j aaos -2008

Recent Developments in theBiology of Fracture Repair

AbstractFracture repair is dependent on local and systemic molecular andcellular processes. During fracture repair, mesenchymal stem cellsare systemically recruited to the fracture site, and cytokines arereleased from the fracture site into the vascular system. In asignificant minority of fractures, healing delays result from adverseclinical factors that interfere with these processes. Extrinsicfactors, such as aging and smoking, adversely affect the molecularand cellular processes occurring locally in the fracture site.Fracture fixation affects healing through local changes in thebiologic signaling within the fracture callus. Current biologictreatment of fractures includes the local application ofosteoinductive bone morphogenetic proteins (ie, BMP-2, BMP-7)and cell-based therapies. Although clinical results with bonemorphogenetic proteins have been satisfactory, they have not beenas impressive as those reported in animal studies. Furtherunderstanding of the biology of fracture repair may lead toimproved treatment modalities.

Fracture healing is a highly effi-cient repair process resulting in

newly formed bone, similar in qual-ity to the original tissue. However,in a small but significant number ofinstances, adverse conditions impairthis process, causing significantmorbidity. Because fractures arecommon in the general population,delayed healing and nonunion aresignificant health care issues. Thus,there remains a need to developtreatment methods that enhancefracture healing and improve out-comes.

Biologic methods of bone regener-ation will continue to have an in-creasing role in the treatment offractures. To further develop thesemethods, knowledge of the patho-logic changes in the fracture repairprocess that lead to delayed union ornonunion is important.

Fracture healing may be seen asboth a local and a systemic process(Figures 1 and 2). There are local mo-lecular and cellular signaling path-ways, and evidence is emerging ofthe systemic recruitment of mesen-chymal stem cells (MSCs) to thefracture site. Extrinsic factors, suchas drugs and aging, influence localprocesses to alter the outcome offractures, while the biomechanics offracture fixation affect the biology offracture healing.

Pathophysiology ofDelayed FractureHealing and Nonunion

Bone regeneration depends on threeessential elements: progenitor cells,growth factors (osteoinduction), andthe appropriate milieu (osteoconduc-tion). Delayed fracture repair and

Francois N. K. Kwong, MD

Mitchel B. Harris, MD

Perspectives on Modern Orthopaedicsarticles provide an objective appraisal ofnew or controversial techniques or areasof investigation in orthopaedic surgery.

Dr. Kwong is Research Fellow, Centerfor Molecular Orthopaedics, Brighamand Women’s Hospital, Harvard MedicalSchool, Boston, MA. Dr. Harris isAssociate Professor of OrthopaedicSurgery, Harvard Medical School, andChief, Orthopaedic Trauma Service,Brigham and Women’s Hospital,Partners Orthopaedic Trauma Service,Boston, MA.

None of the following authors or amember of their immediate families hasreceived anything of value from or ownsstock in a commercial company orinstitution related directly or indirectly tothe subject of this article: Dr. Kwong andDr. Harris.

Reprint requests: Dr. Kwong, Center forMolecular Orthopaedics, Room BLI044, 221 Longwood Avenue, Boston,MA 02115.

J Am Acad Orthop Surg 2008;16:619-625

Copyright 2008 by the AmericanAcademy of Orthopaedic Surgeons.

Perspectives on Modern Orthopaedics

Volume 16, Number 11, November 2008 619

nonunion can result from a lack ofosteoprogenitor cells, insufficientosteoinductive growth factors, or adefective milieu, or, more common-ly, a combination of these factors.Disease or other adverse factors maydelay fracture healing by affectingone or more of these elements.

For example, a critical-sized bonedefect (ie, one that does not healspontaneously because of its size)would result in the absence of hu-man MSCs (hMSCs) within the de-fect, a lack of osteoinductive signals,and a milieu that is nonconducive tohealing. Excess motion at the frac-ture site impairs both the molecularand cellular processes within the de-fect. Hypoxia affects osteogenic dif-ferentiation of hMSCs by reducingthe number of viable cells and alter-ing the molecular signals pro-duced.1 Blood supply is important tofracture repair as blood provides nu-trients and oxygen for cell survival,and blood vessels are the route forinflammatory and osteoprogenitorcells that are recruited to the frac-ture site. Systemic factors may affectfracture repair by reducing the num-ber of osteoprogenitor cells recruited

to the fracture site and/or by affect-ing the local osteoinductive signals.

Nonunions are commonly classi-fied on radiographs as hypertrophicor atrophic. Hypotrophic nonunionis generally thought to be the resultof mechanical instability and istreated by restoring stability, usu-ally by skeletal fixation. Atrophicnonunion is thought to result frombiologic causes, principally poor vas-cularization, and is treated by resto-ration of the osteogenic potential,with resection of fibrous tissue andbone grafting or some other methodof osteoinduction. Both groups ofnonunion contain fibrous tissue, fi-brocartilage, and adipose tissue,which are not normally present inhealing fractures.2 The hypertrophicgroup also has hyaline cartilage andbone in varying proportions.

Local MolecularSignaling

Osteoinductive MoleculesFracture repair is regulated by

several growth factors with varyingosteogenic potential, such as trans-forming growth factor-β, platelet-

Figure 1

Illustration demonstrating molecularsignaling. Fracture repair depends onmolecular processes occurring bothlocally and systemically. BMPs = bonemorphogenetic proteins, IGF-1 =insulin-like growth factor-1, NSAIDs =nonsteroidal anti-inflammatory drugs,TGF-β = transforming growth factor-β,- = inhibits

Figure 2

Illustrations of cellular signaling and the role of mesenchymal stem cells (MSCs) in fracture healing. A, Source of MSCs.B, Action of MSCs within fracture callus. BMPs = bone morphogenetic proteins

Recent Developments in the Biology of Fracture Repair

620 Journal of the American Academy of Orthopaedic Surgeons

derived growth factor, insulin-likegrowth factor-1, and bone morphoge-netic protein (BMP). Of these, BMPappears to be among the most conse-quential. BMP was discovered andnamed by Urist3 in 1965; he also firstdescribed the phenomenon of os-teoinduction. Urist observed newbone formation occurring locally inrodents after they were given intra-muscular implantation of bone cyl-inders decalcified with hydrochloricacid. This phenomenon was attrib-uted to the presence of a protein,BMP, in bone matrix. Since that dis-covery, at least 16 different humanBMPs have been identified. Theseproteins affect cells and tissues in-volved in the repair process in anumber of ways, including the re-cruitment of MSCs from surround-ing tissues to the fracture site, fol-lowed by their proliferation anddifferentiation into chondrocytesand osteoblasts, invasion of bloodvessels, and, ultimately, bone forma-tion. All of these effects are mediat-ed by the binding of BMPs to specif-ic transmembrane receptors onhMSCs, active osteoblasts, and ma-ture chondrocytes as well as to thesubsequent activation of various in-tracellular messenger systems.4

Therapeutic Applicationof BMPs

Extensive animal data have dem-onstrated the potential of BMPs toinduce healing of critical-sized (ie,large) bone defects.5-7 In animal mod-els, BMPs alone (with their carriermatrix) have been shown to inducerapid bone bridging of a defect. Thequality of the repair tissue wasequivalent to or better than that ob-tained with autologous bone graft-ing, the standard treatment for bonedefects and nonunions in clinicalpractice.6,7 At present, only BMP-2(Infuse; Medtronic Sofamor Danek,Memphis, TN) and BMP-7 (OP-1 Im-plant; Stryker Biotech, Hopkinton,MA) have been approved by the USFood and Drug Administration forclinical use. Several clinical studies

have already demonstrated the pos-itive effect of the application ofBMPs on the outcome of fracturesand nonunions. Because a review ofall available evidence is beyond thescope of this article, we will focus onthe treatment of segmental bone de-fects in patients because this allowsa direct comparison with the studiesof BMPs in animals.

In most clinical studies of thetreatment of segmental bone de-fects, BMPs have been used in con-junction with allograft or autograftbone. Jones et al8 demonstrated thata combination of BMP-2 and al-lograft bone was equivalent to autol-ogous bone for the treatment of seg-mental bone defects. In that study,patients with a tibial diaphysealfracture and a residual cortical de-fect were randomly assigned to re-ceive either autogenous bone graftor allograft with an onlay applica-tion of recombinant human BMP-2(rhBMP-2). Radiographic and func-tional outcomes were similar inboth groups. To date, the only pub-lished clinical study on the treat-ment of segmental defects withBMPs alone (with a nonosteocon-ductive carrier matrix only) in hu-mans showed healing of critical-sized fibular defects in patientsundergoing opening wedge high tib-ial osteotomy with fibulectomy.9

RhBMP-7 bound to collagen type1 sponge induced bony union in fiveof six patients with a critical-sizedfibular defect, whereas there was nohealing in any of the six patientstreated with the type 1 collagen car-rier only. Despite these favorable re-sults, no large studies have beendone on the use of BMPs alone inhumans.

The pace of healing of segmentaldefects treated with BMPs differssignificantly among species. Seg-mental defects in large animalstreated with BMPs alone healed in<3 months,5,6 whereas in humans,critical tibial defects treated with al-lograft bone and BMP-2 required ≥6months to achieve bony union.8 The

cause for this difference remains un-clear. Clinically, poor fracture heal-ing in humans may be associatedwith adverse factors not present inanimal studies. For example, soft-tissue coverage of the fracture maynot be adequate. The initial dose ofBMPs given to human subjects (7 mgof BMP-7 and 2 g of collagen carrier,or 12 mg of rhBMP-2 and collagensponge) was much higher than in theanimal studies. Because the releaseof BMP inhibitors depends on the ex-tracellular level of BMPs, it is postu-lated that this higher concentrationof BMPs leads to the expression ofseveral BMP antagonists, which fur-ther limits their efficacy and reduc-es the rate of bone healing. It is alsospeculated that BMP receptors in an-imals and humans are different intheir degree of responsiveness to theBMP molecules.

Role of BMPs andTheir Inhibitors inFracture Healing

The activity of BMPs can be lim-ited by several antagonists, whichbind to them and interfere with theirability to induce receptor activation.One of the most characterized BMPinhibitors is noggin, a protein thatbinds to both BMP-2 and BMP-7 andantagonizes their actions by prevent-ing binding with their membranereceptors.10 Its expression by osteo-blasts is induced by BMP-2,10 imply-ing that BMP-2 and noggin are in-volved in a negative feedback loopduring bone formation. This mayprovide a physiologic mechanismthat prevents overexposure of osteo-blasts to BMP signaling.

The balance between BMPs andtheir inhibitors is likely to be a crit-ical determinant of fracture healing,with a decreased expression of BMPsand/or a relative increase of BMP an-tagonists adversely affecting healing.In a rat model of fracture nonunion,a downregulation of the gene expres-sion of BMPs was demonstrated.11 Inan animal model of atrophic non-union, reversal of this decreased ex-

Francois N. K. Kwong, MD, and Mitchel B. Harris, MD


pression of osteoinductive factors,induced by an early local injection ofrhBMP-7, prevented the develop-ment of nonunion.12 The expressionof chordin, a BMP antagonist with amode of action similar to that ofnoggin, was upregulated in an ani-mal model of fracture nonunion,11

suggesting that downregulation ofchordin in a fracture nonunion hasthe potential of improving bonehealing.

In normally healing fractures, thebalance between BMPs and their in-hibitors can also be manipulated tohasten repair. This would involvethe addition of the osteoinductivefactors (eg, BMP-2), inhibition of theactivity of BMP inhibitors, or a com-bination of both methods. So far, bi-ologic methods of enhancing boneregeneration have centered on thepromotion of osteoinduction via thedelivery of BMPs. However, it wasrecently demonstrated that nogginor chordin suppression can acceler-ate osteogenesis in vitro13,14 and thatnoggin knockdown increased therate of intramembranous ossifica-tion in an animal model.14 We be-lieve that these findings will be ex-tended to fracture healing and thatblockade of the activity of BMP in-hibitors may provide a novel strate-gy for expediting fracture repair.

Local Cellular Signaling

Complete fracture healing requiresthat a sufficient number of hMSCsdifferentiate into chondrocytes andosteoblasts, as well as other cells ofthe mesenchymal lineage, such asadipocytes, and stromal and endo-thelial cells. In addition to differen-tiation, the trophic, or nutritional,activity of MSCs in the repair of oth-er tissues is now well established.15

This refers to the capacity of MSCsto secrete growth factors, whichstimulate blood vessel formationand the proliferation of other localMSCs.15 It is postulated that MSCsexert a trophic activity in the earlystages of fracture repair, although

this has not yet been specificallydemonstrated in fractures. MSCs arethought to be recruited locally fromthe cortex, bone marrow, perios-teum, and external soft tissues (Fig-ure 2). The relative contribution ofMSCs from each tissue is uncertainbut is thought to depend on the localparameters present at the injured tis-sue, such as growth factors, oxygengradient, and mechanical stability.The clinical relevance of muscle as asource of progenitor cells duringfracture repair has been the subjectof several recent studies.16,17

The presence in muscle of a pop-ulation of adult stem cells that candifferentiate into cells of differentlineages has been suspected for sometime based on two observations.First, muscle has the potential toturn into bone, as occurs during het-erotopic ossification. Second, theoriginal description of osteoinduc-tion by Urist3 has been attributed tothe effects of BMPs on progenitorcells within muscle tissue. Howev-er, the isolation of the relevant MSCpopulation from this tissue is rela-tively recent.16 Clinically, the impor-tance of muscle as a source of os-teoprogenitor cells is underlined bythe poor outcome of fractures inwhich muscle has been devitalized,although this poor outcome is oftenattributed to the coexisting damageto the periosteal blood supply. Mus-cle resection significantly reducescallus formation and the biome-chanical properties of the healedbone, while a muscle crush does notsignificantly affect bone healing.18

Conversely, heterotopic ossificationin acetabular fractures has been re-duced as surgeons have becomemore aggressive in débriding injuredand necrotic muscle from the surgi-cal field.19

Systemic Recruitmentof Cells and MolecularSignaling

Traditionally, it was thought thatthe cells involved in fracture repair

were recruited only locally. Howev-er, a systemic mobilization and re-cruitment of osteoblastic precursorsto the fracture site from the periph-eral circulation have now been dem-onstrated in several recent studies.In a rabbit ulnar osteotomy model, itwas demonstrated that some osteo-blasts involved in fracture healingwere systemically mobilized and re-cruited to the fracture from remotebone marrow sites.20 Shen et al21

demonstrated in a murine modelthat, following systemic injection ofMSCs, osteoprogenitor cells local-ized to the fracture callus.

These studies have implicationsfor the development of future cell-based therapies for fracture healing.Cell-based therapies are neededwhen insufficient cells are presentwithin a fracture callus (eg, segmen-tal defect). In such a situation, evenwhen all of the osteoprogenitor cellsat the site of fracture are working tothe maximum, there will be no bonyunion, nor will any osteoinductiveagents be effective because maximalosteogenesis per cell is already oc-curring. In a level III study (case-control), Hernigou et al22 demon-strated the clinical effectiveness oflocal percutaneous injection of bonemarrow aspirate in treating tibialnonunions. We believe that the stud-ies mentioned here suggest that itmight be possible to develop cell-based therapies in which cells aresystemically administered and local-ized to the site of injury.

Evidence is emerging that distantskeletal sites can be affected in re-sponse to a local bone injury.23 Anincreased osteogenic response hasbeen detected in sites distant fromthe fracture in animal models.23 Thismay result from the release ofgrowth factors (eg, transforminggrowth factor-β, insulin-like growthfactor-1) from the fracture site intothe systemic circulation, as shownin a clinical study.24 It is not knownwhether the level of these factors inserum reflects the repair activity ofthe fracture.



Systemic Factors andLocal Fracture Healing

It is widely accepted that extrinsicfactors have an influence on the out-come of fracture healing. However, itis often difficult to isolate the role ofa particular systemic factor in clini-cal situations. For example, impairedfracture healing in the elderly may berelated to age, osteoporosis, drugs,malnutrition, and/or anemia. Evi-dence gained from animal models aswell as recently uncovered cellularand molecular processes have led tobetter understanding of the role ofsystemic factors.

NonsteroidalAnti-inflammatory Drugs

During fracture repair, the en-zyme cyclooxygenase-2 (COX-2) isactivated to produce prostaglandins,which are needed during inflamma-tion and are critical for starting theosteogenic response.25 Nonsteroidalanti-inflammatory drugs (NSAIDs)inhibit COX-2, dramatically reduc-ing prostaglandin production, andtherefore have the potential to nega-tively affect fracture repair. Al-though results in various animalstudies have been conflicting, it isgenerally accepted that NSAIDs im-pair fracture healing in animal mod-els in which a high dose has been ad-ministered.26 This inhibitory effectis most potent during the early phaseof healing, thereby underlining theimportance of the initial inflamma-tory reaction.26

Although it is not known inwhich clinical situations NSAIDs inhigh local concentrations will affectearly-phase fracture healing, admin-istration of NSAIDs has been associ-ated with delayed union and pseud-arthrosis.27 The inhibitory effect ofCOX-2 blockade in vitro has beenshown to be reversed by the admin-istration of BMP-2.28

AgeAging is an independent factor

that negatively affects fracture re-

pair. Delayed fracture healing in theelderly may be caused by differencesin molecular signaling locally with-in the fracture callus as well as tosystemic factors. Meyer et al28 re-ported a decreased expression ofBMP-2 and Indian hedgehog (a factorrelated to endochondral callus for-mation) in fracture calluses of olderrats. Noggin expression was notchanged with age. However, the de-crease in expression of BMP-2 im-plies that the BMP inhibitor predom-inated over BMP-2 in older rats.

