Prevalence of Meniscal Radial Tears of the Knee Revealed by MRI After Surgery

Transcription

1 Downloaded from by on 12/22/17 from IP address Copyright RRS. For personal use only; all rights reserved Thomas Magee 1 Marc Shapiro David Williams Received June 13, 2003; accepted after revision October 21, ll authors: Department of Radiology, Neuroimaging Institute, 27 E Hibiscus lvd., Melbourne, FL ddress correspondence to T. Magee (tmageerad@cfl.rr.com). JR 2004;182: X/04/ merican Roentgen Ray Society Prevalence of Meniscal Radial Tears of the Knee Revealed by MRI fter Surgery OJECTIVE. Meniscal resection decreases the ability of the meniscus to evenly distribute forces placed on it. These forces are oriented centrifugally on the meniscus by normal weightbearing and are distributed by circumferentially oriented fibers. This alteration may predispose the knee to radial tears after surgery. SUJECTS ND METHODS. One of three musculoskeletal radiologists prospectively interpreted 100 consecutive postoperative MRI examinations of the knee. prospective MRI report was generated for the referring orthopedic surgeon, and prospective MRI interpretations were correlated with arthroscopic findings (n = 63). MRI examinations on those patients who underwent second-look arthroscopy were retrospectively reviewed by three musculoskeletal radiologists who reached a consensus on the prevalence of new postoperative meniscal radial tears. MRI criteria for radial tear diagnosis were used as outlined by Tuckman et al.: truncation, abnormal morphology, lack of continuity, absence of the meniscus, or any combination of those criteria on one or more MR images. n additional criterion used was abnormal increased signal in that area on T2-weighted images. RESULTS. Thirty-two of the 100 patients had meniscal radial tears on prospective MRI interpretations. In 29 of these 32 patients, second-look arthroscopy confirmed meniscal radial tears in the areas described on the MRI examinations. Five additional radial tears were shown on second-look arthroscopy that were not seen on prospective MRI interpretations. Two of those additional five radial tears were seen on consensus retrospective MRI review. CONCLUSION. In this study, a 32% prevalence of meniscal radial tears in the postoperative knee was present on prospective MRI interpretations as opposed to a reported 14% prevalence in the nonoperated knee. Meniscal resection decreases the ability of the meniscus to evenly distribute forces placed on it. This circumstance may increase the prevalence of meniscal radial tears in the postoperative knee. New meniscal radial tears are common in patients presenting with pain after knee surgery. M eniscal resection results in altered biomechanics of knee function that can lead to a gradual deterioration of the chondral surface and premature osteoarthritic changes. Meniscal resection may alter the hoop mechanism of the meniscus and decrease the ability of the meniscus to transmit loads [1 4]. This circumstance may predispose the remaining intact meniscus to injury and specifically place the meniscus at risk for radial tears because of the altered hoop mechanism of the knee. Meniscal tears are commonly seen at arthroscopy. Radial tears have been reported to be very difficult to visualize on MR images, but these tears account for a large percentage of meniscal tears missed by MRI. Radial tears involve the meniscal free edge. Thus, keys to interpretation of this injury are recognition of absence or blunting of the inner point of the meniscal triangle [1, 5, 6]. radial tear is perpendicular to the long circumferential axis of the meniscus and can be full-thickness, extending from the apex through the periphery of the meniscus, or partial-thickness. full-thickness radial tear can result in loss of meniscal integrity resulting in either two separate pieces of meniscus or one large meniscal piece attached to the tibia at one end [1, 5]. These tears can be repaired or débrided, but the ability to regain the full function of the meniscus is most likely compromised. The meniscus becomes subluxated peripherally in an anterior, poste- JR:182, pril

