current management of femoro-acetabular impingement mx femoracetabular impingement1.pdfence...

Current Orthopaedics (2008) 22, 300e310

ava i lab le a t www.sc iencedi rec t .com

journa l homepage : www.e lsev ie r . com/ loca te /cuor

HIP

Current management of femoro-acetabularimpingement

M. Hossain*, J.G. Andrew

Department of Trauma and Orthopaedic Surgery, Penrhosgarnedd Road, Ysbyty, Gwynedd, Bangor, LL57 2PW, UK

KEYWORDSFemoro-acetabularimpingement;Hip impingement;Cam impingement;Pincer impingement;Acetabularretroversion;Labral tear;Hip debridement

* Corresponding author.E-mail address: munierh@doctors.

0268-0890/$ - see front matter ª 200doi:10.1016/j.cuor.2008.07.011

Summary

Femoro-acetabular impingement (FAI) is a common cause of hip pain in young adults. The condi-tion can result in labral tears, articular cartilage lesions and eventual osteoarthritis of the hip.FAI results from an abnormal bony shape of the femoral head / neck or the acetabular rim orboth, which causes contact (impingement) between the neck and acetabular rim during move-ment of the hip joint. Two different mechanisms are described, although a combination of thetwo mechanisms is frequent. Cam impingement is caused by an abnormality of the femoral headand neck; principally a reduction in offset between the head and the neck. Pincer impingementis caused by an abnormality on the acetabular side, with either excessive retroversion of theacetabulum or an unduly prominent anterior wall. Either problem is associated with the devel-opment of chondral lesions (especially in the acetabulum) and labral pathology.

Patients with FAI usually present with deep groin pain exacerbated by hip flexion. X-rays typi-cally show an anterior impingement bump on the anterior femoral neck on a horizontal beamlateral, abnormalities of head/neck offset or an excessively prominent anterior acetabularwall. Surgery is the treatment of choice and this involves open or arthroscopic bony resectionto improve femoral head-neck clearance with either resection or refixation of the damagedlabrum. Both the femoral head/neck junction and the acetabular rim may require bony resec-tion. Such surgery yields good symptomatic relief, but whether the surgery prevents the devel-opment of osteoarthritis in the hip in the long term is currently unknown.ª 2008 Elsevier Ltd. All rights reserved.

Introduction

Hip pain in young adults can present a diagnostic challenge.Some patients will have a clear history of recognised pre-disposing factors. Such factors include the various classic

org.uk (M. Hossain).

8 Elsevier Ltd. All rights reserved

forms of developmental dysplasia of the hip, causes ofavascular necrosis or a history of trauma. However, someshape abnormalities of the hip have been recognised formany years as being associated with hip pain (eg the‘‘bullet shaped’’ femoral head and minor slip deformitieson X ray).1 Up until recently there was no systematicunderstanding of the importance of these appearances, andno understanding of how they might be treated. Recently,principally due to the work of Ganz and his group, our

.

mailto:[email protected]

http://www.elsevier.com/locate/cuor

Table 2 FAI in numbers

� 1.5 cm: the distance between the anterior and posteriormargin of the acetabulum. Increased distance issuggestive of increased anteversion, reduced distance issuggestive of retroversion.� 11.6� 0.7 mm: normal anterior offset, vs. 7.2� 0.7 mm

in cam malformation� <8 mm: femoral head-neck offset in cam impingement� 65

�e74

�: a angle in FAI patients vs. 42

�e47

�in controls

� �20�e25

�: residual lateral centre-edge angle to aim for

post-surgery� 30%: maximum circumference of antero-lateral quadrant

of head-neck junction that can be removed without riskof fracture� <0.18:offset ratio in cam impingement

Current management of femoro-acetabular 301

understanding of shape abnormalities has improvedsubstantially.24 The condition of impingement at the hipmargin (femoro-acetabular impingement) has beendescribed and characterised. Femoro-acetabular impinge-ment (FAI) Table 1 is now recognised as one of the commoncauses of hip pain in young adults (Table 1), many of whomdo not have the classical features of hip dysplasia. It is nowpossible to identify and treat many patients with thiscondition (Table 2).

We will begin by reviewing the case history of a typicalpatient, who was initially referred to physiotherapy triageand then on to the orthopaedic clinic after he failed tomake adequate progress with conservative treatment.

A 40-year-old fireman was referred by the physiotherapistwith a long history of left groin pain. He also had pain in theleft sacroiliac joint and the greater trochanteric area. It wasfelt initially that his problems were related to mechanicalback pain. Although back rehabilitation improved thesymptoms at all sites, he still had significant groin pain. Thepain would start after a period of strenuous cycling. It wasparticularly present on prolonged sitting and when he waspushing with his left leg. He reported occasional giving wayof the left hip. The hip had a full range of movement butthere was a positive impingement test on deep flexion andadduction. This test also reproduced the pain about which hecomplained, in both site and character. He also complainedof clicking of the involved hip but had a negative psoas snaptest. His x-rays showed a positive cross-over sign in the lefthip and a fleck of labral ossification on the AP view, witha small impingement bump on a cross table lateral view.A diagnosis of FAI was made.

