Codes:

 

 

 

  

 

  

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

A literature search using MEDLINE was performed through March 2009. The key issue for this policy is whether correction of femoroacetabular impingement (FAI) morphology with open or arthroscopic osteoplasty alters the development of symptomatic cartilage damage and hip osteoarthritis. Given the relatively recent recognition of FAI and development of interventional procedures, neither the natural history of FAI nor the effect of osteochondroplasty on the development of osteoarthritis is known. No randomized controlled trials were identified. Therefore, to evaluate the potential benefit of FAI with the evidence available at this time, studies were reviewed for the following:

 

 

Beck and colleagues reported results of a study of 26 patients with pure pistol-grip deformity and 16 patients with isolated coxa profunda who were identified from 302 hips treated for intra-articular pathology between 1996 and 2001. Only hips with minor radiological changes, with narrowing or osteophytes equivalent to an osteoarthrosis grade less than one according to the classification of Tonnis, were included. Excluded were hips with traumatic or post-traumatic conditions (n=37), avascular necrosis (n=14), and hips that had underdone previous surgery (n=7). Patients with incomplete or inadequate preoperative radiographs were also excluded. For the 26 hips that met the inclusion/exclusion criteria and showed isolated cam impingement on preoperative radiographs, all showed acetabular cartilage damage in the anterosuperior area of the acetabulum with separation between the acetabular cartilage and the labrum. In the 16 hips with isolated pincer impingement, the damage was located more circumferentially, usually including only a narrow strip of the acetabular cartilage. The report illustrated that in carefully selected patients with early stage osteoarthritis and well-defined hip configurations, a strong association exists between specific hip morphology and the pattern of cartilage damage. The intent of the study was “to obtain unequivocal data” on the starting point of joint degeneration with FAI; damage in patients with more complex morphology was not described.

 

Baradakos and Villar retrospectively examined progression of osteoarthritis of 43 patients (43 hips) under 55 years of age with a history of symptomatic idiopathic arthritis, first seen no later than 1997, who exhibited pistol-grip deformity of the femur and mild to moderate osteoarthritis (Tonnis grade 1 or 2) at baseline (Bardakos, 2009). Radiographs taken at least 10 years apart showed progression of osteoarthritis in two-thirds of the patients, with 12 receiving hip replacement or resurfacing after more than 10 years. Logistic regression analysis showed the medial proximal femoral angle and the posterior wall sign as the only significant independent predictors for progression of osteoarthritis in this small sample. A reduction of one degree in the medial proximal angle increased the odds of the osteoarthritis progressing by 21 times, while osteoarthritis in a hip with a positive posterior wall sign (the center of the femoral head located lateral to the outline of the of the posterior acetabular rim) was 10 times more likely to progress than a hip that had a negative posterior wall sign. Of note, one-third of the patients with a pistol-grip deformity did not progress rapidly within the assessment period.

 

In another study, Tanzer and Noiseaux reported that of 38 consecutive patients who were treated arthroscopically and who had a labral tear, 97% were found to have a pistol-grip deformity on preoperative radiographs (Tanzer, 2004).  These authors also reported that in 200 consecutive patients (200 hips) having primary total hip arthroplasty, the underlying etiology of patients’ arthritis was determined by their history and radiographic findings. Anteroposterior pelvis, lateral, and frog lateral hip radiographs were evaluated for abnormalities of the femur and/or acetabulum. All patients without a history or radiographic evidence of underlying hip disease were given the diagnosis of idiopathic hip arthritis. From the 125 cases diagnosed as idiopathic arthritis, 100% exhibited a pistol-grip deformity. Radiographs of the contralateral limb showed that 31% of patients had a healthy hip without a deformity or evidence of osteoarthritis, 14% had a deformity without evidence of arthritis, and 55% had a pistol-grip deformity and radiographic evidence of arthritis. A pistol-grip deformity was associated with arthritis later in life.

