| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Department of Radiology, Hospital for Joint Diseases Orthopaedic Institute,
Bernard Aronson Plaza, 301 E 17th St., 6th Fl., New York, NY 10003.
2 Present Address: Radiology Department, Hospital das Clinicas, University of
Sao Paulo School of Medicine, Av. Eneas de Carvalho de Aguiar, 255, 05403-001,
São Paulo, Brazil.
3 Department of Radiology, Thomas Jefferson University Hospital, Philadelphia,
PA.
4 Department of Orthopedic Surgery, Philadelphia Hand Center, King of Prussia,
PA.
Received February 25, 2004;
accepted after revision September 13, 2004.
Address correspondence to M. Bordalo-Rodrigues.
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. MR images of 34 patients who underwent arthroscopy and had an MRI examination within 1 month of surgery were evaluated by two reviewers for the presence and location of lunate cartilage defects and subchondral edema.
RESULTS. Lunate cartilage defects were seen on MRI in 10 of the 13 patients with chondromalacia, but these defects were also incorrectly noted in three of 21 of patients without chondromalacia. The visible locations for cartilage defects were the ulnar aspect of the proximal lunate bone (n = 3), radial aspect of the proximal lunate bone (n = 4), ulnar aspect of the distal lunate bone (n = 2), and radial aspect of the distal lunate bone (n = 1). Subchondral marrow edema was observed in six of the 10 patients with chondromalacia seen on MRI; in all six patients, the edema was seen in the same quadrant as the cartilage defect. Marrow edema was detected in one patient without chondromalacia.
CONCLUSION. We conclude that lunate chondromalacia can be accurately assessed using routine MRI sequences, although there are occasional false-positive interpretations.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Because clinical assessment gives little information about the integrity of articular cartilage [11], a major obstacle to the study of chondromalacia has been the lack of a reliable noninvasive method by which to directly assess the integrity of the articular cartilage [12]. Much of the ongoing work in this area involves in vitro studies [13–15]. Few in vivo studies on MRI of articular cartilage in joints other than the knee, ankle, or temporomandibular joints have been performed [16–18]. In the wrist, the most common location for hyaline cartilage defects is the lunate bone [19]. Consequently, we sought to study the accuracy of clinical MRI in the assessment of lunate chondromalacia.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Operative reports were reviewed for the presence and location of lunate chondromalacia and also for the frequency with which cartilage loss was related to an adjacent internal derangement, including triangular fibrocartilaginous complex, scapholunate ligament, and lunotriquetral interosseus ligament abnormalities.
We did not directly assess MR images for the presence of triangular fibrocartilaginous complex, scapholunate ligament, or lunotriquetral interosseus ligament tears; instead, we evaluated the surgical diagnosis of a tear in comparison with the MR diagnosis of cartilage defects because the latter was the focus of this study.
Imaging
Imaging was performed on 1.5-T superconducting magnets (Signa, GE
Healthcare) using dedicated wrist coils.
Several pulse sequences were used. Coronal spin-echo T1-weighted (TR range/TE range, 400–600/9–20; number of excitations, 2; matrix size, 256 x 192; slice thickness, 3 mm; interslice gap, 1 mm; field of view, 10 cm), coronal fast spin-echo proton density–weighted (TR/TE, 4,000/40; echo-train length, 8; number of excitations, 2; matrix size, 256 x 256; slice thickness, 3 mm; interslice gap, 1 mm; field of view, 10 cm), and coronal 3D gradient-echo (45/10; number of excitations, 2; flip angle, 10–20°; matrix size, 256 x 128; slice thickness, 1.2 mm; interslice gap, 0 mm; field of view, 10 cm) sequences were performed. In addition, axial spin-echo T1-weighted sequences (TR range/TE range, 400–600/9–20; number of excitations, 2; matrix size, 256 x 192; slice thickness, 3 mm; interslice gap, 1 mm; field of view, 10 cm) and axial fast spin-echo T2-weighted sequences with fat saturation (TR/TE range, 2,500/60–80; echo-train length, 16; number of excitations, 4; matrix size, 256 x 256; slice thickness, 3 mm; interslice gap, 1 mm; field of view, 10 cm) were performed.
In 24 patients, indirect MR arthrography was performed. The protocol was identical except that an additional axial spin-echo T1-weighted fat-suppressed postcontrast sequence (5 min delay after contrast injection) was performed initially. Coronal T1-weighted images were therefore indirect arthrograms.
In nine patients, coronal fast spin-echo T2-weighted images with fat saturation (TR/TE range, 2,500/60–80; echo-train length, 16; number of excitations, 4; matrix size, 256 x 256; slice thickness, 3 mm; interslice gap, 1 mm; field of view, 10 cm) were obtained instead of coronal fast spin-echo proton density–weighted images.
Image Analysis
Two experienced musculoskeletal radiologist reviewers who were blinded to
the initial MR interpretations, surgical results, and patient demographics
evaluated in consensus the images for the presence of lunate chondromalacia.
Chondromalacia was defined as focal disruption of the normal hyaline cartilage
signal intensity. In addition, the location of the cartilage
defects—dividing the lunate bone into proximal and distal radial and
ulnar aspects—was noted (Fig.
1). The scapholunate surface was considered proximal radial, and
the lunotriquetral interosseus surface was considered proximal ulnar. Also,
note was made as to whether any localized subchondral bone marrow edema,
whether cystic or ill-defined, was present even when no cartilage defect was
seen. If the cartilage loss appeared as though it might involve more than one
segment, only the dominant segment was scored. There was no independent
sequence-to-sequence evaluation. Because of the subtlety of the abnormalities
and because this was a pilot study, we believe a global assessment is more
useful.
|
Statistical Analysis
A finding was considered true-positive only if the MRI abnormality was
localized at the exact same location as the visualized arthroscopic
defect.
