A novel radiographic finding for estimation of the cortical-cancellous boundary: A magnetic resonance imaging case series of patients with femoroacetabular impingement

Abstract

Purpose

To determine if comparing alpha angle measurements of the outer cortical margin (preoperative alpha angle) and the inner cortical margin (theoretical postoperative alpha angle) on magnetic resonance imaging (MRI) can estimate the amount of sclerotic bone that needs to be resected to restore the alpha angle to normative values (<55⁰) in patients with symptomatic femoroacetabular impingement (FAI).

Methods

A retrospective review was conducted on patients with FAI undergoing primary hip arthroscopy and CAM osteoplasty from June 2013 to February 2017. Inclusion criteria were radiologist-measured alpha angles >70° on 3T oblique axial MRI, unilateral primary hip arthroscopy, and age ≥18. Exclusions included previous hip surgery, hip trauma, avascular necrosis, advanced osteoarthritis, dysplasia, poor quality images, or missing MRI images. Alpha angles were measured preoperatively using the outer cortical margin and postoperatively using the inner cortical margin of femur. Surgeries were performed by the senior surgeon. Paired t-tests compared preoperative and postoperative alpha angles, and intraclass correlation coefficients (ICC) were used to assess inter-rater and intra-rater reliability.

Results

One hundred patients (100 hips) were included with an average age of 40.8 years (range, 18.7–64.9), with 75 % being male (n = 75). The average BMI was 24.7 ± 3.9 (range, 17.1–38.8). The average maximum alpha angle at the outer cortical margin was 77.0 ± 4.2° (range, 70.0–90.8), with an ICC for inter-rater reliability of 0.861, indicating excellent agreement. The average inner cortical alpha angle was 46.4 ± 3.4° (range, 36.5–55.4), significantly less than 55° (p<0.001). The mean difference between the outer and inner cortical alpha angles was −30.6 ± 5.1° (p< 0.001).

Conclusions

The difference between outer and inner cortical alpha angles on MRI can accurately estimate the amount of sclerotic bone resection needed to restore the alpha angle to normative values in patients with FAI. This information may be directly applied clinically marker to assess adequacy of CAM resection.

1. Introduction

Femoroacetabular impingement (FAI) is a rapidly evolving etiology of hip pain and is the most common indication for hip arthroscopy.¹ FAI can be[ subclassified into three subtypes based on patient morphology: (1) CAM impingement (bony overgrowth on the femoral neck), (2) pincer impingement (bony overgrowth on the acetabulum), or (3) mixed impingement.² It is known that CAM impingement is more likely in athletes, particularly with sports that are associated with repetitive loading of the hip joint in maximum flexion.^3,4 If left untreated, the symptomatic impingement due to FAI progresses and leads to severe degeneration of the hip joint.⁵

The alpha-angle measurement on a plain radiograph is the gold standard for diagnosing FAI with CAM pathology.⁶ The alpha angle lies between the femoral neck axis and the point at which the femoral head deviates laterally from a circle templated onto the radiograph or magnetic resonance imaging (MRI) (Fig. 1).^7,8 There is no universal consensus on the threshold of alpha angle in the orthopedic literature, most agree that alpha angles >55° are defined as pathologic.^9,10

Fig. 1.

The alpha angle is measured by defining the center of the femoral head and then drawing the circumference of a circle. Then one line is drawn from the point at which bone deviates from the circle toward the center of the femoral head; the second is drawn from the center of the isthmus of the femoral neck to the center of the head.

Figure adopted from Philippon et al., 201332, Copyright Sage Publishing 2013.

