Bucket-Handle Meniscal Lesions: Magnetic Resonance Imaging Criteria for Reparability

Bucket-Handle Meniscal Lesions: Magnetic Resonance Imaging Criteria for Reparability Patricia Thoreux, M.D., Frédérique Réty, M.D., Geoffroy Nourissat, M.D., Xavier Rivière, M.D., Patrick Safa, M.D., Sébastien Durand, M.D., and Alain-Charles Masquelet, M.D.

Purpose: The purpose of this study was to determine the accuracy of magnetic resonance imaging (MRI) in predicting knee bucket-handle meniscal tear (BHMT) reparability. Methods: Twenty-eight patients who underwent knee arthroscopy by a single surgeon for BHMT with prior MRI examination were included. BHMTs were diagnosed by MRI based on the association of a displaced meniscal fragment on coronal images and one of the following three signs on sagittal slices: flipped meniscus sign, double posterior cruciate ligament, and meniscal fragment within the intercondylar notch. BHMT patients’ MRIs were retrospectively reviewed independently to search for criteria of reparability by 2 observers with different degrees of experience in musculoskeletal radiology, and disagreements were arbitrated to consensus. The criteria for BHMT reparability were as follows: (1) rim width of less than 4 mm; (2) tear length of 1 cm or greater, regardless of total lesion length; and (3) generation of isosignals by the inner meniscal fragment and peripheral rim compared with the normal contralateral meniscus of the same knee. The first 2 criteria indicate an adequate meniscal lesion length in the vascularized zone (only the peripheral third), enabling meniscal healing after repair; the third criterion guarantees that the meniscus is nondegenerative. Results: Of the BHMTs, 5 (17.9%) were arthroscopically reparable and 23 (82.1%) were not. Interpretation of magnetic resonance images correctly predicted reparability in 4 of 5 reparable BHMTs and irreparability in 22 of 23 irreparable BHMTs (26/28 lesions). Interobserver agreement was good for the prediction of reparability (␬ ⫽ 0.7). Conclusions: These results suggest that knee BHMTs that are predicted to be reparable by MRI would have a high likelihood of actually being reparable. Level of Evidence: Level II, development of diagnostic criteria on basis of consecutive patients and gold standard. Key Words: Bucket-handle meniscal tear—Knee—Magnetic resonance imaging—Reparability prediction—Meniscal repair.

K

nee menisci play major roles in load transmission, shock absorption, joint stabilization, and joint lubrication. Partial or especially total meniscectomy is responsible for the development of degenerative changes in the long term.1-4 Hence, preserving

From the Departments of Orthopedic Surgery (P.T., G.N., S.D., A.-C.M.) and Radiology (F.R., X.R., P.S.), Hôpital Avicenne-University Paris XIII; and the Department of Medical Anatomy, University Paris XIII (S.D.), Bobigny, France. The authors report no conflict of interest. Address correspondence and reprint requests to Patricia Thoreux, Service de Chirurgie Orthopédique, Hôpital Avicenne, 125 Route de Stalingrad, 93009 Bobigny Cedex, France. E-mail: patricia.thoreux@ avc.ap-hop-paris.fr © 2006 by the Arthroscopy Association of North America 0749-8063/06/2209-5418$32.00/0 doi:10.1016/j.arthro.2006.04.111

954

meniscal tissue and preserving its biomechanical function have become essential goals in knee surgery.5 Arthroscopic techniques have improved meniscectomy precision and repair modalities, providing data on meniscal vascularization and healing potential6,7 that have enabled surgeons to determine which meniscal lesions might be amenable to repair.8 Preoperative knowledge of potential meniscal tear reparability is important for the patient and the surgeon. For the patient, physical therapy requirements and return to activity are major factors to be considered in scheduling surgery, and meniscal lesion repair clearly prolongs the interval before one can return to sports and professional activities. For the surgeon, the possibility of meniscal tear repair is taken into account in preoperative planning in terms of procedure duration and equipment requirements. Meniscal lesions are

