Knee Surgery, Sports Traumatology, Arthroscopy

, Volume 23, Issue 4, pp 1128–1134

Do the radiographic findings of symptomatic discoid lateral meniscus in children differ from normal control subjects?

Authors

  • Sang Hee Choi
    • Department of Radiology, Samsung Medical CenterSungkyunkwan University School of Medicine
  • Jin Hwan Ahn
    • Department of Orthopaedic Surgery, Kangbuk Samsung HospitalSungkyunkwan University School of Medicine
  • Kang Il Kim
    • Department of Orthopaedic Surgery, Center for Joint Diseases and RheumatismKyung Hee University Hospital at Gangdong
  • Suk-Kyeong Ji
    • Department of Radiology, Samsung Medical CenterSungkyunkwan University School of Medicine
  • Su Min Kang
    • Department of Radiology, Samsung Medical CenterSungkyunkwan University School of Medicine
  • Jung Suk Kim
    • Department of Orthopaedic Surgery, Center for Joint Diseases and RheumatismKyung Hee University Hospital at Gangdong
    • Department of Orthopaedic Surgery, Center for Joint Diseases and RheumatismKyung Hee University Hospital at Gangdong
Knee

DOI: 10.1007/s00167-014-2924-6

Cite this article as:
Choi, S.H., Ahn, J.H., Kim, K.I. et al. Knee Surg Sports Traumatol Arthrosc (2015) 23: 1128. doi:10.1007/s00167-014-2924-6

Abstract

Purpose

The aim of this study was to quantitatively compare radiographic findings of symptomatic discoid lateral meniscus in children with those of matched controls.

Methods

Seventy-eight consecutive children (91 knees) who underwent arthroscopic surgery for a symptomatic discoid lateral meniscus (discoid group) were included. Another 91 age- and sex-matched controls with normal medial and lateral menisci on the basis of magnetic resonance imaging findings were included in this study (control group). Each plain radiograph was evaluated from the anteroposterior view for the following variables: height of the lateral tibial spine, lateral joint space distance, height of the fibular head, squaring of the lateral femoral condyle, obliquity of the lateral tibial plateau and cupping of the lateral tibial plateau. Lateral femoral condylar notch was evaluated in lateral view. Statistical analyses were used to determine the differences between the two groups.

Results

A significant difference in the mean height of the lateral tibial spine, lateral joint space distance, height of the fibular head, and obliquity of the lateral tibial plateau distinguished the two groups (p < 0.0001). However, there was no statistical difference in the condylar off sign, squaring of the lateral femoral condyle, cupping of the lateral tibial plateau and lateral femoral condylar notch between groups (n.s.). The cut-off values for the height of the lateral tibial spine (6 mm), lateral joint space distance (8 mm), height of the fibular head (14.9 mm) and obliquity of the lateral tibial plateau (17.6°) were determined. With these cut-off values in diagnosing discoid lateral meniscus, the sensitivity and accuracy of height of the fibular head were 78 and 70 %, respectively.

Conclusions

Several plain radiographic findings in symptomatic discoid lateral meniscus in children were significantly different from those in normal control. These findings would be helpful in screening tool of discoid lateral meniscus for children.

Level of evidence

II.

Keywords

KneeDiscoid lateral meniscusPlain radiographChildren

Introduction

The discoid lateral meniscus is an anatomic variant in which the meniscus is thickened and discoid shaped, and the discoid lateral meniscus covers a greater area of the tibial plateau than the normal meniscus. The discoid lateral meniscus has a reported prevalence that ranges from 0.4 to 17 % [5, 25, 26], occurring at a higher rate in Asian populations [1, 4, 9, 14, 20]. Compared with the normal lateral meniscus, the discoid lateral meniscus has a higher frequency of meniscal tears and associated symptoms [4, 10, 26, 33]. Although asymptomatic discoid lateral meniscus does not require treatment, symptomatic discoid lateral meniscus cases require operative treatment when they have an associated tear resulting in pain, clicking or a limited range of motion [3, 5, 12, 14, 29, 31].