Surgical removal of the bony overgrowth in patients with FAI helps restore the hip joint's anatomy and function, reestablish the fluid seal between the labrum and acetabulum, and reduce the alpha angle to normal values (<55⁰).11, 12, 13, 14, 15 However, over-resection of the CAM lesion is a recognized source of failure, leading to micro-instability of the joint and potential femoral neck fracture.¹⁶ Under-resection is also considered problematic, as it can lead to higher rates of reoperation.¹⁶ Both are associated with inferior post-operative outcomes and are common causes of revision surgery.16, 17, 18, 19 However, determining the optimal resection depth intraoperatively remains a technically challenging aspect of the procedure.²⁰

Considering each patient's unique anatomy and sport-specific demands, there is a need for a validated method for quantifying the ideal resection depth of sclerotic bone.²¹ Currently, intraoperative fluoroscopy, arthroscopic robotic computer navigation, and direct visualization with dynamic exam have been cited as techniques to assess CAM resection, however each has limitations.22, 23, 24 Intraoperative fluoroscopy risks over-resection when the plane of the x-ray beam is perpendicular to resection plane.¹⁹ Additionally, Ross et al. reported that the location of the alpha angle observed on fluoroscopy may change depending on the fluoroscopic view, and also that increased femoral anteversion significant changes the location of the maximum alpha angle.²⁵ Arthroscopic computer navigation has the potential to guide resection more accurately, but does not account for concurrent acetabular and femoral resections, potentially leading to CAM over-resection.7, 8, 9 Direct intraoperative visualization and dynamic exam are considered to be the gold standard to assess the depth of resection to determine if sufficient osteoplasty has been performed to adequately address the impingement.²⁶ After osteoplasty of the femoral head-neck junction, the surgeon manipulates the hip in flexion and abduction through the full range of motion.²⁷ The hip joint is directly visualized arthroscopically to ensure that the labrum maintains its fluid seal with the femoral head. This technique allows the surgeon to identify and resect any remaining bone that may be causing impingement.^28,29 However, the direct visualization and dynamic exam is highly dependent on surgeon expertise.^30,31 Achieving an adequate resection depth during osteoplasty for a CAM lesion in patients with femoroacetabular impingement (FAI) due to cam-type impingement continues to be a challenging task for many surgeons.²⁰

The purpose of this study was to assess if the difference between alpha measurements of the outer cortical margin (preoperative alpha angle) and inner cortical margin (theoretical postoperative alpha angle) on MRI can serve as an estimate of the amount of sclerotic bone that must be resected in order to restore the alpha angle to normative values (<55⁰). It was hypothesized that the alpha angle measured using the inner cortical margin, representing operative resection to cancellous bone, would establish a normal alpha angle (<55⁰), and that resection to cancellous bone can function as an intraoperative marker for adequacy of CAM resection.

2. Methods

2.1. Patient selection

This study was approved by the Vail Health Institutional Review Board. A retrospective data registry was queried for all hips between June 2013 and February 2017. Included patients had a radiologist-measured alpha angle >70⁰ on 3T oblique axial MRI, underwent unilateral primary hip arthroscopy for the primary diagnosis of FAI, and were 18 years of age or older. Alpha angles greater than 70⁰ on preoperative MRI were chosen because of the higher susceptibility for these patients to display FAI symptoms (i.e. pain).⁹ Patients were excluded if they had previous hip surgery, hip trauma (e.g. fracture or dislocation), avascular necrosis, advanced osteoarthritis (> Tonnis Grade 2), dysplasia (lateral center edge angle <20⁰), poor quality images, no MRI or missing preoperative and postoperative supine AP pelvis X-ray images. Demographic data, including age at surgery, body mass index (BMI), and sex were collected for all patients. All patients in this study underwent CAM osteoplasty by the senior surgeon (MJP).

2.2. Image acquisition

Images were acquired on a 3-Tesla (3-T) MRI without contrast, using a proton-density-weighted (PDw) fat-suppressed image in the oblique axial plane formatted for measurement of maximal alpha angle.³¹ This series was acquired with a slice thickness/gap of 3.0/0.30 mm and in-plane resolution of approximately 0.25 × 0.25 mm, with the imaging plane aligned with the axis of the femoral neck as viewed in the coronal plane(Fig. 3, Fig. 5, panel A).

