Review Article Femoral Bone Loss in Revision Total Hip ... · bone loss (metaphyseal or...

Femoral Bone Loss in RevisionTotal Hip Arthroplasty: Evaluationand Management

Abstract

Primary total hip arthroplasty (THA) is one of the most effectiveprocedures for managing end-stage hip arthritis. The burden ofrevision THA procedures is expected to increase along with therise in number of primary THAs. The major indications for revisionTHA include instability, aseptic loosening, infection, osteolysis,wear-related complications, periprosthetic fracture, componentmalposition, and catastrophic implant fracture. Each of theseconditions may be associated with mild or advanced bone loss.Careful patient evaluation and bone loss classification guidepreoperative planning and overall patient care. Historically,uncemented fixation has provided the best results, but cementedfixation is required in some cases.

The aging of the population andthe consideration of younger pa-

tients for primary total hip arthro-plasty (THA) has resulted in increas-ing numbers of these proceduresbeing performed annually. The revi-sion burden is expected to increase,as well. The major indications forfemoral revision include aseptic loos-ening, infection, instability, osteoly-sis, periprosthetic fracture, compo-nent malposition, and catastrophicimplant failure.

Femoral revision is often compli-cated by bone loss or the poor integ-rity of the remaining bone stock. Weoffer an approach by which to evalu-ate candidates for femoral revisionand define methods by which to as-sess and manage femoral bone lossencountered during revision surgery.

Preoperative Evaluation

Preoperative evaluation begins witha comprehensive history, physical ex-

amination, and radiographs. Key ele-ments of the history include the loca-tion, character, timing, and durationof pain as well as provocative factorsand associated symptoms.

In all cases of painful hip replace-ment, laboratory tests should bedone, including serum erythrocytesedimentation rate (normal, <20mg/dL) and C-reactive protein level(normal, <10 mg/dL). In most cases,patients with elevated serum inflam-matory markers should undergo pre-operative hip aspiration. Synovialfluid obtained from the hip aspira-tion should be sent for cell countanalysis, including differential, andanaerobic and aerobic cultures. Awhite blood cell count of 2,500 to3,000 and a differential of >60%segmented neutrophils is consideredsuspicious for infection, unless theaspiration is performed in the imme-diate postoperative period.1,2

In the setting of loose femoralcomponents—most commonly loose

Neil P. Sheth, MD

Charles L. Nelson, MD

Wayne G. Paprosky, MD

From the Department ofOrthopaedic Surgery, University ofPennsylvania, Philadelphia, PA(Dr. Sheth and Dr. Nelson), andMidwest Orthopaedics, RushUniversity Medical Center, Chicago,IL (Dr. Paprosky).

Dr. Sheth or an immediate familymember serves as a paid consultantto Zimmer. Dr. Nelson or animmediate family member serves asa paid consultant to CadencePharmaceuticals, GreatbatchMedical, and Zimmer and serves asa board member, owner, officer, orcommittee member of the J. RobertGladden Orthopaedic Society.Dr. Paprosky or an immediate familymember has received royalties fromWright Medical Technology andZimmer, is a member of a speakers’bureau or has made paidpresentations on behalf of Zimmer,serves as a paid consultant toBiomet and Zimmer, and serves asa board member, owner, officer, orcommittee member of The HipSociety.

J Am Acad Orthop Surg 2013;21:601-612

http://dx.doi.org/10.5435/JAAOS-21-10-601

Copyright 2013 by the AmericanAcademy of Orthopaedic Surgeons.

Review Article

October 2013, Vol 21, No 10 601

cemented stems—the proximal femuroften remodels into varus and retro-version (ie, proximal femoral remod-eling). Recognizing the potential forsuch remodeling preoperatively mini-mizes the risk of cortical perforation,intraoperative fracture, and under-sizing of the implant. Extended tro-chanteric osteotomy (ETO) is oftenuseful at the time of revision, partic-ularly in the setting of significantvarus remodeling, a well-fixed unce-mented implant, or a long column ofcement below the stem.3

Plain radiographs often underesti-mate bone loss. CT is occasionally auseful adjunct for further definingthe severity of femoral bone loss.4

We recommend the use of CT forany cases that require further delin-eation of the bone loss pattern andany degree of femoral deformity thatmay influence the plan for femoralreconstruction.

Preoperative Planning

Surgical ApproachThe surgical approach for revisionsurgery is based on surgeon experi-ence and utility of the planned recon-struction. Selection of surgical ap-proach is also influenced byadditional exposure (ie, osteotomy),

degree and location of bone deficits,presence of distorted anatomy (eg,heterotopic ossification), and patientfactors (eg, high risk of instability).The posterolateral approach is mostcommonly used. It affords excellentacetabular and femoral exposure;however, it is associated with higherpostoperative instability.

The locations and configurationsof femoral osteotomy vary consider-ably. Standard single plane, trochan-teric slide, Wagner, and extended tro-chanteric are types of osteotomyperformed about the greater tro-chanter.5

The ETO, which is most com-monly used in the setting of revisionTHA, facilitates acetabular exposureand femoral component removal.6

The results of ETO in femoral revi-sion THA have been favorable. Pa-prosky and Sporer6 evaluated 122 re-vision THAs performed with the useof an ETO and reported a 98%union rate of the osteotomized frag-ment at a mean 2.6-year follow-up.