 

A 2009 study from Asia reviewed records of 843 consecutive patients (978 hips) who underwent primary surgery for osteoarthritis or other diseases of the hip to determine the prevalence of FAI in this population (Takeyama, 2009).  Twenty-six patients (32 hips) were excluded due to insufficient radiographs or records, resulting in a study population of 817 patients (946 hips). The average age at the time of surgery was 54.8 years (range of 12–92 years). All of the patients were Asian. The majority of patients (73%) were diagnosed with osteoarthritis secondary to developmental dysplasia of the hip, another 12% had idiopathic osteonecrosis, and 1.7% had Legg-Calve-Perthes disease. Only 17 patients (1.8%) were considered to have had primary osteoarthritis. Of these, 6 patients (average age, 63 years; range, 32-79) were determined to have FAI from preoperative radiographs, resulting in a possible etiology of FAI for 0.6% of the total population undergoing surgery for osteoarthritis and 35% in the population with primary osteoarthritis.

 

Kim et al reviewed outcomes of 43 patients (mean age, 40 years; range, 18-68 years) who had labral tears and early osteoarthritis (Tonnis grade 0 to 1, average Japanese Orthopedic Association [JOA] scores of less than 1) and symptoms lasting 3 months or more who had been treated with debridement (Kim, 2007).  At an average 50 months’ follow-up (12–96 months), 74% of patients had improved, with 11 showing no improvement. Blinded evaluation of preoperative radiographs and MR arthrograms indicated that 42% of patients had FAI. When treated only with debridement, patients were less likely to improve if early stage osteoarthritis or FAI was present at the time of surgery. For example, on the JOA scale where 0 = severe pain to 3 = no pain, patients without either FAI or osteoarthritis scored 2.6 at follow-up, while patients with FAI scored 1.83 and those with both FAI and osteoarthritis scored 1.67.

 

Evidence on the natural history and long-term effect of treatment is limited due to the relatively recent recognition of this condition. Overall, the retrospective evidence available indicates a relatively strong association between cam-type impingement related to a pistol-grip deformity, labral damage, and the subsequent development of osteoarthritis. The identification of patients with FAI morphology who will progress to osteoarthritis (and perhaps more importantly those who are unlikely to progress) is limited at this time, although some evidence from retrospective studies is beginning to emerge.

 

 

Technology Assessments and Systematic Reviews

In 2006, United Kingdom’s National Institute for Health and Clinical Excellence (NICE) conducted a rapid review of the literature on surgery for femoroacetabular impingement, which included 4 case series and a single non-randomized controlled trial on the open approach.  Specialist advisers’ opinions were included in the assessment, which concluded that there was a lack of evidence to determine whether the procedure successfully slows the rate of progression to osteoarthritis, but that efficacy may be influenced by the degree of pre-existing arthritis or degenerative changes before the intervention. Due to the paucity of data, guidance on the arthroscopic approach could not be issued at that time. In 2007, NICE issued guidance on both open and arthroscopic femoroacetabular surgery for hip impingement syndrome, concluding that current evidence on the safety and efficacy was not adequate for the procedure to be used without special arrangements for consent and for audit or research.

 

Bedi and colleagues performed a systematic review of the literature on labral tears and FAI, with a literature search conducted in May 2008 (Bedi, 2008).  Level I, II, III, or IV study designs were included if the patient population had a labral tear and/or FAI as the major diagnosis. Patients with severe pre-existing osteoarthritis or acetabular dysplasia were excluded. Of the 19 articles included, only 1 met the criteria for level III basis of evidence. The studies reviewed suggested that 65% to 85% of patients will be satisfied with their outcome at a mean of 40 months after surgery. All series reported an increased incidence of failure among patients with substantial pre-existing osteoarthritis. The authors concluded that the quality of literature reporting outcomes of surgical intervention for labral tears and FAI is limited. The larger case series, including those published after the literature search for the systematic review by Bedi et al are described below.

 

 

The March 2009 literature review identified 4 prospective/consecutive case series with at least 100 patients/hips and 1 controlled cohort study. A 2005 report by Sampson that described results from 156 patients was also identified (Sampson, 2005).  However, this retrospective review was primarily a description of the newly developed arthroscopic procedure with preliminary results. The age range of the patients was 14 to 73 years, with symptoms ranging from 3 months to 73 years, although the majority of patients were noted to be in their 20s to 30s. One patient had a pathological fracture and 3 patients converted to total hip arthroplasty (THA) by follow-up of up to 22 months. No other complications were noted. Greater cartilage damage (i.e., greater than grade III chondromalacia or delamination defects) were associated with poorer outcomes and/or longer recovery times. The damage to the articular surfaces observed at arthroscopy had often been undetected by preoperative imaging studies.