Arthroscopy: The Journal of Arthroscopic and Related Surgery, Vol 22, No 9 (September), 2006: pp 954-961

BUCKET-HANDLE MENISCAL LESIONS deemed reparable when they have the theoretic potential to heal after reparative surgery, which explains why such tears are usually traumatic rather than degenerative, and are located within the vascularized zone of the meniscus (i.e., within the peripheral third of the meniscus); notably, the closer the lesion is to the meniscosynovial junction, the better the vascularization. They are most commonly vertical-longitudinal peripheral tears, 1 cm long or greater.7,9,10 The meniscal tear length gives an indication of its instability, which is an important criterion of reparability but varies widely according to the authors and their specialties.8,11-16 For orthopaedic surgeons, meniscal lesion instability can be a dynamic factor to be evaluated intraoperatively by palpation with a hook: the tear is considered unstable when it can be displaced beyond the inferior pole of the femoral condyle or when it remains displaced when the hook is removed (or both).11-14 According to other authors, every meniscal tear exceeding 5 mm in length8 or 10 mm in length15 is deemed unstable. For radiologists, instability is obligatorily associated with a displaced fragment.16 In addition, the damage to the torn and the peripheral segment must be minimal, with no degenerative changes.8 The final decision regarding repair versus meniscectomy is ultimately made based on the arthroscopic assessment of the anatomic characteristics of the meniscal lesion and the clinical setting. Bucket-handle meniscal tears (BHMTs) comprise an ideal subgroup of lesions for the study of reparability criteria because they are exemplary unstable lesions, but they are quite rare (about 10% of meniscal lesions).17,18 A BHMT is usually a vertical or oblique longitudinal tear with an attached fragment displaced away from the periphery of the meniscus.17,18 Most tears begin at or near the posterior insertion of the meniscus onto the tibia and extend beyond the junction of the middle and anterior third, allowing the displacement of the inner segment beyond the equator of the femoral condyle and into the intercondylar notch.19-21 For orthopaedic surgeons, a BHMT can be displaced into the intercondylar notch at insertion of the arthroscope or during palpation with a hook. Magnetic resonance imaging (MRI) criteria correspond exclusively with the presence of a meniscal fragment in the intercondylar notch. MRI is considered to be the best imaging technique for the diagnosis of meniscal tears in the knee22 and, to the best of our knowledge, the only imaging tool used to predict the reparability of these tears.23-25 Although the use of contrast imaging to evaluate the results of meniscal repair has been discussed,26-28

955

FIGURE 1. Coronal T2*-weighted image of a 31-year-old man’s knee showing a displaced meniscal fragment below the PCL (white arrow) in the intercondylar notch, corresponding with the required coronal magnetic resonance image of a displaced meniscal fragment. The rim width was measured between the 2 black arrows. The inset in the upper right corner shows the level of the coronal slice.

MRI appears to be superior, given its ability to detect degenerative meniscal changes and to determine the location of tears with greater potential for healing, such as those at the meniscosynovial junction. The purpose of this study was to evaluate the accuracy of MRI in predicting knee BHMT reparability. METHODS Inclusion Criteria During the 9-year period between 1994 and 2003, all 33 patients at our institution with arthroscopyconfirmed BHMTs and prior MRI examination were considered for this study. The 33 MRI studies were reviewed by 2 radiologists, with consensus being obtained for the diagnosis of BHMT. This diagnosis was established based on the obligatory observation of a displaced meniscal fragment on coronal images and one of the following three findings on sagittal slices: (1) flipped meniscus sign, (2) double posterior cruciate ligament (PCL), and (3) meniscal fragment within the intercondylar notch. On coronal images, a displaced meniscal fragment is in continuity between the anterior and posterior horns observed on contiguous coronal slices (Fig 1). We defined a flipped meniscus