Femoral ComponentRemovalDuring preoperative planning, it isimportant to identify key osseousand functional structures that are atrisk during implant removal. In addi-tion, the surgeon must have at theready implants that would allow re-construction in the event that greaterbone loss than anticipated is discov-ered intraoperatively.

To facilitate stem removal, the fol-lowing instruments should be onhand: manufacturer-specific explanttools, flexible osteotomes, trephines,high-speed burrs (eg, pencil tip, car-bide tip, metal cutting wheel), ultra-sonic cement removal instruments,and universal extraction tools thatallow attachment to the stem or ta-per. The decision to remove a well-fixed implant must be made care-fully.7

Femoral Bone LossClassifications

The American Academy of Ortho-paedic Surgeons introduced a femo-ral bone loss classification based onthe presence of segmental, cavitary,or combined defects8,9 (Table 1). Thisclassification is simple in its organi-zation; however, it is not quantitativeand its practical application is lim-ited. We find the Paprosky classifica-tion to be the most useful for de-scribing femoral bone loss.9-11

The Paprosky classification isbased on the location of femoralbone loss (metaphyseal or diaphy-seal), degree of remaining support ofthe proximal femur (ie, degree ofcancellous bone loss), and theamount of isthmus remaining for di-aphyseal fixation (Table 2). Thesethree variables allow for objective as-sessment of bone loss and provide re-constructive options based on thepattern of femoral bone loss.9,10

With type I femoral bone loss, theproximal metaphyseal bone is main-tained, and proximal femoral remod-eling typically is not exhibited. TypeI defects can be managed with stan-dard length cemented or uncementedimplants. We prefer to manage type Idefects with a standard length exten-sively porous-coated implant9,10 (Fig-ure 1).

Type II femoral defects are themost common type. These demon-strate absent metaphyseal bone losswith an intact diaphysis. Slight prox-imal varus femoral remodeling iscommon. Excellent results have beenreported using extensively porous-coated femoral implants9,10 (Figure2).

Type III defects are subclassified aseither type IIIA or IIIB. Both type IIIdefects exhibit metadiaphyseal boneloss with significant proximal femo-ral remodeling. Type IIIA defectshave ≥4 cm of isthmus remaining for

Table 1

American Academy ofOrthopaedic Surgeons FemoralBone Loss Classification8

Type Description

I Segmental defectII Cavitary defectIII Combined segmental and

cavitary defectIV Femoral malalignment

(rotational or angular)V Femoral stenosisVI Femoral discontinuity

Femoral Bone Loss in Revision Total Hip Arthroplasty: Evaluation and Management

602 Journal of the American Academy of Orthopaedic Surgeons

distal fixation (Figure 3), whereastype IIIB defects have <4 cm remain-ing9,10 (Figure 4). Treatment options fortype III defects include extensivelyporous-coated cylindrical stems, corun-dumized tapered stems with splines (eg,circumferential fluted projectionsaround the tapered stem that confer ro-tational stability of the implant), andcemented stems with impaction bonegrafting. Our preference is to managetype IIIA defects with an extensivelyporous-coated stem and type IIIB de-fects with a modular tapered stem withantirotational splines. In general, thesetapered stems can obtain predictablefixation with only 1 to 2 cm of diaphy-seal bone contact.

Type IV defects exhibit completeloss of the isthmus with little proxi-mal femoral remodeling9,10 (Figure5). Biologic fixation is unlikely, andreconstruction typically requires theuse of an allograft prosthetic com-posite (APC), a long cemented stem,impaction grafting with a long ce-mented femoral component, or prox-imal femoral replacement.

A, Illustration of a Paprosky type I femoral defect. Preoperative (B) andpostoperative (C) AP radiographs of a patient treated with an extensivelyporous-coated stem at the time of two-stage reimplantation to manageperiprosthetic infection and Paprosky type I bone loss. (Panel A courtesy ofDePuy, Warsaw, IN.)

Figure 1

Table 2

Paprosky Classification of Femoral Bone Loss9

Type DefinitionProximal

Metaphysis DiaphysisProximal

RemodelingReconstruction

Options

I Minimal proximal meta-physeal bone loss

Intact Intact None Uncemented fixation; proximalfitting (ie, S-ROM [DePuy]) orextensively porous-coatedstem

II Moderate to severeproximal metaphysealbone loss

Absent Intact Slight Extensively porous-coated stem

IIIA Severe proximal meta-physeal bone loss withdiaphysis intact forsome distance

Absent ≥4 cmof isthmus

Significant Extensively porous-coated stemif <19 mm in diameter. If ≥19mm in diameter, then modulartapered stem.