 

The largest prospective series was by Byrd and colleagues, who provided a brief report on 200 patients (207 hips) from a consecutive group of 220 patients (227 hips) who had been treated with arthroscopy for impingement in 2004-2007 (Byrd, 2009).  Included were 163 hips with cam-type impingement only and 44 hips with combined cam and pincer-type impingement. Twenty hips with pincer impingement only were excluded. The average age of the patients was 33 (range not reported), with symptoms averaging 32 months and no sign of advanced osteoarthritis. There was 100% follow-up of the 207 hips at a minimum of 12 months. At an average of 16 months (range, 12-24 months) after treatment, patients showed an average 20-point improvement (-17 to 60) on the 91-point modified Harris Hip Score (MHHS). Eighty-three percent of patients were considered to be improved by the procedure. One patient (0.5%) required THA within the follow-up period, and 3 patients (1.5%) required revisions for pain.

 

Phillipon et al reported 2.3 year follow-up (range, 2-2.9 years) on 100 of 209 prospectively enrolled consecutive patients who underwent hip arthroscopy for disabling pain (Philippon, 2009).  Patients were included in the study report (n=122) if they underwent arthroscopic treatment for FAI and chondrolabral dysfunction, and did not have bilateral hip arthroscopy, avascular necrosis, or previous surgery. Ten patients did not complete the follow-up questionnaire and were excluded and 12 were lost to follow-up. Of the 100 patients available for follow-up, 90 (90%) improved from an average score of 58 to 84 on the MHHS, and 10 (10%) required THA at a mean of 16 months. Patients with a joint space of less than 2 mm were 39 times more likely to progress to THA.

 

Larson and Giveans reported 10-month follow-up (3 months to 3 years) from a consecutive series of 96 patients (100 hips) who presented with FAI (typically with activity-related groin pain and pain with prolonged sitting) between 2004 and 2007 (Larson, 2008). Preoperative and intraoperative radiographs revealed Tonnis grade 0 to II osteoarthritis with cam impingement in 17 hips, pincer impingement in 28, and a combination of both in 55 hips. The average age was 35 (range, 16-64 years). Following FAI treatment, the impingement test was reported to be better in 86% of patients, with good to excellent results in 75% of patients. Three patients (3%) required THA, and 6 had heterotopic bone formation. Visual analogue scale (VAS) scores for pain improved from 6.7 at baseline to 1.9 at the 3-month to 3-year follow-up. Scores on the SF-12 improved from 60 to 78.

 

A mixed open/arthroscopic approach for treatment of FAI was reported by Laude et al in 97 patients (100 hips) (Laude, 2009).  This technique allows direct visualization of the anterior femoral head-neck junction without dislocation. All patients had a positive impingement test (pain reproduced in flexion, adduction, and internal rotation). All patients had MR arthrography or CT arthrography to analyze the labrum for tears. Nine patients had prior surgery and 3 patients had Tonnis grade 2 osteoarthritis. Thirty patients had grade 1 osteoarthritis. The average age of the patients was 33 (16–56 years old). Ninety-one (94%) were available for follow-up at an average 58 months (range, 29–104 months). Scores on the nonarthritic hip score (NAHS) increased from 55 at baseline to 84 at the last follow-up. One patient had a femoral neck fracture 3 weeks postoperatively, and 13 (14%) required revision due to persistent pain. In 8 of these patients, the damaged part of the labrum was removed and in 6 patients, osteochondroplasty of the head was performed to improve the groove at the head-neck junction. Another patient had heterotopic ossification. Eleven hips (12%) required THA at a mean of 40 months (range, 5–75 months). In the THA group, the acetabular lesions were deeper (10.9 mm vs. 6 mm) and a higher percentage of Beck grade 5 was found (54% vs. 7%). The best results were observed in patients younger than 40 years with a Tonnis grade of 0.