2 Downloaded from by on 12/22/17 from IP address Copyright RRS. For personal use only; all rights reserved Magee et al. rior, or lateral direction resulting in decreased ability of the meniscus to protect the articular surface of the knee. This situation results in premature degenerative change in the affected area [1, 5 7]. Meniscal resection can affect distribution of weight-bearing stresses. These stresses are normally distributed throughout the meniscus by the circumferentially oriented collagen fibers. meniscal resection decreases the ability of the meniscus to distribute these weight-bearing forces. We report findings on the prevalence of meniscal radial tears on MRI examinations in the postoperative knee and the arthroscopic prevalence of meniscal radial tears on second-look arthroscopy in patients with new or persistent knee pain after meniscal surgery. Subjects and Methods One hundred consecutive MRI or MR arthrography examinations on postoperative knee patients with meniscal resection between January 2002 and January 2003 performed at our institution were reviewed. Sixty-eight of these 100 consecutive patients had preoperative MRI examinations available for direct comparison with MRI examinations obtained after surgery. Ninety-one of these patients had operative reports from their first surgery available for review. Those patients with meniscal repairs or those who did not have meniscal surgery were excluded from this study. ll patients in the study had partial meniscal resection. None of the patients underwent total meniscectomies. ll MRI examinations were interpreted prospectively by one of three musculoskeletal radiologists, and a report was generated for the referring orthopedic surgeon. The age range of the patients was years (mean, 39 years). ll patients underwent MRI of the knee in coronal, axial, and sagittal planes on a 1.5-T Symphony scanner (Siemens Medical Systems). Coronal turbo spin-echo T1-weighted (TR/TE, 749/10), coronal and sagittal turbo spin-echo T2-weighted (3,950/51), proton density weighted fat-saturated sagittal (1,800/12), and T2-weighted fast low-angle shot axial images (905/18) were obtained. The field of view was 15 cm on coronal and sagittal images and 16 cm on axial images. Slice thickness on each sequence was 4 mm with a 10% interslice gap. quadrature extremity coil was used. MR arthrograms were obtained for 42 of the 100 patients immediately after the conventional MRI examination when this procedure was requested by the referring orthopedic surgeon. MR arthrography was performed with approximately 25 ml of a mixture of diluted saline and gadopentetate dimeglumine (Magnevist, erlex Laboratories) with a concentration of 0.15 ml of gadopentetate dimeglumine per 20 ml of normal saline. 22-gauge needle was placed beneath the center of the articular surface of the patella. This procedure was performed on the MRI table without the use of fluoroscopy. One of three musculoskeletal radiologists performed the injection. ll injections were successfully performed within the joint space. fter injection of this mixture into the knee joint, patients exercised the knee by extending and bending it continuously for 5 min before rescanning. fter exercise, T1-weighted fat-satu- rated coronal and sagittal images (684/9.5) were obtained; these images were obtained before MR arthrography for direct comparison with images obtained after MR arthrography. The same slice thickness, gap, field of view, and coil used in the conventional MRI examination were used in the MR arthrography examination. Prospective MRI interpretations were correlated with arthroscopic findings for those patients who underwent second-look arthroscopy (n = 63). The patients were selected on the basis of clinical examination findings, MRI examination findings, and the patient s willingness to undergo second-look arthroscopy. Once second-look arthroscopy results were available, the MRI examinations were retrospectively reviewed by consensus of the three reviewers and correlated with arthroscopic results. Consensus was achieved when all reviewers agreed that a meniscal radial tear was present or not on an MR image. MRI findings were retrospectively reviewed in the patients with radial tears on arthroscopy. MR images were assessed for location of radial tears. MRI criteria used for diagnosis of radial tear were truncation or abnormal morphology of the meniscus on one or more MR images or abnormal increased signal in that area on fat-saturated proton density weighted or T2-weighted images [1, 2]. Results One hundred consecutive postoperative MRI and MR arthrography knee examinations from January 2002 to January 2003 were reviewed. Sixty-eight of these 100 consecutive patients had preoperative MRI exam- Fig year-old man with knee pain after surgery. Patient had surgically proven meniscal radial tear., Proton density weighted sagittal MR image (TR/TE, 1,800/12) shows findings consistent with meniscal radial tear (arrow)., T2-weighted sagittal MR image (3,950/51) shows linear radial tear (arrow). 932 JR:182, pril 2004