IIIB Severe proximal meta-physeal bone loss withdiaphysis intact forsome distance

Absent <4 cmof isthmus

Significant Modular tapered stem

IV Complete loss of meta-physeal and diaphy-seal bone

Absent Absent Slight Allograft prosthetic composite,cemented stem, or impactiongrafting plus cemented stem

Neil P. Sheth, MD, et al

October 2013, Vol 21, No 10 603

Clinical Results byReconstruction Method

The goals of femoral component re-vision are to achieve rotational andaxial component stability while re-storing hip biomechanics. Cementedfemoral revisions have demonstratedfailure rates as high as 19%, com-pared with 4% to 6% with unce-mented revisions, which rely on 4 to6 cm of diaphyseal fixation.12,13

When possible, uncemented biologicfixation is the preferred method ofreconstruction (Table 3); however, itmay be necessary to use cementedfixation (Table 4).

Uncemented RevisionProximally Porous-coatedFemoral ComponentsProximally porous-coated uncementedstems can be used for revision THA incases of minimal proximal metaphys-

A, Illustration of Paprosky type II femoral defect. Preoperative (B) andpostoperative (C) AP radiographs of a patient treated with an extensivelyporous-coated femoral stem to manage a loose cemented stem andPaprosky type II bone loss. (Panel A courtesy of DePuy, Warsaw, IN.)

Figure 2

A, Illustration of a Paprosky type IIIA femoral defect. B, Preoperative AP radiograph of a Paprosky type IIIA defect with>4 cm of isthmus remaining for diaphyseal fixation. C and D, Postoperative AP radiographs of a patient treated with asize 20 modular tapered stem to manage an aseptically loose cemented femoral component. An extensively porous-coated stem was not chosen because of the large diameter needed for femoral reconstruction. (Panel A courtesy ofDePuy, Warsaw, IN.)

Figure 3



eal bone loss (ie, Paprosky type I).29

In the setting of proximal bone loss,multiple reports have indicated diffi-culty in obtaining stable proximalmetaphyseal fixation. Berry et al14

assessed 375 femoral revisions per-formed using proximally porous-coated femoral components. At8-year follow-up, the mean survivor-ship was only 52%, using asepticloosening as an end point. Poor sur-vivorship was directly correlatedwith the degree of preoperative boneloss and poor integrity of the remain-ing proximal metaphyseal cancellousbone.

Proximally ModularFemoral ComponentsThe inherent difficulty in achievingadequate initial implant stabilitywith a monoblock proximally coatedstem during femoral revision has re-sulted in increased enthusiasm forthe use of proximally modular femo-ral components such as the S-ROMprosthesis (DePuy). These implantsconsist of a press-fit metaphyseal

sleeve into which a slotted diaphy-seal segment is inserted. This designallows for the metaphysis to be pre-

cisely milled to match the proximalsleeve and accommodates fluted di-aphyseal stems of differing lengths

A, Illustration of a Paprosky type IIIB femoral defect. B, Preoperative APradiograph demonstrating a loose cemented femoral stem with Paproskytype IIIB bone loss. C, AP radiograph following implantation of a modulartapered stem with <4 cm of isthmus remaining for diaphyseal fixation. (PanelA courtesy of DePuy, Warsaw, IN.)

Figure 4

A, Illustration of a Paprosky type IV femoral defect. B, Preoperative AP radiograph demonstrating Paprosky type IVbone loss secondary to periprosthetic infection. C and D, Postoperative AP radiographs following reimplantation withan allograft-prosthesis construct to manage the bone loss. (Panel A courtesy of DePuy, Warsaw, IN.)

Figure 5


October 2013, Vol 21, No 10 605

and configurations. Proximal modu-larity also addresses proximal femo-ral retroversion by allowing versionto be dialed in separately. Unlikeproximally coated nonmodular com-ponents, the results of femoral revi-sions with these prostheses have beenfavorable.

Cameron15 reported on 320 revi-sions performed with S-ROM stems(109 short, 211 long). At a meanfollow-up of 7 years (range, 2 to 12years), there were no reported revi-sions for aseptic loosening in theshort stem group and only 3 in thelong stem group (1.4%). Subsidencewas >5 mm in only two hips (0.6%),

and there were no cases of distal os-teolysis. The author concluded thatproximally coated, proximally mod-ular stems can be used successfully inthe setting of femoral revision.

Extensively Porous-coatedFemoral StemsExtensively porous-coated stems havebeen the workhorse for femoral revisionTHA. In general, 6-inch stems are suf-ficient for most type II and IIIA femo-ral defects. However, when using lon-ger, bowed stems (eg, 8 in, 10 in), it isimportant to ensure adequate hip sta-bility because suboptimal stem antever-sion may be dictated as a consequence

of the femoral bow.Weeden and Paprosky16 evaluated

170 revisions over a mean of 14.2years. The femoral defects were clas-sified as type I (11%), type II (30%),type IIIA (48%), and type IIIB(11%). The overall mechanical fail-ure rate, that is, the percentage ofstems that required revision surgeryor were radiographically unstable,was 4.1%. Eighty-two percent ofhips had radiographic evidence ofbone ingrowth, and 14% had stablefibrous fixation. Four percent wereunstable radiographically. There wasa high rate of failure (21%) withworsening bone loss (type IIIB), and

Table 3

Results of Uncemented Reconstruction by Type of Bone Loss

Study No. of Hips Stem Type

Berry et al14 375 Proximal metaphyseal fitting

Cameron15 320 Proximal modular (S-ROM [DePuy])

Weeden and Paprosky16 170 Extensively porous-coated

Sporer and Paprosky17 51 Extensively porous-coated

Park et al18 62 Modular taperedGarbuz et al19 Modular tapered stem, 31. Extensively

porous-coated stem (Solution [DePuy]),189.