Hormonal differences with agingmay also be a factor. Sera obtainedfrom aged donors are less potent in-ducers of osteoblast differentiation ofhMSC than are sera obtained fromyoung donors.29 These effects werespecific for osteoblast differentiationbecause no donor age differences inthe ability to support differentiationof other cell types were observed.Short-term bone marrow cultures es-tablished from young and old donorscontain similar numbers of hMSCsand exhibit similar proliferationrates.30 In addition, the capacity ofhMSC to differentiate into osteo-blasts and adipocytes was main-tained irrespective of donor age.31

These studies suggest that there areno intrinsic defects in hMSCs withaging and that extrinsic factorspresent in the aging environment ofhMSCs may be responsible for theimpaired osteoblast functions seenwith aging.

These observations need furtherinvestigation. If poor fracture repairin the elderly is related to impairedosteoinductive signals rather than todifferences in the cellular compo-nent of the healing fractures, thenthe elderly patient with trauma ismore likely to benefit from an os-teoinductive agent, such as BMP-2,than from cell-based therapies.

SmokingSmoking has an adverse effect on

fracture healing. In one study, thetime to union for tibial fracturesamong smokers was significantly

(P < 0.05) longer by 4 weeks than innonsmokers.32 This effect may bemediated by either nicotine or someother, yet undefined components incigarette smoke,33 or both.

In animal models, nicotine hasbeen shown to delay cellular differen-tiation into chondrocytes and to slowthe physiologic transition from car-tilaginous callus to bone.34 It is alsohighly likely that the compromise ofmicrocirculation secondary to nico-tine causes a delay in fracture repair.It remains to be seen whether thedeleterious effects of smoking are re-versible with smoking cessation andwhether BMPs can improve healingin smokers. These questions shouldbe investigated using animal modelsof fracture healing and smoking.

Influence of FractureFixation

The local mechanical forces on afracture resulting in movement atthe fracture site are critical factors inthe success of fracture repair. Excessmotion can delay healing; castingand fracture fixation aim to providea mechanical environment in whichstrains are decreased to avoid de-layed union or nonunion. However,some movement, referred to as mi-cromotion, is beneficial to fracturehealing. Different fractures and dif-ferent areas of the skeleton responddifferently to mechanical forces andresultant strains. Yet precisely howthese changes in local mechanicalloading result in a cartilaginous cal-lus or intramembranous ossificationremains speculative. The influenceof the mechanical environment onthe fracture repair can be viewed atthree levels: tissue, cellular, and mo-lecular.

Tissue differentiation requiresmechanical stability. In an animalmodel, Claes et al35 demonstrated astrong association between fracturestability and the spatial distributionof newly formed blood vessels andspecific tissue formation. It was alsoindependently demonstrated in an



animal model that increased stabil-ity affected endochondral ossifica-tion by decreasing the overallamount of cartilage that formed atthe fracture site.36 This was thoughtto be secondary to an increase in therate of maturation of chondrocytesduring the endochondral ossificationstage of fracture healing. Conversely,orthopaedic surgeons often observean increase in the size of fracturecallus with increased movementaround the fracture. Endochondralossification is also affected by thetype of intermittent forces applied tothe fracture site. In a rat model ofhealing osteotomy, intermittent ten-sile strains stimulated endochondralossification, as opposed to compres-sive strains, which favored direct in-tramembranous ossification.37

The method of fracture fixationalso has an effect on the biology ofthe healing callus. In a murine mod-el of femoral fracture healing, thefracture callus was significantly larg-er with intramedullary nail fixationthan with plate fixation.38 Thechanges in gene expression follow-ing each fixation procedure weresimilar and occurred after the samepostoperative time interval. Howev-er, the intramedullary group had sig-nificantly greater expression of genesrelated to cartilage, cell division, andinflammation (P < 0.05 for all), andthere was greater expression of genesrelated to macrophage activity in theplate group than in the nail group(P < 0.001).

Osteoprogenitor cells have the ca-pacity to detect their mechanical en-vironment and modify their rate ofdifferentiation, a process mediatedvia effects on the BMP signal-ing pathway. Cyclic stretching ofhMSCs or osteoblastic precursor cellsin a collagen matrix increased theirproliferation and osteogenic differen-tiation,39,40 a phenomenon associatedwith an increase in BMP-2 pro-duction. This increase in osteogenicdifferentiation may also be mediatedby the downregulation of peroxi-some proliferator–activated receptor

gamma in bone marrow stromalcells, a mediator that favors adipo-genesis over osteogenesis.41 However,compression can also stimulate BMPproduction and decrease noggin pro-duction, thereby stimulating the invitro differentiation of human osteo-blastic cells.42 These studies indicatethat the same changes in molecularsignaling can be induced by differenttypes of forces acting on the progen-itor cells in various circumstances.

Summary

Significant advances have beenmade in the understanding of the bi-ology of fracture healing. In particu-lar, it is now understood that frac-ture repair is not only a localphenomenon but is itself under theinfluence of extrinsic factors. Ad-verse local and systemic clinical fac-tors can affect the molecular and cel-lular processes involved and can leadto delayed fracture repair and non-union. The management of thesehealing problems remains challeng-ing, despite the introduction of ther-apeutic BMPs and other biologicmethods of bone regeneration. Fur-ther understanding of the patho-physiology of fracture repair is need-ed to develop improved treatmentstrategies targeted to the molecularand cellular processes affected inspecific clinical conditions.

References

Evidence-based Medicine: Referenc-es 9, 19, 22, 24, and 32 are level I/IIprospective, randomized studies.The remaining references are case-control cohort studies, basic re-search studies, or expert opinion.

Citation numbers printed in boldtype indicate references publishedwithin the past 5 years.

1. Potier E, Ferreira E, AndriamanalijaonaR, et al: Hypoxia affects mesenchymalstromal cell osteogenic differentiationand angiogenic factor expression.Bone 2007;40:1078-1087.

2. Reed AA, Joyner CJ, Brownlow HC,Simpson AH: Human atrophic frac-ture non-unions are not avascular.J Orthop Res 2002;20:593-599.

3. Urist MR: Bone: Formation by autoin-duction. Science 1965;150:893-899.

4. Kloen P, Di Paola M, Borens O, et al:BMP signaling components are ex-pressed in human fracture callus.Bone 2003;33:362-371.

5. Cook SD, Salkeld SL, Brinker MR,Wolfe MW, Rueger DC: Use of an os-teoinductive biomaterial (rhOP-1) inhealing large segmental bone defects.J Orthop Trauma 1998;12:407-412.

6. Gerhart TN, Kirker-Head CA, KrizMJ, et al: Healing segmental femoraldefects in sheep using recombinanthuman bone morphogenetic protein.Clin Orthop Relat Res 1993;293:317-326.

7. Sciadini MF, Johnson KD: Evaluationof recombinant human bone morpho-genetic protein-2 as a bone-graft sub-stitute in a canine segmental defectmodel. J Orthop Res 2000;18:289-302.

8. Jones AL, Bucholz RW, Bosse MJ, et al:Recombinant human BMP-2 and al-lograft compared with autogenousbone graft for reconstruction of dia-physeal tibial fractures with corticaldefects: A randomized, controlled tri-al. J Bone Joint Surg Am 2006;88:1431-1441.

9. Geesink RG, Hoefnagels NH, BulstraSK: Osteogenic activity of OP-1 bonemorphogenetic protein (BMP-7) in ahuman fibular defect. J Bone JointSurg Br 1999;81:710-718.

10. Abe E, Yamamoto M, Taguchi Y, et al:Essential requirement of BMPs-2/4for both osteoblast and osteoclast for-mation in murine bone marrow cul-tures from adult mice: Antagonism bynoggin. J Bone Miner Res 2000;15:663-673.

11. Niikura T, Hak DJ, Reddi AH: Globalgene profiling reveals a downregula-tion of BMP gene expression in exper-imental atrophic nonunions com-pared to standard healing fractures.J Orthop Res 2006;24:1463-1471.

12. Makino T, Hak DJ, Hazelwood SJ,Curtiss S, Reddi AH: Prevention ofatrophic nonunion development byrecombinant human bone morphoge-netic protein-7. J Orthop Res 2005;23:632-638.

13. Kwong FN, Richardson SM, Evans CH:Chordin knockdown enhances the os-teogenic differentiation of human mes-enchymal stem cells. Arthritis ResTher 2008;10:R65.

14. Wan DC, Pomerantz JH, Brunet LJ,



et al: Noggin suppression enhances invitro osteogenesis and accelerates invivo bone formation. J Biol Chem2007;282:26450-26459.

15. Caplan AI: Adult mesenchymal stemcells for tissue engineering versus re-generative medicine. J Cell Physiol2007;213:341-347.

16. Jankowski RJ, Huard J: Myogenic cel-lular transplantation and regenera-tion: Sorting through progenitor het-erogeneity. Panminerva Med 2004;46:81-91.

17. Gates CB, Karthikeyan T, Fu F, HuardJ: Regenerative medicine for the mus-culoskeletal system based on muscle-derived stem cells. J Am AcadOrthop Surg 2008;16:68-76.

18. Utvåg SE, Grundnes O, Rindal DB,Reikerås O: Influence of extensivemuscle injury on fracture healing inrat tibia. J Orthop Trauma 2003;17:430-435.

19. Rath EM, Russell GV Jr, WashingtonWJ, Routt ML Jr: Gluteus minimusnecrotic muscle debridement dimin-ishes heterotopic ossification afteracetabular fracture fixation. Injury2002;33:751-756.

20. Shirley D, Marsh D, Jordan G, Mc-Quaid S, Li G: Systemic recruitmentof osteoblastic cells in fracture heal-ing. J Orthop Res 2005;23:1013-1021.

21. Shen FH, Visger JM, Balian G, Hur-witz SR, Diduch DR: Systemicallyadministered mesenchymal stromalcells transduced with insulin-likegrowth factor-I localize to a fracturesite and potentiate healing. J OrthopTrauma 2002;16:651-659.

22. Hernigou P, Poignard A, Beaujean F,Rouard H: Percutaneous autologousbone-marrow grafting for nonunions:Influence of the number and concen-tration of progenitor cells. J BoneJoint Surg Am 2005;87:1430-1437.

23. Einhorn TA, Simon G, Devlin VJ,Warman J, Sidhu SP, Vigorita VJ: Theosteogenic response to distant skele-tal injury. J Bone Joint Surg Am 1990;72:1374-1378.

24. Kaspar D, Neidlinger-Wilke C, Hol-bein O, Claes L, Ignatius A: Mitogens

are increased in the systemic circula-tion during bone callus healing.J Orthop Res 2003;21:320-325.

25. Zhang X, Schwarz EM, Young DA, Pu-zas JE, Rosier RN, O’Keefe RJ:Cyclooxygenase-2 regulates mesen-chymal cell differentiation into theosteoblast lineage and is critically in-volved in bone repair. J Clin Invest2002;109:1405-1415.

26. Simon AM, Manigrasso MB, O’ConnorJP: Cyclo-oxygenase 2 function is es-sential for bone fracture healing.J Bone Miner Res 2002;17:963-976.

27. Giannoudis PV, MacDonald DA, Mat-thews SJ, Smith RM, Furlong AJ, DeBoer P: Nonunion of the femoral diaph-ysis: The influence of reaming and non-steroidal anti-inflammatory drugs.J Bone Joint Surg Br 2000;82:655-658.

28. Meyer RA Jr, Meyer MH, TenholderM, Wondracek S, Wasserman R, Garg-es P: Gene expression in older ratswith delayed union of femoral frac-tures. J Bone Joint Surg Am 2003;85:1243-1254.

29. Abdallah BM, Haack-Sørensen M,Fink T, Kassem M: Inhibition of os-teoblast differentiation but not adipo-cyte differentiation of mesenchymalstem cells by sera obtained from agedfemales. Bone 2006;39:181-188.

30. Stenderup K, Justesen J, Eriksen EF,Rattan SI, Kassem M: Number andproliferative capacity of osteogenicstem cells are maintained during ag-ing and in patients with osteoporosis.J Bone Miner Res 2001;16:1120-1129.

31. Justesen J, Stenderup K, Eriksen EF,Kassem M: Maintenance of osteoblas-tic and adipocytic differentiation po-tential with age and osteoporosis inhuman marrow stromal cell cultures.Calcif Tissue Int 2002;71:36-44.

32. Adams CI, Keating JF, Court-BrownCM: Cigarette smoking and open tib-ial fractures. Injury 2001;32:61-65.

33. Skott M, Andreassen TT, Ulrich-Vinther M, et al: Tobacco extract butnot nicotine impairs the mechanicalstrength of fracture healing in rats.J Orthop Res 2006;24:1472-1479.

34. El-Zawawy HB, Gill CS, Wright RW,

Sandell LJ: Smoking delays chondro-genesis in a mouse model of closedtibial fracture healing. J Orthop Res2006;24:2150-2158.

35. Claes L, Eckert-Hübner K, Augat P:The effect of mechanical stability onlocal vascularization and tissue differ-entiation in callus healing. J OrthopRes 2002;20:1099-1105.

36. Le AX, Miclau T, Hu D, Helms JA:Molecular aspects of healing in stabi-lized and non-stabilized fractures.J Orthop Res 2001;19:78-84.

37. Smith-Adaline EA, Volkman SK, Ig-nelzi MA Jr, Slade J, Platte S, Gold-stein SA: Mechanical environment al-ters tissue formation patterns duringfracture repair. J Orthop Res 2004;22:1079-1085.

38. Heiner DE, Meyer MH, Frick SL,Kellam JF, Fiechtl J, Meyer RA Jr: Geneexpression during fracture healing inrats comparing intramedullary fixationto plate fixation by DNA microarray.J Orthop Trauma 2006;20:27-38.

39. Ignatius A, Blessing H, Liedert A, et al:Tissue engineering of bone: Effects ofmechanical strain on osteoblasticcells in type I collagen matrices.Biomaterials 2005;26:311-318.

40. Sumanasinghe RD, Bernacki SH, Lo-boa EG: Osteogenic differentiation ofhuman mesenchymal stem cells incollagen matrices: Effect of uniaxialcyclic tensile strain on bone morpho-genetic protein (BMP-2) mRNA ex-pression. Tissue Eng 2006;12:3459-3465.

41. David V, Martin A, Lafage-ProustMH, et al: Mechanical loading down-regulates peroxisome proliferator-activated receptor gamma in bonemarrow stromal cells and favors os-teoblastogenesis at the expense of ad-ipogenesis. Endocrinology 2007;148:2553-2562.

42. Mitsui N, Suzuki N, Maeno M, et al:Optimal compressive force inducesbone formation via increasing bonemorphogenetic proteins productionand decreasing their antagonists pro-duction by Saos-2 cells. Life Sci 2006;78:2697-2706.



IntroductionExtremity War Injuries:Challenges in DefinitiveReconstruction

The January 2006 AmericanAcademy of Orthopaedic Sur-

geons/Orthopaedic Trauma Associa-tion (AAOS/OTA) Extremity WarInjuries: State of the Art and FutureDirections (EWI) symposium wasthe first combined effort of theAAOS, the OTA, and the UnitedStates military to discuss ortho-paedic injuries sustained as a resultof the Global War on Terror. Thisflagship symposium resulted in aspecial issue of the Journal of theAmerican Academy of OrthopaedicSurgeons (JAAOS) discussing cur-rent care and describing future re-search directions on wound manage-ment, antibiotics and infection,management of segmental bone de-fects, stabilization of long bones, andamputee care.1

The AAOS/OTA/Society of Mili-tary Orthopaedic Surgeons (SOMOS)EWI II: Development of ClinicalTreatment Principles symposium inJanuary 2007 continued the AAOScollaboration with the military to re-view advances in treatment of ex-tremity trauma, examine new chal-lenges faced overseas, and defineprinciples of delivery of forwardcare. An article summarizing EWI IIpresentations was published inJAAOS in October 2007.2

The third Extremity War Injuriessymposium, AAOS/OTA/SOMOSEWI III: Challenges in DefinitiveReconstruction, was held in January2008 and provided a forum for re-view and discussion of research find-ings in the areas of soft-tissue de-fects, segmental bone defects, opentibial shaft fractures, and massive

periarticular reconstruction. In addi-tion, EWI III symposium attendeeswere addressed by high-ranking ad-ministration officials and membersof Congress, including Chairman ofthe Joint Chiefs of Staff ADMMichael Mullen, Assistant Secretaryof Defense for Health Affairs S. WardCasscells, MD, Senator Tom Harkin(D-IA), Representative Tom Latham(R-IA), Representative C. A. DutchRuppersberger (D-MD), and Repre-sentative Tim Walz (D-MN). In addi-tion, National Institute of Arthritisand Musculoskeletal and Skin Dis-eases (NIAMS) Director Stephen I.Katz, MD, PhD, Uniformed ServicesUniversity of the Health SciencesPresident Charles L. Rice, MD, andUnder Secretary of Defense for Per-sonnel and Readiness David S. C.Chu, PhD, attended the event. In hiscomments, Senator Harkin stressedthat the American people have a pro-found moral obligation to supportthose who have served.

