| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Clinical Observations |
1 Department of Radiology, University of Wisconsin-Madison Center for Health
Sciences, 600 Highland Ave., Madison, WI 53792.
2 Radiology Associates of the Fox Valley, Neenah, WI 54956.
Received November 23, 2004;
accepted after revision June 24, 2005.
Presented at the 2004 annual meeting of the American Roentgen Ray Society,
Miami Beach, FL.
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CONCLUSION. On knee MR arthrograms, the signal contacting the surface of a recurrent meniscal tear may be equal to or less than that of adjacent intraarticular gadolinium contrast material.
Keywords: intraarticular contrast material • knee injury • meniscal tears • MR arthrography • MRI
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The criteria for diagnosing a recurrent tear in these four studies using MR arthrography were similar but not identical. Identification of intrameniscal signal intensity contacting the surface, meniscal distortion, or fragments were the most commonly used criteria [1-4]. The relative intensity of the intrameniscal signal contacting the surface was described as either being of increased intensity or equal to that of intraarticular contrast material.
We noted in clinical correlation conferences that we failed to diagnose several recurrent meniscal tears on knee MR arthrography using the criterion that the intrameniscal signal contacting the meniscal surface had to have intensity equal to the adjacent intraarticular gadolinium contrast material on fat-saturated T1-weighted images. Review of these cases suggested that in some patients the meniscal signal contacting the surface in a recurrent tear was increased above the baseline low signal intensity of the meniscus but not equal to the intensity of the adjacent intraarticular gadolinium contrast material. Because of this observation, we undertook a study of our patients who underwent MR arthrography after meniscal surgery to assess the intensity of meniscal signal contacting the meniscal surface in recurrent tears.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Thirty-four patients had either prior medial or prior lateral meniscal surgery, and three had prior surgery on both their medial and lateral menisci (total of 40 menisci with prior surgery). Four lateral and seven medial menisci had been repaired using a suture technique. Seven lateral and 22 medial menisci had undergone resection of the torn portion of the meniscus during the initial surgery. At arthroscopy, recurrent tears were identified in 21 of the 29 menisci with prior resection and in nine of the 11 menisci that had been previously repaired, for a total of 30 recurrent tears and 10 intact menisci. All arthroscopies were performed by fellowship-trained sports medicine orthopedists with more than 8 years of experience each. In those patients who had arthroscopy but did not have a recurrent tear, arthroscopy was performed for the treatment of loose bodies or advanced chondromalacia.
MR Arthrography
All MR arthrography was performed using our standard technique. Under
fluoroscopic guidance, a 1:200 solution of gadodiamide (Omniscan, Amersham
Health) was injected into the knee joint from a lateral approach beneath the
patella. Arthrography on the first 14 postoperative menisci was performed
using 20 mL of contrast material, whereas arthrography on the next 26
postoperative menisci were performed using 40 mL of contrast material. We
switched to the larger injection volume because of a subjective impression
that the degree of joint distention was insufficient with 20 mL. As a second
procedure change with the larger volume, we wrapped an elastic bandage around
the knee in the suprapatellar region after the joint injection to force more
contrast material around the menisci. Before the contrast-enhanced MR images
were obtained, all 37 patients walked 60-80 feet (18-24 m) to distribute the
contrast material within the knee joint.
The MR arthrography included coronal and sagittal T1-weighted spin-echo images obtained with spatial frequency chemical fat saturation (FatSat, GE Healthcare) on a 1.5-T MR unit (Signa, GE Healthcare) using a dedicated phased-array knee coil. Parameters for both the coronal and sagittal images were TR/TE, 600/15-20; field of view, 14-cm; matrix, 256 x 192; and slice thickness, 3 mm with a 1.5-mm interslice gap and one excitation.
Retrospective Review
Two fellowship-trained musculoskeletal radiologists reviewed the MR
arthrograms of these 37 patients in consensus. The radiologists were told only
that these MR arthrograms were obtained on patients who had prior meniscal
surgery. They were not told which menisci had prior surgery or the results of
the follow-up arthroscopy.
The two reviewers assessed each MR arthrogram for the presence of intrameniscal signal that contacted the surface of the meniscus on the fat-saturated T1-weighted images. If there was meniscal signal contacting the surface of a meniscus, the observers counted the total number of images on the coronal and sagittal sequences with such contact for each meniscus. Finally, in each meniscus with signal contacting the surface, they indicated, on the image with the highest signal intensity, whether the signal intensity was as high as or less than that of adjacent intraarticular contrast material. Each of these findings was then correlated with the findings at the second arthroscopy.
Statistical Analysis
The percentage of recurrent tears in which signal intensity contacting the
surface of the meniscus was equal to that of gadolinium was compared for MR
arthrograms using 20 and 40 mL of intraarticular contrast material. We
performed a chi-square analysis with a significant difference defined as
p < 0.05.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
|
|
|
Of the 10 intact menisci, six had no signal contacting the surface, two had signal contacting the surface equal to intraarticular contrast material in intensity, and two had signal contacting the surface that was less than intraarticular contrast material (Figs. 3A, 3B, and 4).
|
|
|
The percentage of recurrent tears showing signal intensity equal to that of intraarticular contrast material did not significantly differ between arthrograms obtained using 20 mL compared with those using 40 mL (p = 0.72).