 

Bardakos and colleagues compared results from 24 patients treated with osteochondroplasty for cam impingement (after 2004) with 47 patients who showed cam impingement but had only the labrum repaired (between 2000 and 2004) (Bardakos, 2008).  The cohorts were matched for age (27–46 years) and for follow-up of 1 year. The number of patients who did not meet the selection criteria was not reported. There was a trend (p=0.11) for improved MHHS outcomes (excellent, good, fair, poor) in patients who were treated for impingement in addition to labral repair in this small study. Post hoc analysis of the percentage of patients in the excellent/good category showed significant improvement for the FAI-treated patients over historical controls (83% vs. 60%, p = 0.043). Results of this study should be interpreted cautiously due to multiple potential sources of bias, including selection bias, limited follow-up, and the small sample size.

 

Seven case series of patients with FAI treated with the open approach and dislocation were identified in the systematic review by Bedi et al (Bedi, 2008).  Two studies reported on 5 patients and 5 studies reported results from 19 to 52 patients, with follow-up ranging from 24 to 60 months. The 5 studies are briefly described here.

 

Beck et al reported outcomes from 19 patients (average age, 36,years; range, 21–52) of 22 who had been selected from their database with confirmed clinical, radiographic, and MR-arthrography diagnosis of FAI, had been treated with surgical dislocation of the hip, and had at least 4 years of follow-up (Beck, 2004).  Three patients were excluded based on a history of prior intra-articular surgery of the involved hip. Of the remaining 19, all had labral damage and 18 had acetabular damage. By 4 to 5 years’ follow-up, 5 patients (26%) had undergone THA, with the failures associated with cartilage damage. Thirteen patients (68%) were reported to have had good to excellent outcomes.

 

Another study selected 52 of 141 consecutive patients to compare the effect of reattaching or removing the labrum during treatment for FAI (Espinosa, 2006).  Patients were selected for age (20–40 years) and no prior surgery; all had preoperative evidence of acetabular damage. Patients were excluded from the study because of incomplete clinical or radiographic documentation (n = 48), open growth plates (n = 4), age of greater than 40 years (n = 29), previous hip surgery (n = 7), or participation in professional athletic activity (n = 1). Independent evaluations of 2-year follow-up indicated improved Merle d’Aubigne scores for both groups, from a baseline of 12 to 15 in the group in which the labrum was resected and from 12 to 17 in the group where the labrum was reattached. The study also found a reduction in progression to osteoarthritis if the labrum was reattached.

 

Peters et al reported on 29 patients (30 hips) in a prospective protocol with minimum 2-years’ follow-up (Peters, 2006).  The specific diagnoses were primary femoroacetabular impingement in 25 patients (26 hips), Legg-Calve-Perthes disease (n = 3), and slipped capital femoral epiphysis (n = 1). The average age of the patients was 31 years (range, 16–51 years). Twenty-nine of the 30 hips had either cam-type impingement (n = 14), or mixed cam and pincer-type impingement (n = 15). Eighteen hips were reported to have had severe cartilage damage that was not seen on MR arthrography. The Harris Hip score improved from 70 at baseline to 87 at an average 32-months’ follow-up. No progression to osteoarthritis was observed in 68% of patients. There was non-union in 8 hips (27%), 5 hips (17%) were expected to convert to THA due to progressive pain, and 4 (13%) had progression to osteoarthritis. Radiographic signs of progression of osteoarthritis and clinical failure requiring conversion to THA were seen only in patients with severe damage to the acetabular-articular cartilage. Two additional retrospective studies (n = 23 and 34) that included patients with severe cartilage damage reported that 50% to 70% of patients improved and 30% to 50% failed (either no improvement or underwent subsequent THA) following open osteochondroplasty with dislocation (Beaulé, 2007) (Murphy, 2004).  

 

Four prospective/consecutive case series with over 100 hips/patients treated for FAI have been published in the last year. These studies show a 20-point improvement on the MHHS at short to mid-term follow-up, indicating a change from marked pain with a serious limitation of activities pre-operatively to mild pain after treatment, or from moderate pain with some limitations of ordinary activity or work to slight or no pain after treatment. Given that the arthroscopic procedure was developed around 2004, long-term follow-up is limited.

