Hip arthroscopy has become an established procedure in adults with well-published results. These techniques have been extended to the teenage population, but the literature remains limited in regard to indications, efficacy, outcomes and complications specific to the paediatric population.
The purpose of this study was to report the short-term outcome of the arthroscopic treatment of cam-type femoroacetabular impingement (FAI) in the skeletally immature adolescent population with open growth plates, with the hypothesis that it is a safe procedure, with good clinical outcomes and low complication rates.
The study design of this article was a multicentre prospective longitudinal case series.
From 2008–2009, 34 consecutive patients who were 18 years or younger at the date of operation (41 hips) with open proximal femoral growth plates had arthroscopic hip surgery for treatment of cam-type FAI and associated lesions by use of the lateral decubitus position by two surgeons. Data were collected prospectively including the modified Harris hip score (MHHS), nonarthritic hip score (NAHS), satisfaction survey and complications. Mean follow-up was 14 months (range 1–2 years).
All hips were available for review. The mean age of patients was 15.7 years with the most common sport activity being Australian Football League, netball and rugby. All patients had cam-type impingement, with the most common associated lesions being acetabular rim lesions (82.9%) and pincer impingement (22%). The labrum underwent debridement in 31.7% of cases, and was repaired in 17.1%. The MHHS and NAHS improved in all patients post-operatively, 77.39–94.15 and 76.34–93.18, respectively. A percentage of 78.1 were able to return to full sporting activity and 88.2% were satisfied with the operation. There were no complications.
Using hip arthroscopy to treat cam-type impingement and associated lesions in the adolescent population, we confirmed our hypothesis with observed high satisfaction levels, return to sports, significant improvement in post-operative hip scores (MHHS and NAHS) and no complications.
Femoroacetabular impingement (FAI) is increasingly being recognized as a leading cause of hip pain in young active adults and a predisposing factor in the development of hip osteoarthritis.[1-3] Open and mini-open techniques have been described to treat FAI.[4, 5] More recently, modern hip arthroscopy techniques has resulted in good to excellent short-term outcomes and a relatively low complication rate in adults.[6-9]
There is increasing clinical evidence that surgical correction of FAI relieves pain and improves function, with proponents advocating early correction procedures to reduce the mechanical cause of articular cartilage damage.[4, 10-12] Hip arthroscopy is thus being cautiously extended to the skeletally immature paediatric population, but with understandable concerns regarding its indications, efficacy and possible disastrous complications including avascular necrosis (AVN), slipped upper femoral epiphysis (SUFE), fracture and growth plate arrest. However, the current literature on the subject remains limited.
The purpose of this study was to report the types of associated intra-articular hip pathologies present in the paediatric population with FAI and the short-term outcomes following arthroscopic hip surgery, including all complications. The hypothesis of this article is that it is a safe procedure in the adolescent population with good clinical outcomes.
Materials and methods
From 2008 to 2009, 41 hip arthroscopic procedures were undertaken in 34 consecutive patients with FAI who were 18 years old or younger with open proximal femoral growth plates as assessed on anteroposterior (AP) pelvis radiograph at the time of surgery. All patients with cam impingement, including those with combined cam and pincer were included. Patients with pure pincers, coxa profunda and protrusio were excluded. Prospective data were collected by two surgeons.
The indication for hip arthroscopy in all patients was recalcitrant hip pain that was not responsive to conservative treatment for at least 6 months. The majority of patients in this group were either unable to play or significantly restricted in their ability to play sport at state or high school level. The most common sporting activities included Australian Rules football, netball, rugby and hockey (Table 1). There were 29 males and 5 females ranging in age from 11–18 (mean age 15.7 years), with no age related difference in clinical presentation, pathology or surgical treatment.
Table 1. Patient demographics
Number of patients
Mean age (range)
Male : female
Number of hips
Left : right
Australian Rules football
All patients had positive pain provocation tests (flexion, adduction and internal rotation) and were investigated with plain radiographs of the hip (AP pelvis (Fig. 1) and lateral Dunn view (Fig. 2)), three-dimensional computed tomography scans and magnetic resonance imaging scans. Cam-type impingement was diagnosed when the α angle greater than 55° on a cross-table lateral radiograph or the femoral head sphericity was lost on computed tomography scan.
Hip arthroscopy was then undertaken with the patient in the lateral decubitus position, through the use of the McCarthy Hip Distractor (Innomed, Inc, Savannah, GA, USA) under general anaesthesia as described. Briefly, the leg was placed in traction with the leg in abduction and slight forward flexion. Two standard portals were used, a viewing portal just above the apex of the greater trochanter and an instrumentation portal 3–4 cm anteriorly. An additional anterior distal portal was established if labral repair was performed. An image intensifier was used to facilitate needle placement and minimize the risk of labral perforation while establishing portals.
An arthroscopic pump was used throughout to maintain constant distension of the joint with Ringer solution, at a pressure of 40–80 mmHg. A 70° arthroscope was used throughout the procedure in both the central and peripheral compartments.