956

P. THOREUX ET AL. performed arthroscopy on all of these patients. The clinical criteria for arthroscopic surgery included a history of locking, a catching sensation, and pain in the involved compartment for more than 6 months with or without a history of trauma and with joint line tenderness associated with a meniscal lesion on preoperative MRI. Of the patients, 12 (42.9%) had right knee involvement, and in 16 (57.1%), the left knee was affected. The lateral meniscus was torn in 6 patients (21.4%), and the medial meniscus was torn in 22 (78.6%); 10 patients (35.7%) had a concomitant anterior cruciate ligament rupture. MRI Technique

FIGURE 2. Proton density–weighted sagittal image of a 40-yearold man’s knee showing a flipped meniscus sign (i.e., the meniscal fragment had flipped anteriorly so that the anterior horn appears large in height and width) (white arrow), corresponding with 1 of the 3 possible sagittal images retained for BHMT diagnosis. The rim width was measured between the 2 black arrows.

sign on sagittal images as a shortened posterior horn associated with an abnormally large anterior horn (ⱖ8 mm in height or width [or both]) (Fig 2). The anterior horn appears larger than normal because the BHMT fragment is seen immediately posterior or superior to the anterior horn.17,29 The double PCL is visualized on sagittal images as a low–signal intensity band parallel and anterior to the PCL (Fig 3).30 A meniscal fragment within the intercondylar notch was defined as a bandlike area of low signal intensity within the notch but not appearing on the same slice as the PCL on sagittal images. Other signs found in the literature were not evaluated because of their low specificity.31 A BHMT was not diagnosed by MRI in 4 patients who had MRI diagnoses of complex meniscal lesions and underwent meniscectomy because the characteristics of their lesions did not satisfy the reparability criteria. In addition, a patient whose interval between MRI and arthroscopy was more than 400 days was excluded from this investigation. Therefore our search for criteria predicting BHMT reparability was based on 28 patients (22 male and 6 female patients) whose mean age was 31 years (range, 14 to 58 years). A single orthopaedic surgeon (P.T.)

For 25 patients, MRI examinations were performed at our institution with a 0.5-T imager (Vectra; GE Medical Systems, Waukesha, WI) and an extremity coil. For the other 3 patients, images were obtained with a 1.5-T imager (Gyroscan Intera; Philips Medical Systems, Best, The Netherlands) because of an upgrading of MRI systems. T2*-weighted images were obtained by use of the following parameters: repetition time (TR) of 700 milliseconds and echo time (TE) of 25 milliseconds for coronal images, TR of 800 milliseconds and TE of 25 milliseconds for sagittal images, and a flip angle of 30°. The proton density– weighted images used a TR of 3,100 to 4,000 milli-

FIGURE 3. Proton density–weighted sagittal image of a 20-yearold man’s knee showing a meniscal fragment lying anterior and parallel to the PCL (arrow), making it like a double ligament. The inset in the upper right corner shows the level of the sagittal slice.