Table 1 Causes of femoro-acetabular impingement

Cam impingement:IdiopathicDevelopmental� Non-spherical femoral head� Coxa vara

Traumatic:� Malunited femoral neck fracture� Post-traumatic retro-torsion of femoral head

Childhood disease:� Perthes’ disease� Slipped capital femoral epiphysis

Iatrogenic� femoral osteotomy

Pincer impingement:IdiopathicDevelopmental� Retroverted acetabulum� Coxa profunda� Coxa vara� Os acetabuli� Protrusio acetabuli� Chronic residual dysplasia of acetabulum

Traumatic� Post-traumatic deformity

Iatrogenic� Over-correction for retroversion in dysplastic hips

This case history illustrates some of the issues arounddiagnosing this disorder. The patient initially presented withsymptoms in a fairly widespread area, including the sacro-iliac joint, and had some mechanical symptoms. His painwas worse with sports. He had features of FAI on X-ray, andhis symptoms have been significantly improved by surgery.

Historical perspective

When assessing a series of patients after periacetabularosteotomy for hip dysplasia, Reinhold Ganz’s group fromBerne observed a sub-group of patients who returned withsymptoms of anterior FAI and progressive pain after whatappeared to be a successful acetabular re-orientationprocedure.2 Subsequently, this group was the first todescribe ‘‘femoro-acetabular impingement ’’as a dynamicpathology where mechanical abutment between the prox-imal femur and the acetabular rim during terminal range ofhip movement resulted in labral and chondral damage.24,25

Other researchers had long suspected that pathologicalterations of bony morphology might have been a contrib-uting factor behind ‘‘idiopathic’’ primary hip osteoarthritis.Stulberg et al was one of the first to recognise that anabnormal morphology of the proximal femur predisposed todegenerative joint disease. His team introduced the term‘‘pistol grip deformity’’ to describe abnormal femoralhead-neck clearance in patients with primary hip osteoar-thritis.3 Harris also subsequently demonstrated the pres-ence of subclinical childhood disease and abnormal bonymorphology in 90% of patients with idiopathic primaryosteoarthritis of the hip.4 There is now growing evidencethat abnormal hip morphology precedes hip osteoarthritis.5

Ganz and his group have developed methods for treatingFAI, principally using open surgical dislocation to exposethe hip. This followed detailed work to examine thefemoral head blood supply, permitting dislocation withminimal risk of subsequent avascular necrosis.26 This openapproach has yielded successful results with good painrelief and improvement of function, and was extremelyhelpful in confirming and understanding the pathology ofFAI. However, open debridement is a substantial surgicalundertaking and other surgeons are now exploring botharthroscopically assisted and arthroscopic hip debride-ment. These latter approaches are less invasive but all

302 M. Hossain, J.G. Andrew

three techniques still require detailed assessment ofoutcomes and complications.

The anatomy of the hip

In broad outline, the hip may be described as consisting of:

- the femoral head, which forms slightly more than halfof a sphere,

- the cylindrical femoral neck which attaches the head tothe femoral shaft, and

- the cup shaped acetabulum.

However, this does not accurately describe the hip indetail. The femoral head is actually a conchoid shape. Theneck is of variable shape but usually narrowest at the mid-part and widest laterally. The ratio between the width ofthe neck and the diameter of the head varies considerably.Furthermore, the head may be placed centrally on theneck (so the centre of the femoral head lies on thegeometrical axis of the femoral neck in both AP and lateralplanes), or it may be displaced (typically so the centre ofthe head lies inferior and particularly posterior to the neckaxis). The upper margin of the neck is usually slightlyconcave. The head-neck clearance (both anteriorly andsuperiorly) allows free motion of the hip withoutimpingement. The femoral head normally appears asa section of a circle in both AP and lateral views and thehead-neck offset is normally more than 9 mm.27 Thearticular surfaces are not completely congruent. Articu-lating surfaces are covered by articular cartilage, which isthickest antero-superiorly in the acetabulum and antero-laterally on the femoral head.

The fibrocartilaginous acetabular labrum is triangular incross-section. The base is attached to the acetabular rimand the free apex overlies the femoral head. The centralarm of the labrum is covered by articular cartilage, which iscontinuous with the cartilage covering the lunate surface.The function of the acetabular labrum is controversial; itappears to act as a shock absorber and also as a fluid sealfor the hip joint to improve lubrication. It also has noci-ceptive and proprioceptive functions.6 The fluid sealprevents fluid flow in and out of the joint and enhancesjoint stability. Disruption of the fluid seal may make thejoint relatively unstable.7 The acetabulum has been rec-ognised to vary in orientation for many years, and devel-opmental hip dysplasia has been described in detail.Recently, several factors predisposing to FAI have beendescribed in the acetabulum, X-rays of which would typi-cally be dismissed as normal. These include acetabularretroversion, coxa profunda and minor degrees of protrusioacetabuli.

Impingement: a dynamic problem

Both the hip and the shoulder joints are ball and socketjoints with three degrees of freedom. The shoulder enjoysgreater mobility than the hip but has a higher risk ofdislocation. The hip joint has to support the body’s weightand also propel the limb forward. There is accordinglygreater bony constraint in the hip joint, and the

acetabulum is much deeper than the glenoid cavity.However, this anatomical arrangement risks significantcoaption between the articulating bony parts at theextremes of movement. Essentially, all hips may experi-ence femoro-acetabular impingement at the extremes ofmovement. It is more likely to be a problem if the anatomyof the hip leads to frequent impingement during activity.This depends both on the patient’s level of activity,including range of hip movement, and their hip anatomy.

