Keywords
femoroacetabular impingement - hip arthroscopy - magnetic resonance imaging - hip pain
Introduction
Femoroacetabular impingement (FAI) is a common cause of groin pain in active young and middle-aged patients resulting from abutment between proximal femur and acetabulum. This disorder can be functional or can result from anatomical abnormalities on either the femoral head–neck (cam-type) or acetabulum (pincer-type), or often a combination (mixed-type).[1]
[2]
[3]
Operative treatment is often indicated to answer the functional requests of these patients. Surgery is aimed at reducing pincer lesions, increasing femoral head–neck offset (osteochondroplasty), and treating associated soft tissue lesions such as labral tears. Recently, arthroscopic procedures have become increasingly popular for the treatment of FAI.[4]
[5]
[6] However, not all patients improve after surgery, and persistent bone impingement has been reported as the most frequent cause for revision.[7]
[8]
[9] Although several criteria have been proposed in the preoperative setting,[4] the postsurgical imaging of FAI has been sporadically discussed in the literature.[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
The purpose of this study was to retrospectively evaluate the preoperative and postoperative radiographic and magnetic resonance imaging (MRI) findings in a group of patients treated arthroscopically for FAI, with particular reference to pain. Our hypothesis was that patients with persistent hip pain after surgery had an insufficient correction of preoperative radiographic abnormalities.
Methods
This study was approved by the local ethic committee and was performed in accordance with the ethical standards of the 1964 Declaration of Helsinki as revised in 2000. The ethic committee waived the need for informed consent for publication since rights and interests of the patients would not be violated and their privacy and anonymity would be assured by this study design.
From a pool of 42 patients undergoing hip arthroscopy, we retrospectively selected 25 patients treated for a mixed-type FAI according to inclusion and exclusion criteria listed in [Table 1].[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24] Among these, 23 patients (23 hips) were available for follow-up evaluation. There were 11 male and 12 female, with an average age of 46.5 (range: 28–67) years. Radiographs and MRIs were performed preoperatively and at a mean of 6.0 ± 1.5 months postoperatively, with the same scanner and standard techniques.
Table 1
Inclusion and exclusion criteria
|
Inclusion criteria
|
Exclusion criteria
|
|
Positive impingement sign[15]
|
Previous surgery on the same hip
|
|
Limited internal rotation of the hip of ∼20 degrees, at 90 degrees of flexion
|
Acetabular retroversion with insufficient posterior wall (positive posterior wall sign)[17] on X-rays
|
|
Cam deformity (α angle > 55 degrees) on axial MRI scans[16]
|
CE angle < 20 degrees[21]
|
|
Pincer deformity (crossover sign,[17] coxa profunda,[18] or protrusio acetabuli[19]) on anteroposterior pelvic X-rays)
|
Sequels of childhood hip disease, such as Perthes' disease or slipped capital femoral epiphysis
|
|
Chondrolabral lesions on MRI scans
|
Hip osteoarthritis > grade 1 according to the Tönnis classification[22]
|
|
Loose bodies
|
Previous infection on the same hip
|
|
No more than 6 mo of failed conservative treatment (rest, activity restriction, NSAIDs) before surgery[20]
|
Autoimmune diseases (i.e., rheumatoid arthritis)
|
|
Written informed consent
|
|
Abbreviations: CE, center-edge; MRI, magnetic resonance imaging; NSAIDs, nonsteroidal anti-inflammatory drugs.
Radiographs included a well-centered anteroposterior (AP) standing pelvis view, unilateral AP view, and cross-table lateral view of the hip.[25] Because magnetic resonance arthrography (MR-A) may inaccurately diagnose the presence of persistent, recurrent, or unhealed labral tears after surgery,[26] standard MRI was performed in all cases using a 1.5-Tesla MR unit (Achieva, General Electric, Milwaukee, Wisconsin, United States). A flexible, wrap-around, receive-only surface coil was used. Coronal T1-weighted fast spin-echo sequence and STIR (short tau inversion recovery) sequence, sagittal T1-weighted fast spin-echo sequence and axial proton density (PD)-weighted fast spin-echo sequence with fat saturation were performed; a transverse oblique (parallel to the long axis of the femoral neck) PD- or T1-weighted fast spin-echo sequences were also performed.