3 Downloaded from by on 12/22/17 from IP address Copyright RRS. For personal use only; all rights reserved MRI of Meniscal Radial Tears of the Knee inations available for direct comparison with postoperative MRI examinations. Ninety-one of these 100 patients had operative reports from their first surgery available for review. Meniscal radial tears were not seen on review of preoperative MRI examinations and were not described in any of the 91 operative reports from the first surgeries. Of these 100 patients, 32 had radial tears reported prospectively (32% prevalence). On prospective MRI interpretation, one radiologist reviewed the MRI examination and generated a report. Twentynine of these 32 patients had second-look arthroscopy confirming meniscal radial tears in the areas described on the MRI examinations (Figs. 1 6). The locations of these arthroscopically detected radial tears were as follows: body of the lateral meniscus (n = 21), posterior horn of the lateral meniscus (n = 5), body of the medial meniscus (n = 2), and the posterior horn of the medial meniscus (n = 1). No radial tears that were not present on arthroscopy (i.e., there were no false-positive MRI interpretations of radial tears) were described in MRI reports. Of the 32 meniscal radial tears seen prospectively on postoperative MRI examination, 28 were in the menisci that had previous partial meniscectomy. f- ter arthroscopy results were available, those patients who underwent arthroscopy had their MRI examinations retrospectively reviewed by consensus of the three reviewers and correlated with arthroscopic results. Five additional radial tears that were not seen on prospective MRI interpretations were shown on second-look arthroscopy. Two of these five additional radial tears seen on arthroscopy were seen on retrospective consensus MRI interpretations (Fig. 7). Otherwise, consensus retrospective MRI interpretations were the same as the prospective MRI interpretations. Fig year-old man with knee pain after surgery. Patient had surgically proven meniscal radial tear., Proton density weighted fat-saturated sagittal MR image (TR/TE, 1,800/12) shows findings suspicious for radial tear (arrow)., T2-weighted sagittal MR image (3,950/51) shows radial tear (arrow) as area of abnormal increased signal touching superior and inferior portions of meniscus. C, T1-weighted fat-saturated sagittal MR image (684/9.5) obtained after arthrography shows radial tear (arrow). C JR:182, pril

4 Downloaded from by on 12/22/17 from IP address Copyright RRS. For personal use only; all rights reserved Magee et al. Fig year-old man with knee pain after surgery. Patient had surgically proven meniscal radial tear., Proton density weighted sagittal MR image (TR/TE, 1,800/12) shows area of abnormal increased signal (arrow) in body of meniscus consistent with radial tear., T2-weighted sagittal MR image (3,950/51) shows linear area of abnormal increased signal (arrow) touching superior and inferior portions of meniscus, consistent with radial tear. Fig year-old woman with knee pain. Patient had surgically proven meniscal radial tear., T2-weighted sagittal MR image (TR/ TE, 3,950/51) shows horizontal meniscal tear (arrow) in posterior horn of medial meniscus. Tear touched inferior articular surface (not seen on this image)., T2-weighted sagittal MR image (3,950/51) obtained after patient had surgery for horizontal meniscal tear shows linear area of abnormal increased signal (arrow) touching superior and inferior portions of meniscus, consistent with radial tear. Fig year-old man with knee pain. Patient had surgically proven meniscal radial tear., Proton density weighted coronal MR image (TR/TE, 2,005/16) obtained at outside institution shows no discrete radial tear. Patient had horizontal meniscal tear (not shown)., T2-weighted coronal MR image (3,950/51) obtained after patient had surgery for horizontal meniscal tear shows linear area of abnormal increased signal (arrow) touching superior and inferior portions of meniscus, consistent with radial tear. 934 JR:182, pril 2004