In order to understand the mechanism of FAI it isimportant to consider the geometry of the hip joint. Theacetabular axis forms an angle of 35

�to 40

�with the hori-

zontal axis. The upper part of the acetabulum thereforeoverhangs the femoral head laterally on erect posture. Theacetabulum is anteverted 15

�to 25

�in normal individuals,

when measured on CT scan at the mid-part of the femoralhead. Both the anterior and posterior margins of theacetabulum overlie the femoral head.

A maintained space between the femoral head-neckjunction and the acetabular rim is essential for an unhin-dered range of hip motion. This requires that there is anoffset between the anterior and superior surfaces of thefemoral neck and the corresponding area of the femoralhead. It is worth emphasising that this femoral neck offsetdoes not describe the head/shaft offset, which is used todescribe hip replacement prostheses. A deviation fromnormal femoro-acetabular spatial relationships may notaffect the mid-range of motion but in the terminal range ofmotion the joint will be vulnerable to mechanical blockage(Fig. 1). In ‘‘pistol grip deformity’’ the normal concavity ofthe femoral head-neck junction is lost. The femoral headextends laterally in a convex shape. Femoral neck offset isreduced in the antero-lateral head-neck junction. Thesepatients also have reduced femoral anteversion. Loss ofsphericity or offset will result in mechanical blockagebetween the femoral head and the acetabular rim interminal range of motion.

Alternatively, excessive acetabular coverage or retro-version of the cup will also affect normal spatial relation-ships between the femoral head and acetabulum.Retroversion of the acetabulum results in antero-lateralover-coverage; the acetabular edge will restrict flexion ofthe femoral neck (Fig. 2). Coxa profunda will increase therelative depth of the acetabulum and predispose tomechanical blockade (coxa profunda is present when thefloor of the acetabular fossa touches or overlaps the ilio-ischial line, cf. protrusio acetabuli, where the femoral headline crosses the ilio-ischial line).28 Ito et al found that thesite of acetabular abnormality was mainly in the supero-medial region of the acetabular rim.8 Finally, patients withnormal hip morphology but excessive range of hip move-ment could also be susceptible to FAI in terminal range ofhip movement due to mechanical abutment.9 Such patientsinclude those with excessive laxity or those engaging inactivities such as dance.

Natural history of FAI

Two distinct mechanisms of femoro-acetabular impinge-ment have been described, although both are present inmost patients. The details of the pathological changes

Figure 2 AP X-ray of the pelvis showing acetabular retro-version with a positive cross-over sign. AW denotes the anteriorwall and PW denotes the posterior wall.

Figure 3 Aspherical femoral head in cam impingement. Arrowsdenote the abnormal extension of the epiphyseal scar. ª AmericanJournal of Radiology, reproduced with kind permission.

Figure 1 Top two images show the normal bony configurationof the hip joint, with sufficient joint clearance. The centre twoimages show pincer impingement due to excessive acetabularover-coverage. The bottom two images show cam impinge-ment, with an aspherical portion of the femoral headeneckjunction getting jammed into the acetabulum. ª AmericanJournal of Radiology, reproduced with kind permission.


caused in the hip may be distinct in the two types ofimpingement, but because of the frequent coexistence ofthe two types it is somewhat difficult to be certain aboutthis. Cam impingement is caused by abnormalities on thefemoral side (Fig. 3). Initially, the abnormality may arisedue to a minor slip of the capital femoral epiphysis(although the evidence for this is weak e it is simply thatthe imaging appearances are strikingly similar to thosefollowing a slip), or an abnormal fusion of the head andtrochanteric epiphyses at the antero-superior border of theneck. Due to absence or loss of the normal femoral head-neck offset, the femoral neck abuts the acetabular rim inhip flexion and internal rotation. Repeated impingementover a period of years leads to reactive new bone formation(‘‘impingement bump’’) on the anterior and superiorfemoral neck, which in turn further reduces the offset andexacerbates the impingement (Fig. 4). This abutment maycause damage to the labrum, including labral tears ordetachment. There is evidence that synovial fluid is pum-ped during this process, with paralabral cysts sometimesoccurring (readily visible on MR arthrogram). At the sametime, a delaminating chondral lesion may occur adjacent tothe antero-superior labrum, where an area of articularcartilage becomes detached from both the acetabular

labrum and the underlying bone. The delaminating lesion ofthe articular cartilage leads to a posteriorly based chondralflap. It is thought that this articular cartilage lesion occursprincipally due to shear from the area of the impingement

Figure 4 Impingement bump (white arrow) readily visible onMRI.

Figure 5 Labral ossification (white arrow) and osseousapposition of the acetabular rim (black arrows) in pincerimpingement. ª American Journal of Radiology, reproducedwith kind permission.


bump, which is out of round. Labral pathology is thought tobe less common in pure cam impingement than in purepincer impingement, and the delamination lesion mayoccur with a normal labrum.

