human shoulder with quantitative magnetic resonance imaging. Osteoarthritis Cartilage 2003;11:475e82. Graichen H, Eisenhart-Rothe R, Vogl T, Englmeier KH, Eckstein F. Quantitative assessment of cartilage

23.

24.

25. 26.

27. 28. 29.

30.

31.

32.

33.

34.

35.

36.

37.

status in osteoarthritis by quantitative magnetic resonance imaging: technical validation for use in analysis of cartilage volume and further morphologic parameters. Arthritis Rheum 2004;50:811e6. Piplani MA, Disler DG, McCauley TR, Holmes TJ, Cousins JP. Articular cartilage volume in the knee: semiautomated determination from three-dimensional reformations of MR images. Radiology 1996;198:855e9. Kladny B, Bail H, Swoboda B, Schiwy-Bochat H, Beyer WF, Weseloh G. Cartilage thickness measurement in magnetic resonance imaging. Osteoarthritis Cartilage 1996;4:181e6. Marshall KW, Guthrie BT, Mikulis DJ. Quantitative cartilage imaging. Br J Rheumatol 1995;34(Suppl 1): 29e31. Raynauld JP, Kauffmann C, Beaudoin G, Berthiaume MJ, 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;11:351e60. Wluka AE, Stuckey S, Snaddon J, Cicuttini FM. The determinants of change in tibial cartilage volume in osteoarthritic knees. Arthritis Rheum 2002;46:2065e72. Cicuttini FM, Wluka AE, Wang Y, Stuckey SL. Longitudinal study of changes in tibial and femoral cartilage in knee osteoarthritis. Arthritis Rheum 2004;50:94e7. Jones G, Ding C, Glisson M, Hynes K, Ma D, Cicuttini F. Knee articular cartilage development in children: a longitudinal study of the effect of sex, growth, body composition, and physical activity. Pediatr Res 2003;54:230e6. Wluka AE, Wolfe R, Davis SR, Stuckey S, Cicuttini FM. Tibial cartilage volume change in healthy postmenopausal women: a longitudinal study. Ann Rheum Dis 2004;63:444e9. Gandy SJ, Dieppe PA, Keen MC, Maciewicz RA, Watt I, Waterton JC. No loss of cartilage volume over three years in patients with knee osteoarthritis as assessed by magnetic resonance imaging. Osteoarthritis Cartilage 2002;10:929e37. Raynauld JP, Martel-Pelletier J, Berthiaume MJ, Labonte F, Beaudoin G, de Guise JA, et al. Quantitative magnetic resonance imaging evaluation of knee osteoarthritis progression over two years and correlation with clinical symptoms and radiologic changes. Arthritis Rheum 2004;50:476e87. Eckstein F, Reiser M, Englmeier KH, Putz R. In vivo morphometry and functional analysis of human articular cartilage with quantitative magnetic resonance imaging e from image to data, from data to theory. Anat Embryol (Berl) 2001;203:147e73. Eckstein F, Winzheimer M, Hohe J, Englmeier KH, Reiser M. Interindividual variability and correlation among morphological parameters of knee joint cartilage plates: analysis with three-dimensional MR imaging. Osteoarthritis Cartilage 2001;9:101e11. 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;43:2543e9. Cicuttini FM, Wluka AE, Forbes A, Wolfe R. Comparison of tibial cartilage volume and radiologic grade of the tibiofemoral joint. Arthritis Rheum 2003;48: 682e8. Burgkart R, Glaser C, Hinterwimmer S, Hudelmaier M, Englmeier KH, Reiser M, et al. Feasibility of T and Z

983

Osteoarthritis and Cartilage Vol. 14, No. 10

38.

39. 40.

41.

42.

43. 44.

scores from magnetic resonance imaging data for quantification of cartilage loss in osteoarthritis. Arthritis Rheum 2003;48:2829e35. Hunter D, Niu J, Zhang Q, McLennan C, LaValley M, Tu X, et al. Cartilage volume must be normalized to bone surface area in order to provide satisfactory construct validity: The Framingham Study. Osteoarthritis Cartilage 2004;12(Suppl B):S2. Bashir A, Gray ML, Hartke J, Burstein D. Nondestructive imaging of human cartilage glycosaminoglycan concentration by MRI. Magn Reson Med 1999;41:857e65. Bashir A, Gray ML, Boutin RD, Burstein D. Glycosaminoglycan in articular cartilage: in vivo assessment with delayed Gd(DTPA)(2-)-enhanced MR imaging. Radiology 1997;205:551e8. Tiderius CJ, Olsson LE, Leander P, Ekberg O, Dahlberg L. Delayed gadolinium-enhanced MRI of cartilage (dGEMRIC) in early knee osteoarthritis. Magn Reson Med 2003;49:488e92. Williams A, Gillis A, McKenzie C, Po B, Sharma L, Micheli L, et al. Glycosaminoglycan distribution in cartilage as determined by delayed gadolinium-enhanced MRI of cartilage (dGEMRIC): potential clinical applications. AJR Am J Roentgenol 2004;182:167e72. Dardzinski BJ, Mosher TJ, Li S, Van Slyke MA, Smith MB. Spatial variation of T2 in human articular cartilage. Radiology 1997;205:546e50. Liess C, Lu¨sse S, Karger N, Heller M, Glu¨er CC. Detection of changes in cartilage water content using MRI

45. 46.

47.

48.

49.

50.

T2-mapping in vivo. Osteoarthritis Cartilage 2002;10: 907e13. David-Vaudey E, Ghosh S, Ries M, Majumdar S. T2 relaxation time measurements in osteoarthritis. Magn Reson Imaging 2004;22:673e82. Dunn TC, Lu Y, Jin H, Ries MD, Majumdar S. T2 relaxation time of cartilage at MR imaging: comparison with severity of knee osteoarthritis. Radiology 2004;232: 592e8. Kurkijarvi JE, Nissi MJ, Kiviranta I, Jurvelin JS, Nieminen MT. Delayed gadolinium-enhanced MRI of cartilage (dGEMRIC) and T2 characteristics of human knee articular cartilage: topographical variation and relationships to mechanical properties. Magn Reson Med 2004;52:41e6. Mendlik T, Faber SC, Weber J, Hohe J, Rauch E, Reiser M, et al. T2 quantitation of human articular cartilage in a clinical setting at 1.5 T: implementation and testing of four multiecho pulse sequence designs for validity. Invest Radiol 2004;39:288e99. Mosher TJ, Liu Y, Yang QX, Yao J, Smith R, Dardzinski BJ, et al. Age dependency of cartilage magnetic resonance imaging T2 relaxation times in asymptomatic women. Arthritis Rheum 2004;50:2820e8. Mosher TJ, Collins CM, Smith HE, Moser LE, Sivarajah RT, Dardzinski BJ, et al. Effect of gender on in vivo cartilage magnetic resonance imaging T2 mapping. J Magn Reson Imaging 2004;19: 323e8.