Characteristic findings of discoid lateral meniscus on plain radiographs are often subtle, but include a widened lateral joint line, squaring of the lateral femoral condyle, cupping of the lateral tibial plateau, mild hypoplasia of the tibial spine and an elevated fibular head [19, 21, 28]. However, these findings have been traditionally observed in adults, with little information on discoid lateral meniscus in children [16, 23, 25]. Several authors reported that plain radiographs were not helpful, and magnetic resonance imaging (MRI) is usually conclusive in the diagnosis of the discoid lateral meniscus, especially for children [6, 11, 23, 25, 27, 30]. However, we believe that plain radiographs also play an important role in the preoperative diagnosis of the discoid lateral meniscus for children. Therefore, the aim of this study was to quantitatively analyse and compare plain radiographic findings of discoid lateral meniscus in children with matched controls. The hypothesis was that the plain radiographic findings of patients with a symptomatic discoid lateral meniscus would be significantly different from those of normal controls.

Materials and methods

From 2000 to 2008, 78 consecutive children (91 knees) who underwent arthroscopic surgery for symptomatic discoid lateral meniscus (discoid group) were enrolled in the study. Control subjects were selected by a one-to-one matching method with matching of age and sex without replacement (control group). The control patients were demonstrated normal medial and lateral menisci on magnetic resonance imaging (MRI) findings, regardless of other intraarticular pathologies. The discoid group consisted of 49 boys and 42 girls, with a median age of 10 years (range 4–15 years). The control group consisted of 49 boys and 42 girls, with a median age of 10 years (range 3–15 years). All of the patients in discoid group were confirmed to have a discoid lateral meniscus on the surgical side based on MRI before operations. Using Watanabe’s classification [32], 84 discoid lateral menisci were classified as complete-type discoid lateral meniscus and seven were classified as incomplete-type discoid lateral meniscus based on the arthroscopic findings.

Radiographic analysis

All patients had non-weight-bearing plain radiographs taken from anteroposterior, lateral, tunnel, and merchant views in single institution, with a tube-to-film distance of 110 cm. The tunnel view was checked with prone position in the knee flexed approximately 40° with the foot supported by a foam block. Each plain radiograph was evaluated from the anteroposterior view for the following variables: height of the lateral tibial spine, lateral joint space distance, height of the fibular head, squaring of the lateral femoral condyle, cupping of the lateral tibial plateau, obliquity of the lateral tibial plateau and lateral femoral condylar notch (Figs. 1, 2). Height of the lateral tibial spine, lateral joint space distance, height of the fibular head and obliquity of the lateral tibial plateau were measured as continuous variables. Squaring of the lateral femoral condyle was regarded as positive when a straight line of the articular surface of the lateral femoral condyle was over 10 mm. Cupping of the lateral tibial plateau was regarded as positive if the depression of the tibial articular surface was over 1 mm from the imaginary mediolateral joint line. Lateral femoral condylar notch was evaluated a lateral view and regarded as positive if the distance from the tangential line of the articular surface of the lateral condyle to the notch was over 1 mm. These characteristic findings were evaluated according to the analysis of Kim et al. [22]. The condylar off sign was also analysed in tunnel view radiographs using the method reported by Ha et al. [15] who developed a method to measure the prominence of the femoral condyle adjacent to the intercondylar notch of the knee. All the measurements were performed on a PACS monitor (Picture Archiving and Communications System; General Electric, Chicago, IL, USA) using a mouse-point cursor and an automated computer calculation for the distance and angle.
https://static-content.springer.com/image/art%3A10.1007%2Fs00167-014-2924-6/MediaObjects/167_2014_2924_Fig1_HTML.jpg
Fig. 1

a Anteroposterior view shows A height of the fibular head, from the imaginary tibial joint line to tip of fibular head, B lateral joint space distance, from the imaginary tibial joint line to lateral femoral condylar joint line at its midportion, and C height of the lateral tibial spine, distance from the imaginary tibial joint line to tip of the lateral intercondylar spine. b Anteroposterior view shows D obliquity of the lateral tibial plateau formed by imaginary tibial joint line and articular line of the lateral tibial plateau

https://static-content.springer.com/image/art%3A10.1007%2Fs00167-014-2924-6/MediaObjects/167_2014_2924_Fig2_HTML.jpg
Fig. 2

a Anteroposterior view shows E squaring of the lateral femoral condyle, distance of straight articular surface of condyle and F cupping of the lateral tibial plateau, from the imaginary tibial joint line to proximal limit of the lateral tibial plateau, positive if measured above 1 mm. b Lateral view shows lateral femoral condylar notch, from the tangential line, which meets smooth contour of articular surface to notch, positive if above 1 mm