Pincer impingement occurs because the arc of availablemotion in the acetabular opening is too small, or because itis oriented in an unsatisfactory fashion (Fig. 1). Theavailable arc will be reduced in coxa profunda and inprotrusio. Conversely, the situation with relative retro-version of the acetabulum is that the available arc ofmotion is biased towards extension and flexion, whichcause impingement. Persistent abutment between thefemoral head-neck junction and the acetabular rim or thelabrum in pincer impingement results in labral damage,intra-substance ganglion formation and labral degenera-tion, and may cause subsequent labral ossification on theanterior or lateral aspect of the joint (Fig. 5). Labralossification will increase over-coverage of the cup andcontribute to worsen the impingement (Fig. 6). Unrelievedimpingement will eventually lever the femoral headposteriorly, resulting in a contre-coup cartilage lesion overthe postero-medial aspect of the femoral head and thepostero-inferior aspect of the acetabulum. Althoughcircumferential, the initial articular cartilage lesion inpincer impingement is limited, as the mechanical contactbetween the femoral neck and the acetabulum is linear.Cartilage loss is also more superficial (mean cartilage loss4 mm, compared to a mean loss of 11 mm in camimpingement).28 Regardless of the type of impingementthe initial zone of impingement is the antero-superiorquadrant of the acetabulum.10 Initial impingement isanterior but with chronicity posterior impingement alsodevelops and is heralded by the presence of postero-infe-rior traction osteophytes. Isolated cam type impingment ismore destructive although less symptomatic. This ispossibly because of more severe damage to the labrum inpincer type impingement, and the labrum has nociceptivefibres. Cam FAI produces antero-superior OA while pincerFAI produces postero-inferior or central OA.29

Who is at risk and when?

The true incidence of FAI is not known. A recent Scandi-navian study found a 6% incidence of cam malformation inmen and 2% in women on radiological assessment ofstandardised AP hip radiographs of patients randomlyselected from the Copenhagen Osteoarthritis study.30 Camtype impingement is said to be more common in youngactive males (M:F 14:1, age range 21e51, mean 32) andpincer type impingement commoner in middle-aged activefemales (M:F 1:3, age range 40e57, mean 40).28 However,most patients actually have a mixed type of impingement.Beck et al reported less than 30% incidence of isolatedtype of impingement in their large series. Stulberg et alreported a 10% incidence of pistol grip deformity in theirhistorical series of patients with frank hip arthritis.3 Theincidence of acetabular retroversion, a recognised causeof pincer type impingement, in published series varies from5% to 20%.11

Clinical presentation of FAI

A young or middle aged patient typically presents with deepgroin pain, frequently exacerbated by athletic activitiesthat demand deep hip flexion, or by prolonged walking,sitting or driving.12 These symptoms overlap with those oflabral tears where patients usually present with an insidiousonset of deep groin pain, which is aggravated by walking,pivoting on the affected side, impact activities or pro-longed sitting.13 Mechanical symptoms from the hip such aspainful locking or giving way are common presentingfeature if a labral tear is present. It is now recognised thatmany patients previously diagnosed to have an isolatedlabral tear may actually have had unrecognised FAI.31

Figure 6 Acetabular over-coverage resulting in pincer impingement and joint degeneration.


Under these circumstances the labral lesion was simply partof a more extensive problem, and resection of the labrallesion alone frequently led to disappointing results. Gaitwill not be affected but range of motion of the hip maybe restricted, especially flexion, internal rotation andadduction.10

The impingement test is usually positive. The hipimpingement test is analogous to the shoulder impingementtest where the articulating surfaces are forcefully jammedagainst each other to reproduce the abnormal contact andresulting pain. Anterior impingement is much more commonthan posterior impingement, but both should be sought.With the patient supine, the hip and knee of the affectedside are passively flexed to 90

�and the hip then further

flexed, adducted and internally rotated. The combinationof deep flexion, adduction and internal rotation jams theedge of the femoral head/neck junction against theacetabular rim. The test is positive if

1. it elicits pain, and2. this pain is similar to that complained of by the patient.

A positive provocation test indicates irritation of thenociceptors of the damaged acetabular rim.

Posterior impingement usually occurs late in impinge-ment, with development of a postero-inferior osteophyteon the acetabular rim. To demonstrate postero-inferiorimpingement, the examiner hyperextends the lower limb ofthe affected side passively by hanging the leg over the endof the couch, while maintaining the opposite limb ina neutral position. The test is positive if passive externalrotation of the extended leg produces pain.

Investigation of FAI

Radiographs may be normal or changes may be subtle. BothAP and lateral radiographs of the hip are essential. The APfilm must be standardised and must avoid malrotation orpelvic tilt to be of value. It should be realised that pelvicflexion (tilt) will automatically increase the prominence of

the anterior acetabular wall and result in a positive cross-over sign. The criteria for an acceptable X-ray have beendefined.28 Neither the AP nor lateral views adequatelyvisualise the shape of the antero-superior femoral neck,and some groups now use CT scans in a plane radial to thefemoral neck to assess prominence in this area. Of the twoplain film views, the lateral view is more valuable and mayshow milder deformities not apparent on the AP view, andmay show the presence of an impingement bump (Fig. 7).The three lateral views found useful by different investi-gators include: the Dunn view in 45

�or 90

�hip flexion-

neutral rotation�20�

abduction, cross-table lateral viewwith 15

�internal rotation of femur, and frog lateral

view.14e16

The radiographic features of impingement Table 3include: a short femoral neck, a non-spherical butcongruent femoral head, a hump deformity of the femoralhead/neck junction, reduced femoral head-neck offset,labral ossification, a positive cross-over sign of a retro-verted acetabulum, and herniation pits on the antero-superior aspect of the femoral head. Herniation pits arethought to be cystic lesions on the femoral neck due tointraosseous ganglia. The cross-over sign is demonstratedby examining the line of the anterior and posterioracetabular rims (Fig. 2). This is readily seen, and ina normal hip the anterior rim is medial to the posterior rimat all points.