BUCKET-HANDLE MENISCAL LESIONS

957

seconds and a TE of 22 to 25 milliseconds. The MRI parameters were as follows: field of view, 16 cm; excitations, 2; and matrix size, 256 ⫻ 192. The section thickness and the between-slice interval were 3 mm and no gap, respectively, for T2*-weighted images and 3 mm and 1 mm, respectively, for proton density– weighted images. Spin-echo images with fat saturation were available for the 3 examinations done on the Philips imager. The fast spin-echo images with fat saturation used a TR of 2,000 milliseconds and a TE of 40 milliseconds with an echo train length of 5. The slice thickness was 3.5 mm with a 0.4-mm betweenslice interval. The knee was placed in a relaxed position at about 10° of flexion and in external rotation; no restraints were applied to the knee during imaging. The mean interval between MRI and knee arthroscopy was 43.5 days (range, 1 to 270 days). Surgical Definition of BHMT and Reparability Criteria The surgical criterion to diagnose a BHMT was a vertical-longitudinal tear in the body of the meniscus that extended at least partially into the anterior half of the meniscus and was displaced or could be displaced beyond the equator of the femoral condyle or into the intercondylar notch. The exact type and location were determined, and the tear length and the remaining rim width were measured with a 3-mm probe. The final determination of meniscal reparability was based solely on the arthroscopic appearance of the tear. The surgical reparability criteria were as follows: (1) tear located in the vascularized peripheral zone of the meniscus, (2) tear length of 1 cm or greater in the vascularized zone, and (3) no degeneration in the peripheral body or the inner portion. Tears within 3 mm of the meniscosynovial junction were considered to be within the meniscal vascular zone. If the tear was 5 mm from the meniscosynovial junction or greater, it was considered to be avascular, unless there was direct evidence of vascularity. According to arthroscopic findings, 5 BHMTs (17.9%) (4 medial and 1 lateral) were identified as reparable and 23 (82.1%) were identified as irreparable (18 medial and 5 lateral). MRI Criteria for Prediction of Meniscal Reparability All images were reviewed retrospectively and independently by 2 observers who were unaware of the patients’ detailed histories, intraoperative findings, and treatment. One was a musculoskeletal radiologist with 5 years of experience in MRI interpretation; the

FIGURE 4. Sagittal T2*-weighted image showing a peripheral meniscal rim less than 4 mm wide (between the 2 black arrows) and a flipped meniscus sign (white arrow) in a 23-year-old man with a reparable BHMT seen on MRI images and at surgery. The inset in the upper right corner shows the level of the sagittal slice.

other was a general radiologist. Discordant findings were arbitrated to consensus. All BHMTs were characterized independently by each examiner with respect to side (medial or lateral). For the radiologists, a tear was considered reparable when it met all of the following criteria: (1) a buckethandle rim segment less than 4 mm wide; (2) tear length of 1 cm or greater, regardless of the total lesion length; and (3) minimal damage to the inner and peripheral meniscal fragments. The rim width was measured on coronal images for the middle third of the meniscus and on sagittal images for the posterior horn. The anterior horn was not evaluated because of the flipped meniscus sign, which prevented distinction between the anterior horn and the displaced meniscal fragment. Rim width was evaluated in millimeter increments with a standard ruler. The measured distances were then compared with the sizing scale shown on each MRI slice to give an overall value in millimeters. BHMTs were considered reparable when the rim was strictly less than 4 mm wide on 4 adjacent images (consistent with tear length ⱖ1 cm) and the inner and peripheral meniscal tear fragments generated low signals or isosignals compared with the normal contralateral meniscus of the same knee, thereby demonstrating the absence of degenerative lesions (Figs 4 and 5). The application of these definitions

958

P. THOREUX ET AL.

FIGURE 5. Sagittal T2*-weighted image showing a flipped meniscus sign (white arrow) and a high-signal peripheral meniscal rim that is more than 4 mm wide (between the 2 black arrows) in a 37-year-old man with an irreparable BHMT seen on MRI images and at surgery.

enabled us to divide the meniscal tears into reparable and irreparable subgroups and to compare MRI results and arthroscopic findings. Statistical Analysis Unweighted ␬ analysis was first undertaken to analyze the concordance between MRI results and arthroscopic findings with regard to the likelihood of reparability. Unweighted ␬ was then applied to determine the coefficient of agreement between observers 1 and 2 for predictions concerning the side of the tear (medial meniscus v lateral meniscus) and its reparability. ␬ Values for the strength of observer agreement were qualitatively defined according to the guidelines of Landis and Koch32 as follows: poor (⬍0.2), fair (0.21 to 0.4), moderate (0.41 to 0.6), good (0.61 to 0.8), and excellent (0.81 to 1.0). The accuracy, sensitivity, specificity, positive predictive value, and negative predictive value of MRI results in accurately predicting reparability were also calculated. RESULTS MRI Evaluation of Reparability of BHMTs Application of the radiologic diagnostic criteria to sagittal images according to the side of the meniscal

