Abstract
Keywords
The pathomechanics of femoroacetabular impingement (FAI) as an important cause of early primary idiopathic osteoarthrosis was recently described by Ganz et al. 13 Structural abnormalities such as “pistol-grip deformity,” “tilt deformity,” and uncorrected pediatric hip deformities are well established. 25,34 By directly visualizing abnormal abutment of the femoral neck against the acetabular rim in more than 600 surgical dislocations of otherwise nondysplastic hips, Ganz et al. enabled the development of multiple open and arthroscopic techniques designed to relieve pain and prevent osteoarthritic progression. 22,24 Because FAI is a relatively new entity, the recent flourish of literature has offered a multitude of treatments with good short-term outcomes, but has yet to define a superior method or demonstrate lasting results.
There are 2 distinct categories of FAI: cam-type and pincer-type impingement. Both are caused by repeated abutment during supraphysiologic ranges of motion or pre-existing anatomical abberations. 13,20 Most patients, however, demonstrate aspects of both. Diagnosis of FAI is by clinical and radiographic examination (Table 1). Intra-articular hip injection with corticosteroids and a local anesthetic is a useful adjunct to confirm an intra-articular source of symptoms. 26 Magnetic resonance arthrography (MRA) with gadolinium in specifically designed sequences is widely used preoperatively to assess cartilage and labral tears. 21
Characteristics of Cam and Pincer-type Impingement 13
Nonoperative treatment of FAI typically includes rest, anti-inflammatory medication, and core strengthening. If symptoms persist, surgical treatment is recommended. The goal of surgery—whether arthroscopic or open—is to correct the cam and pincer deformities with either an osteoplasty of the femoral neck or a trimming of the acetabular rim. The torn labrum can be debrided or repaired. Any delaminated cartilage is debrided to a stable edge; if subchondral bone is exposed, some authors perform microfracture. The open technique is performed using a trochanteric flip osteotomy or a limited anterior approach combined with arthroscopy to address the labral and chondral injuries. 9,13 The total arthroscopic technique is performed in standard fashion for the hip and has been described elsewhere. 15,22,33 The complication rate is low; most large case series have reported no occurrences of femoral head necrosis or femoral neck fracture. 4,5,10,18,19,22,25,26,29 Sampson 32 reported 3 femoral neck fractures in more than 800 treated patients. The complication rate for hip arthroscopy for all indications is less than 1.5% and includes neurapraxia, lateral cutaneous femoral nerve numbness, postoperative portal bleeding, heterotopic ossification, infection, intra-articular instrument breakage, iatrogenic cartilage injury, avascular necrosis of the femoral head, femoral neck fracture, and intra-abdominal fluid extravasation. 17
Although there is a growing body of literature describing the diagnosis and treatment of FAI, there is no consensus on the efficacy of treatment. We therefore performed a systematic review to address 4 questions in regard to the treatment of FAI: (1) Does treatment for FAI succeed in improving symptoms? (2) In which subset of patients should treatment for FAI be avoided? (3) Is labral refixation superior to simple resection? (4) Does treatment for FAI alter the natural progression of osteoarthritis in this group of typically young patients?
Materials and Methods
We performed PubMed, MEDLINE, EMBASE, and CINAHL literature searches for all English language studies up to March 20, 2009 (Table 2). All studies, excluding isolated case series and expert reviews, that directly involved the treatment of FAI with either arthroscopy or open surgical dislocation and reported treatment outcome data were reviewed by 3 authors (V.Y.N., N.A., and X.P.) and the data from each case series was extracted by 2 authors (V.Y.N. and N.A.).
Search Strategy and Results From Major Search Engines
Twenty-three articles were included in this systematic review (Table 3). According to the system described by Wright et al, 36 1 study qualified as level II evidence, 23 2 as level III, 2,12 and 20 as level IV. ¶ The average length of follow-up ranged from 6 months to 5.2 years. The mean age for patients in all studies was 34.8 years (total of 970 cases; 608 male patients, 362 female patients; age range, 11-64 years).