Various radiological indices have been proposed tomeasure the likelihood and severity of impingement. Forcam impingement these include the alpha angle, head-neckoffset and the triangular index (Fig. 8, Table 4).28,30,32

Although hump malformations have been measured by thealpha angle on coronal MR slices, others have claimed thatthe majority of hump malformations can be adequatelyassessed on AP pelvis radiographs alone, using a modifieda angle.30 The alpha angle is defined as the angle formed bytwo lines drawn between the femoral neck axis and a lineconnecting the centre of femoral head with the point ofcommencement of femoral head asphericity.32 The averagealpha angle is 65

�to 74

�for patients compared to 42

�to 47

�

Figure 7 Femoral asphericity, most apparent on the lateral view.


in controls (Fig. 9).14,32 Measurement of the alpha angle hasgood inter-and intra-observer reliability and is useful indetecting femoral head asphericity.14 The head-neck offsetcan be measured on plain films or CT. This examines thedistance between the anterior femoral head and theanterior cortex of the femoral neck.27 Normal anterioroffset is 11.6� 0.7 mm. Patients with cam malformationhave reduced anterior offset of 7.2� 0.7 mm.27

It is not always necessary to obtain MRI investigation ofpatients with impingement, but MRI is useful in attemptingto differentiate between a simple labral tear and one dueto impingement. The triad of abnormal findings demon-strated in MR arthrography include abnormal head-neckmorphology, antero-superior cartilage abnormality, andantero-superior labral abnormality.17 Contrast enhanced MRarthrograms are useful to detect labral tears, impingementcysts, and periarticular lesions. It is not always helpful inassessing the shape of the femoral head-neck junction, astypically the slice thickness may be too great to examine thisproperly. It will also show any alteration of labral substanceand confirm the state of the articular cartilage,18 but maynot show labral detachment or undetached chondral sepa-ration.9 3-dimensional CT scanning has recently been found

Table 3 Radiological features of FAI

Femoral features:� short femoral neck� non-spherical but congruent femoral head� hump deformity of femoral head� reduced femoral head-neck offset� herniation pits on antero-superior aspect of femoral head.

Acetabular features:� labral ossification� positive cross-over sign of retroverted acetabulum� os acetabuli� coxa profunda

useful in assessing femoral offset and may have advantageover MR arthrography as it is non-invasive and gives betterresolution and smaller slice thickness.19

It is often necessary to confirm that the pain arises fromwithin the hip joint and not from some other site. Possiblealternative sources include psoas pain, sportsman’s hernia(‘‘Gilmour’s groin’’), or the lumbar spine. If the diagnosis isin doubt, diagnostic local anaesthetic blocks are invalu-able. Injection of local anaesthetic into the hip should bedone under X-ray control with an arthrogram to confirmthat the needle is placed correctly. Patients should bewarned that the procedure is purely diagnostic and anyimprovement in symptoms will only last for 12e24 hours.

Treatment options

There is little role for non-operative treatment.20 It issensible to advise activity modification to avoid theextremes of range of motion. A trial of non-steroidal anti-inflammatory (NSAID) medication may give temporary reliefof pain. Physiotherapy to improve passive range of motion

Figure 8 Alpha angle and normal head-neck offset in thelateral view. ª American Journal of Radiology, reproducedwith kind permission.

Figure 9 Increased alpha angle and reduced head-neckoffset in cam impingement. ª American Journal of Radiology,reproduced with kind permission.

Table 4 Measurments from the lateral shoot-throughX-rays

How to measure the alpha angle:A line is drawn along the axis of the femoral neck. Anotherline is drawn from the centre of the femoral head to thepoint of commencement of asphericity of the femoral head,superiorly. This point is identified by drawing a circle overthe femoral head and taking the point where the anteriorfemoral head protrudes anterior to the circle. The alphaangle is the angle between these two lines.

How to calculate femoral head-neck offset:Line 1 is drawn along the axis of the femoral neck. Line 2 isdrawn parallel to line 1, but abuts the anterior margin of thefemoral head. Line 3 is drawn parallel to lines 1 and 2, buttransects the point of commencement of asphericity (as formeasurement of the alpha angle). The femoral head-neckoffset is the tangential distance between lines 2 and 3.


or stretching is not helpful.33 If the symptoms continue thenthe patient may be a candidate for surgery. During preop-erative counselling, it should be emphasised that thesurgery has been shown to improve symptoms, but thus farhas not been shown to either delay or prevent the devel-opment of osteoarthritis.

