| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Department of Diagnostic Radiology, The Catholic University of Korea, Kangnam
St. Mary's Hospital, 505 Banpo-Dong, Seocho-Ku, 137-701, Seoul, Korea.
2 Department of Diagnostic Radiology, Yale University School of Medicine, 333
Cedar St., New Haven, CT 06520.
3 Department of Orthopedic Surgery, The Catholic University of Korea, Seocho-Ku,
137-701, Seoul, Korea.
Received March 17, 2000;
accepted after revision June 25, 2002.
Address correspondence to W.-H. Jee.
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. Fast spin-echo MR images obtained at 1.5 T from 110 patients who had meniscal tears identified at arthroscopy were retrospectively and independently classified by two reviewers into five configurations: horizontal, longitudinal, radial, oblique, and complex. MR imaging categorization was compared with arthroscopic results as the standard of reference. Data were also analyzed with longitudinal and oblique tears combined because these usually are reparable, and with horizontal, radial, and complex tears combined because these usually are not reparable. Interobserver and intraobserver agreements were calculated using kappa coefficients.
RESULTS. At arthroscopy, meniscal tears were categorized as
horizontal (n = 44), longitudinal (n = 34), complex
(n = 22), radial (n = 11), and oblique (n = 5).
Sensitivity, specificity, and accuracy of each reviewer for the reparable
tears were 82%, 92%, and 89%; and 59%, 97%, and 84%, respectively.
Interobserver agreements were fair between reviewer 1 and the first and second
interpretations of reviewer 2 (
= 0.25, p < 0.005; and
= 0.21, p < 0.05, respectively). Intraobserver agreement
was substantial ( = 0.71, p < 0.001).
CONCLUSION. MR imaging was accurate for predicting reparable meniscal tears and was sensitive for the determination of nonreparable tears.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The purpose of this study was to evaluate the accuracy of MR imaging for categorizing the configurations of meniscal tears.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MR imaging was performed with a 1.5-T imager (Signa Advantage; General Electric Medical Systems, Milwaukee, WI) and an extremity coil (General Electric Medical Systems). Fast spin-echo pulse sequences were used to obtain proton density-weighted images (TR range/TE range, 3000-4300/16-21) and T2-weighted images (3000-4300/76-108) in the sagittal and coronal planes. In the last 86 patients, imaging was performed with fat suppression because of a change in our clinical imaging protocol. MR imaging parameters were field of view, 14-16 cm; excitations, 2; matrix size, 256 x 192; section thickness, 3 mm; an intersection gap of 1 mm for sagittal images and 1.5 mm for coronal images; and echo-train length, 8 (equivalent to echo-train length of 4 for each echo).
MR images were separately examined by two observers who were unaware of the categorization at arthroscopy. A training session was held to familiarize reviewer 1 and the orthopedic surgeon with the criteria used for categorizing tears, although no training session was held for reviewer 2. One observer reviewed MR images twice during a 3-week interval for intraobserver agreement.
Meniscal tears were classified into one of five configurations using
previously described criteria
[7,
8]. A horizontal tear is
parallel to the tibial plateau and separates the meniscus into upper and lower
parts, a longitudinal tear is vertical (perpendicular to the tibial plateau)
and propagates parallel to the main (circumferential) axis of the meniscus, a
radial tear is vertical and propagates perpendicular to the main axis, an
oblique or parrot-beak tear is vertical and propagates obliquely to the main
axis of the meniscus, and a complex tear comprises two or more tear
configurations (Fig.
1A,1B,1C,1D).
If the meniscal tear was not visible on MR images, the images were classified
as negative for tear. Sensitivity, specificity, and accuracy for the
identification of meniscal tear configurations were determined using the
arthroscopic categorization as the standard of reference. Data were also
analyzed with longitudinal and oblique tears combined because these tear
configurations usually are reparable (reparable group), and with horizontal,
radial, and complex tears combined because these tear configurations usually
are not reparable and require partial meniscectomy (nonreparable group)
[7,8,9].