The EWI III symposium resultedin a Congressional call of support foradditional federal funding for ex-tremity war injury research throughthe US Department of Defense to of-fer our wounded soldiers better heal-ing outcomes and daily living capa-bilities. In a letter circulated to everymember of Congress, Senators Har-kin and Kay Bailey Hutchison (R-TX) and Representatives Latham andRuppersberger urged their colleaguesto sign a letter to the leadershipof the Senate and House DefenseAppropriations Subcommittees re-questing support for an annual oper-ating level of $50 million in fiscal



Symposium


year 2008, through supplemental ap-propriations, for the peer-reviewedOrthopaedic Extremity Trauma Re-search Program. Through programslike the EWI symposium series, theAAOS will continue to work withmembers of Congress to securefunding for research to ensure thatour service men and women are re-ceiving the best possible care.

This paper represents a formalreport of the outcomes of EWI III,including discussions related toresearch updates and funding chal-lenges as well as treatment of seg-mental bone defects, soft-tissue de-fects, open tibial shaft fractures, andmassive periarticular reconstruc-tions.

We would like to like to acknowl-edge the individuals, groups, andsponsors that helped make the sym-posium possible. We would like tothank the EWI III faculty for theirguidance, dedication, and time: LTCRomney C. Andersen, MD; COLMark R. Bagg, MD (Ret); Michael J.Bosse, MD (Ret); COL Paul J. Cut-ting, MD, FACS; COL William C.Doukas, MD; Robert Paul Dunbar,MD; COL Roman A. Hayda, MD;John E. Herzenberg, MD; MAJ Jo-seph R. Hsu, MD; LCDR John J.Keeling, MD; LTC James Keeney,MD (Ret); CPT MC L. Scott Levin,MD, FACS; Ronald W. Lindsey, MD;

CDR Michael T. Mazurek, MD;Michael D. McKee, MD, FRCSC;Theodore Miclau, MD; Sean E.Nork, MD; Regis O’Keefe, MD,PhD; COL Elisha Powell IV, MD;LTC Craig Ruder, MD; Andrew H.Schmidt, MD; Marcus F. Sciadini,MD; LTC Scott B. Shawen, MD;Randy Sherman, MD, FACS; Doug-las G. Smith, MD; Robert J. Spinner,MD; Marc F. Swiontkowski, MD; H.Thomas Temple, MD; J. Tracy Wat-son, MD; and Joshua Wenke, PhD.

We would also like to thank JamesBeaty, MD; E. Anthony Rankin, MD;Joseph D. Zuckerman, MD; and theAAOS Board of Directors, as well asthe AAOS Council on Research,Quality Assessment, and Technology,the AAOS Research DevelopmentCommittee, and AAOS staff leader-ship, including Christy M. P.Gilmour, Erin L. Ransford, KristyGlass, Robert S. Jasak, JD, DavidLovett, JD, Lindsay F. Law, David C.Smith, and Karen Hackett, CAE,FACHE.

The American Academy of Or-thopaedic Surgeons, the OrthopaedicTrauma Association, and the Societyof Military Orthopaedic Surgeonsacknowledge the following industrycontributors and their representa-tives for their financial support ofthe symposium and this article:

• Kinetic Concepts, Inc (GoldLevel)

• Medtronic (Silver Level)• Smith & Nephew Trauma (Sil-

ver Level)• Stryker (Silver Level)• Synthes (Silver Level)• DePuy (Additional Support)

Andrew N. Pollak, MDEWI III Symposium Co-Chair

Chief, Orthopaedic Trauma, RACShock Trauma Center

University of Maryland School ofMedicine

Baltimore, [email protected]

Col. James R. Ficke, MDEWI III Symposium Co-Chair

Orthopaedic Consultant to the U.S.Army Surgeon General

Chief, Orthopaedic SurgeryBrooke Army Medical Center

Fort Sam Houston, [email protected]

References

1. Pollak A, Calhoun J: Extremity warinjuries: State of the art and futuredirections. J Am Acad Orthop Surg2006;14:iii-S214.

2. Ficke JR, Pollak AN: Extremity warinjuries: Development of clinicaltreatment principles. J Am AcadOrthop Surg 2007;15:590-595.


Extremity War Injuries:Challenges in DefinitiveReconstruction

AbstractThe third annual Extremity War Injuries Symposium was held inJanuary 2008 to review challenges related to definitivemanagement of severe injuries sustained primarily as a result ofblast injuries associated with military operations in the Global Waron Terror. Specifically, the symposium focused on the managementof soft-tissue defects, segmental bone defects, open tibial shaftfractures, and challenges associated with massive periarticularreconstructions. Advances in several components of soft-tissueinjury management, such as improvement in the use of free-tissuetransfer and enhanced approaches to tissue-engineering, mayimprove overall care for extremity injuries. Use of distractionosteogenesis for treatment of large bone defects has been simplifiedby the development of computer-aided distraction protocols. Forclosed tibial fractures, evidence and consensus support initialsplinting for transport and aeromedical evacuation, followed byelective reamed, locked intramedullary nail fixation. Managementof open tibial shaft fractures sustained as a result of high-energycombat injuries should include serial débridements every 48 hoursuntil definitive wound closure and stabilization are recommended.A low threshold is recommended for early utilization offasciotomies in the overall treatment of tibial shaft fracturesassociated with war injuries. For management of open tibialfractures secondary to blast or high-velocity gunshot injuries, goodexperiences have been reported with the use of ring fixation fordefinitive treatment. Treatment options in any given case ofmassive periarticular defects must consider the specific anatomicand physiologic challenges presented as well as the capabilities ofthe treating surgeon.

The third annual Extremity WarInjuries Symposium (EWI III),

jointly hosted by American Acad-emy of Orthopaedic Surgeons(AAOS), the Orthopaedic TraumaAssociation (OTA), and the Societyof Military Orthopaedic Surgeons,

was held in Washington, DC, on Jan-uary 23 and 24, 2008. The goal ofEWI III was to review progress onchallenges related to management ofsevere injuries sustained primarilyas a result of blast injuries associat-ed with military operations in the

Andrew N. Pollak, MD

COL James R. Ficke, MD, MC

USA

Extremity War Injuries III

Session Moderators*

Dr. Pollak is Chief, Orthopaedic Trauma,RAC Shock Trauma Center, University ofMaryland School of Medicine, Baltimore,MD. Dr. Ficke is OrthopaedicConsultant to the US Army SurgeonGeneral and Chief, OrthopaedicSurgery, Brooke Army Medical Center,Fort Sam Houston, TX.

*COL Mark R. Bagg, MD (Ret); CPT L.Scott Levin, MD, FACS; CDR Michael T.Mazurek, MD; J. Tracy Watson, MD; LTCRomney C. Andersen, MD; Sean E.Nork, MD; COL Roman A. Hayda, MD;Theodore Miclau, MD; LCDR JohnKeeling, MD; Marc Swiontkowski, MD;and LTC James Keeney, MD (Ret).

The opinions or assertions expressedherein are those of the authors and donot reflect those of the Army, Navy, AirForce, or the Department of Defense.

Reprint requests: Dr. Pollak, Universityof Maryland School of Medicine, RoomT3R54, 22 South Greene Street,Baltimore, MD 21201-1544.



Symposium


Global War on Terror. Specifically,the symposium focused on the man-agement of soft-tissue defects, seg-mental bone defects, and open tibialshaft fractures, as well as challengesassociated with massive periarticu-lar reconstructions.

In addition, joint efforts of theAAOS and the OTA in developingand implementing a Visiting Schol-ars Program were reviewed. Clinicalupdates included reports on the op-erations at the Air Force TheaterHospital in Balad, Iraq, and the Com-bat Support Hospital in Baghdad,Iraq.

Distinguished VisitingScholars Program

The AAOS and OTA have jointly de-veloped and implemented a Distin-guished Visiting Scholars Program.Fifteen orthopaedic trauma special-ists with widely recognized ex-pertise in trauma clinical care andtrauma education have spent a min-imum of 2 weeks each at LandstuhlRegional Medical Center deliveringclinical care and providing educationfor the clinical staff. Their experi-ences in treating wounded warriorsand their interactions with militaryorthopaedic surgeons have also al-lowed them to receive a valuable ed-ucation, with the intention of usingthis instruction to educate their col-leagues in the civilian sector. Thiseducation relates particularly to un-derstanding the nuances and severi-ty of combat-related injury. Evenmore important, this education in-volves understanding the distinc-tions between the treatment of blastinjury and the treatment of injuriesmore typically seen in a civilian en-vironment, as well as the need for

additional research to better under-stand the best methods for clinicalmanagement of war injuries.

Plans for the Distinguished Visit-ing Scholars Program started in July2007. As of the time of EWI III, eightsurgeons had served as distinguishedvisiting scholars. Twenty additionaldistinguished visiting scholars werescheduled to visit Landstuhl overthe course of the next 18 months.Distinguished visiting scholars par-ticipate in all regular patient careactivities, including intake assess-ment, ward rounds, and surgery.They also deliver lectures to ortho-paedic surgical staff and otherspecialists. Applications for theprogram are available [email protected]. Eligibility criteria forselection as a distinguished visitingscholar include 10 years or more ofclinical trauma care experience,demonstrated excellence in traumaeducation, and recognized expertisein trauma clinical care.

The program was modeled after asimilar program for general traumacritical care surgeons developed bythe American College of Surgeonsand the American Association forthe Surgery of Trauma.1 As reportedat EWI III, the response to date fromthe distinguished visiting scholars,as well as that from participants inthe general surgical program, hasbeen uniformly positive. In addition,reaction from the military medicalpersonnel stationed at Landstuhlsuggests that the program has con-tributed to the education of militarysurgeons and to the care of woundedwarriors. Current plans call for theprogram to continue for as long ascasualties related to the conflicts inAfghanistan and Iraq warrant.

Research Update

At the time of the first ExtremityWar Injuries Symposium in January2006, the need for a comprehensivedatabase of injuries and treatmentoutcomes after high-energy extrem-ity trauma related to war injury wasidentified as a priority.2 The MilitaryOrthopaedic Trauma Registry hasnow been established as an effectivemeans of collecting comprehensivedata about injuries sustained by mil-itary personnel in Iraq and Afghani-stan. These data are currently beingcollected in conjunction with theJoint Theater Trauma Registry usingabstractors located at the majormedical centers where these pa-tients are receiving care. Data col-lection will include informationthroughout the entire spectrum ofcasualty care, from initial resuscita-tion and surgery in the combat zonethrough definitive reconstruction. Inaddition, long-term data collectionefforts are under way to help trackoutcomes following treatment ofthese severe injuries. Studies fundedby the Department of Defense areincluded among these efforts to de-termine specific functional andpatient-oriented outcomes followingextremity trauma and limb salvageassociated with blast injury. Onesuch study, Military Extremity Trau-ma and Limb Salvage (METALS), iscurrently examining relative differ-ences between amputation and limbsalvage between 1 and 5 years afterinjury in warriors who have sus-tained high-energy extremity trau-ma.

The Orthopaedic Extremity Trau-ma Research Program (OETRP) is acongressionally funded program fo-cused specifically on improving care

Dr. Pollak or a member of his immediate family has received research or institutional support from Synthes, Stryker, Smith & Nephew, Zimmer,and Wyeth and is an employee or consultant of KCI. Dr. Hayda or a member of his immediate family has received research or institutionalsupport from Howmedica, Synthes, and Smith & Nephew and has received miscellaneous nonincome support from Synthes. None of thefollowing authors or a member of their immediate families has received anything of value from or owns stock in a commercial company orinstitution related directly or indirectly to the subject of this article: Dr. Ficke, Dr. Andersen, Dr. Keeling, Dr. Keeney, Dr. Levin, Dr. Mazurek,Dr. Miclau, Dr. Nork, Dr. Bagg, Dr. Swiontkowski, and Dr. Watson.

Andrew N. Pollak, MD, et al


for extremity war injuries. The pro-gram is administered through the USArmy Institute for Surgical Researchat Fort Sam Houston, Texas; it usesa process similar to that utilized bythe National Institutes of Health toevaluate and grade research propos-als. A total of $19.3 million was ap-propriated by Congress for OETRPin 2006 and 2007, from which 26studies have been funded. In total,156 proposals were reviewed andgiven sufficiently high grades to war-rant funding, suggesting that currentcongressional funding levels are suf-ficient to support only the top 17%of research grant applications.

For 2008, $4.8 million was appro-priated in the regular Department ofDefense Appropriations Act. Addi-tionally, as part of the EmergencySupplemental Appropriations Actfor Defense, 2008, Congress recent-ly approved $276.8 million for a se-ries of specific research topics relat-ed to casualties sustained in Iraq andAfghanistan. Final determination ispending regarding how much of thismoney will be invested in theOETRP and other orthopaedic-related research programs. TheAAOS continues to work with theDepartment of Defense and Con-gress to achieve an annual operatinglevel of at least $50 million for theOETRP specifically and for muscu-loskeletal research in general.

Management ofSoft-tissue Defects

Soft-issue injuries in the war zonerepresent perhaps the greatest chal-lenge in limb reconstruction. Recentadvances in several components ofsoft-tissue injury management mayimprove overall care for extremityinjuries.

One major ongoing concern re-gards the best method of managingopen wounds during transportationfrom the military theater of opera-tions to Landstuhl Regional MedicalCenter in Germany, as well as duringtransportation from Germany back

to the United States. Initial experi-ences suggested that utilization oftraditional wet-to-dry dressings wassafe and sufficiently effective. Disad-vantages of this technique, however,include the inability to change dress-ings in flight, resulting in markedlyprolonged intervals between dressingchanges, with the resultant theoreticincrease in infection risk. Addition-ally, although use of negative-pressure wound therapy (NPWT) hadbeen demonstrated to be safe and ef-fective for wound management bothwithin the theater of military oper-ations and in the military hospitalsin the United States, evidence sup-porting safety and efficacy of thistechnique during aeromedical trans-port was lacking.3 A retrospective re-view of the use of NPWT in aero-medical evacuation from Iraq andAfghanistan to Landstuhl RegionalMedical Center indicates that thetechnique can be safely employed.4

A further prospective evaluation ofsafety is under way evaluating theuse of NPWT during transport fromLandstuhl Regional Medical Centerto the United States.

Improvement in the use of free-tissue transfer for management of se-vere soft-tissue defects has resultedin apparent better utilization of thistechnique for limb salvage. Improve-ments include developing a betterunderstanding of the scope of indica-tions for utilization of free-tissuetransfer techniques, increased un-derstanding of the timing of appro-priate free-tissue transfer in thecombined context of patient andwound concerns, and appreciation oflogistic considerations, includingavailability of an appropriately pre-pared and experienced microvascu-lar surgeon.5,6

The soft tissues represent a com-posite of highly organized structuresthat include skin, muscle, tendon,ligament, nerve, and connective tis-sues. The proper organization ofthese tissues is required for locomo-tion and normal skeletal function.Current work is focused on under-

standing the properties of theindividual components of the softtissues and enhancing tissue-engineering approaches to maximizethe function of these componentswithin the limb. One example istendon reconstruction. Tendon func-tion requires excellent tensile prop-erties with near frictionless gliding.An animal model recently developedto study tendon defect reconstruc-tion shows that coating a freeze-dried tendon allograft with a virusexpressing growth differentiationfactor 5 (GDF-5) markedly improvestendon gliding and accelerateswound healing.7,8 GDF-5 is a mem-ber of the transforming growthfactor-beta/bone morphogenetic pro-tein (TGF-β/BMP) signaling familythat is involved in tendon formation.This and other work shows that ap-proaches utilizing combinations ofbiocompatible materials, cells, andgenes have the potential to meetsome of the challenges of extremityreconstruction in the injured com-batant that involve massive soft-tissue defects.

Segmental BoneDefects

Segmental bone defects represent anongoing challenge in the manage-ment of extremity trauma related tohigh-energy blast injuries. Althoughspecific data regarding the preva-lence of this problem are lacking, an-ecdotal reports from treating sur-geons at all major military medicalinstitutions suggest that defectslacking the capacity for spontaneoushealing are common and that no sin-gle existing treatment option is ap-propriate for all defects. Further-more, a subset of segmental defectscannot be predictably managed wellwith any existing treatment. Furtherdevelopment of treatment optionstherefore seems warranted from aclinical anecdotal standpoint.

