Portal de Periódicos da CAPES

Classifications in Brief: The International Society of Arthroscopy, Knee Surgery, and Orthopaedic Sports Medicine Classification of Meniscal Tears

2021; Lippincott Williams & Wilkins; Volume: 480; Issue: 1 Linguagem: Inglês

10.1097/corr.0000000000001948

1528-1132

Eli T. Sayegh, Elizabeth Matzkin,

Total Knee Arthroplasty Outcomes

History The medial and lateral menisci are fibrocartilaginous, semilunar structures that play a critical role in load transmission, shock absorption, joint congruity, distribution of contact stresses, proprioception, articular cartilage lubrication and nutrition, and secondary stabilization [9, 16, 20, 22]. The smaller lateral meniscus is less likely to tear than the medial meniscus because of its greater mobility, which is the result of fewer capsular attachments and decreased bony conformity in the lateral compartment [15]. Meniscal deficiency is associated with a proportionate increase in contact stresses and predictable progression of secondary osteoarthritis [8, 10, 11, 24]. Familiarity with the biomechanical consequences of observation, abrasion, trephination, partial meniscectomy, meniscal repair, or meniscal transplantation is important when tailoring treatment to the individual patient and meniscal tear. In 1991, Cooper et al. [5] proposed a circumferential zone classification system for meniscal tears in which Zone 0 is the meniscosynovial junction, Zone 1 is the outer third, Zone 2 is the middle third, and Zone 3 is the inner third. This was the first standardized classification scheme to become broadly adopted. The rationale for this classification system is the microvasculature of the meniscus, in which the outer portion is more vascular and the inner portion is avascular. The capillaries traverse a width no deeper than 10% to 25% of the lateral meniscus and 10% to 30% of the medial meniscus [2, 3]. Unfortunately, observed agreement for grading of rim width was low (46% to 54%) in two studies [1, 5]. This classification system also incorporates a radial location scheme, subdividing each of the medial and lateral menisci into thirds labeled with a letter from A to F [5]. Dunn et al. [7] demonstrated higher observed agreement for a radial location scheme subdividing the meniscus into anterior and posterior halves, albeit with lower descriptive utility. In 2015, LaPrade et al. [14] further proposed a classification system of meniscal root tears consisting of five subtypes, which comprise (1) partial root tear, (2) complete radial root tear, (3) complete root tear with bucket-handle extension, (4) oblique root tear, and (5) root avulsion fracture. There is considerable heterogeneity and nonstandard terminology in the reporting of meniscal tears across disciplines and practices [6, 12, 17, 18]. In 2011, the International Society of Arthroscopy, Knee Surgery, and Orthopaedic Sports Medicine (ISAKOS) committee devised and validated an arthroscopic classification system of meniscal tears to standardize arthroscopic tear characterization, documentation, and outcome reporting. Purpose The most obvious aim of the ISAKOS meniscal tear classification system was to achieve sufficient interobserver reliability to pool multicenter clinical trial data to determine the optimal treatment approach for each meniscal tear type and their correlation with long-term results [1]. The use of a detailed, standardized classification system makes this possible by reducing measurement error and improving the validity of results [1]. Pooling results from multiple institutions using the same classification scheme increases statistical power and facilitates well-designed clinical trials, improving the overall quality and level of evidence of the research literature upon which treatment recommendations are devised [1]. Because the characteristics of meniscal tears are thought to be prognostic [8], the implementation of a consistent classification scheme (by use of a reproducible surgical documentation form) allows for standardized outcome reporting and longitudinal assessment of patient outcomes after meniscal tear treatment [7]. Although not a deliberate aim during the inception of this classification scheme, subsequent studies have highlighted the opportunity for intermethod correlation of MRI and arthroscopy [4, 