Modular tapered vs extensively porous-coated

Richards et al20 Modular tapered stem, 103. Extensivelyporous-coated stem, 114.

Modular tapered vs extensively porous-coated

Grünig et al21 38 Nonmodular tapered

Isacson et al22 43 Nonmodular tapered

N/A = not applicablea Paprosky classification unless otherwise noted.b Class I, partial or complete cortical loss above the level of the lesser trochanter; class II, partial or complete cortical loss above a point10 cm below the lesser trochanter; class III, partial or complete cortical loss >10 cm below the lesser trochanter.c Class I, bone below the lesser trochanter is intact, and generally, a primary stem can be used; class II, subtrochanteric bone is damagedsignificantly and requires the use of a long stem; class III, >70 mm of the proximal femur is completely missing, which requires the use of astructural allograft.d Different Paprosky classification: type 1, minimal metaphyseal and diaphyseal loss; type 2A, absent calcar below the intertrochanteric line;type 2B, absent anterolateral metaphyseal bone; type 2C, absent calcar and posteromedial bone; type 3A, B, and C, same as type 2A, B, andC but with diaphyseal extension.



the intraoperative fracture rate withstem insertion was 8.8%.

The limitations of extensivelyporous-coated stems in the setting offemoral revision were identified bySporer and Paprosky17 in a study of51 patients with type III or IV femo-ral bone defects. No failures were re-ported in the 17 femurs with typeIIIA defects. One failure was re-ported in the 15 patients with typeIIIB defects and femoral canals <19mm in diameter (6.7%), and an 18%rate of mechanical failure (ie, revi-sion for aseptic loosening or radio-graphic evidence of instability) wasreported in patients with type IIIB

defects and femoral canals >19 mmin diameter (2 of 11). Additionally,three of eight patients with type IVdefects treated with extensivelyporous-coated stems experiencedmechanical failure. These authorsrecommended use of a modular ta-pered stem or impaction bone graft-ing in type IIIB defects with canals>19 mm in diameter and in type IVdefects.

Tapered StemsEnthusiasm for tapered stems hasgrown in the past decade, and theyarguably have become the newworkhorse for femoral revision sur-

gery in the setting of advanced boneloss. Tapered stems can be nonmod-ular or modular. Final seating ofthese devices is sometimes difficult topredict during bone preparation.Modular designs allow more predict-able restoration of length, offset, andversion. However, concerns exist re-lated to stem fracture at the Morsetaper with these designs.

Modular Tapered StemsModular tapered revision femoralcomponents have been successfullyused in the reconstruction of femurswith moderate to severe proximalbone loss. Park et al18 followed 62

Table 3 (continued)

Results of Uncemented Reconstruction by Type of Bone Loss

Bone Loss ClassificationaMean Clinical

Follow-up Re-revision Rate (%) Survivorship

Minimal, 49; class I, 60; class II, 218;class III, 38;b periprosthetic fractures, 10

4.7 y 16 52% at 8 y. Worse preoperativebone loss correlated withpoorer survivorship.

Class I, 109; classes II and III, 211c 7 y Class I, none; class II and III,1.4

N/A

Type I, 18; type II, 51; type IIIA, 82; typeIIIB, 19

14.2 y Overall failure rate, 4.1. Failurerate with type IIIB bone loss,21.

N/A

Type IIIA, 17; type IIIB, 26 (15 with ≤19mm canal diameter, 11 with >19 mm ca-nal diameter); type IV, 8

4.2 y Type IIIA, none; type IIIB (≤19mm canal diameter), 6.7; typeIIIB (>19 mm canal diameter),27.3; type IV, 25

N/A

Type IIIA, 37; type IIIB, 19 4.2 y None N/AModular group: type I, 4; type II, 5; type

IIIA, 29; type IIIB, 58; type IV, 7. Non-modular group: type I, 1; type II, 15;type IIIA, 60; type IIIB, 31; type IV, 4.

Modular, 37 mo; non-modular, 49 mo

Quality of life measures betterwith modular tapered stems

N/A

Modular group: types IIIB and IV, 65. Non-modular group: types IIIB and IV, 35.

