How I Do It

Primary total hip replacement

G. Bannister
BUPA Glen Hospital, Redland Hill, Durdham Down, Bristol, BS6 6UT

Correspondence to: G Bannister, BUPA Glen Hospital, Redland Hill, Durdham Down, Bristol, BS6 6UT

Introduction Planning Surgical procedure Surgical exposure

Dislocation Acetabular preparation Femur preparation Post-operative management References

Keywords: Patient selection, prostheses, planning, posterior approach
Surg J R Coll Surg Edinb Irel., 1 December 2003, 332-341

Primary hip replacement requires careful patient selection, a knowledge of available prostheses, thorough pre-operative planning, familiarity with surgical exposures and sufficient manual dexterity to carry out the procedure. The aim of hip replacement is to restore the biomechanics of the diseased joint as closely as is practical. Failure to observe these principals results in premature failure of the joint replacement

INTRODUCTION
The majority of patients undergoing primary total hip replacement are aged between 65 and 80. They have a life expectancy of no more than 20 years and 40% will be dead within 15 years. In such patients, a well-tried cemented total hip replacement fulfils their needs. The extremes of age present more of a problem. In patients over the age of 80 instability is a major complication. In my own unit, in the 1980s octogenarians had a 15% dislocation rate becoming chronic in 10%.1 If the hip remained in situ they were in hospital for 27 days but if they dislocated this rose to 71 days. This was with a 22.25mm Charnley prosthesis and a mixture of posterior and direct lateral approaches. In such patients, a larger head or in those with poor balance or neurological disorder a captive cup should be used.

Aseptic cup loosening is the main problem of cemented cups in patients under the age of 65 years and a hemispherical uncemented cup with porous, hydroxyapatite or plasma coating affords at least some osseointegration. The cup selected should have a good locking mechanism for the polyethylene insert and a polished internal surface to prevent backwear.

In young heavy males or patients under the age of 50, wear should be addressed by either ceramic or metal bearings (Table 1).

PLANNING OF PROCEDURE
Total hip replacement is as good as its planning and execution. Planning takes place before surgery to minimise exposure and wound contamination.

The aim of total hip replacement is to restore the biomechanics of the individual’s hip. Implanting a single standard prosthesis into all patients will not achieve this in a significant minority.

Planning should begin with clinical assessment of the patient. Heavy patients need a larger exposure and this is best done at the beginning of the procedure rather than after an unproductive struggle against copious fat or muscle bulk. Tall or small patients require components to match their size (Table 2).

LEG LENGTH DISCREPANCY (TABLE 3)
The AP and lateral views of the pelvis should then be examined (Figure 1 and 2). If the limb is short, the site of the bone loss should be identified. In acetabular bone loss, the centre of rotation should be restored by a distally placed cemented flanged cup, an oversized cementless cup or an acetabular support ring. Femoral shortening is more readily addressed by leaving the stem proud or using a long modular neck (Figure 3). In patients with equal leg length, post-operative lengthening needs to be avoided. Leg length equality after total hip replacement cannot be guaranteed but the surgeon offers the patient the best chance by restoring the centre of acetabular and femoral head rotation. The relation to the centre of rotation of the femoral head to the tip of the greater trochanter should be established to guide neck resection and depth of stem insertion.

Figure 1: Femoral anteversion: lateral view showing acetabular anteversion 

Figure 2: Planning hip replacement in dysplasia: the left hip is dysplastic with lateral subluxation, external rotation and true and apparent valgus alignment of the neck. Shortening measured at the level of the lesser trochanter is predominantly the result of superior acetabular deficiency

Figure 3: Anatomy restored: the socket has been placed in the true acetabulum distally and medially. The offset has been restored by templating the normal right hip. The neck has been resected distally to allow insertion of a standard stem. The relationship of the centre of rotation of the femur has been restored and the shortening corrected

PROTRUSIO ACETABULI (TABLE 4)
Protrusio acetabuli poses two problems: dislocation may be difficult, there is medial bone loss and patients with protrusio have atrophic arthritis and tend to have thin bones. Dislocation by vigorous torsional movements may result in shaft fracture and should be avoided at all costs. The correct approach is to remove any peripheral osteophytes that obstruct the view of the neck, attempt a gentle dislocation and if that fails resect the neck, mobilise the femur, expose the head in the socket and remove this either with a corkscrew or by reaming into it. If there is medial bone loss it can be grafted either structurally with a segment of the femoral head or by impacted cancellous bone taken from the metaphysis of the proximal femur.