Inter- and intraobserver agreements for the detection of meniscal tear
configurations were calculated using kappa coefficients. Kappa value were
interpreted as poor ( = 0), slight ( = 0.0-0.2), fair ( =
0.21-0.40), moderate ( = 0.41-0.60), substantial ( = 0.61-0.80),
and almost perfect ( = 0.81-1.00)
[10].
|
|
|
|
Accuracies for tear characterization by reviewer 1 were determined in patients with associated anterior cruciate ligament tears, discoid menisci, and bucket-handle tears or displaced fragments to determine if the presence of any of these findings significantly changed the accuracy of tear categorization. Significance for differences in correct categorization of tears associated with and not associated with these findings was tested using the chi-square statistic.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Interpretation of MR images resulted in correct categorization of 35 of 44 horizontal (Fig. 2A,2B), 27 of 34 longitudinal (Fig. 3A,3B,3C), eight of 11 radial (Fig. 4A,4B,4C), three of five oblique, and 18 of 22 complex (Fig. 5A,5B) tears for reviewer 1 (Table 1). Ninety-one (78%) of 116 tears were categorized the same at MR imaging by reviewer 1 as at arthroscopy. One hundred three (89%) of 116 were categorized the same at MR imaging and arthroscopy with respect to reparable versus nonreparable groups.
|
|
|
|
|
|
|
|
|
|
|
With MR imaging, the sensitivity, specificity, and accuracy of reviewer 1 for longitudinal tears were 79%, 95%, and 91%; for oblique tears were 60%, 96%, and 95%; for horizontal tears were 80%, 90%, and 86%; for radial tears were 73%, 98%, and 96%; and for complex tears were 82%, 98%, and 95%, respectively. The values for the reparable tear group (longitudinal or oblique configurations) were 82%, 92%, and 89%. Of 25 misinterpreted cases, seven lesions (28%) were incorrectly interpreted as horizontal tears (Fig. 6A,6B,6C), four (16%) as longitudinal tears, four (16%) as oblique tears, two (8%) as radial tears, two (8%) as complex tears, and six (24%) as negative for tear (Table 1).
|
|
|
In the first interpretation of reviewer 2, the sensitivity, specificity,
and accuracy for longitudinal tears were 65%, 95%, and 86%; for oblique tears
were 0%, 100%, and 96%; for horizontal tears were 59%, 90%, and 78%; for
radial tears were 45%, 99%, and 94%; and for complex tears were 68%, 65%, and
66% (Table 2). The values for
the reparable group were 56%, 95%, and 82%. In the second interpretation of
reviewer 2, the sensitivity, specificity, and accuracy for longitudinal tears
were 65%, 98%, and 88%; for oblique tears were 20%, 100%, and 97%; for
horizontal tears were 70%, 92%, and 84%; for radial tears were 45%, 99%, and
94%; and for complex tears were 86%, 72%, and 75%
(Table 3). The values for the
reparable group were 59%, 97%, and 84%. Interobserver agreement was fair
( = 0.25, p < 0.005) between reviewer 1 and the first
interpretation of reviewer 2, and interobserver agreement was fair ( =
0.21, p < 0.05) between reviewer 1 and the second interpretation
of reviewer 2. Intraobserver agreement of reviewer 2 was substantial (
= 0.71, p < 0.001).
|
|
The meniscal tear configuration was correctly interpreted in 14 (74%) of 19 meniscal tears associated with a tear of the anterior cruciate ligament as compared with 77 (79%) of 97 tears not associated with an anterior cruciate ligament tear (p = 0.581). The tear configuration was correctly interpreted in 10 (100%) of 10 tears occurring in discoid menisci compared with 81 (76%) of 106 tears occurring in nondiscoid menisci (p = 0.083). The tear configuration was correctly interpreted in 30 (81%) of 37 tears associated with a bucket-handle tear or displaced fragments compared with 61 (77%) of 79 tears not associated with bucket-handle tear or displaced meniscal fragments (p = 0.747). None of these differences was statistically significant (chi-square test, p > 0.05).