23 mo for each Quality of life measures betterwith modular tapered stems

N/A

Type 1, 1; type 2A, 6; type 2B, 6; type 2C,5; type 3A/B/C, 9d

47 mo 7.5, to manage stem subsi-dence (3 of 40)

N/A

— 25 mo 18.6, due to subsidence andinstability

N/A

N/A = not applicablea Paprosky classification unless otherwise noted.b Class I, partial or complete cortical loss above the level of the lesser trochanter; class II, partial or complete cortical loss above a point10 cm below the lesser trochanter; class III, partial or complete cortical loss >10 cm below the lesser trochanter.c Class I, bone below the lesser trochanter is intact, and generally, a primary stem can be used; class II, subtrochanteric bone is damagedsignificantly and requires the use of a long stem; class III, >70 mm of the proximal femur is completely missing, which requires the use of astructural allograft.d Different Paprosky classification: type 1, minimal metaphyseal and diaphyseal loss; type 2A, absent calcar below the intertrochanteric line;type 2B, absent anterolateral metaphyseal bone; type 2C, absent calcar and posteromedial bone; type 3A, B, and C, same as type 2A, B, andC but with diaphyseal extension.


October 2013, Vol 21, No 10 607

femoral revisions using a fluted mod-ular tapered component for a meanof 4.2 years (range, 2 to 7.8 years).

Thirty-seven (60%) were classified asPaprosky type IIIA, and 19 (31%)were type IIIB. None of the patients

in this cohort required revision dueto mechanical failure at final follow-up.

In similarly designed studies, Gar-buz et al19 and Richards et al20 com-pared the results of tapered, fluted,modular titanium femoral compo-nents with cylindrical nonmodularcobalt chrome stems in revision ar-throplasty. Both studies reported su-perior results with the modular ta-pered components. Richards et al20

found that although the modular ta-pered cohort had worse preoperativebone defects (65% Paprosky typesIIIB and IV femurs versus 35% in thenonmodular group), they had betterclinical outcome scores (ie, WesternOntario and McMaster UniversitiesOsteoarthritis Index, Oxford HipScore, satisfaction), fewer intraoper-ative fractures, and better restorationof the proximal femoral host bone.Overall, modular tapered femoralcomponents have shown excellentpromise in short- to midterm studiesfor revision THA in patients withsubstantial proximal bone loss.These stem designs are widely usedin revision THA, and longer-term

Table 4

Results of Cemented Reconstruction by Type of Bone Loss

Study No. of Hips Type of Reconstruction Stem Type

Ornstein et al23 1,305 Cemented impaction grafting Polished cemented

Blackley et al24 63 Cemented APC Cemented

Safir et al25 50 Cemented APC CementedBabis et al26 57 Cemented APC Cemented

Malkani et al28 50 Cemented Proximal femoral replacementHaentjens et al27 16 Cemented Proximal femoral replacement

APC = allograft prosthesis composite, N/A = not applicablea Paprosky classification unless otherwise noted.b AAOS classification of femoral bone loss: level I, bone loss up to the level of the lesser trochanter; level II, bone loss up to 10 cm distal tothe lesser trochanter; level III, bone loss distal to 10 cm below the lesser trochanter (this also depicts loss of host-prosthesis contact with theneed for structural bone graft).

A, Preoperative AP radiograph demonstrating femoral bone loss secondaryto osteolysis. B, Postoperative AP radiograph following treatment withimpaction grafting.

Figure 6



studies are needed to determine theirefficacy.

Nonmodular Tapered StemsElimination of the modular junctionin femoral component revision de-creases the risk of stem fracture, fret-ting corrosion, metallosis, and resul-tant osteolysis. However, the lack ofmodularity makes proper componentposition and restoration of hip bio-mechanics more difficult. Grüniget al21 evaluated 38 revisions per-formed with one particular nonmod-ular tapered stem. At a meanfollow-up of 47 months, 3 (8%) of38 hips required revision for stemsubsidence, and an additional 16stems had subsided <1 cm by3-month follow-up. This early subsi-dence did not appear to affect clini-cal outcome, and the authors recom-mended protected early weightbearing.

Isacson et al22 reported results us-ing the same type stem in 43 hips. Ata mean follow-up of 25 months, 22of 23 patients (96%) showed abun-dant new bone formation. However,5 of 42 patients (12%) had subsi-dence >20 mm, and 22 (52%) hadsubsidence <5 mm. There were nine

dislocations, of which eight requiredre-revision to manage instability.

Cemented Revision

Impaction GraftingImpaction grafting is performed inan attempt to restore bone stock inyoung or active patients with severeproximal bone loss (ie, Paproskytype IIIB or IV defects). The old stemis extracted, and the canal is dé-brided of all previous cement, neo-cortex, and fibrous tissue or en-dosteal membrane. Deficient corticesare reinforced as necessary with anycombination of wire mesh, strut al-lograft, and cerclage wires. A plug isplaced distally, and morcellized can-cellous allograft is tightly packedinto the canal using cannulatedtamps and broaches over a guiderod. The revision stem (typically apolished tapered stem) is then ce-mented into the reconstituted femur.

Initial reports on impaction graft-ing described variable outcomes.Centers with significant experiencewith this technique reported verygood short- and intermediate-termresults; however, many other centersreported high rates of femoral com-

ponent subsidence as well as intraop-erative or early postoperative femo-ral fracture. Some recent studies havedemonstrated satisfactory long-termresults with a high level of constructsurvivorship.