LATERAL SUBLUXATION
When the head is subluxed laterally it is usually secondary  to posterior and inferior osteophytes. These displace the head laterally and anteriorly and this is often associated with anterolateral cysts in zones 1 and 4 of the acetabulum. These osteophytes disrupt access and orientation of the cup and should be removed early in the procedure. The acetabulum should be reamed posteromedially and the cup aligned on its true rim. The cysts should be located, curetted and grafted.

BONE LOSS
Cysts should be curetted and grafted with cancellous bone, best taken from the femoral metaphysis. If there is an anterior column defect the lateral radiograph should be studied carefully for posteromedial bone stock into which the cup can be inserted. If there is superior bone loss, up to 30% lack of cover can be accepted. An oversized cup can be helpful but the ilium tends to narrow proximally and this needs to be placed distally. The defect can be grafted but if the graft is weightbearing it will fail.

NECK ANGLE (TABLE 5)
The neck angle should be noted. If it is valgus a higher neck resection is indicated and a long neck and modular head or the proud placement of a Monobloc stem will restore the centre of rotation. A varus neck is more difficult. Distal soft tissues must be divided to gain access to the distal neck. Low neck resection should be carried out and an extended offset stem used to restore the anatomy.

Femoral neck anteversion in developmental dysplasia is often too extreme to allow stem insertion after high neck resection (Figure 1). In such cases the neck may need to be resected as low as the lesser trochanter to allow access to the shaft and even then only after removal of the piriform fossa and posterior aspect of the neck.

FEMORAL ROTATION
Many arthritic hips are externally rotated. This is diagnosed radiologically by excessive prominence of the lesser trochanter and valgus alignment of the neck (Figure 2). In externally rotated hips the correct offset of hip replacement should be measured on the normal hip (Figure 3).

MEDULLARY CAVITY
Moving down the femoral shaft the size of the medullary cavity should be noted. If it is wide, it may be difficult to align the cemented prosthesis in neutral and this can be addressed by spacers around the stem. If it is narrow, broad stems do not fit and a narrow stem should be used. In narrow medullary canals a standard sized cement restrictor fragments and needs to be trimmed before insertion. At the end of these qualitative deliberations a template should be taken to act as a guide to the components to be inserted and the nursing staff advised as to the probable range so that they can be ready in theatre before the operation starts. The template should be applied to the femur of the normal hip to correct offset and the acetabulum of the diseased joint.

Figure 4: Short external rotators 

Figure 5: Suture through short external rotators

INFECTION CONTROL
The surgeon can do little about the quality of theatre air provided. Plenum ventilated air carries an increased risk of infection, compared with vertical laminar flow. The surgeon can, however, reduce the risk of infection by wearing a balaclava hat and mask and covering the procedure with systemic antibiotic prophylaxis, supplemented by Gentamicin, and using pulsed jet lavage and disinfectants such as chlorhexidine 0.05%.

SURGICAL PROCEDURE
My personal preference is the posterior approach. It can be extended more proximally, in difficult cases, without damaging the nerve supply to the abductors. Patients walk better and rehabilitate more rapidly with this approach. The principal complication is dislocation and this can be reduced by correct alignment of the prostheses, long posterior wall acetabular components and, on occasions, with large heads or primary captive cups. Recurrent posterior dislocation can again be addressed by extra long posterior wall liners for cementless cups, the Belfast PLAD in cemented implants or a captive cup. This is a more manageable complication than avulsion of the abductors following a direct lateral approach, which is often slow to be recognised and extremely difficult to salvage.

Figure 6: Posterior capsule 

Figure 7: Anterior capsular flap 

Figure 8: Suture through posterior capsular flap

PATIENT PREPARATION
The patient is placed in the lateral position with props posteriorly against the mid-lumbar lordosis and anteriorly against the anterior superior iliac spine. It is the surgeon’s responsibility to position the patient. A prop against the sacrum tends to antevert the pelvis distorting the anatomy and resulting in a retroverted cup. It also limits surgical access posterolaterally.