21]. Therefore, widespread, multidisciplinary implementation of this classification system may allow more accurate patient counseling, treatment planning, and prognostication of meniscal tears. Description The ISAKOS classification incorporates eight arthroscopic findings (Fig. 1): Tear depth categorizes tears as partial if they extend into either the superior or inferior meniscal surfaces and complete if they extend into both surfaces. Rim width, or circumferential location, is estimated using the Cooper zone classification of meniscal tears, depending on how far the tear extends centrally into the meniscus. Tears are categorized as those that extend into Cooper Zone 1 (rim width < 3 mm; outer third), Cooper Zone 2 (rim width 3 mm to < 5 mm; middle third), and Cooper Zone 3 (rim width ≥ 5 mm; inner third). The 3-mm threshold was selected based on evidence of vascularity up to 3 mm from the periphery of the meniscus [1]. If a tear involves multiple zones (such as a radial tear), then it is graded according to the outermost zone involved. Use of historical "red-red," "red-white," and "white-white" terminology is discouraged because of inherent variability in the meniscal vascular anatomy between individuals, which cannot be ascertained with rim width alone. Radial location topographically categorizes tears based on their extension into the posterior, midbody, or anterior zones, dividing the meniscus into thirds. A tear can simultaneously involve multiple zones, such as a complete bucket-handle tear, which would involve all three zones. Alternatively, radial location can be characterized as posterior or anterior, dividing the meniscus into halves, unless a radial tear traverses the mid-body. Lateral meniscal tears are graded as central to the popliteal hiatus if they extend partially or completely in front of the popliteal hiatus. Tear pattern includes longitudinal-vertical (the extension of which is a bucket-handle tear), horizontal, radial, vertical flap, horizontal flap, or complex. The predominant tear pattern should be indicated, except for complex tears, which have two or more tear patterns. Tissue quality refers to whether the meniscal tissue appears degenerative, nondegenerative, or undetermined. The characteristics of degenerative meniscal tissue include cavitations, fibrillations, softening, and/or multiple tear patterns. Tear length (in mm, measured with an arthroscopic ruler or probe) refers to the length of the meniscal tear that extends into its surface. Intermeniscal degeneration or contained tears are not incorporated into the measured tear length. The percentage of the meniscus excised refers to the percentage of meniscal tissue surface area that is arthroscopically removed, which is both numerically estimated and graphically depicted by drawing on a diagram and crosshatching the excised tissue. Fig. 1: A-C This figure shows the ISAKOS criteria for meniscal tears including (A) rim width, (B) tear pattern, and (C) centrality to the popliteal hiatus, in addition to radial location, tear length, tear depth, tissue quality, and the percentage of meniscus excised (not pictured). Reproduced from the Journal of ISAKOS: Joint Disorders & Orthopaedic Sports Medicine, Shah J, Hlis R, Ashikyan O, et al., volume 5, pages 201-207, 2020, with permission from BMJ Publishing Group Ltd. A color image accompanies the online version of this article.Validation Three studies have assessed the reproducibility of the ISAKOS classification [1, 4, 21]. These studies determined interobserver and intraobserver agreement according to the kappa coefficient, which rates the level of agreement between observers [23]. Kappa values range from 0 (indicating agreement due to chance alone) to 1 (indicating complete agreement between observers). Kappa values of 0 to 0.20, 0.21 to 0.40, 0.41 to 0.60, 0.61 to 0.80, and 0.81 to 1.0 denote slight, fair, moderate, substantial, and almost perfect agreement between observers, respectively, according to the criteria of Landis and Koch [13]. The intraclass correlation coefficient (ICC) is used for continuous variables and similarly ranges from 0 to 1.0 (indicating complete agreement) [23], with values greater than 0.75 denoting good interobserver reliability [19]. The Spearman correlation coefficient is used to compare the area of meniscal tissue excised, as graphically