All radiographic studies were reviewed by two experienced musculoskeletal radiologists who were unaware of the arthroscopic findings, clinical history and initial radiographic interpretations. One reviewer was a musculoskeletal radiologist with 15 years of experience, and the other was a musculoskeletal radiologist with experience in a three-year fellowship. Two radiologists who worked in a blinded manner retrospectively reviewed all the images of both groups (session 1). After 3 weeks, all radiographic images were again reviewed by both radiologists (session 2). Our institutional review board approved the study protocol (No. 2012-04-063) and granted exemption for patient consent.

Statistical analysis

The stratified analyses were performed for matched-pair design. All statistical analysis was performed in matched design and considered with strata. A McNemar χ2 test was used to compare the categorical variables between the two groups including for squaring of lateral femoral condyle, cupping of lateral tibial plateau and lateral femoral condylar notch. A paired t test, with or without the Bonferroni correction, was used to compare the height of lateral tibial spine, lateral joint space, fibular head height, obliquity of lateral tibial plateau and condylar off sign. In this study, the cut-off point is defined as the value that discriminates between the two groups. The sensitivity, specificity and positive and negative predictive and accuracy values according to different cut-off points were determined. The cut-off values for continuous variables corresponded to the highest sum of sensitivity and specificity at each level [Youden’s index: maximum (sensitivity + specificity − 1)]. All the analyses were performed using the statistical software package for social sciences (R2 15.0, R Core Team, Vienna, Austria and SAS 9.3, SAS Institute Inc., Cary, NC, USA). p Values <0.05 were considered significant. The reliability of measurements was assessed by examining the interobserver and intraobserver intraclass correlation coefficient (ICC), which quantifies the proportion of the variance of the rating due to variability between measurements [13]. The ICC was interpreted as poor if it was <0.4; as marginal when it was ≥0.4 but <0.75, and as good when it was >0.75.

Results

Table 1 shows the mean, standard deviation, range of each parameter in both groups, as measured in session 1. Significant differences in the height of the lateral tibial spine, lateral joint space distance, height of the fibular head and obliquity of the lateral tibial plateau distinguished the two groups (p < 0.0001, Fig. 3). However, there was no statistical difference in the condylar off sign, squaring of the lateral femoral condyle, cupping of the lateral tibial plateau and lateral femoral condylar notch between groups (n.s.).
Table 1

The mean, range and standard deviation (SD) values of each parameter in the discoid lateral meniscus (discoid) and control groups as measured in session 1

Parameter

Discoid group

(n = 91)

Mean ± SD

(95 % CI)

Control group

(n = 91)

Mean ± SD

(95 % CI)

p value

Height of the lateral tibial spine (mm)

6.6 ± 1.9

(6.3–7.0)

5.9 ± 1.9

(5.5–6.2)

<0.0001

Lateral joint space distance (mm)

8.42.2

(7.9–8.7)

7.2 ± 2.3

(6.8–7.7)

<0.0001

Height of the fibular head (mm)

12.2 ± 3.6

(11.6–12.8)

15.6 ± 3.7

(15.0–16.3)

<0.0001

Obliquity of the lateral tibial plateau (°)

17.7 ± 4.2

(17.1–18.5)

15.8 ± 4.8

(14.8–16.8)

<0.0001

Condylar cut-off sign (ratio)

0.84 ± 0.17

(0.82–0.87)

0.86 ± 0.14

(0.83–0.89)

n.s.

Squaring of the lateral femoral condylea

29 (32 %)

24 (26 %)

n.s.

Cupping of the lateral tibial plateaua

31 (34 %)

39 (43 %)

n.s.