lesion is shown in Table 1. The unweighted ␬ analysis for interobserver agreement was excellent for determination of the meniscus side involved (␬ ⫽ 1). On the basis of arthroscopic findings, 5 lesions (17.9%) were arthroscopically reparable and 23 (82.1%) were not. Among the 5 reparable lesions, MRI results after consensus had good agreement with arthroscopic findings for 4 (Fig 6), with 1 false-negative finding. In this latter patient the rim was 5 mm wide and contained a second vertical tear detected by MRI. Among the 23 arthroscopically irreparable lesions, MRI predictions agreed with arthroscopy for 22 of 23 cases, with 1 false-positive finding. The latter patient had a lateral BHMT with a thin peripheral rim, but the posterolateral portion of the lateral meniscus adjacent to the popliteal tendon was avascular and lacked a synovial fringe. The concordance between MRI-predicted reparability and arthroscopic findings is summarized in Table 2. Hence MRI agreed with arthroscopy for 26 of 28 BHMTs (92.9%). Among the arthroscopically defined irreparable lesions, 3 required the 2 observers to arbitrate to consensus. Interobserver agreement was complete for all other cases. Interobserver agreement concerning MRI-predicted reparability for the 28 diagnosed BHMTs was good (␬ ⫽ 0.7). The accuracy, sensitivity, specificity, positive predictive value, and negative predictive value of MRI results in correctly predicting reparability were 0.85 (95% confidence interval, 0.72-0.98), 0.80 (95% confidence interval, 0.65-0.95), 0.95 (95% confidence interval, 0.87-1), 0.80 (95% confidence interval, 0.650.95), and 0.95 (95% confidence interval, 0.87-1), respectively.

TABLE 1. MRI Diagnostic Criteria for BHMTs on Sagittal Images According to Lesion Location

Diagnostic Criteria Double PCL Flipped meniscus sign Fragment in intercondylar notch Double PCL and flipped meniscus sign Double PCL and fragment in intercondylar notch Flipped meniscus sign and fragment in intercondylar notch Double PCL, flipped meniscus sign, and fragment in intercondylar notch

Medial Meniscus (n ⫽ 22)

Lateral Meniscus (n ⫽ 6)

3 1 4 14

0 0 2 0

0

0

2

4

0

0

BUCKET-HANDLE MENISCAL LESIONS

FIGURE 6. Sagittal T2*-weighted image showing a large meniscal fragment displaced into the intercondylar notch (black arrow) and the absence of a peripheral rim (white arrow) in a 23-year-old man with a reparable BHMT seen on MRI images and at surgery. The meniscal tear was situated just at the meniscosynovial junction.

DISCUSSION We obtained good agreement between MRI findings and arthroscopy for reparable BHMTs, with only 1 reparable lesion being missed by MRI. In that case the rim was 5 mm wide and MRI detected a second vertical tear in it. This second tear was not seen, perhaps because a meniscal fragment had been displaced into the joint, despite the rest remaining within the outer third of the meniscus; however, pertinently, the rim was within the vascularized portion of the meniscus, even though the interval between MRI and arthroscopy was only 6 days, with no known trauma during that time. According to the literature, the degree of peripheral vascular penetration ranges from 10% to 30% of the width of the medial meniscus and from 10% to 25% for the lateral meniscus.6 No relationship could be established for this variation with regard to age, sex, race, or anatomic location, but it is certainly possible that, in some patients, the vascularized portion of the meniscus could extend beyond the peripheral third. Consistent with this possibility, Noyes and Barber-Westin33 reported on arthroscopic repair of meniscal tears extending into the avascular central third in patients aged less than 20 years, with good results in 75%. In contrast, Cannon and Vittori34