Fig. 3.

Single image slice on the preoperative T1 axial proton-density-weighted fat suppressed image at the site of maximal cam morphology used to perform the alpha-angle measurements.

Fig. 5.

In panel A, Oblique axial proton-density-weighted fat suppressed images for one example patient showing the alpha angle measurement slices. In panels B and C the femoral head best-fit circle is shown in white, with the line connecting the center-point of the femoral head and center-point of the femoral neck shown in magenta and the line marking where the femoral head intersects the best-fit circle shown in dotted green. The alpha angle measurement is shown in yellow.³⁸.

2.3. Alpha angle measurement

All MRIs were assessed by two independent fellowship-trained Orthopaedic surgeons. The alpha angle was determined on the preoperative axial oblique T1 MRI at the site of maximal CAM deformity.^32,33 Based on previous techniques, the preoperative alpha angle was determined by first by identifying the center of the femoral head and then drawing a circle of best fit along the circumference of cortical bone lining the femoral head (Fig. 5 B and 4B).^32,34 Then, one line was drawn originating from the center-point of the femoral head to the center-point of the femoral neck (Fig. 5B and C) and a second line is drawn from the point at which sclerotic bone of the outer cortical rim of the femoral neck deviates from the circle of femoral cortical bone. A second alpha angle measurement was obtained similarly, however in this method, the second line now extended to the inner cortical layer of the femur, at the junction of cortical and cancellous bone (Fig. 4C). Using this method, the inner cortical rim line represents theoretical CAM resection of cortical bone to a normal metaphyseal layer (i.e. the theoretic postoperative alpha angle), and the difference between the outer and the inner cortical rim lines represent the theoretical CAM resection (Fig. 4D).

Fig. 4.

(A) The red arrow points to the cam deformity (B) alpha angle at the outer cortical layer of the femoral neck (C) alpha angle at the inner cortical layer of the femoral neck (D) difference between outer cortical to inner cortical alpha angle.

2.4. Surgical technique

All patients met operative indications by the senior author. Patients were placed in the modified supine position on table with two working portals (antero-lateral and anterior).²⁸ With the hip flexed and externally rotated, visualization of the femoral head-neck junction is observed. Osteoplasty of the femoral head-neck junction was performed using the motorized burr (Fig. 2).²⁸ The cam lesion was trimmed to restore the natural shape of the femoral head and remove the bony projection on the femoral neck that was causing impingement on the anterior labrum and acetabulum.³⁵ Adequacy of decompression was confirmed by the senior surgeon through dynamic exam.³⁵ This was achieved by increasing hip flexion beyond 90° and observing the labrum smoothly glide over the femoral head, confirming that the labrum maintained a fluid seal with the femoral head throughout the normal range of motion.^28,29,35

Fig. 2.

An osteoplasty is performed by hip arthroscopists using a burr (B) to resect part of the femoral head (FH) neck junction (C = hip capsule). Over-resection may be a potential complication of treating cam-type FAI.

2.5. Statistical analysis

An a priori power calculation based on our primary hypothesis that there would be a significant difference between outer (70°) and inner (55°) cortical alpha angles found that 20 subjects MRI measurements were needed to detect a significant difference between the outer and inner cortical alpha angles. However, the power calculation based on our secondary hypothesis that the measured inner cortical alpha angle was less than the expected target alpha angle 55° found that a greater number of subjects would be required. Assuming an inner cortical angle of 50° and a standard deviation of 15°, a total of 100 subjects were needed to achieve 90 % power at an alpha level of 0.05 for a two-sided test. This second power analysis was used for the study design.