Lateral femoral condylar notcha

2 (2 %)

2 (2 %)

n.s.

aPositive cases (%) of squaring of lateral femoral condyle, cupping of the lateral tibial plateau and lateral femoral condylar notch

https://static-content.springer.com/image/art%3A10.1007%2Fs00167-014-2924-6/MediaObjects/167_2014_2924_Fig3_HTML.jpg
Fig. 3

a Anteroposterior view shows widened lateral joint space (8.3 mm) and elevated fibular head (12.1 mm). b Coronal MR image shows discoid lateral meniscus of complete type (arrow). c Anteroposterior view shows normal lateral joint space (5.7 mm) and height of the fibular head (18.3 mm). d Coronal MR image shows lateral meniscus of normal shape

The cut-off values of the height of the lateral tibial spine, lateral joint space distance, height of the fibular head and obliquity of the lateral tibial plateau were 6, 8, 14.9 mm and 17.6°, respectively. The sensitivity, specificity, positive predictive value, negative predictive value and accuracy according to the cut-off values in diagnosing discoid lateral meniscus were shown in Table 2. With these cut-off values in diagnosing discoid lateral meniscus, the sensitivity and accuracy of the fibular head height were as high as 78 and 70 %, respectively.
Table 2

The sensitivity, specificity, positive predictive value, negative predictive value and accuracy of each parameter in diagnosing discoid lateral meniscus as measured in session 1

Parameter

Cut-off value

Sensitivity (%)

Specificity (%)

Positive predictive value (%)

Negative predictive value (%)

Accuracy (%)

Height of the lateral tibial spine

6 mm

61.5

52.2

56.3

57.6

56.9

Lateral joint space distance

8 mm

59.3

67.0

64.3

62.2

63.2

Height of the fibular head

14.9 mm

78.0

60.9

66.7

73.5

69.5

Obliquity of the lateral tibial plateau

17.6°

51.1

72.5

65.0

59.7

61.8

Table 3 shows the interobserver ICC for each parameter in both groups for the second session. The interobserver ICC is 0.594–0.834 indicating a high agreement between evaluators, except when measuring obliquity of lateral tibial plateau. The intraobserver ICC for each parameter is 0.863–0.975, indicating a significantly high agreement for among both groups (Table 3).
Table 3

Intraclass correlation coefficient and 95 % CIs for the intraobserver and interobserver reliabilities of each parameter as measured in session 2

Parameter

Interobserver ICC

Intraobserver ICC

Height of the lateral tibial spine (mm)

0.796 (0.737–0.844)

0.912 (0.883–0.938)

Lateral joint space distance (mm)

0.801 (0.743–0.848)

0.924 (0.898–0.947)

Height of the fibular head (mm)

0.826 (0.774–0.867)

0.975 (0.961–0.987)

Obliquity of the lateral tibial plateau (°)

0.594 (0.492–0.680)

0.863 (0.842–0.885)

Condylar cut-off sign (ratio)

0.834 (0.802–0.873)

0.933 (0.914–0.967)

Discussion

The most important findings of the present study were as follows: (1) a significant difference in the mean height of the lateral tibial spine, lateral joint space distance, height of the fibular head and obliquity of the lateral tibial plateau distinguished the discoid lateral meniscus patients from normal controls. (2) Among the cut-off values in diagnosing discoid lateral meniscus, the sensitivity and accuracy of height of the fibular head were as high as 78 and 70 %, respectively. Few reports have described the comparative study of conventional radiograph in between adult discoid lateral meniscus patients and normal controls [15, 22], and to the best of our knowledge, no report has described quantitative radiographic findings of discoid lateral meniscus in children.

Several studies have focused on the characteristic findings of the discoid lateral meniscus with plain radiographs [1, 7, 17, 18, 24]. Nathan and Cole [24] concluded that plain radiographs had little value in the detection of discoid lateral meniscus in adults. Ogut et al. [25] also reported that plain radiographs were not helpful in the diagnosis of the discoid lateral meniscus in children because none of the radiographs (11 cases) showed widening of the lateral joint space. However, Bellier et al. [8] reported that 36.8 % of children with discoid lateral meniscus had radiographic findings such as a widened lateral joint space, increased obliquity of the lateral tibial plateau and a blunted lateral tibial eminence. In recent years, Ahn et al. [2] studied the contralateral knee of 33 patients who underwent an operation for a symptomatic discoid lateral meniscus. Twenty-three (70 %) of 33 contralateral knees showed more than one characteristic radiographic finding of discoid lateral meniscus. Furthermore, statistical analysis revealed good agreement between the radiographic findings, the tear pattern of the lateral menisci and the associated chondral lesions identified on MRI scans. They recommended evaluating radiographs from both knees, even if a discoid lateral meniscus is only unilaterally diagnosed in the clinic. They also commented that MRI should be considered when the characteristic findings of discoid lateral meniscus are seen on the plain radiographs.