depicted, with the estimated numeric percentage [1]. In their international pilot study validating the ISAKOS classification system, Anderson et al. [1] assessed interobserver reliability by having eight international, experienced orthopaedic surgeons grade 37 arthroscopic videos measuring 45 seconds each. They demonstrated substantial agreement for radial location (AP) (κ = 0.65 [95% CI 0.44 to 0.82]) and tear pattern (κ = 0.72 [95% CI 0.63 to 0.81]); moderate agreement for tear depth (κ = 0.52 [95% CI 0.27 to 0.74]), radial location (anterior, middle, and posterior) (κ = 0.46 [95% CI 0.33 to 0.59]), and tissue quality (κ = 0.47 [95% CI 0.35 to 0.60]); and fair agreement for rim width (κ = 0.25 [95% CI 0.18 to 0.33]) and centrality to the popliteal hiatus (κ = 0.36 [95% CI 0.21 to 0.55]). Interobserver reliability using the ICC was good for tear length (ICC 0.83 [95% CI 0.75 to 0.90]) and moderate for the percentage of meniscus excised (ICC 0.65 [95% CI 0.44 to 0.82]). The mean Spearman correlation coefficient for the percentage of meniscus excised was 0.92 (95% CI 0.89 to 0.94). Despite lower interobserver reliability, the ISAKOS committee recommended use of the three-zone radial location scheme (anterior, middle, and posterior) because it is more descriptive than the two-zone alternative. In their two-surgeon study, Shah et al. [21] reported interobserver reliability for lateral meniscal tear classification including tear depth (κ = 0.88 [95% CI 0.42 to 1.0]), Cooper Zone 1 tear location (κ = 0.76 [95% CI 0.29 to 1.0]), Cooper Zone 2 tear location (κ = 1.0 [95% CI 0.57 to 1.0]), Cooper Zone 3 tear location (κ = 1.0 [95% CI 0.57 to 1.0]), anterior-third tear location (κ = 0.88 [95% CI 0.42 to 1.0]), middle-third tear location (κ = 1.0 [95% CI 0.57 to 1.0]), posterior-third tear location (κ = 1.0 [95% CI 0.57 to 1.0]), centrality to the popliteal hiatus (κ = 1.0 [95% CI 0.57 to 1.0]), tear pattern (κ = 0.82 [95% CI 0.54 to 1.0]), tear quality (κ = 0.76 [95% CI 0.29 to 1.0]), and tear length (ICC = 0.96 [95% CI 0.91 to 0.99]). For medial meniscal tears, they reported interobserver reliability for tear depth (κ = 0.87 [95% CI 0.36 to 1.0]), Cooper Zone 1 tear location (κ = 0.87 [95% CI 0.36 to 1.0]), Cooper Zone 2 tear location (κ = 0.87 [95% CI 0.36 to 1.0]), Cooper Zone 3 tear location (κ = 0.87 [95% CI 0.36 to 1.0]), tear pattern (κ = 0.73 [95% CI 0.44 to 1.0]), tear quality (κ = 0.73 [95% CI 0.22 to 1.0]), and tear length (ICC 0.81 [95% CI 0.53 to 0.93]). In their two-surgeon study, Chhabra et al. [4] reported interobserver reliability for lateral meniscal tear classification including tear depth (κ = 0.52 [95% CI 0.09 to 0.96]), Cooper Zone 1 tear location (κ = 0.71 [95% CI 0.29 to 1.0]), Cooper Zone 2 tear location (κ = 0.71 [95% CI 0.29 to 1.0]), Cooper Zone 3 tear location (κ = 0.71 [95% CI 0.29 to 1.0]), anterior-third tear location (κ = 0.81 [95% CI 0.40 to 1.0]), middle-third tear location (κ = 0.81 [95% CI 0.40 to 1.0]), posterior-third tear location (κ = 0.90 [95% CI 0.51 to 1.0]), centrality to the popliteal hiatus (κ = 0.81 [95% CI 0.40 to 1.0]), tear pattern (κ = 0.85 [95% CI 0.61 to 1.0]), tear quality (κ = 0.90 [95% CI 0.51 to 1.0]), and tear length (ICC 0.90 [95% CI 0.77 to 0.96]). For medial meniscal tears, they reported interobserver reliability for tear depth (κ = 0.65 [95% CI 0.25 to 1.0]), Cooper Zone 1 tear location (κ = 0.83 [95% CI 0.44 to 1.0]), Cooper Zone 2 tear location (κ = 0.83 [95% CI 0.44 to 1.0]), Cooper Zone 3 tear location (κ = 0.48 [95% CI 0.06 to 0.89]), tear pattern (κ = 0.82 [95% CI 0.59 to 1.0]), tear quality (κ = 0.91 [95% CI 0.55 to 1.0]), and tear length (ICC 0.94 [95% CI 0.87 to 0.97]). Two studies have evaluated the intermethod reliability of the ISAKOS classification, correlating MRI and arthroscopy [4, 21]. Shah et al. [21] evaluated the intermethod reliability of this classification system using 1.5 T and 3.0 T MRI. Two fellowship-trained musculoskeletal radiologists and two arthroscopists evaluated MRI studies and arthroscopic images, respectively, of 81 knees in 69 patients. A board-certified, fellowship-trained surgeon performed all arthroscopies and captured arthroscopic images with a probe in the field of view. For lateral meniscal tear assessments on 1.5 T MRI, they described the intermethod reliability of tear