959

reported a failure rate of 100% for meniscal repairs when the tear was greater than 4 mm from the meniscosynovial junction. The vascularization of tears 3 to 5 mm from the meniscosynovial junction varies, and measures to enhance healing in this area have been attempted, including insertion of a fibrin clot and stimulation and abrasion of the synovial fringe and the meniscal rim.8 In light of these considerations, even though the length and width of a meniscal tear may concord well between MRI and arthroscopic findings, the preoperative criteria cannot be too strict because the final decision concerning reparability must also be based on other factors, such as patient age and intraoperative evaluation of vascularization and tissue quality. The findings on MRI agreed strongly with the arthroscopic findings for 22 of 23 irreparable BHMTs. The MRI failure was a lateral BHMT with a thin peripheral rim, but because the segment adjacent to the popliteal tendon had no blood supply or synovial fringe, the tear was considered to be irreparable in accordance with Arnoczky and Warren.6 Because the meniscal rim width might be underestimated with our measurement technique, we attempted to achieve consistency by specifying that only the completely black wedge-shaped structure be considered meniscal tissue. Chan et al.9 compared proton density–weighted MRI of the meniscus in cadaveric knees with histologic sections and found that the low-signal wedge-shaped structure corresponded only with the avascular (white) zone of the meniscus whereas its high-signal periphery corresponded with the vascularized (red) zone. However, this zone of enhanced signal was much more evident on proton density–weighted images than on T2-weighted images, where the signal was only minimally heightened over that of the lower-signal central zone. The T2*-weighted images and fat-suppressed T2-weighted images were not assessed. TABLE 2. Retrospective Concordance Between MRI-Predicted Reparability and Arthroscopic Findings After Consensus Arthroscopic Findings MRI Prediction Reparable (n ⫽ 5) Irreparable (n ⫽ 23)

Reparable (n ⫽ 5) 4 (3 MM and 1 LM) 1 (1 MM)

Irreparable (n ⫽ 23) 1 (1 LM) 22 (18 MM and 4 LM)

Abbreviations: MM, medial meniscus; LM, lateral meniscus.

960

P. THOREUX ET AL.

We found only 3 published articles on the ability of MRI to predict meniscal reparability,23-25 but those studies differed widely in terms of the sequences used, the configuration of meniscal lesions, and the reparability criteria. Jee et al.25 analyzed only the tear configuration and classified longitudinal and oblique tears as reparable and horizontal, radial, and complex tears as irreparable; tear location, length, and size were not evaluated. Therefore their MRI criteria for reparability seem to be insufficient. Shiozaki et al.24 examined only the prediction of reparability of isolated semilunar lateral meniscus tears by MRI. Their MRI criterion for reparability was the presence of a longitudinal or oblique high–signal intensity line within 3 mm of the meniscosynovial junction without a high–signal intensity area in the meniscal body. However, MRI is not always useful in predicting reparability of symptomatic isolated lateral meniscus tears because of the difficulty of visualizing peripheral detachment of the meniscus and unstable lateral meniscus tears near the popliteal recess. Matava et al.23 analyzed 115 meniscal tears with regard to morphology. A tear was considered reparable when it met all of the following 4 criteria: at least 10 mm long, within 3 mm of the meniscosynovial junction, passing through greater than 50% of the thickness of the meniscus, and minimal damage to the inner meniscal fragment independent of tear configuration. Combining meniscal tear configuration, location, and length to predict reparability probably lowered the sensitivity and specificity of the MRI reparability criteria. The first limitation of our study is its retrospective design—that is, the MRI observers knew that all of the patients had arthroscopy-proven BHMTs.17 However, that knowledge would not be expected to influence their interpretation of the criteria for reparability.25 Second, the small numbers of reparable BHMTs in this study (5/28 [17.9%]) certainly limits our ability to draw conclusions. However, it must be kept in mind that BHMTs are not very common. In addition, the main published studies on reparable tears mostly addressed the outcomes of the attempted repairs rather than the frequency of reparable lesions. Beaufils et al.12 reported that 14% of reparable lesions and tears warranted arthroscopic intervention, whereas Matava et al.23 repaired 22% of tears, performed meniscal trephination in 13%, and performed meniscectomy in 65%. Finally, the MRI studies were obtained by use of a low field strength (0.5 T) or high field strength (1.5 T), and the influence of field strength cannot be excluded in assessing meniscal lesions. However, as a screening test to determine a reparable tear, MRI may not always be definitive. Additional tears occurring during