The patient’s skin is prepared with chlorhexidine in alcohol and allowed to dry before an adhesive drape is placed between the groin and the affected hip, followed by standard drapes exposing the hip to as far as the iliac crest and leaving the buttock exposed posteriorly for at least six inches.

SURGICAL EXPOSURE OF JOINT
An incision is made in a straight line crossing the greater trochanter at the junction of the anterior two thirds and posterior third, splitting a small amount of gluteus maximus. The fat is divided and left in continuity with the gluteal fascia as this is a fasciocutaneous plane and stripping of the fat simply causes necrosis. Haemostasis is secured, swabs soaked in 0.05% chlorhexidine solution are placed along the cut edge of the fat and held in place with Norfolk and Norwich selfretaining retractors.

The assistant then extends and internally rotates the leg as far as it will move comfortably. The gluteal bursa is identified distally. A pair of Mayo scissors is passed underneath it. It is resected in continuity and takes the surgeon past quadratus femoris, the short external rotators and to the posterior aspect of the gluteus medius. One inch proximal to the tip of the greater trochanter the areolar tissue between the gluteus medius and minimus posteriorly is divided. A Langenbeck retractor is inserted separating the gluteus medius from gluteus minimus. This exposes the silver sheen of the fascia overlying the gluteus minimus and the piriformis tendon which lies posterior to the other three short external rotators (Figure 4). In heavily built young patients with stiff hips it is sometimes difficult to find the plane between gluteus medius and minimis. The piriformis tendon can assist here, because it lies in the plane that is being sought and is relatively easily palpable because of its size and posterior disposition. Sufficient fat is swept away from the short external rotators to expose the piriformis proximally and the inferior gemellus distally and haemostasis secured with diathermy. A plane is then developed by sharp dissection between the piriformis and the gluteus minimus proximally and first Mayo scissors then a blunt Homan retractor are passed along the superior margin of the capsule. Inferiorly, scissors are passed in a perpendicular direction between the inferior gemellus and quadratus femoris, realising the potential space along the inferior capsule, and a blunt Homan inserted. The surgeon can then see two-thirds of the entire hip capsule and has the short external rotators exposed. A suture is placed through the short external rotators as close as possible to their insertion (Figure 5). These are divided and separated posteriorly from the capsule by blunt dissection (Figure 6). The capsule is divided circumferentially as close to the neck as possible and two posterior limbs taken posteriorly just proximal and distal to the neck. This creates a flap (Figure 7) that can be used to repair the iliofemoral ligament to the greater trochanter at the end of the procedure. A suture is placed in this (Figure 8) and it is retracted exposing the femoral head and neck (Figure 9).

DISLOCATION AND REMOVAL OF FEMORAL HEAD 
The neck is extended and internally rotated, then flexed anteriorly gently by the surgeon and gently dislocated (Figure 9). If it will not come easily a Muller skid can be used and failing that the exposure needs to be extended. This is done by placing a suture through the quadratus femoris and dividing it 3-5mm away from its insertion. This exposes a layer of fat anterior into the quadratus in which the principal vessels lie and where they can be easily controlled. The internal rotation and flexion manoeuvre is repeated and if the hip still remains in situ the insertion of the gluteus maximus is divided 3mm from the linear aspera. This removes all the posterior soft tissue restraints. If the hip will still not dislocate the periosteum between the proximal insertion of gluteus maximus and the psoas is divided and elevated posteriorly from the femur. A bone lever is placed in situ exposing the psoas tendon, which is divided from distal to proximal and further proximally still taking the antero-inferior capsule. At this juncture there is nothing further the surgeon can do in terms of soft tissue releases and the neck must be resected, the shaft mobilised and the head removed separately. In practice this is an extremely rare occurrence. The extensile approach should be used without hesitation in heavily muscled patients.