Surgical technique included femoral head-neck reshaping for cam impingement and acetabular rim trimming for pincer impingement using a 5.5 mm burr with labral takedown. Resection of the juxta-articular bone continued until there was no evidence of impingement by the end of the procedure. The depth of resection depended on the shape and size of the cam deformity and the clearance of hip motion. Distally, the resection was completed to create a smooth contour with the anterior cortex of the femoral neck and to avoid notching. There was no alteration in surgical technique in comparison with adult surgery. The growth plate was often seen during the femoral head-neck osteochondroplasty, but no alteration in technique was performed nor was any specific procedure used to avoid growth plate damage (Fig. 3). An intraoperative dynamic evaluation was performed on each case to affirm adequate resection during movement of the hip to 100° flexion and maximum internal rotation. In addition, AP and Dunn view image intensifier radiographs were recorded and maintenance of the labral seal with the femoral head was checked.
Associated labral pathology was treated with either labral detachment and circumferential suture anchor refixation or debrided using a radiofrequency device. The decision to repair or debride the labral pathology was made at the time of arthroscopy. Debridement was performed on frayed, flapped or small labral tears, while all other labral pathology was repaired or refixed to the acetabular rim. No segmental labral resections were performed.
Articular cartilage defects of Outerbridge grades 1–3 were treated with chondroplasty. Tears of ligament teres were debrided using a radiofrequency probe (Vulcan Eflex Ablator Probe, Smith & Nephew, Andover, MA, USA). Microfracture was undertaken in patients with full-thickness cartilage loss (Outerbridge grade 4) at the chondrolabral junction in lesions up to 3 cm2.
Findings at the time of surgery were recorded. The procedure was performed either as a day case or overnight stay, and all patients underwent formal rehabilitation post-operatively using a standard protocol. There was no difference in the surgical technique used in our paediatric population in comparison with that used in adult hip arthroscopy.
Patients were assessed preoperatively and post-operatively with the modified Harris hip score (MHHS) and nonarthrititic hip score (NAHS). These scores were recorded post-operatively at 2 weeks, 6 weeks, 1 year and annually thereafter. A satisfaction survey was completed 1 year after surgery.
Preoperative and post-operative scores were compared using the paired samples t-test. Statistical analysis was performed using SPSS (version 19.0, IBM SPSS, Chicago, IL, USA) software packages. All tests were two-sided and assessed at the 5% level of significance.
All 34 consecutive patients were available for review. The mean duration of follow-up was 14 months (range 1–2 years). All hips had a minimum follow-up of 1 year. There were 21 left hips and 20 right hips affected.
All 41 hips had cam impingement with nine (22%) having an associated pincer impingement (Table 2). There were no patients with isolated pincer impingement. Labral pathology, which underwent debridement, was present in 13 hips (31.7%). Rim lesions (i.e. lesions of the articular cartilage found at the chondrolabral junction) were present in 34 hips (82.9%). These lesions were assessed using the method described by Haviv et al. Of these, 10 (24.4%) hips had cartilage softening at the edge of the chondrolabral junction (equivalent to grade 1 Outerbridge changes), 8 (19.5%) had less than 30% cartilage loss, 1 (2.2%) had greater than 30% cartilage loss and 15 (36.6%) had labral separation. The percentage of full-thickness articular cartilage loss was measured on a line from the acetabular edge to the acetabular fossa, and 30% articular cartilage loss approximates a 1-cm cartilage defect. Two patients underwent acetabular microfracture for grade 4 lesions. Six (14.6%) hips had partial tears of the ligamentum teres, which underwent radiofrequency debridement.
Table 2. Operative findings and details
Ligamentum teres tears
Acetabular chondral rim lesions
Cartilage softening at edge
<30% cartilage loss
>30% cartilage loss
Femoral head cartilage damage (chondromalacia)
There was no variation in pathological findings with age. None of the patients had femoral head chondral damage, extra-articular hip pathology, developmental dysplasia of the hip or Perthes disease.
Both the mean MHHS and NAHS showed statistically significant improvement improved by 16.76 and 16.84, respectively (Table 3).
Table 3. Preoperative and post-operative outcome scores
SD, standard deviation; NAHS, nonarthritic hip score; MHHS, modified Harris hip score.
Thirty-two of the 34 patients were actively engaged in competitive sporting activities. Of those, 25 (78.1%) were about to return to full sporting activity, 4 (12.5%) were able to return to sport but at a lower level, while 3 (8.8%) were unable to return to sport. The two patients who did not participate in sports were able to return to full activity. Overall, 30 of the 34 patients (88.2%) of the patients were satisfied and 31 (91.2%) would have the surgery again.
There were no cases of AVN, acute SUFE, physeal growth arrest, wound infections or any other complications. Also, none of the patients went on to subsequent surgery at the time of latest follow-up.
In this study, we describe our early experience of performing hip arthroscopy in the older juvenile and adolescent skeletally immature population. In our cohort of 34 patients, 41 hip arthroscopies were performed for the treatment of FAI and associated labral pathologies and ligamentum teres tears. None of the patients were treated for Perthes, hip dysplasia, AVN or SUFE.