 

What can be ascertained from the current literature:

 

What cannot be ascertained from the literature:

 

Based on 1) the intraoperatively established relationship between FAI morphology and damage to the acetabulum, 2) the consistent improvement in symptoms reported in large prospective case series, and 3) the potential for continued and irreparable cartilage damage if FAI pathology is not addressed, open or arthroscopic treatment of FAI meets Primary Coverage Criteria of effectiveness when specific criteria are met.  Because of the differing benefits and risks of open and arthroscopic approaches, patients should make an informed choice.

 

The evidence is insufficient to permit conclusions concerning the effect of this procedure on the development of osteoarthritis. Therefore, treatment of FAI morphology in the absence of symptoms does not meet Primary Coverage Criteria of effectiveness.

 

 

This update mostly focuses on FAI after slipped capital femoral epiphysis (SCFE).  Sink et al. reported a retrospective review from 2 U.S. centers on 36 patients (39 hips) with stable SCFE who were treated with open surgical hip dislocation for chronic symptoms (Sink, 2010).  The degree of slip was considered to be mild in 8, moderate in 19, and severe in 11 patients, and the average time between in situ pinning and surgical hip dislocation was 20 months (range 6-48 months). The majority of patients had partial or complete relief of symptoms immediately after initial pinning followed by a recurrence of symptoms that were consistent with impingement. All but one patient had either labral or cartilage injury, with labral injury observed in 34 of 39 hips and cartilage injury in 33 or 39 hips (5 grade 1, 10 Grade 2, 4 Grade 3, 10 Grade 4, and 4 Grade 5); the average depth of cartilage damage was 5 mm (range 2-10 mm). There was no correlation between slip severity or duration of symptoms and the type of cartilage injury.

 

Dodds et al. examined the prevalence of FAI in 36 patients (49 hips) who returned for clinical evaluation at an average 6 years after SCFE (Dodds, 2009).  There was no difference in the grade of slip between those patients who were available for follow-up and the total cohort treated for SCFE. The average age at presentation was 12.2 years, and at the time of evaluation all patients had reached skeletal maturity. Postoperative radiographs were reviewed for the grade of slip, Southwick slip angle, Loder’s classification of physeal stability, and the anterior head-neck offset (alpha) angle. Pain and impingement were found in 30% of the 30 hips with Grade 1 slips, 25% of the 8 hips with Grade 2 slips, and 0% of the 4 hips with Grade 3 slips. None of the radiographic factors including the grade of slip were predictive of subsequent impingement; the alpha angle was the most influential variable in regression analysis. Together, these results indicate that it is difficult to predict which patients with SCFE will develop FAI, but that all children should be followed into adulthood and monitored for impingement.

 

The literature search identified several articles from specialized centers on the treatment of symptomatic FAI in children with developmental hip disorders. The largest series on SCFE was a joint retrospective review from the Swiss group of Ganz and Leunig, together with the Children’s Hospital Boston, with 1- to 8-year follow-up on 40 patients (between 9 and 18 years of age) with moderate to severe SCFE who were treated by capital realignment with surgical dislocation (Ziebarth, 2009). The primary aim of the article was to determine whether this capital realignment technique was feasible and repeatable, and would restore hip anatomy and function while avoiding osteonecrosis. Dislocation was not performed in SCFE with a slip angle of less than 30 degrees, in which trimming of the anterior metaphysis was considered sufficient to restore the anterior offset without weakening the femoral neck. No patients from either institution developed osteonecrosis, infection, deep venous thrombosis or nerve palsies. Three patients developed delayed unions, none developed non-unions. Five patients required additional surgery for heterotopic ossification, residual impingement, or breakage of screw or wire fixation. The short-term postoperative clinical outcomes were found to be near normal, with similar scores between the operative and non-operative hips. Stability and the duration of symptoms of SCFE (1 day to 3 years) were associated with the severity of acetabular cartilage damage observed at the time of surgery.

 

From the same U.S. institution was a 2006 report of 19 patients (12 – 43 years of age) who underwent either femoral neck osteoplasty or osteoplasty with intertrochanteric osteotomy via Ganz-type surgical dislocation (Spencer, 2006). Out of 12 patients with a history of SCFE (12-38 years of age), 9 were found to be improved at 8-25 months follow-up. Out of the 7 patients (17-43 years of age) without SCFE who underwent open surgical dislocation for pistol grip deformities, 5 had worse symptoms or minimal relief. Outcomes for patients with a chondral flap were worse than for patients without a chondral flap. For example, function scores on the WOMAC improved from a baseline of 26 to 10 in patients without a chondral flap, but did not improve (25 to 24) in patients with chondral flap damage.