Sensitivity, specificity, positive predictive value, negative predictive value, and accuracy with 95% confidence intervals for the detection of lunate chondromalacia were calculated. Fisher's exact test was used to assess the correlation of lunate chondromalacia with the presence of each type of ligament and triangular fibrocartilaginous complex tear. Surgical reports were used as the standard of reference for the diagnosis of lunate chondromalacia and determining whether scapholunate ligament, lunotriquetral interosseus ligament, or triangular fibrocartilaginous complex tears were present.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
On MRI, chondromalacia was seen in 10 of the 13 patients with chondromalacia; however, MR findings were interpreted as showing chondromalacia in three of 21 patients without chondromalacia (false-positive findings).
The locations for chondromalacia seen on MRI were the ulnar aspect of the proximal lunate (n = 3), radial aspect of the proximal lunate (n = 4) (Fig. 2A, 2B, 2C), ulnar aspect of the distal lunate (n = 2), and radial aspect of the distal lunate (n = 1). The three false-positive MR interpretations were in the radial aspect of the proximal lunate (n = 2) (Fig. 3A, 3B) and ulnar aspect of the proximal lunate (n = 1).
|
|
|
|
|
There were three false-negative MR examinations. At arthroscopy, minor proximal cartilage fraying in the ulnar aspect of the proximal lunate was noted in two cases and in the radial aspect of the proximal lunate in one case (Fig. 4).
|
Ill-defined subchondral marrow edema was observed in six of the 10 patients with chondromalacia seen on MRI; in all six patients, the edema was in the same location as the cartilage defect. None of the three patients with false-negative MR findings for chondromalacia had subchondral edema. Also, ill-defined bone marrow edema was detected in one patient with a false-positive MR study (Fig. 3A, 3B).
Triangular fibrocartilaginous complex tears were seen at surgery in six patients with chondromalacia, scapholunate ligament tears in three patients with chondromalacia, and lunotriquetral interosseus ligament tears in three patients with chondromalacia. These frequencies are similar to those of the patients without chondromalacia (Table 1).
|
The sensitivity, specificity, positive predictive value, negative predictive value, and accuracy of MRI for visible cartilage defects are provided in Table 2.
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The most frequent location for hyaline cartilage loss in the carpal bones is the lunate bone because of its biomechanical centrality [19]. According to several theories proposed to explain wrist motion [21–24], the lunate bone is an intercalated segment between the scaphoid and triquetral bones and between the radius and capitate bones (Fig. 5). In intrinsic intercarpal ligament dysfunction, nonsynchronous motion between and within the carpal rows may result in carpal instability and, consequently, in osteoarthritis [25, 26]. However, in our population there was no statistical association of lunate chondromalacia with surgically proven ligament tears. This result was likely seen because of the limited number of cases studied. End-stage triangular fibrocartilaginous complex tears can result in ulnar impaction on the lunate bone especially with positive ulnar variance [27]. Because MRI is not performed in the position to evaluate for ulnar variance, this assessment was not made in our study. All these mechanical disorders can lead to lunate chondromalacia. We found a fairly high incidence of lunate chondromalacia in our population. This finding is likely related to the strong interest of our referring hand surgeon in cartilage loss.
|
Lunate cartilage loss is in the clinical differential diagnosis for both radial- and ulnar-sided wrist pain. Because of the mechanics and clinical importance, lunate chondromalacia was chosen for evaluation in this pilot study. We found that chondromalacia of the lunate bone predominates proximally. All cases of chondromalacia accompanying triangular fibrocartilaginous complex tears were seen proximally, almost equally divided between radial and ulnar sides. In other parts of the body, cartilage loss is seen in association with ligament injuries [4, 28]. However, we found no association with specific internal derangements.
Although MRI was surprisingly accurate for the detection of chondromalacia, especially in the region of the proximal lunate, we missed a quarter of these lesions, usually minor ones. We also had a 14% false-positive rate, which was probably related to the hypersensitivity of the reviewers and the focused nature of this study.
We did not objectively compare sequences as the patients were assessed globally. Also, indirect MR arthrography and coronal T2-weighted imaging were not performed in all patients. However, our subjective impression is that the proton density–weighted images and indirect arthrographic images showed cartilage defects best.
Bone marrow edema has been suggested to be a predictive sign of worsening of chondropathy in osteoarthritis of the knee [29]. Another study of the frequency and spectrum of abnormalities in the bone marrow of the wrist postulated that 34% of cases were attributed to focal osteoarthritis [30]. In our study, ill-defined marrow edema was seen in half of all cases of lunate chondromalacia. However, in one false-positive case, ill-defined marrow edema seen on the MR study corresponded to cystic changes at surgery, with no overlying chondromalacia.
We acknowledge several limitations to this study. First, this study was retrospective. Second, although our images were typical of clinical high-field-strength images, certain proposed cartilage-specific sequences such as 3D double-echo steady state [31] and spoiled gradient-recalled acquisition in the steady state [32] were not performed. Third, because of our fairly rigid inclusion criteria, a limited population was studied.
We conclude that lunate chondromalacia can be accurately assessed by typical high-field-strength MRI sequences, although detection of early changes is limited and there is a modest false-positive rate.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  N. Saupe, C. W. A. Pfirrmann, M. R. Schmid, T. Schertler, M. Manestar, and D. Weishaupt MR Imaging of Cartilage in Cadaveric Wrists: Comparison between Imaging at 1.5 and 3.0 T and Gross Pathologic Inspection Radiology, April 1, 2007; 243(1): 180 - 187. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
