1 Introduction

Femoroacetabular impingement (FAI) syndrome occurs in patients with abnormal morphology of the hip joint, such as cam and/or pincer lesions, resulting in premature and abnormal bony contact during joint motion.^1–3 Repetitive abnormal contact, coupled with morphological changes of the femoral head and acetabulum, results in abnormal contact forces, asymmetric loading patterns, and shear and crush injuries of the labrum, articular cartilage, and the chondrolabral junction.⁴ These pathologic cam and pincer deformities in FAI can lead to hip pain, stiffness, and reduced range of motion and function.² FAI has been implicated as an important cause of early-onset hip osteoarthritis, highlighting the need for early intervention in symptomatic patients.⁵

When nonoperative management for these patients fails to achieve adequate symptomatic relief, hip arthroscopy is considered. The main goals of arthroscopy are to relieve patients’ symptoms, improve function, and delay or even halt the degenerative process. Modern hip arthroscopy has been associated with up to a 42% relative risk reduction in arthritis progression in patients with FAI at midterm follow-up studies.^6,7 However, the presence of degenerative changes, including more advanced radiographic (higher Tönnis grade),^8–10 arthroscopic (higher Outerbridge grade) osteoarthrosis, and more severe labral damage are associated with poor patient-reported outcome measures (PROMs) and subsequent risk of conversion to total hip arthroplasty.^9,11–14 Since these factors can negatively impact the outcome of arthroscopy, detection and further characterization of the labral and chondrolabral abnormalities have become even more critical prior to surgery.

While the prognostic importance of articular cartilage and labrum damage has been explored independently.^8,15 the association with pre-operative patients’ symptoms remains under-investigated. The chondrolabral junction is the transition zone cartilage on the acetabulum that is within 5 mm of labral attachment¹⁶ and is a continuous transition between the fibrocartilaginous labrum and hyaline articular cartilage. Owing to its unique histological features, mainly composed of a unique crimped type 1 collagen, the chondrolabral junction can facilitate force transmission and structural continuity under high shear stresses.¹⁷ It has limited healing potential following injury via scar extension from the vascularized capsular aspect of the labrum.¹⁸ The chondrolabral junction is essential to sustain the hip suction seal and its disruption can compromise this seal.^19–22

Studies have emphasized surgical techniques that preserve the chondrolabral junction to maintain blood flow to the labrum and reduce the rates of capsular adhesions.^11,12,23,24 The incidence of chondrolabral junction pathology with concomitant labral and articular cartilage damage seen during arthroscopy has also been studied.^25,26 To the best of our knowledge, few studies have specifically examined the correlation between the degree of labral and chondrolabral junction pathology with preoperative functional disability and pain levels in patients with FAI. Therefore, the purpose of this study was to assess the correlation of preoperative clinical symptom severity, preoperative diagnostic imaging for degree of labral abnormalities and chondrolabral junction pathology, and intraoperative findings in patients undergoing hip arthroscopy for FAI. We hypothesized that preoperative clinical symptoms severity would be positively associated with higher grades of labral and chondrolabral pathology on preoperative magnetic resonance imaging (MRI) and arthroscopic evaluation, and that a positive correlation would be found between intraoperative pathology and preoperative MRI findings of labral and chondrolabral abnormalities.

2 Methods

2.3

2.3 MR imaging and imaging analysis

Preoperative MRIs were independently evaluated by three fellowship-trained musculoskeletal radiologists with 10-, 8- and 2-years’ experience who were blinded to preoperative clinical and arthroscopic findings. To establish consistency, readers attended a training session led by the first author, with reference material summarizing the grading schemes provided. Images were viewed on a dedicated picture archiving and communication system workstation utilizing either Visage 7 (Visage Imaging, San Diego, CA) or cloud-based Joints ® PACS (Medstrat, Downers Grove, IL). MR images reviewed included coronal proton density, sagittal proton density with fat suppression, axial T2-weighted with fat suppression, and axial oblique proton density images (Fig. 1). Only morphologic sequences were reviewed (i.e., no sequences such as dGEMRIC, T2 mapping or T1rho were included).