depth (κ = 0.71 [95% CI 0.23 to 1.0]), Cooper Zone 1 tear location (κ = 0.53 [95% CI 0.05 to 1.0]), Cooper Zone 2 tear location (κ = 0.35 [95% CI -0.12 to 0.83]), Cooper Zone 3 tear location (κ = 0.94 [95% CI 0.49 to 1.0]), anterior-third tear location (κ = 0.76 [95% CI 0.29 to 1.0]), middle-third tear location (κ = 0.83 [95% CI 0.35 to 1.0]), posterior-third tear location (κ = 1.0 [95% CI 0.57 to 1.0]), centrality to the popliteal hiatus (κ = 0.76 [95% CI 0.29 to 1.0]), tear pattern (κ = 0.91 [95% CI 0.65 to 1.0]), tear quality (κ = 0.35 [95% CI -0.12 to 0.82]), and tear length (ICC 0.76 [95% CI 0.33 to 0.93]). For medial meniscal tear assessments on 1.5 T MRI, they described the intermethod reliability of tear depth (κ = 0.60 [95% CI 0.08 to 1.0]), Cooper Zone 1 tear location (κ = 0.40 [95% CI -0.10 to 0.90]), Cooper Zone 2 tear location (κ = 0.27 [95% CI -0.23 to 0.76]), Cooper Zone 3 tear location (κ = 0.67 [95% CI 0.15 to 1.0]), tear pattern (κ = 0.73 [95% CI 0.44 to 1.0]), tear quality (κ = 0.60 [95% CI 0.08 to 1.0]), and tear length (ICC 0.51 [95% CI -0.05 to 0.83]). Use of 3.0 T MRI was associated with more reliable lateral meniscal zone identification and correlation of tear lengths. However, when compared with arthroscopy, this modality overestimated the lengths of medial meniscal tears (9.74 mm [95% CI 4.22 to 15.25]; p = 0.001) but not lateral meniscal tears (4.04 mm [95% CI -3.05 to 11.12]; p = 0.26). Similarly, 1.5 T MRI overestimated medial meniscal tear lengths when compared with arthroscopy (11.01 mm [95% CI 3.64 to 18.39]; p = 0.004). Using isotropic three-dimensional user-defined reconstructions, which are more commonly available with 3.0 T than with 1.5 T MRI scanners [21], Chhabra et al. [4] measured intermethod agreement for individual components of the ISAKOS classification system. A fellowship-trained surgeon performed all arthroscopies of 39 patients with 44 meniscal tears and captured arthroscopic images with a probe in the field of view. A second fellowship-trained surgeon reviewed and assessed tear characteristics 4 months afterward. Two fellowship-trained musculoskeletal radiologists independently evaluated MRI studies. For lateral meniscal tears, they described the intermethod reliability for tear depth (κ = 0.54 [95% CI 0.18 to 0.90]), Cooper Zone 1 tear location (κ = 0.53 [95% CI = 0.19 to 0.87]), Cooper Zone 2 tear location (κ = 0.64 [95% CI 0.31 to 0.97]), Cooper Zone 3 tear location (κ = 0.61 [95% CI 0.29 to 0.93]), anterior-third tear location (κ = 0.58 [95% CI 0.17 to 0.98]), middle-third tear location (κ = 0.43 [95% CI 0.01 to 0.85]), posterior-third tear location (κ = 0.91 [95% CI 0.69 to 1.0]), centrality to the popliteal hiatus (κ = 0.49 [95% CI 0.11 to 0.87]), tear pattern (κ = 0.91 [95% CI 0.78 to 1.0]), tissue quality (κ = 0.33 [95% CI -0.04 to 0.7]), and tear length (ICC 0.74 [95% CI 0.45 to 0.89]). For medial meniscal tears, they described the intermethod reliability for tear depth (κ = 0.29 [95% CI -0.21 to 0.79]), Cooper Zone 1 tear location (κ = 0.39 [95% CI 0.06 to 0.71]), Cooper Zone 2 tear location (κ = 0.38 [95% CI -0.10 to 0.87]), Cooper Zone 3 tear location (κ = 0.61 [95% CI 0.30 to 0.91]), anterior-third tear location (κ = 0.56 [95% CI 0.19 to 0.94]), middle-third tear location (κ = 0.72 [95% CI 0.24 to 1.0]), tear pattern (κ = 0.54 [95% CI 0.31 to 0.78]), tissue quality (κ = 0.78 [95% CI 0.52 to 1.0]), and tear length (ICC 0.75 [95% CI 0.03 to 0.85]). Of note, tear lengths were larger on MRI than on arthroscopy (mean difference 9.74 mm for meniscal tears and 4.04 mm for lateral meniscal tears; p = 0.034). Limitations The ISAKOS classification system was formally devised as an arthroscopic assessment (Fig. 2). Studies utilizing this classification system that use either arthroscopic videos or preoperative MRI for research purposes are therefore subject to inherent biases because they lack the advantages of tactile feedback and multiple arthroscopic viewing approaches available to the surgeon. Lastly, the procedures included in these validation studies may represent a best-case scenario, without the time constraints of the operating room, and may not represent real-world procedures in which visualization is sometimes poor or probing of the lesion is not feasible.Fig. 2: A-D These case examples demonstrate application