Descriptions of Studies Included in Systematic Review
FAI, femoroacetabular impingement. C, cam; M, mixed lesions (pincer and cam); P, pincer.
An Excel worksheet (Microsoft, Redmond, Washington) was created to compile all relevant data. The pertinent details of these studies were juxtaposed in tabular form (Tables 3 through 5) to facilitate further analysis and review. Effect size, defined as the difference between the 2 means (M1 − M2) divided by the within-group standard deviation, was used to measure the magnitude of a treatment effect across different studies and measures. The accuracy of the estimated effect size depended on the sample size. The sampling error of the estimated effect size was quantified using its 95% confidence interval. If the confidence interval of the estimated effect size included zero, it was interpreted that the difference between the 2 groups was not significantly different from zero.
Summary of Patient Outcomes and Labral Treatment Methods a
CI, confidence interval; HHS, Harris hip score; WOMAC, Western Ontario and McMaster Osteoarthritic Index; UCLA, University of California at Los Angeles shoulder score; SF12, SF-12 Health Survey; THA, total hip arthroplasty; NA, not available; MHHS, modified HHS; VAS, visual analog scale; NAHS, nonarthritic hip score; HOS, hip outcome score; ADL, activities of daily living.
Effect size was calculated using the standard deviation given or estimated from similar study.
Summary of Cartilage Treatment Approaches and Progression of Osteoarthritis
Grade 0 (no signs of osteoarthritis [OA]), I (increased sclerosis, slight joint-space narrowing, no/slight loss of head sphericity), II (small cysts, moderate joint-space narrowing, moderate loss of head sphericity), III (large cysts, severe joint-space narrowing, severe deformity of head).
Grade I (softening, swelling of cartilage), II (partial-thickness fissuring), III (full-thickness fissuring), IV (exposed subchondral bone).
Classification scheme not specified.
Results
Multiple different outcome scores were used, including the Western Ontario and McMaster Osteoarthritic Index (WOMAC), the Harris hip score (HHS), the modified HHS (which includes only the pain and function portion of the original HHS), the visual analog scale (VAS), the SF-12 Health Survey (SF-12), the nonarthritic hip score (NAHS), and the Merle d’Aubigné hip score. The reported outcome scores improved after treatment for FAI in all studies, and the effect size was significant for improvement in patient outcome (confidence interval larger than zero) (Table 4).
Although treatment for FAI consistently improved mean hip function and severity of symptoms as measured by multiple clinical outcome scoring systems, patient satisfaction was not universally positive. Although Stahelin et al 33 reported increased flexion and internal rotation and Eijer et al 10 found improvement in general range of motion, Espinosa et al 12 and Beck et al 5 reported no significant change. Postoperative patient reporting of pain was documented in most studies. Mean improvement in pain ranged from 25.1% to 100%, and patient dissatisfaction rates and patients with no improvement of symptoms ranged from 0% to 31.2% (Table 6). Rates ranged from 0% to 30% among the studies documenting patients with eventual conversion to total hip arthroplasty (THA) at most recent follow-up. 5,22,25,26,30
Summary of Pain / Satisfaction Rates a
HHS, Harris hip score; VAS, visual analog scale; WOMAC, Western Ontario and McMaster Osteoarthritic Index. Includes studies that specifically reported pain symptoms or patient satisfaction.
