1 Introduction

The acetabular labrum is a fibrocartilaginous structure which forms a rim surrounding the acetabulum. The labrum serves to deepen the socket similar to the function of the glenoid labrum in the shoulder. It may enhance stability of the joint by partially sealing the joint and providing negative intra-articular pressures. Additionally, this seal has been shown to redistribute applied force to the hip more evenly across the acetabular cartilage surface.¹

Magnetic resonance imaging (MRI) remains a useful tool for evaluating lesions and soft tissue structures including the labrum and articular cartilage in patients with femoroacetabular impingement (FAI).^2,3 The ability to accurately couple imaging with clinical findings is critical for preoperative planning. Accurate measurement of the size of the acetabular labrum is valuable in the diagnosis of hip dysplasia. Studies have shown labral thickening in patients with acetabular dysplasia as compensation to protect their under covered hip joint.^4–7 While MRI measurements are often interpreted as factual representations of the underlying anatomy, there is limited evidence to support its accuracy in predicting true labral thickness in hips with femoroacetabular impingement. The goal of this study is to determine whether there is a true correlation of labral thickness between imaging via MRI arthrogram and arthroscopic visualization in patients undergoing surgery for femoroacetabular impingement.

2 Methods

Patients undergoing hip arthroscopy for symptomatic FAI and labral tears between September 2017 to January 2019 by two fellowship-trained hip arthroscopists were prospectively enrolled in the study. Exclusion criteria included revision surgery and lack of essential preoperative imaging (radiographs and MR arthrogram). Baseline patient demographic information was collected from the electronic medical record and outlined in Table 1. Pre-operative radiographs were reviewed by three independent reviewers to assess for radiographic parameters, including the lateral center edge angle (LCEA), anterior center edge angle (ACEA), cross-over sign, alpha angle, and Tonnis grade. This study was approved by an institutional review board.

Pre-operative MR arthrograms were reviewed by three independent reviewers (2 orthopaedic surgeons and 1 radiologist). Hip labral widths were measured at three different zones within the acetabulum that correlate with the anterior-superior portion (zone 2), mid-superior portion (zone 3), and posterior-superior portion (zone 4) as described by Ilizaliturri et al.⁸ Cross-referenced orthogonal imaging was used to accurately identify the labrum within zones 2, 3, and 4. MRA images within the axial plane were used to assess labral widths for zone 2 and zone 4, while images within the coronal plane were used to assess labral widths for zone 3. This is depicted in Fig. 1. Intra-operative hip labral thickness was measured by one of two fellowship-trained orthopaedic surgeons who perform a high volume of hip arthroscopy. These measurements were made at the same three anatomical zones of the acetabulum as previously described, using an arthroscopic probe with 5 mm etchings to assess size (Fig. 2).

Close

Fig. 1 Examples of MRA images used to assess labral size. Images within the coronal plane were used to assess labral widths for zone 3 (A) and images within the axial plane were used to assess labral widths for zone 2 (B) and zone 4 (C).

Export to PPT

Close

Fig. 2 Examples of intra-operative images used to assess labral size at zones 3 (A), 2 (B), and 4 (C).

Export to PPT

Hip labral sizes as measured on MRA were then averaged between the three reviewers. A Kappa co-efficient between reviewers was then calculated to report inter-observer reliability of labrum measurements. The association between MRI and intra-operative hip labrum measurements was then compared using two separate methods. The first method compared labral thickness using a student paired t-test. The second method used a Pearson correlation to determine the correlation coefficient between the MRA and intra-operative measurements.

Secondarily, patients were grouped according to those with and without hip dysplasia. Hip dysplasia was defined as having either an LCEA <25° or ACEA <20°. Labral sizes, using either MRA or arthroscopically based measurements, were then compared between those with and without dysplasia using a student's t-test.

3 Results

117 patients were enrolled in the study. 70% of the patients were female with 61% undergoing surgery for their right hip. This cohort had an average age 39.1 ± 13.3 years old and an average body mass index (BMI) of 26.5 ± 5.4. All patients underwent an osteoplasty with 92% undergoing labral repair vs 8% undergoing labral debridement.

There was a strong inter-rater reliability in MRA-based hip labrum measurements between the three reviewers, with correlation coefficients >0.98 (p < 0.0001) in all zones of the acetabulum (Table 2). The average labral sizes based on intraoperative measurements were 6.85 mm in zone 2, 7.45 mm in zone 3, and 7.29 mm in zone 4. The average labral sizes based on MRA were 6.95 mm in zone 2, 7.24 mm in zone 3, and 6.71 mm in zone 4. There was a poor correlation however between MRA-based measurements and intraoperative measurements in zones 2 and 3 (zone 2: R = 0.171, p = 0.065; zone 3: R = 0.335, p = 0.00022) and no significant correlation in zone 4 (R = −0.22, p = 0.82) (Table 3).