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Configuration of meniscal tears and their locations and sizes are important in determining the type of treatment [11, 12, 15]. Four alternatives exist for treatment of meniscal tears: no meniscal surgery, meniscal repair, partial meniscectomy, and complete meniscectomy. In patients undergoing meniscal surgery, orthopedic surgeons attempt to conserve as much meniscal tissue as possible because of the increased development of degenerative changes in the knee after the removal of large amounts of meniscus [8, 15]. Although repair of the meniscus has become more popular in recent years, often repair is not possible, and partial meniscectomy, or rarely, complete meniscectomy is performed [15,16,17]. Longitudinal and oblique tears usually are amenable to repair, whereas horizontal, radial, and complex tears usually cannot be repaired and most often require partial meniscectomy [11, 12, 15]. Thus, determining meniscal configuration can allow orthopedic surgeons to preoperatively advise patients of the likelihood of performing repair versus meniscectomy. Preoperative categorization of tear configuration not only affects the prognosis of treatment but also affects the recovery time and the likelihood of reoperation (both of which are greater for meniscal repair than for partial meniscectomy) [15].
After meniscal repair, patients are usually non- or partially weight-bearing on crutches for 3-4 weeks, whereas with partial meniscectomy they can resume full weight-bearing immediately. Identification of potentially reparable tear configurations can also influence the timing of arthroscopy. Meniscal repairs have an improved outcome when they are performed within 8 weeks of injury [18]. In athletes wishing to complete their season, identification of potentially reparable menisci could influence their decision because continued sports activities could convert a reparable tear into a nonreparable tear. In addition to facilitating preoperative planning, MR imaging categorization of tear configuration can obviate arthroscopy because small peripheral longitudinal tears often do not require surgery [8, 9, 12, 15].
Limitations of this study include that the MR imaging reviewers knew that all patients had arthroscopically confirmed tears, which may have increased the sensitivity of MR imaging for revealing tears; however, that knowledge would not be expected to influence the categorization of tears. Fast spin-echo proton density—weighted images can result in blurring, which could lower agreement between reviewers and between the MR interpretations and arthroscopic findings. Blurring likely was a small effect because an echo-train length of 4 was used, and high accuracy for meniscal tear evaluation has been shown for imaging with echo trains of less than 5 [3]. The addition of fat suppression to the imaging protocol in the midst of the study could be another limitation, although we believe the addition of fat suppression would be unlikely to greatly influence the categorization of meniscal tear configuration. The location, extent, and size of the tear, and other factors that affect the treatment planning of meniscal tears, were not evaluated in this study. The orthopedist had reports of clinical MR imaging available at the time of the arthroscopy that might have biased the arthroscopic categorization of the tears. However, the configuration of tears was not described in the reports of clinical MR imaging when this study was performed, and thus the availability of the MR reports likely did not greatly bias the arthroscopic categorization of tears. The sample sizes in this study were small for oblique and radial tears.
In conclusion, MR imaging was accurate for predicting reparable meniscal tears and sensitive for the determination of nonreparable tears.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  M. Englund, A. Guermazi, D. Gale, D. J. Hunter, P. Aliabadi, M. Clancy, and D. T. Felson Incidental Meniscal Findings on Knee MRI in Middle-Aged and Elderly Persons N. Engl. J. Med., September 11, 2008; 359(11): 1108 - 1115. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. E. Gold, R. F. Busse, C. Beehler, E. Han, A. C. S. Brau, P. J. Beatty, and C. F. Beaulieu Isotropic MRI of the Knee with 3D Fast Spin-Echo Extended Echo-Train Acquisition (XETA): Initial Experience Am. J. Roentgenol., May 1, 2007; 188(5): 1287 - 1293. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. M. Brody, H. M. Lin, M. J. Hulstyn, and G. A. Tung Lateral meniscus root tear and meniscus extrusion with anterior cruciate ligament tear. Radiology, June 1, 2006; 239(3): 805 - 810. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Boxheimer, A. M. Lutz, M. Zanetti, K. Treiber, L. Labler, B. Marincek, and D. Weishaupt Characteristics of Displaceable and Nondisplaceable Meniscal Tears at Kinematic MR Imaging of the Knee Radiology, January 1, 2006; 238(1): 221 - 231. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. W. Harper, C. A. Helms, H. S. Lambert III, and L. D. Higgins Radial Meniscal Tears: Significance, Incidence, and MR Appearance Am. J. Roentgenol., December 1, 2005; 185(6): 1429 - 1434. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