Close

Fig. 1 Examples of labrum and chondrolabral junction pathology on MRI which were confirmed by arthroscopy. A. 29-year-old patient with femoroacetabular impingement (FAI) – Sagittal proton density (PD) with fat suppression MR image shows anterosuperior chondrolabral separation detected by two readers. B. 44-year-old patient with FAI- Sagittal PD with fat suppression MR image shows anterosuperior cartilage delamination that was detected by all three readers. C. 51-year-old patient with FAI – Coronal PD with fat suppression MR image shows superolateral chondral flap which was detected only by one reader. D. 19-year-old patient with FAI – Axial oblique PD with fat suppression MR image shows anterosuperior intrasubstance labral tear which was characterized as complex by one reader and intrasubstance tear by the other two readers. E. 29-year-old patient with FAI - Sagittal proton density (PD) with fat suppression MR image shows linear tear pattern of the anterosuperior labrum which was characterized as linear tear pattern by all three readers. F. 49-year-old patient with FAI - Sagittal proton density (PD) with fat suppression MR image shows anterosuperior multidirectional complex labral tear which was characterized as complex by two readers and linear tear pattern by one reader.

Export to PPT

MRIs were graded for presence of labral fraying, chondrolabral separation, labral tears, chondral flaps, and cartilage delamination. Labral tears were further characterized based on the tear patterns into intrasubstance (signal abnormality does not reach the surface of the labrum), linear (linear signal reaches the surface of the labrum), or complex (two or more patten of signal abnormality reach surface of the labrum) tears (Fig. 1). Grading for labral tears was based on the highest-grade labral damage identified within the arthroscopically accessible anterosuperior quadrant (from the 11 o'clock to 3 o'clock positions). Ligamentum teres (LT) pathology was also graded as degeneration/partial tear or complete tear.

3 Results

3.1

3.1 Patient demographics

There were 109 consecutive patients comprising 114 hips (5 patients had bilateral hip surgery) who were screened for inclusion in the study. Of these hips, 18 were excluded due to history of prior hip arthroscopy (10 hips), incomplete or poor quality preoperative MRI (6 hips), and presence of avascular necrosis of the affected hip (2 hips), resulting in a total of 96 hips of 93 patients included in the final analysis (Fig. 2). Patient demographics are presented in Table 1. The cohort was majority female (57 patients, 61.3%) with an average age of 34.4 years old (range 16-63).

Close

Fig. 2 Flow chart of hips included in analysis.

Export to PPT

Table 1 Patient demographics and clinical characteristics.

Sex
Male	36 patients (38.7%), 38 hips (39.6%)
Female	57 patients (61.3%), 58 hips (60.4%)
Age (years) a	Mean and SD: 34.4 ± 11.9Range: 16 – 63
Laterality
Right	46 hips
Left	50 hips
PROMs b
mHHS	Mean and SD: 63.0 ± 19.7
NAHS	Mean and SD: 59.4 ± 19.2

a Calculated out of 93 unique patients.

b Scores available for 88 out of 96 hips (91.7%).

4 Discussion

The results demonstrate that preoperative MRI in patients with FAI underestimated the presence of labral and chondrolabral junction pathology identified intraoperatively. The most common labral tear patterns were linear and complex on MRI and at arthroscopy with up to fair agreement between MRI and arthroscopic findings of labral tear pattern. There was no significant agreement between MRI and arthroscopic identification of remaining labral and chondrolabral junction pathology. Inter-reader agreement between the three radiologists ranged from fair to moderate for labral and chondrolabral pathology. No significant association was identified between preoperative MRI findings and PROMs, nor between arthroscopic findings and PROMs.

A diagnosis of FAI necessitates a patient be symptomatic with objective clinical signs and imaging findings.³¹ Interestingly, our results demonstrate that preoperative MRI significantly underdiagnosed labral and chondrolabral pathology when compared to arthroscopic evaluation. Furthermore, inter-reader agreement between three radiologists for MRI findings (fair to moderate) was slightly lower than in some other studies with moderate to substantial agreement for 3T MRI.^32–34 The inter-method and inter-reader discrepancies may both, in part, relate to the different variables of labral and chondrolabral junction pathology that were assessed. For example, we specifically assessed chondrolabral separation, chondral delamination and flaps as separate entities as opposed to assessing only the presence or absence of labral tears. While most MRI studies in our cohort were done with 3T scanners we also included 1.5T MRI studies which

are known to have inferior performance than 3T in detection of labral pathology.³⁵ Our results also highlight the challenges of assessment of the hip labral and chondrolabral junction pathology on MRI for known reasons, such as deep location of the hip joint, the oblique orientation of the joint, the highly curved and closely aligned articular surfaces, and small size of those structures. Two systematic reviews of diagnostic value of MRI for assessing hip labral tears found significant variation in the diagnostic performance of unenhanced MRI for detection of hip labral tears with both sensitivity and specificity ranged from zero to 100%.34, 36 with few studies demonstrating sensitivity and specificity greater than 80%. The same studies have shown overall higher performance of direct MR arthrography against a surgical reference standard in the assessment of labral tears.³⁴