5 Downloaded from by on 12/22/17 from IP address Copyright RRS. For personal use only; all rights reserved MRI of Meniscal Radial Tears of the Knee Fig year-old man with knee pain after surgery. Patient had surgically proven meniscal radial tear., T2-weighted coronal MR image (TR/TE, 3,050/105) shows linear area of abnormal increased signal (arrow) in body of meniscus, consistent with radial tear., T1-weighted fat-saturated coronal MR image (684/9.5) obtained after arthrography shows radial tear (arrow). Discussion Persistent or recurrent pain after knee surgery is common. MRI is useful for visualization of some common extrameniscal causes of postoperative knee pain such as severe degenerative arthrosis, chondral injuries, or avascular necrosis. MRI is also useful in the diagnosis of new meniscal tears in postoperative patients. The sensitivity of conventional MRI in the diagnosis of new meniscal tears in the postoperative knee is less than that in knees that have not had surgery. MR arthrography may be more sensitive than conventional MRI for detection of new meniscal tears in the postoperative knee [8]. In 42 patients, MR arthrography was performed in addition to conventional MRI at the request of the referring orthopedic surgeon. The depiction of gadolinium traversing an area of abnormal signal in the meniscus is diagnostic of a meniscal tear in the postoperative knee [8]. high prevalence of radial tears exists on MRI examination in postoperative patients who had previous meniscal resection. In our series, the prevalence was 32% on MRI examination and even higher in those patients who underwent second-look arthroscopy 32 of 100 knee MRI examinations and 34 of 63 second-look arthroscopies. In an earlier study performed on patients who had not undergone previous knee surgery, the prevalence of meniscal radial tears was 14% [2]. In this study, the prevalence of meniscal radial tears on MRI examinations of the postoperative knee was more than double the prevalence on MRI examinations of knees without surgery. The significantly higher prevalence of radial tears after meniscal resection may be explained by an alteration in the biomechanics of knee function after resection of a portion of the meniscus. This alteration in biomechanics is well known to predispose postoperative patients to premature osteoarthritis [3] and may also redistribute stresses to the meniscus so that it is more susceptible to radial tears. Forces may be distributed to the remaining intact portions of the meniscus that Fig year-old woman with knee pain after surgery. Patient had surgically proven meniscal radial tear., T2-weighted coronal MR image (TR/TE, 3,050/105) shows linear area of abnormal increased signal (arrow) in body of meniscus. This signal was not prospectively interpreted as radial tear., T1-weighted fat-saturated coronal MR image (684/9.5) obtained after arthrography shows area of gadolinium abnormally traversing area of abnormal signal (arrow) seen on T2-weighted images. This signal was interpreted prospectively as meniscal tear, not as radial tear. JR:182, pril

6 Magee et al. Downloaded from by on 12/22/17 from IP address Copyright RRS. For personal use only; all rights reserved make them more prone to radial tears. The diagnosis of a radial tear is important because it can be a source of pain for a patient, and the meniscal free edge can be treated surgically with débridement or resection. One potential limitation of this study is that retrospective review of the images was performed by consensus of the three reviewers rather than as an independent review of the images by each of the three reviewers. In conclusion, meniscal radial tears are very common in patients presenting with new or persistent pain after knee surgery. In this series, the prevalence of meniscal radial tears on MRI was 32%. Radial tears may be difficult to visualize, especially in a postoperative knee; therefore, radiologists need to be aware of MRI signs of radial tears such as meniscal truncation, abnormal meniscal morphology, lack of continuity of the meniscus on one or more MR images, and linear areas of abnormal increased signal on T2-weighted MR images. careful analysis of postoperative knee MRI examinations for meniscal radial tears is essential because these tears are a common and often treatable cause of knee pain in the postoperative patient. References 1. Tuckman G, Miller WJ, Remo JW, Fritts HM, Rozansky MI. Radial tears of the menisci: MR findings. JR 1994;163: Magee T, Shapiro M, Williams D. MR accuracy and arthroscopic incidence of meniscal radial tears. Skeletal Radiol 2002;31: oyd KT, Myers PT. Meniscus preservation: rationale, repair techniques and results. Knee 2003;10: Jones RS, Keene GC, Learmonth DJ, et al. Direct measurement of hoop strains in the intact and torn human medial meniscus. Clin iomech 1996;11: Rubin D, Paletta G Jr. Current concepts and controversies in meniscal imaging. Magn Reson Imaging Clin N m 2000;8: Kaplan P, Dussault R, Helms C, et al. Knee. In: Kaplan P, ed. Musculoskeletal MRI. Philadelphia, P: W Saunders, 2001: Lambert S, Helms C, Higgins HD. Radial meniscal tears: incidence and MR appearance. (abstr) Radiology 2000;217(P): Totty WG, Matava MJ. Imaging the postoperative meniscus. Magn Reson Imaging Clin N m 2000;8: JR:182, pril 2004