the interval before surgery cannot be ruled out, and the length of that interval may indeed affect the predictability of treatability.24 CONCLUSIONS Our results suggest that an unstable BHMT predicted to be reparable by MRI based on a rim segment less than 4 mm wide and 1 cm long or greater with a low-signal inner meniscal fragment has a relatively high likelihood of actually being reparable. Our good interobserver agreement (␬ ⫽ 0.7) also suggests that detection of these signs should be reproducible for other observers. Our MRI reparability criteria can be useful when applied to BHMTs or longitudinal meniscal tears, except for those situated in the popliteal area. REFERENCES 1. DeHaven KE, Black KP, Griffiths HJ. Open meniscus repair. Technique and two to nine year results. Am J Sports Med 1989;17:788-795. 2. Johnson RJ, Kettelkamp DB, Clark W, Leaverton P. Factors effecting late results after meniscectomy. J Bone Joint Surg Am 1974;56:719-729. 3. Dejour H, Dejour D, Ait Si Selmi T. Chronic anterior laxity of the knee treated with free patellar graft and extra-articular lateral plasty: 10-Year follow-up of 148 cases. Rev Chir Orthop Reparatrice Appar Mot 1999;85:777-789 (in French). 4. Neyret P, Donell S, Dejour H. Results of partial meniscectomy related to the state of the anterior cruciate ligament. J Bone Joint Surg Br 1993;75:36-40. 5. Eggli S, Wegmüller H, Kosina J, Huckell C, Jakob RP. Longterm results of arthroscopic meniscal repair. An analysis of isolated tears. Am J Sports Med 1995;23:715-720. 6. Arnoczky SP, Warren RF. Microvasculature of the human meniscus. Am J Sports Med 1982;10:90-95. 7. Arnoczky SP, Warren RF. Microvasculature of the meniscus and its response to injury: An experimental study in the dog. Am J Sports Med 1983;11:131-141. 8. DeHaven KE, Arnoczky SP. Meniscal repair. J Bone Joint Surg Am 1994;76:140-152. 9. Chan PS, Kneeland JB, Gannon FH, Luchetti WT, Herzog RJ. Identification of the vascular and avascular zones of the human meniscus using magnetic resonance imaging: Correlation with histology. Arthroscopy 1998;14:820-823. 10. King D. The healing of semilunar cartilages. J Bone Joint Surg Br 1936;18:333-342. 11. Saragaglia D, Tourne Y, Chamseddine A, Butel J. Meniscal sutures combined with the reconstruction of the anterior cruciate ligament. Comparative results between sutures of chronic and recent lesions: 102 cases. Rev Chir Orthop Reparatrice Appar Mot 1990;76:170-176 (in French). 12. Beaufils P, Bastos R, Wakim E, Cho SH, Petit-Jouvet C. Meniscal injury in the plastic reconstruction of the anterior cruciate ligament. Meniscal suture or abstention. Rev Chir Orthop Reparatrice Appar Mot 1992;78:285-291 (in French). 13. Dandy DJ. The arthroscopic anatomy of symptomatic meniscal lesions. J Bone Joint Surg Br 1990;72:628-633. 14. Newman AP, Daniels AU, Burks RT. Principles and decision making in meniscal surgery. Arthroscopy 1993;9:33-51.