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This finding was suspected in view of our clinical experience. However, the reason for the signal intensity contacting the meniscal surface being less than that of gadolinium in a recurrent tear is not obvious. One would assume that because contrast material should enter the tear, the signal should have the intensity of gadolinium. We speculate that there is only a small amount of the contrast material within the slitlike nature of a tear so that volume averaging minimizes the overall signal intensity. There may also be loose granulation tissue within the tear limiting the amount of contrast material within the tear. We remember that when we performed conventional knee arthrography, vigorous varus and valgus stressing of the knee was often necessary to force contrast material to enter a meniscal tear. Presumably, the amount of contrast material present within the tear is minimal because the knee is not being stressed during the knee MR examination.
Previous studies on the use of MR arthrography for the diagnosis of a recurrent tear used differing descriptions of the intensity of the signal contacting the surface, describing it as "an increase in intrameniscal signal intensity" [1], "fluid entering the meniscal substance" [2], "communication of the contrast mixture from the joint into the substance of the meniscus" [3], and "increased intrameniscal signal intensity extending to the meniscal articular surface" [4]. In addition, three review articles indicated that the criteria for diagnosing a recurrent tear included "extension of the gadolinium mixture into the residual meniscal fragment or into the site of the repair" [5], "intrameniscal gadolinium entering the crevice of the meniscal tear" [6], and "signal intensity similar to that of gadolinium on T1-weighted images...or to that of fluid on T2-weighted images that extends into the meniscus" [7]. Using the criterion that the meniscal signal intensity had to equal that of intraarticular contrast material, we failed to diagnose a number of recurrent tears on our original MR interpretations.
It is interesting that increasing the volume of intraarticular contrast material did not significantly increase the number of recurrent tears showing gadolinium signal intensity. Previous studies evaluating the accuracy of MR arthrography for recurrent tears have used a range of injected volume of contrast material including 25 mL [3], 20-40 mL [2], 30-50 mL [4], and 40-50 mL [1]. Although we found no significant difference in our accuracy when using 20 mL or 40 mL of contrast material, we thought the higher volume provided better delineation of the meniscal surfaces. For this reason, we have continued our recent protocol of injecting 40 mL of contrast material and wrapping the suprapatellar region of the knee with an elastic bandage after the injection.
An unexpected finding was that a tear was far more likely if there were two or more images with signal contacting the meniscal surface than if only one image had signal contacting the meniscal surface. Twenty-three (96%) of 24 menisci with signal intensity contacting the surface on two or more images had recurrent tears. In contrast, only four (57%) of the seven menisci with one image with signal contacting the surface had recurrent tears. This finding has not been previously noted in studies on MR arthrography for recurrent tears. However, our findings are similar to those noted in a study of the accuracy of conventional MRI in menisci without prior surgery [8]. In that study of 164 meniscal tears, the positive predictive value of more than two images with surface contact was 93% for a meniscal tear, and the positive predictive value of only one image with surface contact was only 46%.
Intrameniscal signal was seen contacting the meniscal surface on a single image in three intact menisci and on three images of one intact meniscus. The reason that intact menisci would have signal extending to the surface remains unclear. Perhaps some of these were not intact menisci but had small recurrent tears that were not seen at the second arthroscopy. Another explanation is suggested from a nonarthrographic MR study of an animal model [9]: It may be that some postoperative menisci have scars that have increased signal intensity on fat-saturated T1-weighted images on MR arthrography, but this was infrequently the case in our study.
The limitations of our study include the study size and its retrospective nature. However, our study size of 40 previously operated menisci with arthroscopic correlation is similar to the prior studies, which included 13 [2], 28 [4], 41 [1], and 57 postoperative menisci [3] with MR arthrography and operative correlation. In addition, our study was of a sufficient size to identify 11 torn menisci in which the intensity of the signal was not equal to that of the gadolinium contrast material.
The second limitation of our study is its retrospective nature. Our MR arthrograms were originally interpreted as showing a recurrent tear only if the intrameniscal signal intensity contacting the meniscal surface was equal to the intensity of the gadolinium contrast material. We do not know how many patients with recurrent tears were misdiagnosed by MR arthrography as having no tears and thus did not go on to arthroscopy. Similarly, we do not know how many intact menisci had increased signal intensity contacting the surface. Because of this unavoidable limitation of patient selection, we did not calculate sensitivity and specificity as part of this study. In addition, we do not know if the high likelihood of a recurrent tear when a meniscus has two or more images with signal contacting the surface will apply prospectively.
Nevertheless, our study has proven that the intensity of intrameniscal signal contacting the meniscal surface in a recurrent tear need not be equal to that of gadolinium contrast material on fat-saturated T1-weighted images. As a result of our findings, we now consider an MR arthrogram positive for a recurrent tear if the meniscus has signal contacting the meniscal surface whether it is equal to or less than gadolinium contrast intensity. In addition, we diagnose a meniscus as torn if signal contacting the surface is present on two or more images or as possibly torn if signal contacting the surface is present on only one image.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  S Ostlere Imaging the knee Imaging, September 1, 2007; 19(3): 249 - 268. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