Figure 9: Femoral head dislocated posteriorly 

Figure 10: 360° view of acetabulum 

Figure 11: Acetabulum showing ligamentum teres and fossa acetabulare

Having dislocated the hip it should be possible to internally rotate it beyond neutral and, if this is not possible, the approach should be extended in the manner previously described and the anterior inferior capsule divided along with the psoas which will afford another 20° of internal rotation. The head is resected at a pre-determined level depending on its angle and at this point approximately half the acetabulum can be seen. Posteriorly, the view is obstructed by the labrum and anteriorly by the capsule and labrum. The superior blunt Homan usually lies over the anterior inferior iliac spine and the inferior one in the obturator foramen between the inferior capsule and the distal tissues. The labrum is removed posteriorly, superiorly and anteriorly, along with any thickened capsule anteriorly. The ischial tuberosity is palpated and a Judd or Charnley pin placed into it angling it posteriorly as it is driven home. A further Judd or Charnley pin is placed superiorly between the gluteus minimus and capsule, 1cm proximal to the superior rim of the acetabulum. This gives rim exposure of the superior, posterior and inferior acetabulum and the anterior rim is exposed by placing a straight sharp Homan retractor underneath the anterior rim, angling it distally so that it controls the proximal femur and allowing the internally rotated femur to return to neutral. This serves to take the femur anteriorly giving 360° vision of the acetabulum (Figure 10). The only structure obstructing access at this juncture is the inferior capsule which is under tension, has a blunt Homan separating it from underlying muscles and can be resected. There is usually a small vessel anteriorly. The last structure that needs to be removed is the ligamentum teres (Figure 11). The ligamentum teres is structurally thick laterally and has a vessel passing underneath the transverse ligament medially. The best way of removing it is to grasp it with a Kocher’s forceps, pull it proximally placing it under tension, divide it with a long handled knife and when the last of the structural fibres yield take a Charnley spoon and curette from inferior to superiorly against the rim of the fovea acetabulare. This dissects out the vessels which can then be diathermied and divided leaving a bare fovea which demarcates the inner cortex of the acetabulum and leaves a blood free field. The acetabulum is then reamed to accept the implant selected at pre-operative planning.

Figure 12: Uncemented cup with ceramic liner 

Figure 13: 360° exposure of cut neck of femur 

Figure 14: True axis of femur established with box osteotome introduced posterolaterally

ACETABULAR PREPARATION
The acetabulum is prepared with hemispherical reamers. If the bone is particularly hard, a rotatory movement with a small reamer breaks down the subchondral bone. In acetabula with overhanging osteophytes these can either be removed with an osteotome or reamed as the hemispherical reamer is inserted into the acetabulum. The reamer needs to be rotating before it hits the posteroinferior osteophytes and held firmly as they are removed. The practice of ramming a reamer deeply into the socket then turning on the drill results in stalling and is a wasted manoeuvre. The hemispherical reamer is often mounted on a universal drill with variable gears to allow reaming and drilling. Reaming is slow with powerful torque and drilling is fast with lesser torque. In very atrophic or deficient acetabula, the powerful torque of the reamer may cause a fracture and this can be avoided in such cases by turning the driver reamer into drill mode, which is much gentler on this type of bone. Once the acetabulum has been reamed to the selected size, three 1cm keyholes are placed in the ilium and one each in the pubis and ischium. The ischium is identified by palpation. It is best to make a small hole and expand it with a curette rather than a large one posteriorly, which simply allows cement to extrude and potentially damage the sciatic nerve. Anteriorly, the pubis falls away vertically and again should be palpated, a small hole placed into it and the space expanded. The acetabulum is brushed, pressure lavaged and dried, a flanged cemented cup is cut to fit the true rim and orientated in anatomical anteversion. Cement is mixed and when in dough form, at approximately three minutes, taken by the surgeon who in a single movement removes the swab in the acetabulum, inserts it, compresses it manually into the keyholes then uses a cement compressor. If the acetabulum is 52mm or less the cup needs to be inserted early. If the cup is cavernous, 54-60mm, cement should be left for a further minute so that the prosthesis does not bottom out. The cup should be flush with the true acetabular rim. Having implanted the cup, the inserting device is removed, the position of the cup is checked, adjusted manually if required and then held with a pusher with a hemispherical tip. The cement should dry within 10 minutes. Loose cement is removed between three-five minutes and particular attention made to cement extrusion under the transverse ligament of the acetabulum which should be checked digitally. 

If a cementless cup is used a trial should be inserted first, peripheral osteophytes resected to allow implantation of the liner, the shell then inserted, tested for stability, supplemented with screws, if necessary, and the liner then implanted. The liner should fit flush with the rim of the shell and if it does not there is impingement and more tissue should be resected (Figure 12).