BUCKET-HANDLE MENISCAL LESIONS 15. Pierre A, Hulet C, Locker B, Schiltz D, Delbarre JC, Vielpeau C. Outcome of 95 stable meniscal tears left in place after reconstruction of the anterior cruciate ligament. Rev Chir Orthop Reparatrice Appar Mot 2001;87:661-668 (in French). 16. Vande Berg BC, Poilvache P, Duchateau F, et al. Lesions of the menisci of the knee: Value of MR imaging criteria for recognition of unstable lesions. AJR Am J Roentgenol 2001; 176:771-776. 17. Wright DH, De Smet AA, Norris M. Bucket-handle tears of the medial and lateral menisci of the knee: Value of MR imaging in detecting displaced fragments. AJR Am J Roentgenol 1995;165:621-625. 18. Magee TH, Hinson GW. MRI of meniscal bucket-handle tears. Skeletal Radiol 1998;27:495-499. 19. Trillat A. Traumatic lesions of the internal meniscus of the knee. Anatomical classification and clinical diagnosis. Rev Chir Orthop Reparatrice Appar Mot 1962;48:551-560 (in French). 20. Shakespeare DT, Rigby HS. The bucket-handle tear of the meniscus. A clinical and arthrographic study. J Bone Joint Surg Br 1983;65:383-387. 21. Dandy DJ. The arthroscopic anatomy of symptomatic meniscal lesions. J Bone Joint Surg Br 1990;72:628-633. 22. Reicher MA, Hartsman S, Duckwiler GR, Bassett LW, Anderson LJ, Gold RH. Meniscal injuries: Detection using MR imaging. Radiology 1986;159:753-757. 23. Matava MJ, Eck K, Totty W, Wright RW, Shively RA. Magnetic resonance imaging as a tool to predict meniscal reparability. Am J Sports Med 1999;27:436-443. 24. Shiozaki Y, Horibe S, Mitsuoka T, Nakamura N, Toritsuka Y, Shino K. Prediction of reparability of isolated semilunar lateral meniscus tears by magnetic resonance imaging. Knee Surg 2002;10:213-217.

961

25. Jee WH, McCauley TR, Kim JM, et al. Meniscal tear configurations: Categorization with MR imaging. AJR Am J Roentgenol 2003;180:93-97. 26. Davis KW, Tuite MJ. MR imaging of the postoperative meniscus of the knee. Semin Musculoskelet Radiol 2002;6:35-45. 27. Hantes ME, Zachos VC, Zibis AH, et al. Evaluation of meniscal repair with serial magnetic resonance imaging: A comparative study between conventional MRI and MR arthrography. Eur J Radiol 2004;50:231-237. 28. van Trommel MF, Potter HG, Ernberg LA, Simonian PT, Wickiewicz TL. The use of noncontrast magnetic resonance imaging in evaluating meniscal repair: Comparison with conventional arthrography. Arthroscopy 1998;14:2-8. 29. Haramati N, Staron RB, Rubin S, Shreck EH, Feldman F, Kiernan H. The flipped meniscus sign. Skeletal Radiol 1993; 22:273-277. 30. Singson RD, Feldman F, Staron R, Kiernan H. MR imaging of displaced bucket-handle tear of the medial meniscus. AJR Am J Roentgenol 1991;156:121-124. 31. Dorsay TA, Helms CA. Bucket-handle meniscal tears of the knee: Sensitivity and specificity of MRI signs. Skeletal Radiol 2003;32:266-272. 32. Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics 1977;33:159-174. 33. Noyes FR, Barber-Westin AD. Arthroscopic repair of meniscal tears extending into the avascular zone in patients younger than twenty years of age. Am J Sports Med 2002;30:589-600. 34. Cannon WD, Vittori JM. The incidence of healing in arthroscopic meniscal repairs in anterior cruciate ligament-reconstructed knees versus stable knees. Am J Sports Med 1992;20: 176-181.