Ornstein et al23 evaluated 1,188 re-visions performed with impactiongrafting using a polished Exeter stem(Stryker). The original cohort con-sisted of 1,305 revisions. Clinicaland radiographic follow-up rangedfrom 5 to 18 years. Only 70 cases re-quired re-revision (5.9%). The survi-vorship at 15 years was 94.0% forwomen and 94.7% for men, usingany reason for revision as an endpoint. Survivorship at 15 years was99.1% for aseptic loosening, 98.6%for infection, 99% for subsidence,and 98.7% for fracture. Overall,long-term results of impaction graft-ing are encouraging, but proper pa-tient selection is required, and theprocedure is labor intensive, requir-ing adequate surgeon experience(Figure 6).

Allograft Prosthetic CompositeThe use of an APC should be consid-ered in the setting of severe circum-ferential femoral bone loss. With this

Table 4 (continued)

Results of Cemented Reconstruction by Type of Bone Loss

Bone Loss ClassificationaClinical Follow-up in

Years Re-revision Rate (%) Survivorship

— Range, 5–18 5.4 94% for women and 94.7% formen at 15 y

Level II and III bone lossb Mean, 11 27 (graft resorption in 13 of 48hips)

77% at final follow-up

— Mean, 16.2 14 N/AType IV, 55 Mean, 12 33 69% at 10 y. Survivorship de-

creased with worse bone loss(ie, type IV).

Type IV, 33 Mean, 11.1 32 64% at 12 yNot reported Mean, 5 Not reported N/A

APC = allograft prosthesis composite, N/A = not applicablea Paprosky classification unless otherwise noted.b AAOS classification of femoral bone loss: level I, bone loss up to the level of the lesser trochanter; level II, bone loss up to 10 cm distal tothe lesser trochanter; level III, bone loss distal to 10 cm below the lesser trochanter (this also depicts loss of host-prosthesis contact with theneed for structural bone graft).


October 2013, Vol 21, No 10 609

technique, the deficient proximal fe-mur is osteotomized and removed. Along-stem prosthesis then is ce-mented into the bulk proximal femo-ral allograft, and this APC is matedto the host bone while the distal partof the stem is press-fit or cementedinto the host femoral canal. Stablefixation between the APC and host isenhanced by press-fitting the distalstem and by creating a step cut at theAPC-host interface, thereby increas-ing the surface area for creeping sub-stitution. Cerclage wires are alsoused to enhance junctional stability.

Advantages of reconstruction withan APC over a proximal femoralprosthesis include the ability to re-store bone stock in young patientsand the ability to reattach host softtissues. One disadvantage is the po-tential for disease transmission. Therisk of viral transmission with fresh-frozen, unprocessed allograft is ap-proximately 1 in 500,000 (range,440,000 to 600,000).30 There is alsoa risk of secondary bacterial infec-tion. Other disadvantages includedifficulty obtaining the appropriateallograft, the risk of nonunion orgraft resorption, and greater techni-cal demands.31

Several studies, most from the or-thopaedic oncology literature, havereported encouraging results withthe use of this technique in revisionTHA, although most demonstrateshort-term clinical follow-up. Black-ley et al24 reported on 63 consecutiverevisions using an APC construct.With a mean follow-up of 11 years(range, 9 years 4 months to 15years), the success rate was 77% forthe 45 patients who were alive at thelatest follow-up. The nonunion ratewas 6% at the host-allograft junc-tion, and all patients required treat-ment with autograft. Allograft re-sorption was seen in 27% of patientswho were followed for at least 9years. Four of the 63 hips dislocated,and 2 of these required acetabular

revision. Five hips developed deep in-fection, all of which required reoper-ation. There were only three cases ofaseptic loosening, and all occurred atthe implant-cement interface. Theaverage time to loosening was morethan 10 years.

Safir et al25 recently published aretrospective study with an averageclinical and radiographic follow-upof 16.2 years (range, 15 to 22 years).They reported the results of 50 hipsout of an original cohort of 93.Seven APC constructs were revisedfor any reason. The authors con-cluded that proximal femoral al-lograft in revision THA is a durableoption for most patients with severefemoral bone loss.

Babis et al26 evaluated the use of APCin the setting of complex revision THAfor severe proximal femoral bone loss.A total of 57 hips was available at amean follow-up of 12 years (range, 8to 20 years). APC survivorship at 10years was 69%, with 19 hips (33.3%)requiring revision at a mean follow-upof 44.5 months (range, 11 to 153months). Survivorship was significantlyaffected by the degree of preoperativebone loss (ie, Paprosky type IV) (P =0.019), the number of previous hip sur-geries exceeding two (P = 0.047), andthe length of the APC graft (P = 0.005).Satisfactory results were seen with theuse of APC to manage severe proximalbone loss in revision THA.

Megaprosthesis (ProximalFemoral Replacement)The use of proximal femoral–replac-ing prostheses (ie, megaprostheses)has substantial disadvantages andshould be limited to elderly and low-demand patients with massive boneloss for whom the alternative is re-section arthroplasty. Disadvantagesinclude problems with fixation andearly loosening,11,32,33 instability sec-ondary to inadequate soft-tissue at-tachment,27,34 severe stress shielding

and late fatigue fracture, limb-lengthdiscrepancy, sciatic nerve palsy, andcost.35,36 Additionally, further loss ofbone stock makes subsequent revi-sion more challenging. The one ad-vantage is that implantation can bedone quickly, which makes this anattractive reconstructive option forelderly patients in poor health.