 

Due to the unclear balance of risks and benefits, questions regarding whether, when and how to treat symptomatic FAI in children with SCFE are difficult. Although the impact of not treating FAI is established, there is limited evidence on treatment outcomes in pediatric patients. The open dislocation procedure is technically demanding with a high risk of serious complications and has not been shown to be safe and effective outside of a few highly specialized centers. In addition, questions remain concerning selection criteria and the appropriate timing and approach for FAI treatment in patients with developmental hip disorders. In a 2009 review of SCFE, surgeons from Children’s Hospital Boston considered subcapital correction osteotomy with surgical dislocation to be an emerging treatment, stating that “Currently, we recommend that this type of treatment should be restricted to few select specialized centers until the availability of long-term results and outcome. Also, this type of treatment has a steep learning curve, and it is advised to learn this surgical technique at a specialized center” (Gholve, 2009).  Since this approach has not been shown to be safe and effective outside of a few specialized centers, surgical treatment of FAI in pediatric patients does not meet primary coverage criteria that there be scientific evidence of effectiveness.  

 

 

 

 

 

 

 

 

 

 

 

 

  

 

 

 

 

 

 

 

 

 

 

In 2016, Oner and colleagues conducted a retrospective study to determine the prevalence of FAI as an etiologic factor for OA in the hip joint among patients who had undergone THA (Oner, 2016). Radiographs of 1004 patients who had undergone THA between 2005 and 2010 were reviewed by 3 authors. Intra- and inter-observer consistencies were calculated. The predisposing etiologic factor leading to end-stage degenerative hip disease was undetermined in 26 of the radiographs. Among the remaining 978 patients, 99 patients were diagnosed with FAI by all 3 reviewers, 83 with a cam-type FAI and 16 with pincer-type FAI. Inter-observer agreement was high, with a contingency coefficient of 0.71 for the diagnosis of FAI between observers.

 

 

 

 

Nwachukwu and colleagues published a systematic review and meta-analysis comparing open with arthroscopic surgical techniques for the treatment of FAI (Nwachukwu, 2016). The literature search included studies published through October 2014, which had a mean follow-up of at least 3 years. Sixteen studies met inclusion criteria: 9 open surgical hip dislocation studies and 7 hip arthroscopy studies. Pooled cohort analyses were conducted on data from 600 hips and mean follow-up of 58 months from the open surgery studies and 1484 hips and a mean follow-up of 51 months from the arthroscopy studies. Conversion to total hip arthroplasty (THA) was the outcome endpoint, with an overall survival rate of 93% for patients undergoing open surgery and 90.5% for patients undergoing arthroscopy (p=0.06). The 12-Item Short Form, an instrument for measuring general health-related quality of life, showed significantly better scores among patients undergoing arthroscopy. Direct comparison of other outcomes was limited by outcome instrument heterogeneity. Both surgical techniques demonstrated favorable outcomes in their respective measuring systems.

 

 

A mini-open surgical technique performed on 118 patients with FAI was described by Chiron and colleagues (Chiron, 2012). Fifty-eight percent had cam-type impingement and 42% had mixed type impingement. Average follow-up was 2.2 years. NAHS scores, internal rotation, and alpha angles significantly improved following surgery. Eight revisions were performed, 2 patients experienced residual pain and eventually underwent TKA, and 2 progressed rapidly to OA.

 

 

 

 

 

 

The evidence for revision arthroscopic surgery for patients with residual FAI symptoms consists of 2 systematic reviews of observational studies. The observational studies, although low-quality, showed consistent favorable functional outcomes following revision surgery. Evidence for revision arthroscopic surgery for children consists of 1 observational study. Results show that children receiving revision surgery have comparable functional outcomes compared with children receiving primary arthroscopic surgery. Evidence is sufficient to determine the net health outcomes for the use of revision surgery on patients with residual FAI symptoms following primary surgery.