Developing indications for treatment of FAI and identifying which patients do not perform well with arthroscopic or open surgical debridement are important for future clinical use of these techniques. Beaule et al 4 reported 5 of 6 patients with an unsatisfactory outcome had substantial delamination of articular cartilage (Beck type 4) at the time of the index operation, but they were unable to correlate cartilage damage to clinical outcome because many patients with good outcomes had similarly severe damage. All 3 hips that underwent subsequent THA in the report by Larson and Giveans 22 had Outerbridge grade IV chondral delamination greater than 2 cm at the time of surgery that was associated with Tonnis grade I or II on preoperative radiographs. Peters and Erickson 26 noted that severe damage (Outerbridge grade III or IV) to the acetabular cartilage that was not fully evident on preoperative MRA or MRI was seen intraoperatively in each of their 4 patients with failed results. In the series by Beck et al, 5 2 of the 3 hips that showed mild to no preoperative radiographic changes that had deteriorated and required THA had extensive cleavage cartilage lesions on the intraoperative examination. Two additional hips had Tonnis grade II changes at the time of surgery and required THA. Murphy et al 25 reported 3 early failures; all of the patients had Tonnis grade II or III osteoarthritis preoperatively as well as additional hip pathology, such as a circumferential osteophyte or untreated acetabular dysplasia. Stahelin et al 33 reported significantly poorer results in pain relief, NAHS, range of motion, and clinical examination findings for impingement in patients with preoperative Tonnis grade I or II osteoarthritis compared with those with no signs of degenerative joint disease on radiographs.
In the context of these results, it should also be considered that not all patients with advanced chondral lesions have poor outcomes. Philippon et al 27 performed a variety of arthroscopic microfracture procedures, thermal chondroplasty, or no treatment of cartilage disease in 45 professional athletes (21 of whom had Outerbridge grade IV acetabular chondral defects) and reported an overall 93% (42 of 45) return rate to competitive sport. Laude et al 23 noted that older patients (40.3 vs 32 years) were more likely to undergo THA. Philippon et al 28 found that poor overall cartilage, a longer time from onset of symptoms to surgery, and lower joint-space measurements correlated with lower postoperative Harris hip score.
Although multiple studies involved either the debridement of labral defects or reattachment of the intact peripheral portion of the labrum following osseous rim resection, only 2 studies examined the effect of labral refixation on patient outcome. 12,23 In the study by Espinosa et al, 12 2 groups of patients under 40 years old with no significant preoperative differences in chondral lesions, pain, radiographic signs of osteoarthritis, or Merle d’Aubigné scores were compared as a retrospective case series. Patients who underwent labral resection had significantly worse clinical outcomes at both 1- and 2-year follow-up than those who received labral refixation. At 2 years, the results of the former group were 28% excellent, 48% good, 20% moderate, and 4% poor, while the latter results were 80% excellent, 14% good, 6% moderate, and 0% poor. In addition, the average pain score closely correlated with clinical outcome and improved with labral refixation by 73% compared with 59% with simple labral resection. Laude et al 23 found no significant difference in outcome comparing patients with and without labral refixation, although there was a tendency for higher NAHS scores in patients with labral refixation. Nevertheless, the authors thought that failure of labral refixation in 8 patients from early weightbearing may have been the cause of persistent pain necessitating revision surgery. 23
Many studies recorded the extent of preoperative osteoarthritis, but only 8 reported postoperative findings. Rates of progression of osteoarthritis ranged from 0% to 33.3% (Table 7). Espinosa et al 11 found that labral resection was associated with more severe osteoarthritic change compared with labral refixation, and Peters and Erickson 26 reported a high incidence of grade IV cartilage damage in those patients with osteoarthritic progression.
Summary of Radiologic Osteoarthritis (OA) Progression a
HHS, Harris hip score. Includes studies that specifically report postoperative OA findings.