Few reports describe the use ofmegaprostheses in the setting of fem-oral revision THA. Malkani et al28

published long-term results using aproximal femoral replacement for non-neoplastic disorders. Thirty-three of 50hips were revised to address femoralbone loss. The mean clinical follow-upwas 11.1 years (range, 5.1 to 18.8years). Overall survivorship was 64%at 12 years. Sixteen components in 12patients required revision for any rea-son (32%). Eleven hips dislocated(22%), 4 of which required re-revision.Additionally, 27% of patients hadmoderate to severe pain, and 48% ofpatients had a severe limp or were un-able to walk. Harris Hip scores im-proved from 46 points preoperativelyto 76 points at latest follow-up.

Haentjens et al27 evaluated 16 pa-tients treated with proximal femoralreplacement for salvage of a failedTHA. At a mean follow-up of 5years (range, 2 to 11 years), all pa-tients reported pain relief, but all pa-tients also required an assistive de-vice for ambulation. Four patientssustained an intraoperative fracture,seven had a dislocation, and two haddeep infection. Given the high rate ofcomplications and limited postopera-tive function provided by the mega-prostheses, revision with this type ofconstruct should be considered onlyas a salvage procedure.

Summary

Management of femoral bone lossduring revision THA begins withproper preoperative evaluation. In



patients who require femoral compo-nent revision, bone loss should beclassified to help determine an ap-propriate prosthesis and fixationstrategy. For Paprosky types I, II,and IIIA bone loss, cylindrical fullyporous-coated uncemented femoralcomponents have been associatedwith predictable long-term fixation.In our experience, type IIIB defectsare generally best managed withmodular tapered fluted stems; how-ever, some centers have had goodlong-term results with impactionbone grafting. APC and megapros-theses should be considered part ofthe armamentarium for managingtype IV femoral bone loss.

References

Evidence-based Medicine: Levels ofevidence are described in the table ofcontents. In this article, references 2,4, 6-29, and 31-36 are level II stud-ies. Reference 1 is a level III study.Reference 3 is level V expert opinion.

References printed in bold type arethose published within the past 5years.

1. Ghanem E, Parvizi J, Burnett RS, et al:Cell count and differential of aspiratedfluid in the diagnosis of infection at thesite of total knee arthroplasty. J BoneJoint Surg Am 2008;90(8):1637-1643.

2. Della Valle CJ, Sporer SM, Jacobs JJ,Berger RA, Rosenberg AG, PaproskyWG: Preoperative testing for sepsis be-fore revision total knee arthroplasty.J Arthroplasty 2007;22(6 suppl 2):90-93.

3. Della Valle CJ, Paprosky WG: The femurin revision total hip arthroplasty evalua-tion and classification. Clin OrthopRelat Res 2004;(420):55-62.

4. Puri L, Wixson RL, Stern SH, Kohli J,Hendrix RW, Stulberg SD: Use of helicalcomputed tomography for the assess-ment of acetabular osteolysis after totalhip arthroplasty. J Bone Joint Surg Am2002;84(4):609-614.

5. Archibeck MJ, Rosenberg AG, BergerRA, Silverton CD: Trochanteric osteot-omy and fixation during total hip arthro-plasty. J Am Acad Orthop Surg 2003;11(3):163-173.

6. Paprosky WG, Sporer SM: Controlledfemoral fracture: Easy in. J Arthroplasty2003;18(3 suppl 1):91-93.

7. Maloney WJ, Herzwurm P, Paprosky W,Rubash HE, Engh CA: Treatment of pel-vic osteolysis associated with a stableacetabular component inserted withoutcement as part of a total hip replace-ment. J Bone Joint Surg Am 1997;79(11):1628-1634.

8. D’Antonio J, McCarthy JC, Bargar WL,et al: Classification of femoral abnormal-ities in total hip arthroplasty. ClinOrthop Relat Res 1993;(296):133-139.

9. Paprosky WG, Aribindi R: Hip replace-ment: Treatment of femoral bone lossusing distal bypass fixation. Instr CourseLect 2000;49:119-130.

10. Pak JH, Paprosky WG, Jablonsky WS,Lawrence JM: Femoral strut allografts incementless revision total hip arthro-plasty. Clin Orthop Relat Res 1993;(295):172-178.

11. Aribindi R, Barba M, Solomon MI, ArpP, Paprosky W: Bypass fixation. OrthopClin North Am 1998;29(2):319-329.

12. Kavanagh BF, Ilstrup DM, FitzgeraldRH Jr: Revision total hip arthroplasty. JBone Joint Surg Am 1985;67(4):517-526.

13. Pellicci PM, Wilson PD Jr, Sledge CB,et al: Long-term results of revision totalhip replacement: A follow-up report.J Bone Joint Surg Am 1985;67(4):513-516.