Hip arthroscopies were performed in the lateral position and under general anesthesia. Three portals (anterolateral, posterior paratrochanteric, and superior) were commonly used, with further portals being added if required. Cam-type impingement lesions were excised with arthroscopic burr (osteochondroplasty); pincer-type impingement lesions were treated by partial detachment of the labrum, acetabuloplasty, and labrum suture with anchors; loose bodies were removed and other intra-articular procedures were performed as needed. Osteochondroplasty was judged adequate when by dynamic examination and with direct visualization, an impingement-free internal rotation of at least 30 degrees in the 90-degree flexed hip position was achieved. After surgery, weight-bearing as tolerated on two crutches for 6 weeks was prescribed.
The same well-trained musculoskeletal radiologist, who was unaware of the clinical results, measured in a blinded fashion preoperative and postoperative center-edge (CE) angle[23] on AP view of the pelvis and α angle[18] on axial MRI scans. Alpha angles were measured on MRI because on plain radiographs, the asphericity of the femoral head would be underestimated.[25]
[27]
[28]
[29]
[30]
[31]
[32]
Descriptive statistics were used to summarize the characteristics of the study group and subgroups, including mean and standard deviation of all continuous variables. An unpaired two-sided Student's t-test was used to compare objective outcomes. Significance was defined as p < 0.05. Statistical analyses were performed with SPSS v.15.0 (IBM, Chicago, Illinois, United States).
Results
For the whole group of patients, the mean preoperative CE angle at radiographic evaluation was 38.6 ± 5.2 degrees, whereas on MRI scans, the mean preoperative α angle was 67.3 ± 7.2 degrees. Preoperative MRI also revealed that 20 of 23 hips had a cartilage abnormality that involved the anterosuperior acetabulum. One patient had an anterosuperior cartilage defect at surgery that was not visible at imaging. Twenty-one hips had an anterosuperior labral tear.
At follow-up, 17 patients were pain–free, and imaging studies documented no pathological bone, chondral, or labral finding. In this group of patients, the mean preoperative and postoperative CE angles were 38.1 ± 5.6 and 32.6 ± 4.8 degrees, respectively ([Fig. 1]), whereas the mean preoperative and postoperative α angle on MRI scans were 66.3 ± 7.9 and 47.9 ± 8.9 degrees, respectively ([Fig. 2]).
Fig. 1 Standard anteroposterior view of a 36-year-old female patient presenting without persistent hip pain at follow-up. Center-edge angle measured 33.35 degrees preoperatively (A) and 26.95 degrees postoperatively (B).
Fig. 2 Axial magnetic resonance imaging views of the same patient reported in [Fig. 1] showing reduction of the α angle after surgery from 63.2 degrees preoperatively (A) to 54.3 degrees postoperatively (B).
Among the six patients with persistent hip pain at follow-up, MRI evidenced persistence of bone edema and signs of acetabular labrectomy (labral borders irregularity with persistent fragmentation and chronic synovial flogosis) in four cases. In the two remaining cases in which FAI was associated to loose bodies, a reduction in their number was found without persistent edema associated to acetabular labrectomy. In this group of patients with persistent hip pain, the mean preoperative and postoperative CE angles were 39.8 ± 3.6 and 35.8 ± 3.1 degrees, respectively ([Fig. 3]), whereas the mean preoperative and postoperative α angles were 70.0 ± 3.9 and 58.8 ± 2.6 degrees, respectively ([Fig. 4]).
Fig. 3 Standard anteroposterior view of a 36-year-old male patient presenting with persistent hip pain at follow-up. Center-edge angle measured 48.44 degrees preoperatively (A) and 45 degrees postoperatively (B).
Fig. 4 Magnetic resonance imaging scans of the same patient reported in [Fig. 3] showing an insufficient reduction of the α angle after surgery from degrees 71.5 preoperatively (A) to 60.9 degrees postoperatively (B).
On comparing patients with and without persistent hip pain after arthroscopy, we did not observe any significant difference in terms of gender, age, and preoperative CE and α angles. Mean values of the analyzed radiological parameters improved significantly after surgery, except mean CE angle in patients with residual hip pain (p = 0.1). Patients without pain showed significantly lower postoperative α angle (p = 0.008) and lower CE angle (p = 0.07) than those with residual pain.