Several techniques have beenused in civilian populations to treatbone defects, with varying degrees of

Extremity War Injuries: Challenges in Definitive Reconstruction


success. Vascularized bone graftshave been successfully employed incivilian settings for the treatment ofosteonecrosis of the femoral headand, anecdotally, in the reconstruc-tion of segmental defects resultingfrom trauma, infection, and tumorresection.9,10 Limitations of thistechnique for treatment of traumat-ically acquired segmental defectsconcern structural adequacy of fibu-lar grafts and donor site morbidity.Patients are understandably reluc-tant to consent to the harvest of tis-sue from an uninjured extremity inthe context of limb-threatening,contralateral lower extremity trau-ma. Consensus opinion suggestedthat vascularized structural graftingwas best suited for upper extremitydefects.11

Early clinical experience with ti-tanium mesh cages in conjunctionwith osteoinductive and osteocon-ductive grafting has been promising.12

Additional study is necessary to bet-ter understand the role for this tech-nology in trauma and posttraumaticreconstruction. A major potentialdrawback to this technique is the in-corporation of a foreign body into theendoskeleton of the bone in the con-text of an open injury with high pro-pensity for both short- and long-terminfectious complications.

A substantial body of animal data,including studies with subhumanprimates, suggests that use of recom-binant BMPs (particularly BMP-2and BMP-7) may be extremely effec-tive in promoting healing of bone de-fects of critical size.13-15 Several hu-man clinical comparative studies areavailable, and promising results al-ready have been published. Prospec-tive randomized analysis demon-strates efficacy in the use ofrecombinant BMP-2 for supplemen-tal grafting of open tibial fractures atthe time of surgical débridement andwound closure.16 Preliminary data inextremity war injuries also appearpromising. More comparative stud-ies would be helpful in better defin-ing the role of recombinant proteins

combined with structural and osteo-conductive materials.

An informal survey of militarysurgeons involved in the treatmentof high-energy lower extremity warinjuries suggests that distraction os-teogenesis is the technique most fre-quently employed for treatment ofbone defects that exceed approxi-mately 8 cm in length, particularlyin the tibia. Use of this modality hasbeen simplified by the developmentof computer-aided distraction proto-cols that allow for correction oflength and alignment without theneed for major frame modifica-tion.17 As a result, utilization ofthese devices in the civilian sectorhas increased along with surgeon fa-miliarity and comfort, a develop-ment that has likely led to a recipro-cal increase in utilization and aprogressive expansion of indications.However, two major drawbacks tothe use of this treatment modalityhave been described. The first is theneed for the patient to wear theframe for a prolonged period (typical-ly 3 days for every millimeter ofbone defect addressed).18 Second, thering fixator that is typically em-ployed limits access to the woundfor flap coverage or dressing care.This often results in initiation of thedistraction process being delayed un-til soft-tissue defects have been ad-dressed, thereby further contributingto the substantial length of the treat-ment period.

One technique described for em-ploying distraction osteogenesis inthe treatment of segmental bone de-fects uses internal rather than ex-ternal fixation.19,20 Utilizing theintramedullary skeletal kinetic dis-tractor involves creation of a corti-cotomy using an intramedullarysaw, stabilization of the bone withan intramedullary nail, and trans-port over the nail using internal mo-tors. There are no published seriesdescribing the use of this techniquefor trauma or posttraumatic defects,but the potential to address large de-fects without the need for prolonged

use of external fixation while main-taining access to the limb for woundcare seems promising. A major po-tential drawback remains the needto rely on an internal fixation devicefor a prolonged period in the contextof a high potential for infectiouscomplications.

Although high-quality compari-son studies are lacking, the gold stan-dard for treating small bony defects(<5 cm for segmental defects) re-mains autogenous cancellous graft.21

Potential sources include iliac crestsand proximal tibia for large defects.For smaller defects, the distal radiusand distal tibia are potentially valu-able donor sites. Obvious drawbacksto the technique include limitationsin the size of defect that can be ade-quately addressed and donor sitemorbidity, whose significance hasvaried widely in the available litera-ture. For war-related injuries in par-ticular, another major drawback con-cerns the potential need to addressmultiple bone defects associatedwith multiple limb trauma withinthe same patient.

A simple option for addressingsegmental defects is bony shortening.In the femur and tibia, subsequentlengthening can be done after initialshortening and healing. In the hu-merus, more substantial shorteningcan be considered as definitive treat-ment than in the lower extremity.

Management of TibialShaft Fractures

More so than other topics covered atEWI III, the issue of management oftibial shaft fractures is one that canbe guided by a substantial volume ofmedical literature associated withvarying levels of scientific evidence.In the context of understanding theunique challenges posed by war inju-ries, discussions of the evidence,which is largely based on civilian in-juries, yielded several recommenda-tions for the treatment of tibialfractures sustained in a combat envi-ronment.



For closed tibial fractures, evi-dence and consensus support initialsplinting for transport and aeromed-ical evacuation, followed by electivereamed, locked intramedullary nailfixation.22 Antibiotic prophylaxis us-ing a first-generation cephalosporinpreoperatively and for 24 hours post-operatively is recommended.23,24

Consensus opinion was that, unlikeopen tibial shaft fractures, closedfractures sustained in a combat envi-ronment are similar to those typical-ly sustained and treated in a civilianenvironment, and they can thereforebe treated similarly.

Management of open tibial shaftfractures sustained as a result ofhigh-energy combat injuries is morecontroversial. Recommendationswere developed based on a combina-tion of best-available medical evi-dence and consensus expert opinionbased on experience in treating sim-ilar injuries in both Iraq and Afghan-istan as well as at higher level treat-ment facilities in Germany and theUnited States. Evidence supportingearly débridement of open tibial frac-tures in civilian settings is marginal.This early initial débridement ofopen tibial war injuries is recom-mended within 6 hours of injurywhen possible, based on logistic con-siderations, and when safe, based onphysiologic considerations.25,26 Seri-al débridements every 48 hours un-til definitive wound closure and sta-bilization are recommended, despitethe lack of supporting evidence oth-er than expert opinion.

Prophylaxis against infection us-ing a first-generation cephalosporinis recommended on initial casualtypresentation and for 24 hours aftereach surgical débridement. Althoughbroader spectrum coverage often hasbeen recommended for high-energyinjuries sustained in civilian settings,studies are lacking comparing first-generation cephalosporins alone tobroader coverage in war injuries.27

Open wounds can be managed effi-caciously using NPWT dressings.3

Other options for wound manage-

ment also may be effective but, aswith other components of injurymanagement, specific comparisonstudies are lacking. Based purely onintuitive analysis, the combination ofantibiotic beads with NPWT is un-likely to effectively deliver high-concentration local antibiotic thera-py; however, the combination mayrepresent an effective means of avert-ing complication in the event ofNPWT device failure.

For multiple reasons, a lowthreshold is recommended for utili-zation of fasciotomies in the overalltreatment of tibial shaft fractures as-sociated with war injuries.28 First,many of these injuries result fromhigh-energy blast mechanisms, thusincreasing the risk for developmentof the complication. Second, logisticconsiderations put the injured war-rior at risk for development of com-partment syndrome during transitfrom one level of care to anotherwhen surgical decompression of thecompartments involved may not bepossible for a prolonged period.Third, anecdotal reports suggest thattransportation of the injured warriorat altitude, such as occurs in a mede-vac helicopter or during fixed-wingtransport, may further increase therisk of developing compartment syn-drome. Finally, particularly for mul-tiply injured patients, serial exami-nation may be compromised, thusincreasing the likelihood that earlysigns of developing compartmentsyndrome may be missed.

The ongoing protocol for tempo-rary stabilization of open tibial shaftfractures before definitive stabiliza-tion involves application of a mono-lateral external fixator in Iraq or Af-ghanistan before air evacuation toGermany and the United States.29,30

No studies have been publishedcomparing the safety or efficacy ofearly definitive stabilization com-pared with staged treatment.

Definitive treatment of open tib-ial fractures sustained in civilian en-vironments with intramedullary nailstabilization is supported by prospec-

tive randomized studies.31,32 No sim-ilar studies are available comparingtreatment options for battle-relatedinjuries, but differences at least seemto be likely, based on differences inthe injury mechanism. Opinions varyabout the best method for manage-ment of open tibial fractures second-ary to blast or high-velocity gunshotinjuries, but concerns have beenraised regarding anecdotally high in-fection rates in fractures treated withintramedullary nailing; good experi-ences have been reported with ringfixation for definitive treatment ofthese injuries.

Massive PeriarticularReconstructions

Treatment of large defects of bone,cartilage, and soft-tissue associatedwith periarticular combat injuriesrepresents a substantial, ongoingchallenge for surgeons managingwounded warriors. Although somepublished reports have addressedtreatment options, none includescontrol groups to allow for meaning-ful comparisons between treatmentoptions. Furthermore, most seriescontain small numbers of patientsand minimal outcome data.7,8,33-41

The consensus recommendationfor management of massive periar-ticular defects was that treatmentoptions used in any given case mustconsider the specific anatomic andphysiologic challenges presented aswell as the capabilities of the treat-ing surgeon. Specific considerationsshould include soft-tissue envelopeviability, presence or absence of ev-idence of infection, support of sur-rounding musculature, and availablereconstruction options for relevantassociated ligamentous or tendinousdeficiencies. Age and activity de-mands of the patient, particularlywith emphasis on the long-term ef-fects on longevity of the reconstruc-tion, also must be considered.

Documentation of injury charac-teristics, treatments employed, andoutcomes achieved will allow im-



portant long-term future assess-ments of factors associated withsuccessful management of thesecomplex problems. This informa-tion can provide a good long-termbasis for designing future prospec-tive trials of treatment options forthese injuries.42,43

Conclusions

The consensus at EWI III was that theopportunity to carefully review com-plicated management problems re-lated specifically to war injuries wasextremely valuable for the partici-pants and, therefore, for injured war-riors. In addition, substantial addi-tional research directed at thetreatment of high-energy extremitywar injuries is critical and likely willlead to dramatic improvements in ourmilitary colleagues’ ability to achievethe best possible outcomes from thesedevastating injuries.

References


1. Moore EE, Knudson MM, SchwabCW, Trunkey DD, Johannigman JA,Holcomb JB: Military-civilian collab-oration in trauma care and the seniorvisiting surgeon program. N Engl JMed 2007;357:2723-2727.

2. Pollak AN, Calhoun JH: Extremitywar injuries: State of the art and futuredirections. Prioritized future researchobjectives. J Am Acad Orthop Surg2006;14:S212-S214.

3. Leininger BE, Rasmussen TE, SmithDL, Jenkins DH, Coppola JR: Experi-ence with wound VAC and delayedprimary closure of contaminated softissue injuries in Iraq. J Trauma 2006;61:1207-1211.

4. Pollak AN, Flaherty SF, Cooper EO,Fang R, Powell ET, Ficke JR: Abstract:Use of negative pressure wound ther-apy during aeromedical evacuation.Presented at the annual meeting ofthe Society of Military OrthopaedicSurgeons, Vail, Colorado, December12, 2007.

5. Levin LS: New developments in flaptechnique. J Am Acad Orthop Surg2006;14:S90-S93.

6. Lawson R, Levin LS: Principles of freetissue transfer in orthopaedic prac-tice. J Am Acad Orthop Surg 2007;15:290-299.

7. Basile P, Dadali T, Jacobson J, et al:Freeze-dried tendon allografts astissue-engineering scaffolds for Gdf5gene delivery. Mol Ther 2008;16:466-473.

8. Hasslund S, Jacobson JA, Dadali T, etal: Adhesions in a murine flexor ten-don graft model: Autograft versus al-lograft reconstruction. J Orthop Res2008;26:824-833.

9. Adani R, Delcroix L, Innocenti M, etal: Reconstruction of large posttrau-matic skeletal defects of the forearmby vascularized free fibular graft.Microsurgery 2004;24:423-429.

10. Levin LS: Vascularized fibula graft forthe traumatically induced long-bonedefect. J Am Acad Orthop Surg 2006;14:S175-S176.

11. Ring D, Jupiter J, Toh S: Transarticu-lar bony defects after trauma and sep-sis: Arthrodesis using vascularizedfibular transfer. Plast Reconstr Surg1999;104:426-434.

12. Cobos JA, Lindsey RW, Gugala Z: Thecylindrical titanium mesh cage fortreatment of a long bone segmentaldefect: Description of a new tech-nique and report of two cases.J Orthop Trauma 2000;14:54-59.

13. Einhorn TA, Majeska RJ, MohaideenA, et al: A single percutaneous injec-tion of recombinant human morpho-genetic protein-2 accelerates fracturerepair. J Bone Joint Surg Am 2003;85:1425-1435.

14. Salkeld SL, Patron LP, Barrack RL,Cook SD: The effect of osteogenicprotein-1 on the healing of segmentalbone defects treated with autograft orallograft bone. J Bone Joint Surg Am2001;83:803-816.

15. Yasko AW, Lane JM, Fellinger EJ, et al:The healing of segmental bone de-fects, induced by recombinant humanbone morphogenetic protein (rhBMP-2): A radiographic, histological, andbiomechanical study in rats. J BoneJoint Surg Am 1992;74:659-670.

16. Govender S, Csimma C, Genant HK,et al: Recombinant human bone mor-phogenetic protein-2 for treatment ofopen tibial fractures: A prospective,controlled, randomized study of fourhundred and fifty patients. J BoneJoint Surg Am 2002;84:2123-2134.

17. Rozbruch SR, Pugsley JS, FragomenAT, Ilizarov S: Repair of tibial non-unions and bone defects with the Tay-lor Spatial Frame. J Orthop Trauma2008;22:88-95.

18. Paley D: Problems, obstacles, andcomplications of limb lengthening bythe Ilizarov technique. Clin OrthopRelat Res 1990;250:81-104.

19. Hankemeier S, Gösling T, Pape HC,Wiebking U, Krettek C: Limb length-ening with the Intramedullary Skele-tal Kinetic Distractor (ISKD). OperOrthop Traumatol 2005;17:79-101.

20. Hankemeier S, Pape HC, Gösling T,Hufner T, Richter M, Krettek C: Im-proved comfort in lower limb length-ening with the intramedullary skele-tal kinetic distractor: Principles andpreliminary clinical experiences.Arch Orthop Trauma Surg 2004;124:129-133.

21. Khan SN, Cammisa FP, Sandhu HS,Diwan AD, Girardi FP, Lane JM: Thebiology of bone grafting. J Am AcadOrthop Surg 2005;13:77-86.

22. Bone LB, Johnson KD: Treatment oftibial fractures by reaming and in-tramedullary nailing. J Bone JointSurg Am 1986;68:877-887.

23. Gillespie WJ, Walenkamp G: Antibi-otic prophylaxis for proximal femoraland other closed long bone fractures.Cochrane Database Syst Rev 2001;1:CD000244.

24. Boxma H, Broekhuizen T, Patka P,Oosting H: Randomized controlledtrial of a single dose antibiotic prophy-laxis in surgical treatment of closedfractures: The Dutch Trauma Trial.Lancet 1996;347:1133-1137.

25. Pollak AN: Timing of débridement ofopen fractures. J Am Acad OrthopSurg 2006;14:S48-S51.

26. Werner CML, Pierpont Y, Pollak AN:The urgency of surgical débridementin the management of open fractures.J Am Acad Orthop Surg 2008;16:369-375.

27. Zalavras CG, Patzakis MJ: Open frac-tures: Evaluation and management.J Am Acad Orthop Surg 2003;11:212-219.

28. Bumbaširevic M, Lesic A, MitkovicM, Bumbaširevic V: Treatment ofblast injuries of the extremity. J AmAcad Orthop Surg 2006;14:S77-S81.

29. Cammuso MR: Far-forward fracturestabilization: External fixation versussplinting. J Am Acad Orthop Surg2006;14:S118-S123.

30. Ficke JR, Pollak AN: Extremity warinjuries: Development of clinicaltreatment principles. J Am AcadOrthop Surg 2007;15:590-595.

31. Tornetta P III, Bergman M, Watnik N,Berkowitz G, Steuer J: Treatment ofgrade-IIIb open tibia fractures: A pro-spective randomized comparison ofexternal fixation and non-reamed



locked nailing. J Bone Joint Surg Br1994;76:13-19.

32. Henley MB, Chapman JR, Agel J, Har-vey EJ, Whorton AM, SwiontkowskiMF: Treatment of type II, IIIA, andIIIB open fractures of the tibial shaft:A prospective comparison of un-reamed interlocking intramedullarynails and half-pin external fixators.J Orthop Trauma 1998;12:1-7.

33. Culpan P, Le Strat V, Piriou P, Judet T:Arthrodesis after failed total ankle re-placement. J Bone Joint Surg Br 2007;89:1178-1183.

34. Dean GS, Hollinger EH IV, UrbaniakJR: Elbow allograft for reconstructionof the elbow with massive bone loss:Long-term results. Clin OrthopRelat Res 1997;341:12-22.

35. Feibel RJ, Uhthoff HK: PrimaryIlizarov ankle fusion for nonrecon-structable tibial plafond fractures.

Oper Orthop Traumatol 2005;17:457-480.

36. Garberina MJ, Fitch RD, HoffmannED, Hardaker WT, Vail TP, Scully SP:Knee arthrodesis with circular exter-nal fixation. Clin Orthop Relat Res2001;382:168-178.