Means, standard deviations, and ranges were recorded for alpha angle measurements. A paired t-test was used to determine if there was a significant difference between outer and inner cortical alpha angles. A one-sample t-test was used to determine if the inner cortical alpha angle was different from 55°. All tests were two-sided. P values<0.05 were considered significant (alpha level = 0.05). A two-way random effects model was used to estimate intraclass correlation coefficients (ICC) for inter- and intra-rater reliability.

3. Results

One hundred patients (100 hips) were included with an average age of 40.8 years (range, 18.7–64.9). Seventy-five percent of the patients were male (n = 75). The average body mass index for all patients was 24.7 ± 3.9 (range, 17.1–38.8). The average maximum alpha angle measured at the outer cortical margin was 77.0 ± 4.2° (range, 70.0–90.8) (Fig. 3, Fig. 4). The ICC for inter-rater reliability for the outer cortical alpha angle was 0.86, indicating excellent agreement among raters. The average inner cortical alpha angle was 46.4 ± 3.4° (range, 36.5–55.4). This value of 46.4 is significantly less than 55° (p < 0.001) (Fig. 4). The mean difference obtained from outer cortical to inner cortical alpha angle was −30.6 ± 5.1° (p < 0.001). The ICC for inter-rater reliability for the inner cortical alpha angle was 0.84, also indicating excellent agreement.

4. Discussion

While several methods exist for intraoperative determination of CAM resection, no single method is universally accepted. Leaving the assessment of adequate resection depth up to surgeon preference. The present study determined that hypothetical resection of the cortical layer on CAM deformity via MRI measurements may reliably restore the alpha angle to normative values (<55) and provides a proof-in-concept that resection to the cancellous layer is likely of sufficient depth. Which would result in lowering complications due to over or under-resection. This information may be directly applied clinically marker to assess adequacy of CAM resection. Consequently, using a preoperative imaging guide may help surgeons establish an adequate amount of bony resection intraoperatively and therefore avoiding the complications of under or over-resection. MRI may be useful in establishing the amount bone to debride during surgery to restore native anatomy and function.

Studies from non-pathologic cadavers have provided orthopedic surgeons guidelines for the safe depth of resection (at the site of largest bony deformity as a surrogate marker of adequate depth of resection).³⁶ Resulting in wide acceptance that resections involving less than 30 % of the femoral head-neck diameter are considered safe in terms of minimizing the risk of fracture.³⁶ However, a study conducted by Atkins and colleagues suggests that CAM impingement leads to cortical thickening throughout the entire proximal femur.⁷ Their research utilized volumetric segmentation data from computed tomography images of patients with CAM lesions and age-matched controls. The findings revealed that the median thickness of cortical bone was significantly greater in patients with cam deformity when compared to controls, both in the area of the cam and in the proximal femur.⁷ The authors propose that the impingement caused by cam type FAI may induce hypertrophy of the bone in response to increased load at the primary site of impingement (Wolff's law). If CAM type FAI patients have thicker cortical bone, then the current guidelines for resection (which are based on studies of patients with normal anatomy) may be too conservative. Henceforth, resulting in underresection.⁷ Atkins et al. conducted another study that confirms the efficacy of using this bony boundary as a tool for restoring femoral head-neck anatomy and supports its application as a guide to limit resection depth during osteochondroplasty for cam-type deformities.^20,36 Before surgery, the median difference between patients with cam deformity and those without was 1.8 mm which decreased to 0.2 after resection.³⁶

Mansor et al. reports that CAM over resection has more consequences than undersection.¹⁶ Their prospective study included 120 patients who initially presented for revision hip arthroscopy, and grouped them according to level of resection from previous surgery (over-resected, under-resected, or neutral). At mean 39.6 month follow up, it was found that patient reported outcomes (modified Harris Hip score and the Nonarthritic Hip Score) were lower at initial presentation in the over resected group compared to under resected. Still at two years follow up, the mHHS was lower than the under resected group.¹⁶ Conversation to a total hip arthroplasty was more common in patients with over resection (30 % vs 0 %).¹⁶ Over-resection may compromise the labral seal, which plays a crucial role in maintaining hip fluid dynamics and intra-articular stability, as noted by the authors.¹⁶ In effect, reestablishment of the fluid seal through the labrum is guided by the amount of resection that is made on the femoral head neck junction.