Discussion
Imaging after surgical treatment of FAI is demanding due to controversies in the selection of the appropriate diagnostic protocol and in the differentiation between normal and pathological findings. Direct MR-A is considered the gold standard in the preoperative setting; however, it is unable to accurately diagnose the presence of persistent, recurrent, or unhealed labral tears after surgery.[26] Furthermore, MR-A requires fluoroscopic or sonographic guidance for the intra-articular injection of the contrast medium, has a potential risk of infection, and is not always well accepted by the patients.[33] Conventional MRI may be used, but the imaging protocol is not widely reproducible in the average clinical setting.[34]
[35]
On plain radiographs, α angle values higher than 50 to 55 degrees have been commonly used as a two-dimensional threshold indicating cam-type impingement.[18]
[27]
[36] However, the values of this angle have been demonstrated to vary between radiographic views, leading to unreliability in the clinical practice.[37] MRI has been used to analyze the three-dimensional α angles, aiming to a better evaluation of the sphericity of the femoral head,[25]
[27]
[28]
[29]
[30]
[31]
[32] but some authors[38]
[39] questioned its ability to improve the diagnosis of cam-type FAI.
There is an increasing evidence that an insufficient correction of femoral and acetabular bony abnormalities is a cause of persistent groin pain and poor postoperative outcomes.[7]
[9]
[11]
[15]
[16] On the other hand, an excessive correction of these deformities may lead to femoral neck fracture and hip instability.[40]
[41] In patients in whom the correction of the α angle to less than 55 degrees would lead to excessive resection, limiting reduction by no more than 20 degrees should be considered.[42]
de Sa et al[15] included a total of 14 studies (12 case series, 1 prospective cohort study, and 1 case–control study) in their systematic review and meta-analysis, with a follow-up between 1 and 4 years. Thirteen of these studies reported a postoperative α angle less or equal to 55 degrees. Only one of these studies[13] was a prospective ongoing study on 106 patients (118 hips) treated with a minimally -invasive anterolateral approach, at an average follow-up of 2.2 years. The authors reported a significant improvement on the mean α angle from 61.8 to 36.9 degrees, with a significant improvement in all range of motions after surgical correction of the α angle to less than 55 degrees.[13]
Other papers, not included in the aforementioned review, reported on postoperative imaging in FAI.[11]
[12]
[14] Philippon et al[11] in their prospective study reported on 58 hips a mean reduction of CE angle of 3.9 (range: 0 – 17) degrees. They showed that change in CE angle could be estimated from millimeters of acetabular bony resection, and the CE angle is a reliable radiographic diagnostic tool for the measurement of acetabular coverage. Bedi et al,[12] in their prospective nonrandomized study on 60 male patients younger than 40 years treated with either an arthroscopic or an open approach, reported that the imaging results of arthroscopic osteochondroplasty were comparable to open surgical dislocation for anterior and anterosuperior cam and focal rim impingement deformity. However, the authors reported that open technique resulted in greater correction of posterosuperior loss of femoral offset and may be favorable for FAI patterns that demonstrate considerable proximal femoral deformity on AP radiographs. Zingg et al,[14] in their prospective comparative study on 28 patients treated with either an open or arthroscopic approach, at 6 weeks follow-up, reported on MRI that arthroscopic osteochondroplasty removed more bone than necessary in some positions, resulting in significant lower α angles. The authors complained that this overcorrection is unwelcomed since irreversible and may be an issue in terms of joint sealing. We think that the results of this study should be taken with caution; in fact, most of the scrutinized patients refused to be included into the study, and most of the enrolled patients were not randomized.
In our study, mean values of all the analyzed radiological parameters improved significantly after surgery, except mean CE angle in patients with residual hip pain (p = 0.1). On comparing the two subgroups, significant lower postoperative α angles were observed in patients without pain (p = 0.008). CE angle was lower in patients without pain, albeit difference did not reach statistical significance (p = 0.07).
This study has some strengths. Patients were selected according to strict inclusion and exclusion criteria, and a homogeneous group of mixed-type FAI patients was analyzed. Imaging studies were acquired with the same standardized protocol and analyzed by the same well-trained musculoskeletal radiologist in a blinded fashion. Radiological parameters were considered both on the femur and the acetabulum.
Several limitations have to be recognized. Certainly, this is a retrospective study with a small sample size; results are reported at a short-term follow-up, still in the range of previously reported literature (6 weeks to 45 months).[10]
[11]
[12]
[13]
[14]
[15] Finally, preoperative and postoperative functional scores were not reported, but this was beyond this study.
In conclusion, in case of persistent pain after arthroscopic treatment of FAI, a set of imaging studies should be performed. The recognition of the anatomical changes following arthroscopic treatment is an important step to avoid a wrong clinical evaluation after therapy. Persistent pain after arthroscopic treatment of FAI may be due to an insufficient correction of preoperative radiological abnormalities.