Over the last 15 years, our understanding of the causes of early-onset hip arthritis in adolescents and young adults has evolved with the recognition and treatment of FAI and labral pathologies. There is also increasing recognition that symptoms of impingement may not arise until early adulthood when the articular cartilage and labrum may already have sustained significant damage.
In the adult population, open and arthroscopic procedures are aiming to restore normal hip biomechanics, and have been shown to correct the problem predictably and safely in the short and medium term.[2, 6-10] Indicating that treating impingement early may be beneficial to the longevity of the hip. For that reason, FAI ideally should also be suspected, diagnosed and treated in paediatric patients. In our study cohort, the paediatric patients presented with similar symptoms and signs to adults with FAI, with no variation in presentation and pathological findings with patient age. Preventing chondro-labral damage with the inevitable deterioration to osteoarthritis will require earlier diagnosis and treatment of the problem in children and adolescents. However, potential serious risks exist in the paediatric hip, which are not present once the physes close.
Treatment of FAI, whether open or arthroscopic, should have two aims: in the short term, the relief of preoperative symptoms and, in the longer term, the aim of prevention of repetitive injury to the articular cartilage and labrum, thereby reducing the risk of developing osteoarthritis.
The anatomy and pathophysiology of childhood hip disorders differs from adults. Conditions such as acetabular retroversion, hip dysplasia, Perthes disease and slipped upper femoral epiphyses lead to abnormal femoral head-neck offset and impingement. Such disorders are being treated early with osteotomies for dysplasia, containment for Perthes, early in situ pinning for SUFE, with the goal of keeping the femoral head spherical and well centred in the acetabulum, thus maintaining joint congruity and attempting to avoid the early onset of hip arthritis. In this study, none of the patients had a past history of predisposing childhood hip disorders.
Specific concerns regarding hip femoral osteochondroplasty procedures in children with open physes, include AVN, acute SUFE and proximal physeal growth arrest. Head-neck recontouring is primarily antero-lateral, sparing the critical posterior and posterosuperior ascending vessels, thereby reducing the risk of AVN.
Acute SUFE may potentially arise if the perichondrial ring is disrupted, while proximal physeal arrest may lead to femoral head-neck deformity. Fortunately, this risk is minimized in this patient cohort as most of the patients in this study present in the later teenage years when remaining growth is limited. None of these specific complications were seen in this group of patients.
The evolution of newer instruments and techniques in hip arthroscopy has increased the indications of hip arthroscopy in adults. In paediatrics, authors have described use in small series of congenital dislocation of the hip, Perthes disease, SUFE and neuropathic subluxation, chondroplasty of unstable osteochondral lesions, synovial biopsy and loose body removal.[20-23]
Kocher et al. reported on 54 hip arthroscopies on 42 patients 18 years old and younger, mainly for the treatment of isolated labral tears, Perthes and hip dysplasia with labral tear after prior periacetabular osteotomy. With a minimum 1-year follow-up, the MHHS improved overall and for each of the three main diagnoses.
Philippon et al. reported on 16 patients (mean age 15) who underwent hip arthroscopy for FAI with a mean follow-up time of 1.36 years (range 1–2 years). This study also reported excellent improvement in function and a high level of patient satisfaction in the short term.
Singh and O'Donnell reported on 24 Australian Football League (AFL) players with a mean age of 22 years who underwent hip arthroscopy with median follow-up of 22 months with comparable incidence of cam and rim lesions to our study. Both the mean MHHS (86–94) and NAHS (81–82) had improved at 12 months. Our paediatric population also had a large number of patients participating in AFL, with lower preoperative outcome scores, and higher overall improvement. Patients involved in sports such as AFL, netball and rugby, with repetitive kicking, twisting and side-to-side movements seem to be most as risk for rim and labral lesions.
As with any study, there are significant limitations in this study. Firstly, the inherent limitations with retrospective designs including possible selection bias, lack of concurrent control subjects and limited sample size. Other limitations include the lack of post-operative physical examination and post-operative radiographic data. There are also shortcomings with the outcome measurement tools used. The MHHS, although extensively used, is not a validated outcome tool. The NAHS has been validated for preoperative reproducibility, internal consistency and validity, but not for post-operative monitoring of changes. Furthermore, no scores have been validated for use in the paediatric population.
This study is also limited in short follow-up. However, we considered that 1 year is a reasonable time at which to assess the early results of symptomatic relief and complications. Longer-term follow-up data are currently being collected.
Arthroscopic treatment of FAI in the paediatric population is a new technique and whether it contributes to the prevention of arthritis will require long-term follow-up. Our initial application of these surgical methods has been cautious but with promising results. We have confirmed our hypothesis that the arthroscopy treatment of cam-type FAI and associated lesions is a safe reliable technique leading to improved short-term outcomes, high patient satisfaction and return to sport. The specific paediatric complications of osteonecrosis, growth arrest and acute SUFE were not encountered.