14. Berry DJ, Harmsen WS, Ilstrup D,Lewallen DG, Cabanela ME: Survivor-ship of uncemented proximally porous-coated femoral components. ClinOrthop Relat Res 1995;(319):168-177.

15. Cameron HU: The long-term success ofmodular proximal fixation stems in revi-sion total hip arthroplasty. J Arthro-plasty 2002;17(4 suppl 1):138-141.

16. Weeden SH, Paprosky WG: Minimal11-year follow-up of extensively porous-coated stems in femoral revision total hiparthroplasty. J Arthroplasty 2002;17(4suppl 1):134-137.

17. Sporer SM, Paprosky WG: Revision totalhip arthroplasty: The limits of fullycoated stems. Clin Orthop Relat Res2003;(417):203-209.

18. Park YS, Moon YW, Lim SJ: Revisiontotal hip arthroplasty using a fluted andtapered modular distal fixation stemwith and without extended trochantericosteotomy. J Arthroplasty 2007;22(7):993-999.

19. Garbuz DS, Toms A, Masri BA, DuncanCP: Improved outcome in femoral revi-sion arthroplasty with tapered flutedmodular titanium stems. Clin OrthopRelat Res 2006;453:199-202.

20. Richards CJ, Duncan CP, Masri BA,Garbuz DS: Femoral revision hip arthro-plasty: A comparison of two stem de-signs. Clin Orthop Relat Res 2010;468(2):491-496.

21. Grünig R, Morscher E, Ochsner PE:Three-to 7-year results with the unce-mented SL femoral revision prosthesis.Arch Orthop Trauma Surg 1997;116(4):187-197.

22. Isacson J, Stark A, Wallensten R: TheWagner revision prosthesis consistentlyrestores femoral bone structure. IntOrthop 2000;24(3):139-142.

23. Ornstein E, Linder L, Ranstam J, LewoldS, Eisler T, Torper M: Femoral impac-tion bone grafting with the Exeter stem:The Swedish experience. Survivorshipanalysis of 1305 revisions performedbetween 1989 and 2002. J Bone JointSurg Br 2009;91(4):441-446.

24. Blackley HR, Davis AM, Hutchison CR,Gross AE: Proximal femoral allograftsfor reconstruction of bone stock in revi-sion arthroplasty of the hip: A nine tofifteen-year follow-up. J Bone Joint SurgAm 2001;83(3):346-354.

25. Safir O, Kellett CF, Flint M, BacksteinD, Gross AE: Revision of the deficientproximal femur with a proximal femoralallograft. Clin Orthop Relat Res 2009;467(1):206-212.

26. Babis GC, Sakellariou VI, O’Connor MI,Hanssen AD, Sim FH: Proximal femoralallograft-prosthesis composites in revi-sion hip replacement: A 12-yearfollow-up study. J Bone Joint Surg Br2010;92(3):349-355.

27. Haentjens P, De Boeck H, Opdecam P:Proximal femoral replacement prosthesisfor salvage of failed hip arthroplasty:Complications in a 2-11 year follow-upstudy in 19 elderly patients. Acta OrthopScand 1996;67(1):37-42.

28. Malkani AL, Settecerri JJ, Sim FH, ChaoEY, Wallrichs SL: Long-term results ofproximal femoral replacement for non-neoplastic disorders. J Bone Joint SurgBr 1995;77(3):351-356.

29. Meding JB, Ritter MA, Keating EM,Faris PM: Clinical and radiographic eval-uation of long-stem femoral componentsfollowing revision total hip arthroplasty.J Arthroplasty 1994;9(4):399-408.

30. Tomford W: Current concepts review:Transmission of disease through trans-plantation of musculoskeletal allografts.J Bone Joint Surg Am 1995;77(11):1742-1754.

31. Chandler H, Clark J, Murphy S, et al:Reconstruction of major segmental lossof the proximal femur in revision totalhip arthroplasty. Clin Orthop Relat Res1994;(298):67-74.

32. Morris HG, Capanna R, Del Ben M,


October 2013, Vol 21, No 10 611

Campanacci D: Prosthetic reconstructionof the proximal femur after resection forbone tumors. J Arthroplasty 1995;10(3):293-299.

33. Sim FH, Chao EY: Hip salvage by proxi-mal femoral replacement. J Bone JointSurg Am 1981;63(8):1228-1239.

34. Malkani AL, Sim FH, Chao EY:Custom-made segmental femoral re-placement prosthesis in revision total hiparthroplasty. Orthop Clin North Am1993;24(4):727-733.

35. Clarke HD, Berry DJ, Sim FH: Salvageof failed femoral megaprostheses withallograft prosthesis composites. ClinOrthop Relat Res 1998;(356):222-229.

36. Renard AJ, Veth RP, Schreuder HW,Schraffordt Koops H, van Horn J, KellerA: Revisions of endoprosthetic recon-structions after limb salvage in musculo-skeletal oncology. Arch Orthop TraumaSurg 1998;117(3):125-131.



Review Article Femoral Bone Loss in Revision Total Hip ... · bone loss (metaphyseal or...

Documents

Transcript of Review Article Femoral Bone Loss in Revision Total Hip ... · bone loss (metaphyseal or...