FEMUR PREPARATION AND PROSTHESIS INSERTION
The cut neck of the femur is exposed by placing a sharp Homan retractor between the neck and the psoas medially with the hip internally rotated and extended (Figure 13). The hip is then flexed and adducted to expose the anterior aspect which is levered forward with a blunt Homan retractor. The abductor ‘mechanism’ is retracted laterally by placing a sharp Homan from anterior to posterior under the gluteus medius to lever it posterolaterally, exposing the insertion of the piriformis. The stump of piriformis is resected and the shape of the proximal femur will present as a rectangle with a posterior indentation that is the piriform fossa. The piriform fossa is an aberration that obscures the true landmarks and must be removed. The true axis of the femur begins posterolateral to the anterior wall of the piriform fossa and this area is located on the greater trochanter.

A spiked box osteotome is then used to remove the bone medial to the tip of the greater trochanter and the wall of piriform fossa extending medially into the cancellous bone of the neck. The box osteotome is inserted in a line from lateral to medial and then anteriorly. This manoeuvre will remove approximately 2.5cm of clean cancellous bone giving the correct axis to the sagittal plane of the femur (Figure 14). A pencil reamer is then passed. If the initial bone resection by the box osteotome has been carried out correctly, the pencil reamer will pass from lateral to medial and into the correct coronal axis of the femur. Cancellous bone is broken up with a pencil reamer then removed with a curette. Removal with a curette is better than using a broach because it reduces the amount of fat disseminated into the systemic circulation. Cancellous bone should be removed to the peripheral 3mm (Figure 15). Cancellous bone tends to be deposited in the trochanters which are posteriorly disposed. The best way of removing cancellous bone is to take a Charnley spoon with a T handle, place the sharp end of the spoon distal to the trochanteric bone and engage the posterolateral and posteromedial cancellous bone and remove it. This reproduces the canal flare which gives additional stability to the cement bone interface. 

Having curetted away bone, the broaches are then introduced. If a broach will not pass it means that the entry point is either too medial or too anterior and this should be corrected by reusing the box osteotome or removing the piriform fossa with nibblers. The broach sizes selected should be bigger than the actual prosthesis inserted to allow for a complete cement mantle, a spigot is placed on the broach and a trial head on the spigot. The centre of rotation of the trial head should reproduce the centre of rotation of the true femoral head in relation to the tip of the greater trochanter. The hip is then reduced by longitudinal traction, flexion and external rotation. If the hip does not reduce the tissue tension is too tight, a shorter head should be tried and failing reduction then a smaller broach should be inserted more deeply into the femoral shaft. The position of the knees will have been checked before surgery and this should be reproduced after the trial femoral component has been inserted. When this is done the neck of the prosthesis is marked on the greater trochanter with the diathermy, the broach removed, the bone prepared by brushing, pressurised lavage and drying (Figure 16). A Hardinge cement restrictor is inserted 16cm from the tip of the greater trochanter. The medullary cavity is packed with an open swab or gauze roll and cement is mixed in a gun. Palacos R with Gentamicin cement had the best results in the Swedish Arthroplasty Survey. It handles well and can be gunnedin at one minute and 40 seconds after commencement of mixing at room temperature. To avoid a head of blood contaminating the proximal cement-bone interface a venting tube is placed distally adjacent to the cement restrictor and the cement is gunned-in, beginning 10cm distal to the greater trochanter. This compresses cement into the nutrient foramen that provides the arterial blood supply to the proximal femur. The cement forces any blood that has escaped between removal of the pack and insertion of cement onto the venting tube and a dry interface is obtained over the proximal 10cms in the majority of cases. Having filled the canal fully (Figure 17) with cement a proximal restrictor is applied. This is best made of rubber that conforms to the irregular anatomy of the proximal femur and, when adequate pressure is obtained, fat extrudes from the outer surface of the femoral neck and metaphysis. After two minutes the cement is sufficiently viscous that back bleeding will not displace it and a prosthesis can be inserted.2 The preferred prosthesis is a polished collarless double tapered device with a spacer and a stable introducer. The point of entry is posterolaterally. A thumb will usually give better and more versatile proximal restriction than proprietary devices. The prosthesis is inserted slowly in anatomical anteversion until the neck reaches the diathermy mark on the greater trochanter. The introducer is then removed.  Redundant cement is removed from the greater trochanter and the posterior aspect of the neck, whilst the hip is held flexed and internally rotated. The hip is then extended, still internally rotated, exposing the anteromedial aspect of the neck giving access to the remaining extruded cement. Pressure is maintained on the prosthesis and the cement as the latter expands by some 10% during curing. At the end of the procedure the cement mantle should be thickest anteromedially and thinnest posteriorly indicating that the prosthesis has been aligned correctly in the sagittal plane of the femur.