37. Gross AE, Shasha N, Aubin P: Long-term follow-up of the use of fresh os-teochondral allografts for posttrau-matic knee defects. Clin OrthopRelat Res 2005;435:79-87.

38. Kharrazi FD, Busfield BT, KhorshadDS, Hornicek FJ, Mankin HJ: Osteoar-ticular and total elbow allograft re-construction with severe bone loss.Clin Orthop Relat Res 2008;466:205-209.

39. Lai D, Chen CM, Chiu FY, Chang MC,Chen TH: Reconstruction of juxta-articular huge defects of distal femurwith vascularized fibular bone graft

and Ilizarov’s distraction osteogene-sis. J Trauma 2007;62:166-173.

40. McAuliffe JA, Burkhalter WE, Ouel-lette EA, Carneiro RS: Compressionplate arthrodesis of the elbow. J BoneJoint Surg Br 1992;74:300-304.

41. Mears DC, Velyvis JH: Acute total hiparthroplasty for selected displaced ac-etabular fractures: Two to twelve-yearresults. J Bone Joint Surg Am 2002;84:1-9.

42. Bosse MJ, MacKenzie EJ, Kellam JF, etal: An analysis of outcomes of recon-struction or amputation after leg-threatening injury. N Engl J Med2002;347:1924-1931.

43. Mackenzie EJ, Bosse MJ, Pollak AN,et al: Long-term persistence of disabil-ity following severe lower-limb trau-ma: Results of a seven-year follow up.J Bone Joint Surg Am 2005;87:1801-1809.



Advancements in AnkleArthroscopy

AbstractImportant progress has been made during the past 30 years inarthroscopic ankle surgery. Ankle arthroscopy has graduallychanged from a diagnostic to a therapeutic tool. Most arthroscopicprocedures can be performed by using the anterior working areawith the ankle in dorsiflexion or plantar flexion; there is no needfor routine ankle distraction. Anterior ankle problems, such asanterior impingement syndrome, are approached by anteromedialand anterolateral portals and, if necessary, an accessory portal.Most osteochondral defects can be reached from anterior with theankle in plantar flexion. For a far posterior location, theosteochondral defect can be approached from posterior. The two-portal hindfoot endoscopic technique (ie, both arthroscopic andendoscopic surgery), with the patient in the prone position,provides excellent access to the posterior ankle compartment andto posteriorly located extra-articular structures.

Although Burman1 in 1931 foundthe ankle joint unsuitable for

arthroscopy because of its typicalanatomy, Tagaki2 and, later, Wa-tanabe3 made considerable contribu-tions to arthroscopic surgery. Wa-tanabe3 in 1972 published the resultsof a series of 28 ankle arthroscopies.Numerous studies followed, andduring the past 30 years, arthroscopyof the ankle joint has become an im-portant procedure for the detectionand treatment of chronic and post-traumatic problems. The main indi-cations for anterior arthroscopy arefor treatment of anterior impinge-ment syndrome and talar osteochon-dral defects (OCDs).4,5

Endoscopic surgery (ie, both ar-throscopic and endoscopic surgery),offers the possible advantages of di-rect visualization of structures, im-proved assessment of articular carti-lage, less postoperative morbidity,

faster as well as functional rehabili-tation, earlier resumption of sports,and outpatient treatment.6,7 The val-ue of diagnostic arthroscopy is lim-ited.4,8 Some authors advocateroutine mechanical distraction com-bined with a 2.7-mm arthroscope.9

In most procedures, however, anklearthroscopy can be performed moreeffectively without routine joint dis-traction.4,10

Posterior ankle problems pose adiagnostic and therapeutic challengebecause of their nature and deep lo-cation. By means of a two-portalhindfoot approach with the patientin the prone position, posterior an-kle joint problems (eg, loose bodies,ossicles, osteophytes, OCDs) can betreated.11 In the case of a posteriorimpingement syndrome, bony im-pediments (eg, os trigonum) can bedetached and removed via the two-portal hindfoot approach.11

C. Niek van Dijk, MD, PhD

Christiaan J. A. van Bergen, MD

Dr. van Dijk is Professor and Head,Department of Orthopaedic Surgery,Academic Medical Center, Amsterdam,The Netherlands. Dr. van Bergen isResearch Fellow, Orthopaedic ResearchCenter Amsterdam, Department ofOrthopaedic Surgery, AcademicMedical Center, Amsterdam.

None of the following authors or amember of their immediate families hasreceived anything of value from or ownsstock in a commercial company orinstitution related directly or indirectly tothe subject of this article: Dr. van Dijkand Dr. van Bergen.

Reprint requests: Dr. van Bergen,Orthopaedic Research CenterAmsterdam, Academic Medical Center,G4-262, Meibergdreef 9,1100 DDAmsterdam, The Netherlands.




Anterior AnkleArthroscopy

Indications andContraindications

Ankle problems that can be man-aged by means of routine anteriorankle arthroscopy include soft-tissue and bony impingement, syno-vitis, loose bodies, ossicles, andOCDs.5 Certain ankle fractures(eg, Weber type B distal fibular frac-ture, Tillaux fracture) also can besuccessfully treated by means ofarthroscopy-assisted (open) reduc-tion and internal fixation, whichoffers the advantage of direct visual-ization and treatment of concomi-tant intra-articular injuries.12-14 Oth-er procedures include arthroscopicankle stabilization by means ofradiofrequency and arthroscopy-assisted ankle arthrodesis. In case ofmultiple disorders (eg, a symptomat-ic OCD or ankle impingement withconcomitant ankle instability), acombined treatment is often possi-ble.

Absolute contraindications areinfection and severe degenerativechanges. Relative contraindications

are degenerative changes with di-minished range of motion, narrow-ing of the joint space, vascular dis-ease, and edema.15

DiagnosisThe value of diagnostic ankle ar-

throscopy without a preoperative di-agnosis is limited; only 26% to 43%of patients benefit from the proce-dure.4,8 Hence, a diagnosis should beestablished preoperatively by physi-cal examination and plain radio-graphs. If the diagnosis remains un-clear, additional radiographs (eg, heelrise view, anteromedial impinge-ment view) can be obtained16,17 (Fig-ure 1). Furthermore, local infiltra-tion can be performed. Temporaryrelief from pain after intra-articularinfiltration suggests intra-articularpathology, such as soft-tissue im-pingement or an OCD. When anOCD, a loose body, or an ossicle issuspected, the lesion may be dis-closed by magnetic resonance imag-ing (MRI) or spiral computed tomog-raphy (CT).17 If the CT scan isnegative and the preoperative diag-nosis remains unclear, it is unlikely

that the patient will benefit from di-agnostic arthroscopy.4,8

Routine Fixed DistractionVersus Dorsiflexion

Previous investigators have re-ported the routine use of ankle jointdistraction.9 The available noninva-sive distraction devices are sterile andattach to the operating table. A strapconnected to this distraction deviceis placed around the ankle. The sur-geon using one of these devices standsbeside the patient to perform the ar-throscopic procedure. Ankle arthros-copy without distraction, however, al-lows the surgeon to stand at thebottom end of the operating table. Inthis position, the surgeon can leanagainst the patient’s foot, therebybringing the ankle joint into the dor-siflexed position (Figure 2).

Several important factors favorthe use of dorsiflexion rather thandistraction. First, distraction of thejoint may result in tightening of the

Figure 1

Lateral (A) and anteromedial impingement (B) radiographs of the right ankle of a32-year-old woman who reported progressive pain in the right ankle for 1 year. Onpalpation, a recognizable tenderness on the anteromedial distal tibia was noted.In the lateral view, no pathologic structures can be visualized. In the anteromedialimpingement view, talar and tibial osteophytes are clearly visible (arrows).

Figure 2

Positioning of the patient duringanterior arthroscopy. The affected heelrests on the very end of the operatingtable, making it possible for thesurgeon to fully dorsiflex the ankle jointby leaning against the sole of the foot.

Advancements in Ankle Arthroscopy


anterior capsule, leading to a reduc-tion of the anterior working area18

(Figure 3). Second, loose bodies andosteophytes are usually located inthe anterior compartment of the an-kle joint. Dorsiflexion creates an an-terior working area, which makes re-moval easy. Introduction of salinesolution opens the anterior workingarea. In the case of a loose body anddistraction, the loose body may fallinto the posterior aspect of the joint,which makes removal more diffi-cult. Third, in the dorsiflexed posi-tion, the talus is concealed in thejoint, thereby protecting the carti-lage from potential iatrogenic dam-age.18

Mechanical distraction and use ofa small-diameter arthroscope may bebeneficial in some situations. Theseinclude treatment of ossicles, a soft-tissue impediment, a loose bodycaught in the joint space betweenthe fibula and tibia (the intrinsicsyndesmotic area), an OCD locatedin the posterior tibial plafond, andposterior ankle problems. An impor-tant alternative for the treatment ofposterior ankle problems is a two-portal posterior arthroscopy.11

Should distraction be indicated, aresterilizable noninvasive distrac-

tion device enables the surgeon tochange quickly from the dorsiflexedposition to the distracted positionand vice versa19 (Figure 4).

Surgical TechniqueThe anterior dorsiflexion proce-

dure4 is performed as outpatient sur-gery under general or epidural anes-thesia. The patient is placed in thesupine position with slight elevationof the ipsilateral buttock. A tourni-quet is placed around the upperthigh. The heel of the affected footrests on the very end of the operatingtable, thus making it possible for thesurgeon to fully dorsiflex the anklejoint by leaning against the sole ofthe patient’s foot (Figure 2). The twoprimary anterior portals used for an-terior ankle arthroscopy are the an-teromedial and anterolateral, locatedat the level of the joint line. Whentheir use is indicated, accessory an-terior portals are located just in frontof the tip of the medial or lateralmalleolus. Some surgeons combinethe anterior portals with a postero-lateral portal.20

The anteromedial portal is madefirst. After a skin incision has beencreated just medial to the tibialisanterior tendon, the subcutaneous

layer is bluntly divided with a hemo-stat. The 4-mm, 30°-angle arthro-scope, which we use routinely, is in-troduced in the fully dorsiflexedposition. Saline solution is then in-troduced into the joint. Under ar-throscopic control, the anterolateralportal is made by inserting a spinalneedle lateral to the peroneus tertiustendon while respecting the superfi-cial peroneal nerve (Figure 5).

Depending on the procedure per-formed, the instruments can be ex-changed between portals. Afterremoval of the instruments, the ar-throscopic incisions and accessoryportals are closed with Ethilon su-tures (Ethicon, Piscataway, NJ) to pre-vent sinus formation.

ComplicationsSeveral complications have been

described, including injury to neu-rovascular structures, instrumentbreakage, articular surface damage,neuroma formation, infection, andreflex sympathetic dystrophy.20-23

The superficial peroneal nerve is athighest risk, and injury to this nerveis associated with the anterolateralportal.20

Reports of complications in anklearthroscopy vary widely. With the useof either invasive or manual constantdistraction, complication rates of 9%to 17% have been reported.20-23

Figure 3

Schematic lateral view of the ankle joint. A, In dorsiflexion, the anterior working areais enlarged. B, Distraction of the ankle joint (arrows) results in tightening of theanterior capsule, reducing the anterior working area.

Figure 4

A resterilizable distraction device,which permits the surgeon to move theankle quickly from the dorsiflexedposition to the distracted position andvice versa.

C. Niek van Dijk, MD, PhD, and Christiaan J. A. van Bergen, MD


In the largest series published to date,nearly 50% of complications wereneurologic.20 The use of constantdistraction might be indicative ofhigher complication rates. In a recentsurvey performed in our departmentin which 1,300 consecutive patientswith ankle arthroscopy withoutroutine joint distraction were in-cluded, the overall percentage ofcomplications was 3.4%. This fig-ure includes 1.4% for hindfoot en-doscopy.24

Osteochondral Defects

An OCD is a lesion involving artic-ular cartilage and subchondral bone.The incidence of OCDs of the talardome in patients with acute lateralankle ligament ruptures is 4% to7%.25,26 Talar OCDs are usually pos-teromedial (58%) or anterolateral(42%).27 Medial lesions are typicallydeep and cup-shaped; lateral lesionsare shallow and wafer-shaped.28 In-appropriate treatment of OCDs may

eventually result in osteoarthritis ofthe ankle.28

EtiologyPrevious trauma to the ankle joint

is reported in 93% of lateral lesionsand 61% of medial lesions.27 In laterallesions, the trauma mechanism isusually a combination of inversionand dorsiflexion; in medial lesions,the combination is inversion, plantarflexion, and rotation.29 In nontrau-matic OCDs, possible causes are ge-netic, metabolic, vascular, endocrine,and degenerative factors as well asmorphologic abnormalities.29,30

Clinical PresentationPatients with a chronic lesion

typically experience persistent or in-termittent deep ankle pain during orafter activity, sometimes accompa-nied by swelling and limited range ofmotion. Often, on examination, fewabnormalities are found. Affectedankles may have a normal range ofmotion with the absence of swellingand no recognizable tenderness onpalpation.

DiagnosisRoutine radiographs consist of

weight-bearing anteroposterior andlateral views of both ankles. The ra-diographs may show an area of de-tached bone surrounded by radio-lucency. Initially, the damage maybe too small to be visualized on aroutine radiograph.

A heel rise mortise view may re-veal a posterior defect.17 For furtherdiagnostic evaluation, CT and MRIhave demonstrated similar accura-cy.17 A multislice helical CT scan ispreferred because it is more helpfulfor preoperative planning (Figure 6).

Classification and StagingSeveral classifications based on

radiography, CT, MRI, and arthros-copy have been proposed.29,31-33 Thefirst and most frequently used classi-fication was that of Berndt andHarty:29 stage I, a small compressionfracture; stage II, incomplete avul-

Figure 5

Anterior arthroscopy of a left ankle. A, External view. B, Arthroscopic view. Underarthroscopic control, the anterolateral portal is made by introducing a spinal needlelateral to the peroneus tertius tendon. AJ = anterior joint capsule, Lat = lateral,Med = medial, SN = spinal needle (guiding needle for a lateral portal), Tal = talus,Tib = tibia

Figure 6

Coronal (A) and sagittal (B) computed tomography reconstructions of aposteromedial osteochondral defect of the left talus in a 15-year-old girl. The patientpresented with persistent deep pain on the anterior side of the left ankle and lockingand giving way. One year before her visit, she had had a supination trauma, whichhad been treated nonsurgically.



sion of a fragment; stage III, com-plete avulsion of a fragment withoutdisplacement; and stage IV, displacedfragment. Scranton and McDer-mott34 later added stage V, represent-ing cystic lesions. None of the cur-rent grading systems, however, issufficient to direct the choice oftreatment.

TreatmentVarious surgical techniques for the

treatment of symptomatic osteochon-dral lesions have been published.These are generally based on one ofthe following three principles:30 (1) dé-bridement and bone marrow stimu-lation (microfracturing, drilling, abra-sion arthroplasty); (2) securing a lesionto the talar dome (fragment fixation,retrograde drilling, bone grafting); and(3) development or replacement of hy-aline cartilage (autologous chondro-cyte implantation [ACI], osteochon-dral autograft transplantation [OAT;ie, mosaicplasty], allografts). Thechoice of treatment depends on thepatient’s age, symptoms, duration ofcomplaints, and location and size ofthe defect, as well as whether itconcerns a primary or secondaryOCD.30,35

Asymptomatic or low sympto-matic lesions are treated nonsurgi-cally for a trial period of 6 months,consisting of rest, ice, temporarilyreduced weight bearing, and, in caseof giving way, an orthosis.30,36 Symp-tomatic lesions are treated primari-ly by débridement and bone marrowstimulation.27 With this technique,all unstable cartilage is removed,including the underlying necroticbone. Any cysts underlying the de-fect are opened and curetted. Thesclerotic-calcified zone that is mostcommonly present is perforated bymeans of microfracturing into thevascularized subchondral bone. Theunderlying intraosseous blood ves-sels are disrupted and growth factorsare released, leading to the forma-tion of a fibrin clot in the created de-fect. The formation of local newblood vessels is stimulated, marrow

cells are introduced into the OCD,and fibrocartilaginous tissue isformed.37 In case of a cystic defect≥15 mm in size, a cancellous bonegraft may be placed in the defect.35

Retrograde drilling, combined withcancellous bone grafting when neces-sary, may be performed for primaryOCDs when there is intact cartilagewith a large subchondral cyst.32 Whenprimary treatment fails, OAT or ACIare options.38,39 Although both tech-niques are promising, results are notyet widely published, and the num-ber of patients included in studies issmall.

OAT consists of harvesting one ormore osteochondral plugs in alesser–weight-bearing area of theknee and transplanting them intothe talar defect.38 Although mostreports show excellent results,the technique is associated withdonor site morbidity, and a medialmalleolar osteotomy is often re-quired.40-42 ACI is the implantationof in vitro–cultured autologouschondrocytes, using a periosteal tis-sue cover after expansion of isolatedchondrocytes. Despite excellent re-sults reported by some investiga-tors,39,43 disadvantages include thetwo-stage surgery, high cost, and re-ported donor site morbidity.40,43

Fragment fixation with one ortwo lag screws is preferred in acuteor semiacute situations in which thefragment is ≥15 mm. In adolescents,fixation of an OCD always should beconsidered following failure of a6-month period of conservativetreatment.