The aim of operative treatment is to improve femoro-acetabular clearance and prevent recurrence of mechan-ical blockage at the limits of the desired range of hipmotion. Both arthroscopic and open debridement of thejoint has been attempted. A major advantage of the openapproach is that it gives an unrestricted view of the hipjoint. The surgical approach for open debridement asdescribed by Ganz et al is also known as the trochantericflip approach.26 This approach allows surgical dislocation ofthe joint and affords a 360

�view of the acetabulum and the

femoral head. The vascular supply to the head of the femuris at risk during hip dislocation, but detailed examination ofthe blood supply has demonstrated that the relevantvessels pursue a predictable course. Ganz’s group hasshown that the predominant blood supply to the femoralhead is via the deep branch of the medial circumflexfemoral artery. One or two vessels pass obliquely along theposterior surface of the neck as the retinacular vessels, andeventually enter the femoral head at the upper surface ofthe head/neck junction, just posterior to the midline. Theydescribe a technique to avoid damage to the medialcircumflex femoral branch by keeping the short externalrotators intact, performing an osteotomy which leaves theoverhanging posterior tip of the trochanter in place withapproximately 1 centimetre of the gluteus medius stillattached. This splits the piriformis tendon. A Z-shapedcapsulotomy is made antero-laterally along the femoralneck axis. The cut must remain anterior to the lessertrochanter to avoid damage to the medial circumflexfemoral artery. The authors did not report any cases ofavascular necrosis in their series from this approach. Duringthe osteochondroplasty on the femoral neck, care must betaken to avoid straying too far posterior on the uppersurface of the head/neck junction and risking damage tothe retinacular vessels. There is a reported risk of avascularnecrosis of less than 1:1000 with this approach. At the endof the procedure, the Z-capsulotomy is repaired with one ortwo sutures, and the greater trochanter reattached usingtwo or three 4.5 mm lag screws.

Alternatively, a Smith-Petersen or Huetter anteriorapproach can also be used. It is important to undertakecareful soft tissue dissection after skin incision to avoidinjury to the lateral cutaneous nerve of the thigh if thisapproach is used. With this approach, excellent visual-isation of the anterior and superior part of the head/neckjunction is obtained, together with the anterior rim of theacetabulum. It is not necessary to dislocate the hip, asa good view of intra-articular structures can be obtained ifthe procedure is performed on a traction table with trac-tion applied at the appropriate point to permit insertion ofan arthroscope. Adequate visualisation of the acetabulumis important because of the prevalence of cartilagedelamination lesions.

Hip arthroscopy is minimally invasive but because oflimited space may not provide adequate visualisation.There is also a steep learning curve associated with hiparthroscopy. Hip joint distraction involves considerableforce and leaves a few patients with persistent post-operative groin pain. The procedure can be performed withthe patient in either the supine or lateral position, butincreasingly it is performed supine. Image intensifier guid-ance is required. A 70� scope is the standard instrument forhip arthroscopy, and use of cannulated introducers andswitching sticks is essential. Hip arthroscopists describe thehip joint as being divided into central and peripheralcompartments. The central compartment includes thelabrum, ligamentum teres, articular surfaces and medialstructures. This compartment is viewed using hip jointdistraction. The peripheral compartment is viewed withoutdistraction, and includes all intra-capsular structureslateral to the labrum. Arthroscopy is useful to deal withlabral tears and chondral damage. Achieving adequatespace to provide visualisation at the front of the hiprequires resection of a section of the anterior capsule andsynovium with a shaver or radio-frequency probe; this maybe time consuming. Reattachment of labral tears to bone isdifficult (but possible) by arthroscopic means, requiring theuse of suture anchors and being similar in principle to anarthroscopic Bankart repair. Resection of the acetabularrim may be performed arthroscopically. A combinedarthroscopic followed by open approach is now preferred

Figure 10 Os acetabuli in cam impingement. ª AmericanJournal of Radiology, reproduced with kind permission.


by some surgeons.12 Initial arthroscopy allows an assess-ment of the severity of any pathology that then dictates theextent of exposure. A focal cam impingement lesion may besuitable for limited open debridement without jointdislocation.34

Whichever approach to the problem is used, there areseveral goals during the operation. First, the surgeon aimsto re-create the normal concave anatomy of the femoralhead/neck junction by sequentially removing bone fromthe femoral neck. Second, labral and chondral damagemust be identified and treated. Third, excessive anteriorprojection of the anterior acetabular rim should be cor-rected (so called ‘‘rim trim’’). Finally, the surgeon shouldcheck that impingement has been adequately corrected.

The anterior impingement osteophyte is probably themost obvious radiological sign of impingement. Many ormost patients prove to have a prominence at the anteriorand superior femoral neck which requires trimming. Thiscan be performed using gouges or a burr. It is surprising howmuch bone needs to be resected in many patients. Per-operatively in open surgery, a spherical template can beused to identify a non-spherical portion of the femoralhead/neck junction. Osteochondroplasty can be performedat the point where the template abuts against the head.The area involved in FAI may be covered by frayed hyalinecartilage and be obvious to naked eye examination. Incontrast to unblemished white hyaline cartilage the areawill have a reddish or bluish hue.33 The surgeon needs to becareful while removing slices from the superior aspect ofthe neck in order to preserve the retinacular vessels.Excision osteoplasty is continued gradually until impinge-ment free motion is confirmed intra-operatively. The risk ofneck fracture can be avoided if the total amount resectedremains less than 30% of the antero-lateral quadrant of thehead/neck junction.21 Patients with reduced anteversionand a varus neck may also be helped by a flexion-valgusintertrochanteric osteotomy.33

Acetabular over-coverage can be treated with acetab-ular rim resection after elevating the labrum. The aim ofresection is to achieve a residual lateral centre-edge angleof �20