Kim et al. [22] performed a retrospective review of plain radiographs of arthroscopically confirmed discoid lateral meniscus, with comparison to the normal controls. They analysed characteristic features of 68 discoid lateral meniscus and 70 normal knees with various indirect signs on plain radiographs, such as a squared-off appearance of the lateral femoral condyle, widened femorotibial joint space, cupping of the lateral tibial plateau, obliquity of the lateral tibial plateau articular surface, high fibular head and hypoplasia of the lateral intercondylar spine. Widened lateral joint space and high fibular head showed statistically significant differences between the two groups. However, they only included adult patients, and a relatively low sensitivity and specificity were found. In this series, only children with symptomatic discoid lateral meniscus were studied and compared to age- and sex-matched controls with normal medial and lateral menisci. Our result showed a significant difference in the mean height of the lateral tibial spine, lateral joint space distance, height of the fibular head and obliquity of the lateral tibial plateau between the two groups. However, there was no statistical difference in other signs. This is the first report to compare the radiographic features of discoid lateral meniscus and normal child knees with various radiographic findings.

Ha et al. [15] described the condylar cut-off sign, a radiographic sign in knees with a discoid lateral meniscus, for diagnosis of discoid lateral meniscus. They retrospectively reviewed tunnel view radiographs of 100 knees, including 50 with discoid lateral meniscus and 50 with normal menisci in adults. Using a cut-off ratio, this sign showed a major difference between the two groups and found the ratio had 76 % sensitivity, 96 % specificity, 95 % positive predictive value and 80 % negative predictive value. They concluded that the condylar cut-off sign, as seen on tunnel view radiographs, is useful to help diagnose a complete discoid lateral meniscus in skeletally mature patients. However, our study found that there was no statistical difference between the discoid lateral meniscus and normal group in skeletally immature knees. This conflicting result might be related to the difference of the prominence of the femoral condyle between mature and immature knees. We believe that this sign is only helpful for diagnosing discoid lateral meniscus in skeletally mature knees.

The strength of the study is the relatively large number of subjects, in spite of the symptomatic discoid lateral meniscus being an uncommon condition in children. Another strength is that the control group consisted of age- and sex-matched controls with normal medial and lateral menisci on the basis of MRI findings to reduce the selection bias.

Nevertheless, there are several limitations. First, this study did not distinguish between complete- and incomplete-type of discoid lateral meniscus, although only seven incomplete discoid lateral menisci were enrolled. However, we think that the type of discoid lateral meniscus may add a small bias to the results, as all cases included in the arthroscopic diagnosis were with symptomatic discoid lateral meniscus for 16 years or less. In addition, tear patterns of discoid lateral meniscus in MRI or arthroscopic findings were not evaluated. However, this study was only focused on findings of plain radiographs. The group of symptomatic discoid lateral meniscus cases was compared with age-.and sex-matched group. Second, interobserver ICC for the obliquity of lateral tibial plateau was marginal (0.594). This could be related to the fact that it is difficult to make the line in immature tibia because tibia plateau in skeletally immature knee is rounder than that in mature knee. However, interobserver ICC for other parameters and intraobserver ICC for each parameter were high. Finally, this study did not calculate the normalized knee dimensions that were the absolute values of each knee dimension divided by each individual’s interepicondylar distance of the distal femur. However, Kim et al. [22] showed that there were no differences in the results of absolute dimensions and normalized dimensions. Furthermore, both groups were enrolled age- and sex-matched.

The hypothesis of this study was that the plain radiographs would be used with discoid lateral meniscus in children. We recommend that MRI scans be considered in child patients with knee joint pain in case of characteristic radiographic findings of discoid lateral meniscus, especially elevated fibular head (<15 mm) or widened lateral joint space (>8 mm). These radiographic findings could offer a cheap and efficacious pre-screening tool for discoid lateral meniscus prior to utilizing MRI.

Conclusion

Several plain radiographic findings in symptomatic discoid lateral meniscus children were significantly different from those in the normal control. These findings would be helpful in screening tool of discoid lateral meniscus for children.

Conflict of interest

The authors report no conflict of interest/financial disclosure.

Copyright information

© Springer-Verlag Berlin Heidelberg 2014