The size and location of the lesiondetermine whether to use a standard4.0-mm arthroscope and treat theOCD in the anterior working area byfull plantar flexion of the ankle, or touse a 2.7-mm arthroscope in combi-nation with mechanical distrac-tion.4,36 In patients with unlimitedplantar flexion, all defects located inthe anterior half of the talus or in theanterior part of the posterior half canbe reached and treated by anterior ar-throscopy.4,36 Other options are to

approach the defect from posteriorby means of a two-portal hindfootapproach or to proceed by means ofa medial malleolar osteotomy.11,44

Surgical Technique andRehabilitation

A 4.0-mm scope and a 4.5- or 5.5-mm shaver are routinely used. Incase of synovitis, a local synovecto-my is performed with the ankle inthe dorsiflexed position. The lesionis identified in the forced plantar-flexed position by palpating the car-tilage with a probe or hook (Figure7). During this part of the procedure,a soft-tissue distractor can be applied(Figure 4). If possible, the full-radiusresector is introduced into the de-fect. In doubtful cases, identifyingthe defect by introducing a spinalneedle, probe, or curet can be usefulbefore introduction of the resector.The key to success is identifying theanterior part of the defect and re-moving the unstable cartilage andsubchondral necrotic bone. The in-struments are subsequently broughtinto the defect to treat the posteriorpart. Every step in the débridementprocedure is checked by regularlyswitching portals. After full débride-ment, the sclerotic zone is penetrat-ed by means of a microfracture probeor a Kirschner wire (Figure 7). Post-operatively, a compression dressingis applied.

Active plantar flexion and dor-siflexion are encouraged. Partialweight bearing (ie, egg shell pressure)is allowed as tolerated. We allowprogress to full weight bearing with-in 2 to 4 weeks in patients with cen-tral or posterior lesions of up to 1cm. Larger lesions and anteriorlesions require partial weight bear-ing up to 6 weeks. Running oneven ground is permitted after 12weeks.30 Full return to normal andsporting activities is usually possible4 to 6 months after surgery.

ResultsIn the treatment of osteochondral

lesions, nonsurgical therapy yields a



45% success rate.27 However, a trialperiod of nonsurgical treatment doesnot adversely affect the outcome ofsurgery.44

In a systematic review of the lit-erature, treatment by débridementand bone marrow stimulation wassuperior to other methods, with amean of 86% good or excellent re-sults in 21 studies (total, 272 pa-tients).27 However, OAT and ACIwere not included because of the fewnumber of studies using these meth-ods, and sizes of the treated lesionswere not described.

A recent randomized controlledtrial that compared chondroplasty,microfracture, and OAT showedsimilar results among these methodsat 2-year follow-up.45 However, thechondroplasty and microfracturetechniques are recommended be-cause of less postoperative pain. Fur-thermore, costs are lower comparedwith those of other techniques.40

Anterior AnkleImpingement

Chronic anterior ankle pain is com-monly caused by formation of tibialor talar osteophytes at the anteriorpart of the ankle joint. Morris46 andlater McMurray47 named this condi-tion athlete’s ankle or footballer’sankle; these terms later were re-placed by anterior ankle impinge-ment syndrome. The condition pre-dominately affects soccer players,but it also occurs in runners, balletdancers, high jumpers, and volley-ball players.10

EtiologyVarious theories exist regarding

the causes of anterior impingement:traction, trauma, recurrent micro-trauma, and chronic ankle insta-bility.7,47-49 The hypothesis of trac-tion is not plausible because theanterior joint capsule is attached

more proximally to the site wherethe tibial spurs originate.50-52

In trauma secondary to or associ-ated with supination stress, damageto the non–weight-bearing cartilagerim often occurs.26 A repair reactionis initiated, with cartilage prolifera-tion, scar tissue formation, and cal-cification. Ankle sprains resultingfrom chronic instability as well asforced dorsiflexion enhance this pro-cess.49,53 The pain in anterior ankleimpingement likely is caused by theinflamed soft tissue along the anteriortibiotalar joint line, which is com-pressed by the talar and tibial osteo-phytes during forced dorsiflexion.50,54

Clinical PresentationAnterior ankle impingement is

characterized by anterior pain at thelevel of the ankle joint, swelling af-ter activity, and, occasionally, limit-ed dorsiflexion.7 Athletes with re-current ankle sprains are prone to

Figure 7

Arthroscopic images of débridement and drilling of a lateral osteochondral defect (OCD) of the right talus. A, The arthroscopeis in the anteromedial portal. With the ankle in the neutral position, the OCD is out of the arthroscopic view. B, By bringing theankle into the forced plantarflexed position, the OCD can be seen. C, The size of the lesion is identified by palpating the cartilagewith a probe (arrows), which is inserted through the anterolateral portal. D, Next, a shaver is introduced for débridement of thedefect. E, With the use of a Kirschner wire (K-wire), small holes are drilled in the subchondral bone. F, Arthroscopic view, afterswitching portals, of the microfractured lesion. G, During loosening of the tourniquet, sufficient hemorrhage in the defect ischecked.



this condition, and sporting activi-ties are often reduced because of thepain.

On palpation, recognizable ten-derness is noted on the anterior boneat the joint level. Tenderness on pal-pation medial to the tibialis anteriortendon indicates anteromedial im-pingement; tenderness lateral to theperoneus tertius tendon indicatesanterolateral impingement54 (Figure8). Forced hyperdorsiflexion mayprovoke the pain, but this test is of-ten false-negative.

DiagnosisStandard anteroposterior and lat-

eral radiographs may not detect thepresence of osteophytes. When an-teromedial osteophytes are suspect-ed, an oblique radiograph should beadded because some osteophytesmay be undetected by standard ra-diographs.16 Anteromedial tibial ortalar osteophytes may be over-projected by the anterolateral borderof the distal tibia or by the lateralpart of the talar neck and body, re-spectively. In the oblique anterome-dial impingement view, the beam istilted in a 45° craniocaudal direc-tion, with the leg in 30° external ro-tation and the foot in plantar flexionin relation to the standard lateral ra-diograph position (Figure 9). The an-teromedial impingement radiographhas a high sensitivity for detectinganteromedial osteophytes: 93% fortibial and 67% for talar osteo-phytes16 (Figure 1).

Treatment andRehabilitation

Although nonsurgical treatmentsuch as intra-articular injections andheel lifts is recommended in the ear-ly stage of anterior ankle impinge-ment, outcome is frequently disap-pointing. The arthroscopic approachis performed as described above.With the ankle in the dorsiflexed po-sition, the contour of the anteriortibia is identified by shaving awaythe tissue just superior to the osteo-phyte. A 4-mm chisel and/or motor-

ized shaver system is subsequentlyused to remove them.

A compression bandage is ap-plied, and partial weight bearingfor 3 to 5 days is permitted as toler-ated. Active dorsiflexion is encour-aged.

ResultsOpen resection and arthroscopic

resection of osteophytes were com-pared by Scranton and McDer-mott.6 The average length of hospi-talization and time to recovery wereshorter in the arthroscopic group.

In prospective studies, successrates varied from 73% to 96%.51,55,56

A significant difference in outcomeswas seen between patients with nor-mal joint spaces (90% success) andthose with joint space narrowing(50% success).51 This finding waslater confirmed in two long-termfollow-up studies.52,57 Because the al-ternative in these osteoarthritic pa-tients is arthrodesis, a 50% successrate is acceptable. Both series52,57

also reported a high rate of recur-rence of osteophytes. However, nostatistical correlation was seen be-tween recurrence of osteophytes andreturn of symptoms.

Hindfoot Endoscopy

HistoryThe deep location of hindfoot

structures makes direct access diffi-cult. Historically, the hindfoot wasapproached by a three-portal tech-nique (ie, anteromedial, anterolater-al, posterolateral), with the patientin the supine position.20 The tradi-tional posteromedial portal is associ-ated with potential damage to thetibial nerve, the posterior tibial ar-tery, and local tendons.15

A two-portal endoscopic approachwith the patient in the prone positionwas introduced in 2000.11 This tech-nique has been shown to provide ex-cellent access to the posterior anklecompartment, the subtalar joint, andextra-articular structures.11

IndicationsThe main indications for endos-

copy in the posterior compartmentof the talocrural joint are posterior-

Figure 8

Localization of anterior impingement.When a patient recognizes the localtenderness on palpation lateral to theperoneus tertius tendon (A) or medialto the tibialis anterior tendon (B), thediagnosis of anterolateral oranteromedial impingement is made,respectively. AC = anterocentralregion, AL = anterolateral region,AM = anteromedial region

Figure 9

In the oblique anteromedialimpingement view, the beam is tilted45° in a craniocaudal direction, with theleg in 30° external rotation and thefoot in plantar flexion in relation to thestandard lateral radiograph position.



ly located OCDs of the ankle joint,loose bodies, ossicles, posttraumaticcalcifications or avulsion fragments,posterior tibial rim osteophytes,chondromatosis, and chronic syno-vitis.24 In the posterior compartmentof the subtalar joint, the main indi-cations are osteophytes, loose bod-ies, an intraosseous talar ganglion,and subtalar arthrodesis. Otherextra-articular structures that canpotentially be treated are the poste-rior tibial tendon, the flexor hallucislongus tendon, the peroneal tendons,the Achilles tendon, the deep por-tion of the deltoid ligament, and asymptomatic os trigonum or hyper-trophic talar process.24

Posterior AnkleImpingement

Posterior ankle impingementsyndrome is a clinical diagnosis in

which the patient experiences painin the hindfoot when the ankle isforced into a plantarflexed position(Figure 10). It is caused by overuse ortrauma and mainly occurs in balletdancers, downhill runners, and soc-cer players.54

After an injection of lidocaine,the forced hyperplantar flexion testshould be negative. On plain radio-graphs and CT, an os trigonum orhypertrophic posterior talar processmay be detected. Treatment consistsof resection of a symptomatic ostrigonum, reduction of a prominentposterior talar process, or removal ofa soft-tissue impediment.

Surgical Technique andRehabilitation

The procedure is performed asoutpatient surgery under general orepidural anesthesia. The patient isplaced in a prone position. A tourni-quet is applied around the upper leg,and a small support is placed underthe lower leg, making it possible tomove the ankle freely (Figure 11). Asoft-tissue distraction device can beused when indicated.19 For irriga-tion, normal saline with gravity flowis suitable. A 4.0-mm, 30° arthro-scope is routinely used for posteriorankle arthroscopy. In addition to thestandard excisional and motorized

instruments for treatment of osteo-phytes and ossicles, a 4-mm chiseland small periosteal elevator can beuseful.

With the ankle in the neutral po-sition, a line is drawn from the tip ofthe lateral malleolus to the Achillestendon, parallel to the foot sole. Theposterolateral portal is situated justabove this line, in front of the Achil-les tendon (Figure 12, A). After avertical stab incision is made, thesubcutaneous layer is split by a mos-quito clamp. The mosquito clamp isdirected anteriorly, pointing in thedirection of the interdigital webspace between the first and secondtoe (Figure 13). When the tip of theclamp touches the bone, it is ex-changed for a 4.0-mm arthroscope.The direction of view is 30° to thelateral side.

The posteromedial portal is nowmade at the same level (Figure 12, B).After making a vertical stab incisionin front of the medial aspect of theAchilles tendon, a mosquito clampis introduced and directed towardthe arthroscope shaft in a 90° angle.When it touches the shaft of the ar-throscope, it is moved anteriorly inthe direction of the ankle joint, allthe way down, touching the arthro-scope shaft until it reaches the bone(Figure 14). The arthroscope is now

Figure 10

The forced hyperplantar flexion test.The patient is sitting with the kneeflexed in 90°. The test is performed byrepetitive, quick, passive hyperplantarflexion movements (arrow). This may berepeated in slight external or internalrotation of the foot. The investigatormay apply this rotational movement onthe point of maximal plantar flexion,thereby “grinding” the posterior talarprocess or os trigonum between thetibia and calcaneus.

Figure 11

During hindfoot endoscopy the patient is placed in the prone position. A tourniquet(T) is applied around the upper leg, and a small support (S) is placed under thelower leg, which makes it possible to move the ankle freely.



pulled slightly backward until thetip of the mosquito clamp comesinto view. The clamp is used tospread the extra-articular soft tissuein front of the tip of the lens. In sit-uations where scar tissue or adhe-sions are present, the mosquitoclamp is exchanged for a 4.5-mmfull-radius shaver.

After removal of the very thin jointcapsule of the subtalar joint by a fewturns of the shaver, the posterior com-partment of the subtalar joint can bevisualized. At the level of the anklejoint, the posterior tibiofibular andtalofibular ligaments are identified.The posterior talar process can befreed of scar tissue, and the flexor hal-lucis longus tendon is identified. Theflexor hallucis longus tendon is animportant landmark to prevent dam-age to the medial neurovascular bun-dle (Figure 15). One should always

Figure 12

Portal placement in a right ankle during hindfoot endoscopy. A, The posterolateralportal (PL) is made just above the line from the tip of the lateral malleolus (LM) to theAchilles tendon (AT), parallel to the foot sole, just in front of the Achilles tendon.B, The posteromedial portal (arrow) is made at the same level as the posterolateralportal, just above the line from the tip of the medial malleolus, in front of the medialaspect of the Achilles tendon.

Figure 14

Schematic transverse section of the right ankle joint at the level of the arthroscope.A = the posterolateral portal is made first. The arthroscope is directed toward thefirst web space. B = the posteromedial portal is made at the same level. Themosquito clamp is directed toward the arthroscope shaft at a 90° angle. C = whenthe mosquito clamp touches the shaft of the arthroscope, the clamp is movedanteriorly in the direction of the ankle joint, all the way down, touching thearthroscope shaft until it reaches the bone (curved arrow). All following instruments(eg, shaver) are introduced in the same manner.

Figure 13

The posterolateral instrument (2) isdirected anteriorly, pointing in thedirection of the interdigital web spacebetween the first and second toe (1).



stay lateral to this tendon and movemedially only when release of theneurovascular bundle is indicated (eg,tarsal tunnel syndrome). After re-moval of the thin joint capsule of theankle joint, the ankle joint can be en-tered and inspected.

Removal of a symptomatic os tri-gonum or treatment of a nonunion ofa posterior talar process fracture in-volves partial detachment of the pos-terior talofibular ligament, release ofthe flexor retinaculum, and release ofthe posterior talocalcaneal ligament(Figure 16). The os trigonum can belifted from the subtalar joint bymeans of a small-sized bone elevator(Figure 17) and removed with agrasper. At the end of the procedure,hemorrhage is controlled by electro-cautery, and the skin is closed withEthilon sutures. A sterile compres-sion dressing is applied.

Postoperative treatment is func-tional and consists of weight bearingon crutches as tolerated for 3 days,after which the dressing is removed.The patient is advised to start range-of-motion exercises as soon as possi-

ble after surgery. If necessary, phys-iotherapy is begun.

ResultsWe performed 146 endoscopic

hindfoot procedures on 136 consec-utive patients.24 The main indicationswere bony impingement, OCDs, andflexor hallucis longus tendinitis.Treatment was successful in mostpatients; two minor complications(1.4%) were seen (ie, an area of dimin-ished sensation over the heel pad ofthe hindfoot in both cases).24 Similarresults have been published by othersurgeons.58 Furthermore, the tech-nique described is considered to be asafe method, according to a recentanatomic study.59 It is recommendedthat the procedure be performed by anexperienced arthroscopist who haspracticed this type of surgery in a ca-daveric setting at arthroscopy cours-es.60,61

Summary

Over the last three decades, the fieldof arthroscopic foot and ankle sur-

gery has progressed significantly. Ar-throscopy of the ankle joint has be-come the procedure of choice for thetreatment of chronic and posttrau-matic pathologies. The diagnosisshould be established before surgery.When necessary, CT and additionalradiographs (eg, heel rise view, an-teromedial impingement view) canbe obtained to confirm a clinical di-agnosis and for preoperative plan-ning. When the dorsiflexed positionis used with anterior arthroscopy,

Figure 15

A and B, Hindfoot endoscopy of the right ankle. A combination of two separateprocedures is shown. The flexor hallucis longus (FHL) is the important landmark inposterior ankle arthroscopy. Instruments should always stay lateral to this tendon.Viewing medial to the FHL (panel A) reveals the neurovascular bundle (tibial nerve[N], and the posterior tibial artery and veins [V]). This medial view is indicated onlywhen a tarsal tunnel release is performed. Lat = lateral, M = mosquito clamp inlateral portal, Med = medial

Figure 16

Endoscopic view of a left ankle.Removal of a symptomatic os trigonumor treatment of a nonunion of aposterior talar process fracture involvespartial detachment of the posteriortalofibular ligament (PTFL), release ofthe flexor hallucis longus (FHL)retinaculum, and release of theposterior talocalcaneal ligament (TCL).Lat = lateral, Med = medial

Figure 17

Endoscopic view of a right ankle. Theos trigonum can be lifted from thesubtalar joint with a small boneelevator. Lat = lateral, Med = medial



ankle distraction is necessary onlyin a minority of cases.