Discussion
Femoroacetabular impingement as the cause of primary idiopathic osteoarthritis in young patients is a relatively new concept. Therapeutic techniques, although well described, are still in their formative stages. This systematic review of available literature was designed to address 4 clinically important questions: First, does treatment for FAI succeed in improving patient symptoms? By both outcome scores and effect size, both arthroscopic and open treatment methods for FAI significantly improved mean postoperative scores. Second, in which patients should standard treatment for FAI be avoided? Although mean outcome measures invariably improve, not all patients have satisfactory results. Between 0% and 30% of patients eventually required THA, 5,22,25,26,30 and 0% to 31.2% of patients were dissatisfied with the procedure or had no improvement of their pain. Outerbridge grade III or IV cartilage injury seen intraoperatively and osteoarthritis greater than Tonnis grade I on preoperative radiographs were associated with poorer outcomes. Third, is labral refixation superior to labral resection? The studies examined varied in their treatment of labral injury. Only Espinosa et al 12 and Laude et al 23 directly compared labral refixation with simple resection and debridement. The former study found that hip function scores, pain scores, and delay of radiographic progression of osteoarthritis improved more with labral refixation. The latter study demonstrated no significant difference in outcome, but did note a tendency for higher outcome scores with labral refixation. Last, does treatment for FAI affect the progression of osteoarthritis? Because all case series were relatively recent and degenerative joint disease is generally a gradual process, it was difficult to elucidate the effect of surgical correction of FAI on the progression of osteoarthritis with radiographic imaging. In the studies that reported osteoarthritic radiographic change, only a minority of the patients had progression of their disease.
There remain many gaps in the available literature reporting treatment of FAI and most studies have relatively limited follow-up. In addition to a lack of high level of evidence studies, there has yet to be a direct comparison between open and arthroscopic methods. When treatment of FAI was compared with simple arthroscopic labral debridement and subchondral microfracture, Bardakos et al 2 found excision of the impingement lesion produced higher postoperative scores and more good to excellent results. Bedi et al 6 examined the outcomes of open and arthroscopic treatment of labral tears and FAI in a systematic review and found insufficient evidence to support open surgical dislocation, the historic gold standard, as a superior procedure.
The findings here are in agreement with the general conclusions by Bedi et al 6 that both open and arthroscopic procedures can achieve satisfactory clinical outcomes. The varying outcome measures used across the literature provide a challenge in assessing the true value of each technique in reducing patient symptoms. The optimal management of severe acetabular cartilage lesions in the context of FAI is unknown. Patients with advanced osteoarthritic change as shown on preoperative radiographs will likely benefit less from treatment of FAI. The true extent of chondral lesions is difficult to assess preoperatively, even with MRA. Magnetic resonance imaging has been shown to be only 22% sensitive for cartilage delamination. 1 Identifying these patients and developing definite contraindications to the treatment of FAI will be necessary to significantly improve success rates.
There is now a general consensus that labral preservation and restoration are beneficial; this shift is evident in the comparative study by Espinosa et al 12 in which their 2 cohorts were a result of evolving methods rather than intentional randomization. Studies with higher level of evidence are needed before labral refixation is likely to be widely adopted. The longest mean follow-up, 5.2 years, was reported by Murphy et al 25 in a study that included techniques now considered suboptimal. 26 Clearly, longer-term results with large cohorts are needed to formulate specific evidence-based recommendations and to determine whether treatment for FAI alters its natural course. Bardakos and Villar 3 found that 65% of patients with an untreated pistol-grip deformity and Tonnis grade I or II degenerative changes had progression of osteoarthritis after 10 years.
Early evidence in the treatment of FAI has shown that hip function and pain are improved in the majority of patients. Advanced chondral lesions and osteoarthritic change are associated with worse patient outcome, but the full extent of cartilage defects is difficult to assess preoperatively. Preliminary results support labral refixation, and it is too soon to conclude whether treatment of FAI can delay or halt the progression of osteoarthritis.
An online CME course associated with this article is available for 1 AMA PRA Category 1 Credit™ at http://ajsm-cme.sagepub.com. In accordance with the standards of the Accreditation Council for Continuing Medical Education (ACCME), it is the policy of The American Orthopaedic Society for Sports Medicine that authors, editors, and planners disclose to the learners all financial relationships during the past 12 months with any commercial interest (A ‘commercial interest’ is any entity producing, marketing, re-selling, or distributing health care goods or services consumed by, or used on, patients). Any and all disclosures are provided in the online journal CME area which is provided to all participants before they actually take the CME activity. In accordance with AOSSM policy, authors, editors, and planners’ participation in this educational activity will be predicated upon timely submission and review of AOSSM disclosure. Noncompliance will result in an author/editor or planner to be stricken from participating in this CME activity.