�to 25

�.32 Periacetabular osteotomy should be

considered if the anterior acetabular over-coverage isassociated with marked deficiency of posterior coverage, orif there is dysplasia identifiable on the AP X-ray. Acetabularcartilage must be relatively well preserved for peri-acetabular osteotomy to be the preferred procedure.33

During acetabular rim resection, the anterior labrum islifted from the underlying bone. An adequate amount ofbone is then resected, together with the underlyingacetabular cartilage. This may make a ‘‘rim trim’’ anattractive option if there is a small chondral delaminationlesion in the acetabulum. Ossified labrum is excised.Degenerate labrum is resected but this should be limited tothe area of definite degeneration, only to preserve as muchintact labrum as possible for refixation. In the early stagesof the disease labral tears result from localised chronicirritation and the tip remains uninvolved. Labral fixationfollowing resection of a degenerate portion is a feasibleoption.22 Viable torn labrum can be reattached to theacetabular rim with suture anchors. The early results oflabral refixation have proved better compared to labralresection alone.17

At the end of the procedure the surgeon should assessthe impingement-free range of motion available at thejoint. Most surgeons use bone wax on the debrided area ofthe femoral neck to minimise the risk of ectopic boneformation, which is a common problem after this proce-dure. The use of NSAIDs to minimise this problem shouldalso be considered. Weight-bearing is restricted post-operatively after open surgery, with crutches being used for2 months to minimise the risk of detachment of the greatertrochanter. In semi-open or arthroscopic surgery, thepatient is asked to use crutches for 2 weeks to protect thefemoral head. Physiotherapy to work on range of motionexercises may be helpful from about 2 to 3 weeks; abductorexercises should be avoided for 2 months in patients witha trochanteric osteotomy. Most patients are back to activ-ities of daily living by 6 to 8 weeks, and back to sportingactivities by 4 to 6 months.

Outcomes

Both arthroscopic and open debridement surgery forimpingement provide good symptomatic relief. There isa high risk of heterotopic ossification following opena debridement, although the risk of avascular necrosis islow.26 We do not yet have long term longitudinal studies toassess the role of debridement surgery in eventualprevention of hip arthritis. It is important to have a detaileddiscussion with the patient pre-operatively so that expec-tations remain realistic. Although surgery will curemechanical impingement it will not recreate normal jointanatomy. Patients with advanced chondral damage may bebetter off with replacement arthroplasty. The worst prog-nostic indicator for failure of FAI surgery is significantosteoarthritis. Hips with additional risk factors, likeuntreated dysplasia and subluxation, are also at risk offailure.23 Although early results are encouraging, we need


larger series with longer term follow-up to determine ifmechanical correction of pathomorphic abnormalities willprevent long term progression of osteoarthritis.

Future research directions

� Long term studies with larger numbers of patientsare needed to examine the effects of impingementsurgery in the potential prevention of osteoar-thritis of the hip joint.� Guidelines need to be developed covering how

early specialist referral should be made and whensurgery should be undertaken.� Better and more predictive imaging modalities are

needed, such as three-dimensional modelling, toassess the severity of joint degeneration and tohelp plan surgery.� There is a need to develop appropriate patient

orientated outcome measures and functionalscores in relation to FAI surgery.

Practice points

� There is growing evidence that femoro-acetabularimpingement results in osteoarthritis of the hipjoint.� Patients with FAI usually present with deep groin

pain exacerbated by deep hip flexion.� Pincer impingement is more common in middle

aged athletic females.� Cam impingement (Fig. 10) is commoner in young

active males.� Labral damage rarely occurs alone and may signify

ongoing impingement.� Radiology may reveal a non-spherical femoral

head and/or retroverted acetabulum.� Contrast enhanced MRI is useful to visualise

a damaged labrum.� Surgery is the treatment of choice and is effective

in alleviating symptoms of FAI.

References

1. Law W. Osteoarthritis of the hip. London: Butterworth; 1952.2. Myers SR, Eijer H, Ganz R. Anterior femoroacetabular

impingement after periacetabular osteotomy. Clin OrthopRelat Res 1999;363:93e9.

3. Stulberg SD, Cordell LD, Harris WH, et al. Unrecognisedchildhood disease : a major cause of idiopathic osteoarthritisof the hip. In: The hip. Proceedings of the Third OpenScientific Meeting of the Hip Society. St Louis, MO: CV Mosby;1975. p. P212e28.

4. Harris WH. Etiology of osteoarthritis of the hip. Clin OrthopRelat Res 1986;213:20e33.

5. Ecker TM, Tannast M, Puls M, et al. Pathomorphologic alter-ations predict presence or absence of hip osteoarthrosis. ClinOrthop Relat Res 2007;465:46e52.

6. Kim YT, Azuma H. The nerve endings of the acetabular labrum.Clin Orthop Relat Res 1995;320:176e81.

7. Crawford MJ, Dy CJ, Alexander JW, et al. The biomechanics ofthe hip labrum and the stability of the hip. Clin Orthop RelatRes 2007;465:16e22.

8. Ito K, Minka MA, Leunig M, et al. Femoroacetabular impinge-ment and the cam-effect. A MRI-based quantitative anatomicalstudy of the femoral head-neck offset. J Bone Joint Surg Br2001;83:171e6.