Most OCDs can be treated by an-terior arthroscopic débridement andmicrofracturing with the ankle inplantar flexion. The anterior im-pingement syndrome is treated byarthroscopic excision of osteo-phytes. Posterior ankle problems (eg,posterior impingement syndrome)can be effectively treated by meansof a two-portal hindfoot approachwith the patient in the prone posi-tion. This approach offers excellentaccess to the posterior ankle com-partment, the subtalar joint, andextra-articular structures.

Acknowledgment

Peter A. J. de Leeuw, PhD fellow, isgratefully acknowledged for the kindpreparation of Figures 1, 4 through 8,10 through 13, and 15 through 17.

References

Evidence-based Medicine: There arethree level I/II studies (references 17,27, and 45); the remainder are levelIII/IV and level V, including two on-line course descriptions.


1. Burman MS: Arthroscopy or the di-rect visualization of joints: An exper-imental cadaver study. J Bone JointSurg Am 1931;13:669-695.

2. Tagaki K: The arthroscope. J JPN Or-thop Assoc 1939;14:349-411.

3. Watanabe M: Sefloc-Arthroscope(Watanabe no. 24 arthroscope):Monograph. Tokyo, Japan: TeishinHospital, 1972.

4. van Dijk CN, Scholte D: Arthoscopyof the ankle joint. Arthroscopy 1997;13:90-96.

5. Ferkel RD, Small HN, Gittins JE:Complications in foot and ankle ar-throscopy. Clin Orthop Relat Res2001;391:89-104.

6. Scranton PE Jr, McDermott JE: Ante-rior tibiotalar spurs: A comparison ofopen versus arthroscopic debride-ment. Foot Ankle 1992;13:125-129.

7. Cutsuries AM, Saltrick KR, Wagner J,

Catanzariti AR: Arthroscopic arthro-plasty of the ankle joint. Clin PodiatrMed Surg 1994;11:449-467.

8. van Dijk CN, Verhagen RAW, Tol JL:Arthroscopy for problems after anklefracture. J Bone Joint Surg Br 1997;79:280-284.

9. Boynton MD, Parisien JS, Guhl JF,Vetter CS: Setup, distraction, and in-strumentation, in Guhl JF (ed): Footand Ankle Arthroscopy, ed 3. NewYork, NY: Springer, 2004, pp 87-97.

10. Tol JL, van Dijk CN: Anterior ankleimpingement. Foot Ankle Clin 2006;11:297-310.

11. van Dijk CN, Scholten PE, Krips R: A2-portal endoscopic approach for diag-nosis and treatment of posterior anklepathology. Arthroscopy 2000;16:871-876.

12. Takao M, Uchio Y, Naito K, FukazawaI, Kakimaru T, Ochi M: Diagnosis andtreatment of combined intra-articulardisorders in acute distal fibular frac-tures. J Trauma 2004;57:1303-1307.

13. Ono A, Nishikawa S, Nagao A, Irie T,Sasaki M, Kouno T: Arthroscopicallyassisted treatment of ankle fractures:Arthroscopic findings and surgicaloutcomes. Arthroscopy 2004;20:627-631.

14. Panagopoulos A, van Niekerk L: Ar-throscopic assisted reduction and fix-ation of a juvenile Tillaux fracture.Knee Surg Sports Traumatol Ar-throsc 2007;15:415-417.

15. Ferkel RD, Fischer SP: Progress in an-kle arthroscopy. Clin Orthop RelatRes 1989;240:210-220.

16. Tol JL, Verhagen RAW, Krips R, et al:The anterior ankle impingement syn-drome: Diagnostic value of oblique ra-diographs. Foot Ankle Int 2004;25:63-68.

17. Verhagen RAW, Maas M, DijkgraafMGW, Tol JL, Krips R, van Dijk CN:Prospective study on diagnostic strat-egies in osteochondral lesions of thetalus: Is MRI superior to helical CT?J Bone Joint Surg Br 2005;87:41-46.

18. Golano P, Vega J, Pérez-Carro L,Götzens V: Ankle anatomy for the ar-throscopist: Part I. The portals. FootAnkle Clin 2006;11:253-273.

19. van Dijk CN, Verhagen RAW, Tol JL:Resterilizable noninvasive ankle dis-traction device. Arthroscopy 2001;17:1-5.

20. Ferkel RD, Heath DD, Guhl JF: Neu-rological complications of ankle ar-throscopy. Arthroscopy 1996;12:200-208.

21. Guhl JF: New concepts (distraction)in ankle arthroscopy. Arthroscopy1988;4:160-167.

22. Barber FA, Click J, Britt BT: Compli-cations of ankle arthroscopy. FootAnkle 1990;10:263-266.

23. Unger F, Lajtai G, Ramadani F, Aitzet-müller G, Orthner E: Arthroscopy ofthe upper ankle joint: A retrospectiveanalysis of complications [German].Unfallchirurg 2000;103:858-863.

24. van Dijk CN: Hindfoot endoscopy.Foot Ankle Clin 2006;11:391-414.

25. Bosien WR, Staples OS, Russell SW:Residual disability following acuteankle sprains. J Bone Joint Surg Am1955;37:1237-1243.

26. van Dijk CN, Bossuyt PMM, MartiRK: Medial ankle pain after lateral lig-ament rupture. J Bone Joint Surg Br1996;78:562-567.

27. Verhagen RAW, Struijs PAA, BossuytPMM, van Dijk CN: Systematic re-view of treatment strategies for osteo-chondral defects of the talar dome.Foot Ankle Clin 2003;8:233-242.

28. Canale ST, Belding RH: Osteochon-dral lesions of the talus. J Bone JointSurg Am 1980;62:97-102.

29. Berndt AL, Harty M: Transchondralfractures (osteochondritis dissecans)of the talus. J Bone Joint Surg Am1959;41:988-1020.

30. Zengerink M, Szerb I, Hangody L,Dopirak RM, Ferkel RD, van Dijk CN:Current concepts: Treatment of os-teochondral ankle defects. FootAnkle Clin 2006;11:331-359.

31. Anderson IF, Crichton KJ, Grattan-Smith T, Cooper RA, Brazier D: Os-teochondral fractures of the dome ofthe talus. J Bone Joint Surg Am 1989;71:1143-1152.

32. Taranow WS, Bisignani GA, TowersJD, Conti SF: Retrograde drilling of os-teochondral lesions of the medial ta-lar dome. Foot Ankle Int 1999;20:474-480.

33. Hepple S, Winson IG, Glew D: Osteo-chondral lesions of the talus: A re-vised classification. Foot Ankle Int1999;20:789-793.

34. Scranton PE Jr, McDermott JE: Treat-ment of type V osteochondral lesionsof the talus with ipsilateral knee os-teochondral autografts. Foot AnkleInt 2001;22:380-384.

35. Giannini S, Buda R, Faldini C, et al:Surgical treatment of osteochondrallesions of the talus in young activepatients. J Bone Joint Surg Am2005;87(suppl 2):28-41.

36. Schuman L, Struijs PAA, van DijkCN: Arthroscopic treatment for os-teochondral defects of the talus: Re-sults at follow-up at 2 to 11 years.J Bone Joint Surg Br 2002;84:364-368.

37. O’Driscoll SW: Current concepts re-



view: The healing and regeneration ofarticular cartilage. J Bone Joint SurgAm 1998;80:1795-1812.

38. Hangody L, Fules P: Autologous os-teochondral mosaicplasty for thetreatment of full-thickness defects ofweight-bearing joints. J Bone JointSurg Am 2003;85(suppl 2):25-32.

39. Baums MH, Heidrich G, Schultz W,Steckel H, Kahl E, Klinger HM: Autol-ogous chondrocyte implantation fortreating cartilage defects of the talus.J Bone Joint Surg Am 2006;88:303-308.

40. Giannini S, Vannini F: Operativetreatment of osteochondral lesions ofthe talar dome: Current concepts re-view. Foot Ankle Int 2004;25:168-175.

41. Baltzer AWA, Arnold JP: Bone-cartilage transplantation from the ip-silateral knee for chondral lesions ofthe talus. Arthroscopy 2005;21:159-166.

42. Reddy S, Pedowitz DI, Parekh SG,Sennett BJ, Okereke E: The morbidityassociated with osteochondral har-vest from asymptomatic knees for thetreatment of osteochondral lesions ofthe talus. Am J Sports Med 2007;35:80-85.

43. Whittaker JP, Smith G, Makwana N,et al: Early results of autologous chon-drocyte implantation in the talus.J Bone Joint Surg Br 2005;87:179-183.

44. Alexander AH, Lichtman DM: Surgi-cal treatment of transchondral talar-dome fractures (osteochondritis disse-cans): Long-term follow-up. J BoneJoint Surg Am 1980;62:646-652.

45. Gobbi A, Francisco RA, Lubowitz JH,

Allegra F, Canata G: Osteochondrallesions of the talus: Randomized con-trolled trial comparing chondroplas-ty, microfracture, and osteochondralautograft transplantation. Arthros-copy 2006;22:1085-1092.

46. Morris LH: Report of cases of athlete’sankle. J Bone Joint Surg Am 1943;25:220.

47. McMurray TP: Footballer’s ankle.J Bone Joint Surg Br 1950;32:68-69.

48. Tol JL, Slim E, van Soest AJ, van DijkCN: The relationship of the kickingaction in soccer and anterior ankleimpingement syndrome: A biome-chanical analysis. Am J Sports Med2002;30:45-50.

49. Biedert R: Anterior ankle pain insports medicine: Aetiology and indi-cations for arthroscopy. Arch OrthopTrauma Surg 1991;110:293-297.

50. Tol JL, van Dijk CN: Etiology of theanterior ankle impingement syn-drome: A descriptive anatomicalstudy. Foot Ankle Int 2004;25:382-386.

51. van Dijk CN, Tol JL, Verheyen CCPM:A prospective study of prognostic fac-tors concerning the outcome of ar-throscopic surgery for anterior ankleimpingement. Am J Sports Med1997;25:737-747.

52. Tol JL, Verheyen CP, van Dijk CN: Ar-throscopic treatment of anterior im-pingement in the ankle: A prospectivestudy with a five- to eight-year follow-up. J Bone Joint Surg Br 2001;83:9-13.

53. Krips R, van Dijk CN, Halasi T, et al:Anatomical reconstruction versus te-nodesis for the treatment of chronicanterolateral instability of the ankle

joint: A 2- to 10-year follow-up, mul-ticenter study. Knee Surg SportsTraumatol Arthrosc 2000;8:173-179.

54. van Dijk CN: Anterior and posteriorankle impingement. Foot Ankle Clin2006;11:663-683.

55. Amendola A, Petrik J, Webster-Bogaert S: Ankle arthroscopy: Out-come in 79 consecutive patients.Arthroscopy 1996;12:565-573.

56. Baums MH, Kahl E, Schultz W,Klinger HM: Clinical outcome of thearthroscopic management of sports-related “anterior ankle pain”: A pro-spective study. Knee Surg SportsTraumatol Arthrosc 2006;14:482-486.

57. Coull R, Raffiq T, James LE, StephensMM: Open treatment of anterior im-pingement of the ankle. J Bone JointSurg Br 2003;85:550-553.

58. Willits K, Sonneveld H, Amendola A,Giffin JR, Griffin S, Fowler PJ: Out-come of posterior ankle arthroscopyfor hindfoot impingement. Arthros-copy 2008;24:196-202.

59. Lijoi F, Lughi M, Baccarani G: Posteri-or arthroscopic approach to the ankle:An anatomic study. Arthroscopy2003;19:62-67.

60. Arthroscopy Association of NorthAmerica. Master courses: Foot/ankle.Available at: http://www.aana.org/cme/MastersCourses/descriptions.aspx#Foot/Ankle. Accessed August 28,2008.

61. Amsterdam Foot & Ankle Platform.Available at: http://www.ankleplatform.com/page.php?id=854. AccessedAugust 28, 2008.



RadiocarpalFracture-dislocations

AbstractRadiocarpal fracture-dislocations most often are caused by high-energy trauma. These difficult, uncommon injuries involvesignificant soft-tissue and osseous trauma, requiring meticulousreduction and fixation. The mechanism of injury is generally asevere shear or rotational insult. Anatomically, the dislocationresults in disruption of the radiocarpal ligaments and, usually, boththe radial and the ulnar styloid. Understanding the anatomy of theradiocarpal joint is central to understanding the osseous and soft-tissue constraints that are disrupted with a radiocarpal dislocation.Diagnosis can be reliably made on physical examination andradiographic evaluation. Radiocarpal fracture-dislocation injuriesmust be differentiated from Barton fractures. Associated injuriessuch as open fractures, neurovascular involvement, and distalradioulnar dislocations also must be taken into account. Closedreduction can be obtained relatively easily, but open reduction andinternal fixation is typically necessary to ensure accurate anatomicrestoration of injured bone and ligaments.

Radiocarpal fracture-dislocationsare complex injuries character-

ized by dislocation of the radiocarpaljoint. The dislocation may be in ei-ther a dorsal or a volar direction andcan be associated with fractures ofthe cortical rim of the distal radius,the radial styloid, and the ulnar sty-loid. Prior to the descriptions of Pou-teau1 in 1783 and Colles2 in 1814, adeformed wrist after injury was of-ten considered to be dislocated.3

However, these early authors, with-out the benefit of radiographic as-sessment of the injury, in most cas-es were making an educated guess.On occasion, a postmortem exami-nation of the affected wrist wasdone. According to Malgaigne,4 frac-tures of the distal radius accountedfor <10% of the fractures seen atl’Hotel Dieu in Paris between 1827and 1830. However, Dupuytren,5

who worked at the same institutionas did Malgaigne, believed that frac-tures of the distal radius were com-mon. Both agreed that the previous-ly held view of deformed wristsrepresenting a dislocation ratherthan a fracture was incorrect.3

The radiocarpal dislocation injurywas truly first recognized and de-scribed by Malle in 1838, when heidentified a volar radiocarpal fracture-dislocation.3 Shortly thereafter, Mar-jolin3 and Voillemier6 identified andreported dorsal radiocarpal fracture-dislocations. All of these observa-tions were made from examinationof postmortem specimens.

Radiocarpal fracture-dislocationsare estimated to account for just0.2% of all wrist injuries.7 Few largeseries exist, and most accounts inthe literature consist of casereports.8-19 Destot8 reported the first

Asif M. Ilyas, MD

Chaitanya S. Mudgal, MD

Dr. Ilyas is Director, Temple HandCenter, and Assistant Professor,Department of Orthopaedic Surgery andSports Medicine, Temple UniversityHospital, Philadelphia, PA. Dr. Mudgal isInstructor in Orthopaedic Surgery,Harvard Medical School, MassachusettsGeneral Hospital, Boston, MA.

None of the following authors or amember of their immediate families hasreceived anything of value from or ownsstock in a commercial company orinstitution related directly or indirectly tothe subject of this article: Dr. Ilyas andDr. Mudgal.

Reprint requests: Dr. Mudgal, YawkeyCenter, Suite 2100, 55 Fruit Street,Boston, MA 02114.




radiographically documented case ofa radiocarpal fracture-dislocation in1926. The patient had an open inju-ry and eventually succumbed to sep-sis. The few series that are availablesuggest that these injuries are the re-sult of severe high-energy traumaand that they occur most common-ly in young men.20-25

These injuries are associated witha spectrum of injury patterns. Dorsaldislocations are more common thanvolar dislocations20,22,24,25 (Figure 1).The soft-tissue disruption can leadto radiocarpal instability, resultingin ulnar translocation and multidi-rectional radiocarpal instability.19,26

Radiocarpal fracture-dislocationinjuries must be differentiated frommarginal or rim fractures of the dis-tal radius. These latter injuries havebeen eponymously associated withJohn Rhea Barton,27 who in 1838provided what is considered to bethe earliest description of a margin-al shearing fracture of the distal endof the radius. Radiocarpal fracture-dislocations represent a high-energyshear and rotational injury to the

wrist with or without a fracture ofthe radius or ulna. In contrast, Bar-ton fractures are compression inju-ries in which the articular surface ofthe fractured distal radius remains incontact with the proximal carpalrow holding the intact radiocarpalligaments (Figure 2). In addition, thedisplaced articular fragment in a Bar-ton fracture forms a substantial partof the entire distal radial articularsurface, which is in contrast to thesmaller cortical rim or styloid frac-tures that typically occur with radio-carpal fracture-dislocations.