In a study detecting adequacy of resection, Ross et al. found that maximum alpha angle measured on preoperative CT scans (67°) was comparable to the maximum alpha angle assessed on intraoperative fluoroscopy (65°).³⁷ This present study used the same methodology to measure alpha angle.³⁷ We found that that the maximum alpha angle on pre-operative MRI was 77.0°. Which is 10° higher than Ross et al. but still this difference may be due to differences in patient population, cam impingement severity, etc. Regardless, both alpha angles are significantly above the asymptomatic values. The limitation of our study compared to Ross et al. is the lack of postoperative assessment of alpha angle. Given that we used the same methodology as Ross et al., it could be assumed that our results would be similar. Additionally, MRI has advantage over computer tomography by not exposing the patient to ionizing radiation.

5. Limitations

The primary limitation of this retrospective cohort study is the lack of comparison to postoperative imaging to identify whether resection approximated the preoperative measured alpha angle to the cancellous bone. Therefore, it is not known whether the cortico-cancellous junction noted on imaging correlates to the intraoperative findings of the cortico-cancellous layer. Second, this study operated on the definition that a normal alpha angle is to <55°. However, it remains controversial whether resection to <55° is necessary to improve function in patients with FAI. ^22,23,24 Additionally, the study populations consisted of subjects who were radiographically screened using α angles in a single radiographic view. Given that the study population was predominantly male (75%), it is uncertain if this methodology is satisfactory in female patients. Additionally, sport specific demographic information was not recorded in these patients. Future studies should examine the effectiveness of resection to the cancellous bone in comparing pre-and-postoperative images as well as patient reported outcomes.

6. Conclusion

The difference between outer and inner cortical alpha angles on MRI can accurately estimate the amount of sclerotic bone resection needed to restore the alpha angle to normative values in patients with FAI. This information may be directly applied clinically marker to assess adequacy of CAM resection.

Credit author statement

Madeleine Grace DeClercq: Writing - Original Draft, Resources, Writing – review & editing, Validation, Project administration, Methodology, Data collection. Rui W. Soares: Writing - Original Draft, Data collection. Adam M Johannsen: Writing - Review & Editing, Validation, Project administration, Methodology, Mentorship. Toufic R. Jildeh: Writing - Review & Editing, Validation, Project administration, Methodology, Mentorship. Carly A. Lockard: Writing - Review & Editing, Methodology, Mentorship. Lauren A. Pierpoint: Data anylysis and methodology. Maitland D. Martin: IRB writing and literature review. Marc J Philippon: Supervision, Conceptualization, Methodology, Supervision, Writing – review & editing, Supervision, Project administration.

Ethical statement

This manuscript has been prepared in full compliance with the relevant laws and institutional guidelines. The procedures followed were approved by the Institutional Review Board (IRB) of Vail Health. This study was submitted for an exemption consideration. This categorization was based on the research involving minimal risk to participants and the data collected being either anonymous or sufficiently de-identified to protect privacy rights. No identifiable personal data has been disclosed in this manuscript. The exemption was approved on January 28, 2020 (IRB No. 2020-03). Therefore, the nature of the research did not require informed consent from participants. Measures have been taken to protect the privacy rights of human subjects, ensuring that all data was anonymized and confidentiality maintained.

Guardian/patient's consent

The Institutional Review Board (IRB) of Vail Health Hospital has reviewed the study's methods and granted an exemption on January 28, 2020 (IRB No. 2020-03), indicating that the scope of the research involved minimal risk to participants. Given the exempt status of the study, in alignment with the regulations and ethical considerations, the IRB determined that informed consent was not a requisite for participants involved in the research.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.