Our study design was designed to evaluate the correlation of imaging and arthroscopic findings with severity of the symptoms on the baseline PROMs. The lack of significant correlation between pre-operative MRI findings and baseline clinical symptoms is in agreement with few other studies. Grace et al. studies correlation of PROMs with MRI–based measures of the cartilage composition including T1ρ and T2 cartilage relaxation times in patients undergoing hip arthroscopic surgery for symptomatic FAI.³⁷ The labral abnormalities in their study were, however, only assessed during hip arthroscopy. They reported that worse PROMs were correlated with progressive femoral cartilage damage, as indicated by increasing T1ρ and T2 relaxation times in the anterosuperior femoral head region, while there were no correlations with acetabular cartilage damage found on quantitative MRI or labral tearing identified arthroscopically. Westermann et al. also showed that baseline PROM scores did not correlate with cartilage and labral damage seen during hip arthroscopy for patients with FAI.³⁸ There was no imaging assessment of labral or chondrolabral junction pathology in their study.

Stelzeneder et al. studied correlation of MRI morphological and dGEMRIC changes of 21 hips with FAI on seven radial MR images divided to 7 anatomical regions.³⁹ The labrum was characterized as normal, partial tear, or full-thickness tear/complete detachment from acetabular rim. Acetabular cartilage damage was graded as normal, focal defect, or generalized defect. Authors finally compared eight sum scores for each type of MRI finding (cartilage damage, labrum damage, paralabral cyst, acetabular rim bone cyst, acetabular center bone cyst, femoral bone cyst, acetabular rim osteophyte, femoral osteophyte). The authors found only moderate correlation of cartilage damage and Western Ontario and McMaster Universities (WOMAC) pain score in FAI patients. Labrum damage did not correlate with the WOMAC pain score.

Gillinov et al. assessed the association of chondrolabral injuries detected at arthroscopy on PROMs preoperatively and at 24-month follow-up postoperatively. Severe chondrolabral injuries were associated with worse functional outcomes and pain levels preoperatively and up to 24 months after hip arthroscopy. Patients with severe chondrolabral injury experienced greater mean improvements in PROMs and achieved clinically meaningful outcomes at similar rates to patients with mild chondrolabral injury.⁴⁰ They concluded that patients with severe chondrolabral injury in the absence of significant osteoarthritis can still benefit substantially from hip arthroscopy.

Our studies differ in the respect that Gillinov et al. used the Beck classification of transition zone cartilage and labrum injury to grade the damage during arthroscopy^4,41 without correlation with MRI findings. Our intraoperative findings failed to show a significant association with the baseline PROMs. Gillinov et all also evaluated the severity of chondrolabral junction injury and suggested that patients with chondrolabral injury and minimal articular cartilage damage (Tönnis grade <2) but unusually high pain levels and functional limitations may have symptoms attributable to severe chondrolabral injury based on their arthroscopic grading.⁴⁰ Therefore, the result of our study that showed no correlation between chondrolabral separation and baseline PROMs may be in part because we did not assess the severity of chondrolabral injury.

In another study by Freke et al., the association between preoperative PROMs and intraarticular pathology identified at the time of hip arthroscopy was evaluated. The authors found that severe cartilage damage of the femoral head and large labral tears had the greatest number of significant associations with preoperative PROMs; however, at best, only explained 22% of the variability in preoperative PROMs.⁴² In a recent study, Jardon et al. assessed the correlation of pre-operative symptoms with pre-operative imaging and intraoperative cartilage loss in patients undergoing hip arthroscopy for FAI.⁴³ Their study showed similarly that MRI underestimated the cartilage loss with very weak negative to no correlation of imaging cartilage findings and weak positive correlation with intra-operative cartilage findings with baseline symptoms. Based on our and other studies’ results, it is plausible that progressive damage to the labrum and chondrolabral junction in FAI may thus occur at a subclinical level for a period before symptoms from these pathologies worsen preoperative PROMs. Furthermore, preoperative PROMs and severity of symptoms are likely influenced by a variety of factors, including and independent of intraarticular pathology.