Eight patients within this cohort were found to have hip dysplasia (Table 4). Labrum size was not significantly different between patients with or without dysplasia, regardless of whether the labrum was measured intra-operatively or using MRI.

4 Discussion

This is the first study to investigate the accuracy of MRA to measure the thickness of the acetabular labrum. Other studies have compared MRI to arthroscopic measurements of tendon thickness in the knee. Bickel et al. showed that cross sectional area of hamstring tendons on MRI corresponded with sufficient autograft size for ACL reconstruction at the time of surgery.⁹ Another study showed that MRI cross sectional area can identify patients with inadequate intraoperative graft size for ACL reconstruction preoperatively with 79% sensitivity and 74% specificity.¹⁰ However, cartilage lesions in the knee have been shown to be 65% smaller on MRI than the defect area after arthroscopic debridement.¹¹ This is of significant value for orthopaedic surgeons as their preoperative plans often rely on lesion size as suggested by MRI.

The literature is inconclusive regarding the accuracy of MRI with respect to osteochondral lesions of the talus (OLTs). Bae et al. found a good correlation between MRI grading of OLTs and arthroscopic classification.¹² However, Yasui et al. showed that MRI overestimated area and diameter of OLTs compared with arthroscopy.¹³

Additionally, MRI has been shown to be excellent in diagnosing ACL and meniscal tears. Kulkarni et al. reported MRI to be 88% accurate in diagnosing ACL tears and 85% for meniscal tears.¹⁴ A systematic review comprised of 59 articles comparing MRI to the gold standard arthroscopy showed that the MRI-based diagnosis of medial and lateral meniscus tears and ACL lesions to be 85% accurate.¹⁵

Our study found no increase in labral thickness in patients with acetabular dysplasia. This is contrary to current literature, which suggests the labrum hypertrophies to provide added stability for a dysplastic hip. Arthroscopy has shown labral size to be significantly larger in dysplastic hips (LCEA ≤ 25°).⁴ MRI has also shown increased labral size in patients with dysplasia.⁵ Petersen et al. reported that lateral labral length increases as LCEA decreases and Kubo et al. echoed these findings.^5,7 This study showed no increase in labral thickness on either MRI or intraoperative measurements. This is likely due to a sample size of patients with acetabular dysplasia too small (n = 8) to compare with our cohort of patients without hip dysplasia.

This is the first study to compare the accuracy of MRA measurements of acetabular labral thickness with intraoperative measurements. However, it is not without its limitations. The methods used to determine labral widths using MRA imaging is not a validated method, despite being previously described and published.³ Cross-sectional images were used to confirm proper location of measurements, however this relies on perfect cross-sectional images of each labrum and some of the imaging may have been off axis, particularly when performed at an outside institution. Intra-operative measurements were also made by each surgeon without overlap, which limits the reliability of these measurements. Finally, there were a limited number of patients with acetabular dysplasia, due in part by the nature of surgical indications for hip arthroscopy, which limits the power to compare labral size in patients with and without dysplasia as described in our secondary outcomes.

5 Conclusion

This study demonstrates a poor correlation in labral measurements between magnetic resonance arthrogram and intraoperative measurements. This suggests that MRA may not provide sufficient detail to accurately predict labral size.

Authors’ contribution list

Matthew J. Hartwell: Acquisition, analysis, and interpretation of data, critical revisions/addition of intellectual content to manuscript, final approval of version to be published, and agrees to be accountable for all aspects of the work.

Ryan S. Selley: Acquisition and analysis of data, critical revisions/addition of intellectual content to manuscript, final approval of version to be published, and agrees to be accountable for all aspects of the work.

Steven R. Dayton: Interpretation of data, drafting of the manuscript, final approval of version to be published, and agrees to be accountable for all aspects of the work.

Sarah H. Ishamuddin: Interpretation of data, drafting of the manuscript, final approval of version to be published, and agrees to be accountable for all aspects of the work.

Ksheeraja Ravi: Acquisition of data, critical revisions/addition of intellectual content to manuscript, final approval of version to be published, and agrees to be accountable for all aspects of the work.

Michael A. Terry: Acquisition of data, critical revisions/addition of intellectual content to manuscript, final approval of version to be published, and agrees to be accountable for all aspects of the work.

Vehniah K. Tjong: Acquisition of data, critical revisions/addition of intellectual content to manuscript, final approval of version to be published, and agrees to be accountable for all aspects of the work.