of the ISAKOS classification system: (A) Sagittal T2-weighted MRI sequence and (B) arthroscopic image demonstrating a complete medial meniscal tear involving Zone 3 of the posterior and middle thirds, with vertical flap tear pattern, with degenerative tissue quality, measuring approximately 1 cm, for which approximately 20% of the meniscus was excised. (C) Sagittal T2-weighted MRI sequence and (D) arthroscopic image demonstrating a complete lateral meniscal tear involving Zone 1 of the middle third, not central to the popliteal hiatus, with radial tear pattern, with nondegenerative tissue quality, measuring approximately 1.5 cm, for which approximately 20% of the meniscus was excised (following debridement of the central portion and repair of the residual tear). Reproduced from the Journal of ISAKOS: Joint Disorders & Orthopaedic Sports Medicine, Shah J, Hlis R, Ashikyan O, et al., volume 5, pages 201-207, 2020, with permission from BMJ Publishing Group Ltd.It should be noted that this classification system was not designed to correlate MRI and arthroscopy, nor to be an independent MRI classification scheme. Nonetheless, correlation of MRI to arthroscopy may be obfuscated by lag time to arthroscope because tears may either partially heal or progress and create a discrepancy in the anticipated tear size. MRI may overestimate tear length (perhaps by amplifying intrasubstance tears that might be easily missed), may not be conducive to probing, or may be partially healed or embedded with fibrous tissue at the time of arthroscopy [4]. Conversely, arthroscopy may underestimate the tear length relative to MRI because of incomplete visualization of certain tears while also introducing measurement error related to impreciseness of the surgical probe [21]. Applying this classification system to preoperative MRI may not be useful for predicting tear reparability [21]. The findings reviewed here are consistent with those of prior studies noting the more challenging diagnosis of lateral meniscal tears [25] as well as lower reliability between surgeons in measuring rim width [7]. Certain components of the ISAKOS classification symptom, including rim width (or Cooper zone classification) and centrality to the popliteal hiatus, demonstrated lower reliability than others, which may be a function of inherent variation in root attachment morphology and other anatomic factors across patients [4]. In addition, the large width of the 95% confidence intervals reported for many of the classification components implies inconsistent reliability and/or methodological limitations (such as insufficient sample size) in the published evidence to date. Conclusion The ISAKOS classification of meniscal tears is a validated arthroscopic assessment of tear depth, rim width, radial location, centrality to the popliteal hiatus, tear pattern, tissue quality, tear length, and percentage of meniscus excised. This classification system was devised to improve the consistency of arthroscopic classification and documentation of meniscal tears, thereby improving preoperative planning, standardizing outcome reporting, and enabling an assessment of treatment options across studies. Subsequent studies have demonstrated that this classification system is reproducible using 1.5 T and 3.0 T MRI, and these results are in sufficient agreement with arthroscopic findings, which suggests a role for bridging and standardizing communication between multiple disciplines tasked with the care of meniscal tears. The ISAKOS classification system has not yet gained widespread acceptance among surgeons and trainees, which may be due to lower familiarity compared with the system of Cooper et al. [5], greater inherent complexity, and a perception that it has broader applicability for research purposes than clinical decision-making. The ultimate goal of this classification system is to facilitate multicenter data pooling and longitudinal outcome tracking of patients with meniscal tears so that the optimal treatment approach to each tear type can be better understood. Broader adoption of this classification system may improve interdisciplinary communication as well as shared decision-making among surgeons and their patients along the continuum of care, while improving the quality and level of evidence of the literature on meniscal tears.