The exact mechanism behind pain generation in the hip joint remains under investigation. The articular cartilage has no neural supply and is considered an unlikely direct source for pain44. However, it is believed other structures such as the synovium, the joint capsule and the subchondral bone are responsible for pain generation associated with cartilage damage.^44,45 Assessment of these structures was beyond the scope of our study, but they could potentially function as pain generators in our cohort. Although we found no correlation between labral or chondrolabral junction pathology and PROMs, labral tears are recognized as a cause for pain in the hip joint.^15,46–48 Studies have shown that the superior labrum has the highest density of the nociceptor among all labral regions. A recent systematic review concluded that the anterior capsule and the superior labrum appear to be the primary pain generators, given their higher density of nociceptors.⁴⁹ Therefore, patients may experience different symptoms based on the location of labral damage. Several studies have shown MRI evidence of labral tears in hip joints of up to 70% of asymptomatic patients.^50–53 There is, however, no specific information on the exact location of the labral tears in these studies.

Several limitations of this retrospective study should be considered. All patients in our cohort were symptomatic and studied before and during arthroscopy; therefore, including a truly asymptomatic control group was not possible. Nevertheless, this has not affected the goal of this study, which was to correlate the severity of patient symptoms with the extent of labral and chondrolabral junction pathology. Commensurate with all retrospective studies, our findings are not sufficient to establish a causal link between labral and chondrolabral injury identified on MRI or at the time of hip arthroscopy and baseline PROMs. Furthermore, other unmeasured factors may have affected the correlation, such as cartilage damage, bony pathology, or a higher prevalence of mixed-type (cam plus pincer morphology) FAI in those with higher grades or more complex labral damage. However, to address this, our study only included patients with symptomatic FAI who underwent arthroscopy with mild osteoarthritis (Tönnis grade <2) that controlled for fixed effects, including Tönnis grade and FAI type. We did not assess other demographic factors such as mental health status, activity level or smoking as potential factors that can impact baseline PROMs as suggested by other studies.

In conclusion, in patients with FAI who underwent hip arthroscopy, preoperative MRI underestimated the presence of labral and chondrolabral junction pathology found intraoperatively. Furthermore, MRI and intraoperative findings of labral and chondrolabral junction pathology did not correlate with preoperative symptom severity, suggesting that patients’ symptoms may be more affected by other structural or degenerative changes of the joint until progressive damage to the labrum and chondrolabral junction reaches a certain threshold, the extent of which will require further investigation.

Data availability statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Authors’ contributions

Mohammad Samim MD: Conceptualization; Data curation; Formal analysis; Investigation; Methodology; Project administration; Validation; Visualization; Roles/Writing – original draft; Writing – review & editing. Meghan Jardon MD: Conceptualization; Data curation; Formal analysis; Investigation; Methodology; Project administration; Validation; Visualization; Roles/Writing – original draft; Writing – review & editing. Caroline Vonck, MD: Formal analysis; Investigation; Methodology; Validation; Visualization; Roles/Writing – original draft; Writing – review & editing. Dhruv S. Shankar MD: Data curation; Formal analysis; Investigation; Methodology; Validation; Visualization; Roles/Writing – original draft; Writing – review & editing. Charles C. Lin MD: Formal analysis; Investigation; Methodology; Validation; Visualization; Roles/Writing – original draft; Writing – review & editing Joshua P. Castle MD: Formal analysis; Visualization; Roles/Writing – original draft; Writing – review & editing. Thomas Youm MD: Formal analysis; Investigation; Visualization; Roles/Writing – original draft; Writing – review & editing Christopher Burke MD: Conceptualization; Data curation; Formal analysis; Investigation;

Patient consent statement (if applicable)

Approval from the Institutional Review Board was obtained, and HIPAA requirements were followed. Requirement for informed consent was waived. None of the data presented in this manuscript is individually identifiable.

Permission to reproduce material from other sources (if applicable)

N/a.

For clinical trials (if applicable)

N/A.

Level of evidence

Level IV.

Ethics Approval and Consent

Ethics Approval and Consent to Participate This study was approved by the institutional review board at our institution. This study was granted a waiver of informed consent as this study solely involved retrospective chart review.