Whilst the cement is setting, the threads through the short external rotators and posterior capsule are separated into the proximal and distal two. Two drill holes are placed through the greater trochanter, the proximal through the tip and the distal just proximal to the quadratus femoris and the threads passed through with either a proprietary passer, the eye of a sinus probe or a skin hook. The trial head is then applied, the hip reduced, leg length checked again, the hip dislocated, the definitive head applied and reduced for the last time.

Figure 15: 3mm cancellous rim 

Figure 16: Lavaged clean dry cancellous bone 

Figure 17: Cement compressed on dry clean cancellous bone

SURGICAL CLOSURE
Following definitive reduction the hip is extended and externally rotated and the posterior capsule and external rotators tied down firmly to their origin on the greater trochanter. If an extensile approach has been used the tendon of gluteus maximus and the muscle belly of quadratus femoris are repaired before this manoeuvre is carried out.

Closure is then carried out with a deep and superficial drain and continuous 2 Vicryl to the gluteal fascia, interrupted sutures to fat and clips to skin. Adhesive dressings applied after an adhesive drape often cause blistering and should be avoided. Preferred closure is Telfa, pads or wool and elastoplast.

POST-OPERATIVE MANAGEMENT
Drains are left in for 24 hours and the patient allowed to stand and walk either the following day or the day after. Most patients are able to climb stairs with the aid of a stick within five to six days and octogenarians are sent to an orthogeriatric rehabilitation unit as they generally take two to three weeks before they can get home. Clips are removed after two weeks, patients are reviewed with an x-ray after three months. If at that time the prosthesis is well-aligned, there has been a good functional recovery, patients over 80 years are discharged as it is extremely unlikely that aseptic loosening will take place. If the cementing technique has been poor, patients are reviewed after five years. Younger patients with less well-tried prostheses are reviewed after five years and patients between the ages of 60 and 79 reviewed after seven years, as it is extremely unlikely that cup loosening or wear will occur before that time.

REFERENCES
1. Newington DP, Bannister GC, Fordyce M. Primary total hip replacement in patients over 80 years of age. J Bone Joint Surg 1990; 72B: 450-52.
2. Majkowski RS, Bannister GC, Miles AW. The effect of bleeding on the cement-bone interface: an experimental study. Clin Orthop 1994; 299: 293-97.

Copyright: 28 April 2003

Applications are invited for:

GENERAL RESEARCH (PUMP-PRIMING) GRANTS Including one grant sponsored by the JP Morgan Fleming Foundation

Applications are invited for research grants (up to £6,000) from surgical trainees and recently appointed consultants who are Fellows/Members of the College in good standing. Research must be undertaken in the UK.

The closing date for receipt of applications is Friday 13 February 2004

MAURICE WOHL RESEARCH FELLOWSHIP IN SURGERY/DENTAL SURGERY (£30,000)

Applications for funding (for one year) are invited from Fellows/Members of the College in good standing, undertaking research projects in the UK.

The closing date for receipt of applications is Friday 13 February 2004

OPHTHALMOLOGY RESEARCH GRANTS (sponsored by the Royal Blind Asylum & School and the Scottish National Institution for the War Blinded)

(i) Major project grants (up to £30,000) (ii) Pump priming grants for ongoing research (up to £6,000)

Applications for funding under the above categories are invited from Fellows/Members of the College in good standing, undertaking research projects in the UK.

The closing date for receipt of applications is Friday 27 February 2004

Further details and application forms can be obtained from: Awards & Grants Secretary The Royal College of Surgeons of Edinburgh, Nicolson Street, Edinburgh EH8 9DW 
Tel: 0131 527 1618 Fax: 0131 527 1667 Email: e.cook@rcsed.ac.uk