9. Crawford JR, Villar RN. Current concepts in the management offemoroacetabular impingement. J Bone Joint Surg Br 2005;87:1459e62.

10. Kubiak-Langer M, Tannast M, Murphy SB, et al. Range of motionin anterior femoroacetabular impingement. Clin Orthop RelatRes 2007;458:117.

11. Kim WY, Hutchinson CE, Andrew JG, et al. The relationshipbetween acetabular retroversion and osteoarthritis of the hip.J Bone Joint Surg Br 2006;88:727e9.

12. Jaberi FM, Parvizi J. Hip pain in young adults: femo-roacetabular impingement. J Arthroplasty 2007 Oct;22(7Suppl. 3):37e42.

13. Burnett RS, Della Rocca GJ, Prather H, et al. Clinical presen-tation of patients with tears of the acetabular labrum. J BoneJoint Surg Am 2006;88:1448e57.

14. Clohisy JC, Nunley RM, Otto RJ, Schoenecker PL. The frog-leglateral radiograph accurately visualized hip cam impinge-ment abnormalities. Clin Orthop Relat Res 2007 Sep;462:115e21.

15. Meyer DC, Beck M, Ellis T, et al. Comparison of six radiographicprojections to assess femoral head/neck asphericity. ClinOrthop Relat Res 2006;445:181e5.

16. Eijer H, Myers SR, Ganz R. Anterior femoroacetabularimpingement after femoral neck fractures. J Orthop Trauma2001;15:475e81.

17. Kassarjian A, Yoon LS, Belzile E, et al. Triad of MR arthro-graphic findings in patients with cam-type femoroacetabularimpingement. Radiology 2005;236:588e92.

18. Espinosa N, Rothenfluh DA, Beck M, et al. Treatment of fem-oroacetabular impingement: preliminary results of labralrefixation. J Bone Joint Surg Am 2006;88:925e35.

19. Tannast M, Kubiak-Langer M, Langlotz F, et al. Noninvasivethree-dimensional assessment of femoroacetabular impinge-ment. J Orthop Res 2007;25:122.

20. Jager M, Wild A, Westhoff B, et al. Femoroacetabularimpingement caused by a femoral osseous head-neck bumpdeformity: clinical, radiological, and experimental results.J Orthop Sci 2004;9:256e63.

21. Mardones RM, Gonzalez C, Chen Q, et al. Surgical treatmentof femoroacetabular impingement: evaluation of the effect ofthe size of the resection. J Bone Joint Surg Am 2005;87:273e9.

22. Ito K, Leunig M, Ganz R. Histopathologic features of theacetabular labrum in femoroacetabular impingement. ClinOrthop Relat Res 2004;429:262e71.

23. Murphy S, Tannast M, Kim YJ, et al. Debridement of the adulthip for femoroacetabular impingement: indications andpreliminary clinical results. Clin Orthop Relat Res 2004;429:178e81.

Further reading

24. Ganz R, Parvizi J, Beck M, et al. Femoroacetabular impinge-ment: a cause for osteoarthritis of the hip. Clin Ortop 2003;417:112e20.

25. Beck M, Kalhor M, Leunig M, et al. Hip morphology influencesthe pattern of damage to the acetabular cartilage:


femoroacetabular impingement as a cause of early osteoar-thritis of the hip. J Bone Joint Surg Br 2005;87:1012e8.

26. Ganz R, Gill TJ, Gautier E, et al. Surgical dislocation of theadult hip. A technique with full access to the femoral head andacetabulum without the risk of avascular necrosis. J Bone JointSurg Br 2001;83:1119e24.

27. Eijer H, Leunig M, Mahomed N, et al. Cross-table lateralradiographs for screening of anterior femoral head-neck offsetin patients with femoroacetabular impingement. Hip Int 2001;11:37e41.

28. Tannast M, Siebenrock KA, Anderson SE. Femoroacetabularimpingement: radiographic diagnosisdWhat the radiologistshould know. Am Journalism Rev 2007;188:1540e52.

29. Ganz R, Leunig M, Leunig-Ganz K, Harris WH. The etiology ofosteoarthritis of the hip: an integrated mechanical concept.Clin Orthop Relat Res 2008 Feb;466(2):264e72.

30. Gosvig KK, Jacobsen S, Palm H, et al. Magnusson E.A newradiological index for assessing asphericity of the femoral headin cam impingement. J Bone Joint Surg Br 2007 Oct;89(10):1309e16.

31. Beck M, Leunig M, Parvizi J, et al. Anterior femoroacetabularimpingement: part II. Mid-term results of surgical treatment.Clin Orthop Relat Res 2004;418:67e73.

32. Notzli HP, Wyss TF, Stoecklin CH, et al. The contour of thefemoral head-neck junction as a predictor for the risk ofanterior impingement. J Bone Joint Surg Br 2002 May;84(4):556e60.

33. Lavigne M, Parvizi J, Beck M, et al. Anterior femoroacetabularimpingement: part I. Techniques of joint preserving surgery.Clin Orthop Relat Res 2004;418:61e6.

34. Maheshwari A, Malik A, Dorr LD. Impingement of the nativehip joint. J Bone Joint Surg Am 2007;89:2508e18.

current management of femoro-acetabular impingement mx femoracetabular impingement1.pdfence...

Documents