Anatomy andPathophysiology

The articular surface of the distal ra-dius is biconcave and triangular,with the radial styloid process form-ing the apex of the triangle. The sig-moid notch forms the base and artic-ulates with the head of the ulna. Theextrinsic radiocarpal ligaments, cap-sule, and the scaphoid and lunatefossae of the distal radius providestability to the radiocarpal joint.28

The volar surface of the distal ra-dius is relatively flat. However, thevery distal margin slopes volarly inthe form of a ridge from which thestrong volar radiocarpal ligamentsoriginate. The short radiolunate lig-ament begins at the ulnar volar mar-gin of the lunate facet and inserts onthe volar surface of the lunate. Theshort radiolunate ligament is the pri-mary soft-tissue restraint againstvolar translation of the carpus.29

More radially, the radioscapholu-nate ligament, the long radiolunateligament, and the stout radioscapho-capitate ligaments take their originalong the volar rim of the distal radi-us.29 The stout radioscaphocapitateprovides restraint against ulnartranslation of the carpus.30 On theulnar side of the wrist, the ulnolu-nate and ulnotriquetral ligamentsoriginate on the volar side of thetriangular fibrocartilage complex,which in turn inserts into the base ofthe ulnar styloid.31

The dorsal surface of the distal ra-dius is convex and serves as the floorof the dorsal extensor compart-

Figure 1

Posteroanterior (A) and lateral (B) views of a dorsal radiocarpal fracture-dislocation.

Figure 2

Lateral view of a Barton fracturedepicting the large articular fracturefragment (arrows) that, in contrast toa radiocarpal fracture-dislocation,remains in continuity with the carpus.

Radiocarpal Fracture-dislocations


ments. When viewed from the later-al aspect, the most distal dorsal edgeof the radius extends past the distaledge of the volar surface, providingthe site of origin of the dorsal radio-carpal ligaments.

Radiocarpal fracture-dislocationsmay be envisaged as an internal dis-articulation of the wrist joint. Thisglobal injury of varying degrees is aproduct of several factors: position ofthe radiocarpal joint at impact, thestrength of the radiocarpal liga-ments, the strength of the bonystructures, and the magnitude andrate of deforming forces. Bohler,32

who originally postulated the mech-anism of injury for dorsal radiocarpalfracture-dislocations, stated that acompressive and rotational force oc-curs against a hyperextended andpronated wrist.

Our understanding of injuries in-volving the carpus, its ligaments,and the distal radius changed consid-erably after the comprehensive de-scription of injury mechanisms byMayfield et al33 in 1980. These de-scriptions clarified the spectrum ofinjury that can occur with similarloading patterns. These authors werethe first to suggest that the rotation-al force, which is an essential featurein the causation of radiocarpalfracture-dislocations, is in effect an“intercarpal supination.” They fur-ther demonstrated that wrist exten-sion and ulnar deviation producedtension on the volar radiocarpal lig-aments, which caused avulsions ofthe volar radial lip or radial styloid.Thus, it appears that in addition to arotational component, radiocarpalfracture-dislocations involve a shearand angulatory component, withtranslation of the carpus resulting inpeeling off of the carpus from the ra-dius and ulna. Depending on themagnitude and direction of the force,the wrist deforms, with a variableamount of bony and soft-tissue inju-ry. The most commonly avulsedfracture fragments include the radi-al styloid by the radioscaphocapitateligament,30 the volar lunate facet by

the short radiolunate ligament,29 andthe ulnar styloid.

Clinical Features

The patient typically presents witha painful, swollen, and deformedwrist. The most common mecha-nisms of injury are falls from aheight, motor vehicle injuries, andindustrial injuries.20-25 Consequently,open wounds and associated injuriesare common (Figure 3). Recognitionof this injury demands a completetrauma evaluation for injuries ofall extremities and organ systems.In one series of open radiocarpalfracture-dislocations, an associatedfracture or injury to other organ sys-tems was found in every patient.22

Neurologic deficits of the injuredextremity are common and are oftenassociated with vascular insufficiencyof the hand.20-24 Arterial occlusionsecondary to the deformity may re-sult in ischemia, which should be cor-rected by expeditious reduction of thejoint with longitudinal traction.21

Neurologic injury is also common,particularly with open injuries.22,24

The median nerve is more often in-volved than is the ulnar nerve.24 Lesscommonly, a radiocarpal dislocationmay be associated with an irreducibledistal radioulnar joint dislocationwhen soft tissue (eg, tendon, nerves)or osteoarticular fragments becomeincarcerated within the joint.24,34,35

RadiographicEvaluation

Plain radiographs consisting of pos-teroanterior and lateral views of thewrist are obtained initially to assistin making the diagnosis. Evaluationof fracture geometry is facilitatedwith radiographs obtained after pro-visional reduction with longitudinaltraction. Examination of the radio-graphs should begin with identifica-tion of fractures of the distal radius,distal ulna, and carpal bones. An ob-lique view can aid in identifyingfracture fragments. Fractures of the

radial or ulnar styloid are examinedcarefully, with emphasis placed ontheir size and location.

On the posteroanterior view,alignment of the carpus is evaluatedby examining the position of the lu-nate relative to the radius. Normal-ly, the lunate is in alignment withthe ulnar column of the distal radi-us, with a minimum of two thirds ofthe lunate articulating with the dis-tal radius.36 With complete radiocar-pal ligament disruption, the carpustends to translocate ulnarly downthe radioulnar slope of the distal ra-dius (Figure 4). The posteroanteriorradiograph is also carefully exam-ined for intercarpal ligament injury,such as scapholunate or lunotrique-tral dissociation. The relationship ofthe radiocarpal and midcarpal jointsis assessed by evaluating the align-ment of the capitate, the lunate, andthe articular surface of the lunatefossa of the distal radius. The threeGilula arcs—radiocarpal, proximalmidcarpal, distal midcarpal—shouldbe colinear.37 Disruption of the Gilu-la arcs or overlapping of normallyequally spaced carpal bones is high-ly suggestive of injury to the sup-porting ligaments, the carpal bones,or both.37 Scapholunate ligament in-juries in particular must be suspect-ed when a radial styloid fractureexits at the interval between thescaphoid and the lunate fossae38 (Fig-ure 5).

Figure 3

Open wound associated with dorsalradiocarpal fracture-dislocation.

Asif M. Ilyas, MD, and Chaitanya S. Mudgal, MD


Loss of colinearity of the lunatewith the articular surface of the radi-us on the lateral view indicates dis-ruption of the radiocarpal joint (Fig-ure 6). In general, the lateral viewdemonstrates the direction of the ra-diocarpal dislocation. Marginal rimfractures are best evaluated on the

lateral view. The so-called teardropview or 10° proximal view is helpfulin evaluating fractures of the volarrim and lunate facet.36

With technological advances, theroutine use of computed tomographyfor wrist injuries is increasing. Al-though not mandatory, computed to-

mography can aid in the evaluationof cortical rim fractures, fracture de-pression of the articular surface, andthe relationship of the carpus anddistal radioulnar joint (Figure 7).Magnetic resonance imaging may beuseful in studying soft-tissue inju-ries, particularly in evaluating thescapholunate and lunotriquetral lig-aments. Occult injury to the inter-carpal ligaments has been suggestedto result in late intercarpal disrup-tion.16,25

Classification

Classification of radiocarpal frac-ture-dislocations ideally should en-compass the identification of allbone and soft-tissue injuries, gradingthe risk of instability, and subse-quently directing treatment. Twoclassification schemes have beendiscussed in the literature.

Moneim et al21 classified radio-carpal fracture-dislocations into twotypes based on the presence or ab-sence of concurrent injury to the in-tercarpal articulations (Table 1). In atype I injury, the normal carpal anat-

Figure 4

Ulnar translation of the carpus afterreduction of a radiocarpal dislocation.Note that the lunate is positioned overthe distal ulna and not the lunate fossaof the distal radius.

Figure 5

Scapholunate ligament injurypresenting with mild scapholunatediastasis. Note the radial styloidfracture exiting at the level of thescapholunate interval, indicating a highrisk of a scapholunate ligamentdisruption. Such a disruption wasconfirmed intraoperatively.

Figure 6

Lateral view of dorsal dislocationillustrating loss of colinearity of thecapitate, the lunate, and the articularsurface of the radius.

Figure 7

A, Plain radiograph demonstrating radiocarpal fracture-dislocation. Anterior (B) andlateral (C) three-dimensional computed tomography scans illustrating ulnar anddorsal subluxation of the radiocarpal joint associated with a large radial styloidfracture fragment.



omy is maintained, with dislocationof the radiocarpal joint. Type II inju-ries involve intercarpal injuries (spe-cifically, scapholunate or lunotrique-tral intercarpal ligament injuries) inaddition to the radiocarpal disloca-tion. This second type represents amore complex pattern and could beconsidered a variation of a perilu-nate fracture-dislocation as de-scribed by Mayfield et al.33

Dumontier et al25 also classifiedradiocarpal fracture-dislocations in-to two groups, but their system isbased on the extent of radial styloidinvolvement (Table 2). Group 1 inju-ries include pure ligamentous radio-carpal dislocations or dislocationswith only a small cortical or radialstyloid avulsion fracture. Group 2injuries include dislocations with alarge radial styloid fracture fragmentinvolving at least one third of thescaphoid fossa of the distal radius.On the dorsal surface, the ligamen-tous injury represents more of a cap-suloperiosteal avulsion than a truetear of the dorsal radiocarpal liga-ments. Group 1 injuries consist ofglobal ligamentous injuries, whichhave the potential for multidirec-tional instability; as such, they posea greater treatment challenge thando group 2 injuries.25 In group 2 inju-ries, the radiocarpal ligaments re-main attached to the fractured largeradial styloid fragment.30 Stabilitycan more reliably be restored withsecure anatomic fixation of the frac-ture than can occur with the morepure ligamentous injuries seen ingroup 1.

Management

Successful management of radiocar-pal fracture-dislocations requires eval-uation and treatment of the columnsof the wrist, while taking into ac-count the direction of dislocation andthe presence of any intercarpal inju-ries. Three treatment principles arerecommended: (1) concentric reduc-tion of the radiocarpal joint, (2) iden-tification and treatment of intercar-

pal injuries, and (3) stable repair of theosseous-ligamentous avulsions.

To better direct surgical treat-ment and to address all aspects ofthe injury, we have adopted the co-lumnar concept of the carpus as de-scribed by Navarro39 and modifiedby Taleisnik,40 as well as the col-umns of the distal radius and ulna asdescribed by Rikli and Regazzoni.41

Each column of the distal radius andulna, namely, the radial, the inter-mediate, and the ulnar, is ap-proached separately in a stepwisefashion to achieve radiocarpal stabil-ity.41 Concomitantly, the columns ofthe carpus, which include the mo-bile lateral column (ie, scaphoid), theflexion-extension central column (ie,lunate, distal carpal row), and therotatory medial column (ie, tri-quetrum), are evaluated for intercar-pal ligamentous injury and resultantcarpal instability.40 By addressing ev-ery column individually, radiocarpaland intercarpal stability can beachieved.

Although closed reduction andcast immobilization have been re-ported to yield satisfactory results inthe management of radiocarpal dis-locations,10,11,14,21 we consider theseinjuries to be complex and unstableconditions that routinely warrantsurgical reduction and fixation to at-tain a stable, concentric, and con-

gruent wrist. All irreducible dislo-cations, open injuries, and casesinvolving neurovascular embar-rassment require surgical treat-ment.21,22,24,35

The steps in surgical treatment ofradiocarpal fracture-dislocation are(1) provisional radiocarpal joint re-duction, (2) decompression of neu-rovascular structures, (3) exposureand débridement of the joint, (4)treatment of intercarpal injuries, and(5) fracture fixation and/or soft-tissue repair (Figure 8). We recom-mend the use of general anesthesia.

The wrist is provisionally reducedwith longitudinal traction. An exter-nal fixator may be applied to holdthe joint reduced. An extensile volarapproach ulnar to the flexor tendonsand median nerve is used so thatboth the carpal tunnel and Guyoncanal can be decompressed as need-ed. The radiocarpal joint is examinedthrough the volar capsular site ofdisruption. The joint is irrigated anddébrided of any loose cartilage orbone fragments. Stay sutures or su-ture anchors are then placed in thearea of capsular and ligament disrup-tion but are not tied down. Fluoros-copy is used to identify any carpalfractures or interosseous ligamentinjuries, particularly of the scapho-lunate or lunotriquetral ligaments.Intercarpal ligament injuries are

Table 1

Moneim et al21 Classification of Radiocarpal Fracture-dislocation

Type I Radiocarpal fracture-dislocation without associatedintercarpal dissociation

Type II Radiocarpal fracture-dislocation with an associatedintercarpal dissociation

Table 2

Dumontier et al25 Classification of Radiocarpal Fracture-dislocation

Group 1 Radiocarpal fracture-dislocation that is purely ligamentousor involves only a small cortical avulsion fracture off theradius

Group 2 Radiocarpal fracture-dislocation associated with a largeradial styloid fracture fragment (involving at least onethird of the scaphoid fossa)



confirmed and treated through a sep-arate dorsal capsular incision. A sub-periosteal approach through thefloor of the third extensor compart-ment is used. The columns of thejoint are approached sequentially.

Beginning with the radial col-umn, the fractured radial styloid isaccurately reduced and internallyfixed. Fixation options include aKirschner wire, compression screw,or plate application (Figure 9). Eithera Kirschner wire or screw fixationcan provide stable fixation of the ra-dial styloid. Screw fixation providesthe added benefit of compression, as-suming that the styloid fragment islarge enough to accommodate ascrew without requiring later re-moval and pin-tract complications.Volar, radial, or dorsal plating is se-lected based on the fracture person-ality and surgeon preference. Mov-ing to the intermediate (ie, central)column, fractures of the lunate facetthat are amenable to fixation shouldbe repaired with internal fixation us-

Figure 8

Treatment algorithm for radiocarpal fracture-dislocations. DRUJ = distal radioulnar joint, ExFix = external fixation, ORIF = openreduction and internal fixation

Figure 9

Anteroposterior (A) and lateral (B) radiographs demonstrating open reduction andinternal fixation of the radial and intermediate columns of a radiocarpal fracture-dislocation. Note the screw fixation of the radial styloid and anchor fixation withinthe distal radius (ie, volar ligament repair).



ing screws or a tension band wireloop.42 When fractures of the radialstyloid and/or the lunate facet arenot amenable to fixation, soft-tissuerepair is undertaken by direct suturerepair or with suture anchors (Figure9). Stay sutures that previously wereplaced to repair the extrinsic volarligaments are tied. The origins of theshort radiolunate and radioscapho-capitate ligaments are repaired inparticular to avoid late volar sublux-ation or ulnar translocation, respec-tively. Reduction and stability of thefixation is confirmed both visuallyand radiographically.

The ulnar column is approachedin the presence of injury to the distalradioulnar joint and ulnar supportligaments (ie, ulnolunate, ulnotri-quetral) or when instability persistsafter fixation of the radial and inter-mediate (ie, central) columns. Largeulnar styloid fractures require inter-nal fixation with screws or tensionband wiring. This procedure usuallyrestores a concentric distal radioul-nar joint. In the presence of persis-tent instability, the distal radioulnarjoint is examined and evacuated ofany interposed tissue, followed byrepair of the ulnocarpal ligaments.Persistent instability can be ad-dressed by pinning the distal radioul-nar joint in midsupination.

Additional stability to the con-struct can be provided with the useof an external fixator or radiolunatepin. The external fixator is especiallyuseful in situations in which dailycare of an open wound is needed. Ap-plication of a radiolunate pin can beused intraoperatively to maintain sta-ble reduction of the radiocarpal jointwhile fracture fixation and soft-tissuerepair are undertaken. If necessary,the external fixator or radiolunate pinmay be left in situ postoperatively for4 to 6 weeks to reinforce reduction ofthe radiocarpal joint.

Outcome

Despite being a complex wrist inju-ry, radiocarpal fracture-dislocation

can achieve a satisfactory outcomeprovided the surgeon follows thesetreatment principles: concentric re-duction of the radiocarpal joint,treatment of intercarpal injuries,and sound repair of the osseous-ligamentous injury. There are fewlarge series on this subject in the lit-erature and, to our knowledge, onlythree with more than eight pa-tients.22,24,25

Mudgal et al24 reported on a seriesof 12 patients who presented with ra-diocarpal fracture-dislocation. Fourcases were open injuries, seven hadneurologic compromise, two had anintercarpal ligamentous injury, andfive had an associated injury. Exclud-ing patients with concomitant inter-carpal injury, mean wrist motion onfollow-up assessment consisted of53° of extension, 59° of flexion, 82° ofpronation, and 74° of supination.These results are consistent with theother large series in the literature,which indicate that an overall 30%to 40% decrease in total arc of wristflexion-extension can be expectedfollowing successful open treat-ment.22,25

Using the criteria of Knirk and Ju-piter,43 Mudgal et al24 identified 3 of12 patients as having evidence of ra-diocarpal arthritis. Dumontier etal23 reported that 3 of 27 patients de-veloped radiocarpal arthritis, whileSchoenecker et al44 reported thatfour of the six patients in their seriesdeveloped arthritis.

Factors predictive of an inferioroutcome include open injury, com-plete radiocarpal ligamentous injury,associated nerve injury, and intercar-pal ligamentous injury. Intercarpalinjury, particularly of the scapholu-nate ligament, can significantlycompromise outcome. Moneim etal21 used the presence of such an in-jury to classify and guide treatment.In their series, all three patients weretreated surgically; unfortunately, allhad an inferior outcome. Often, thenegative impact of an intercarpal in-jury manifests late because the inju-ry is initially missed and goes un-

treated, resulting in late midcarpaland radiocarpal instability.16,25

The presence of associated inju-ries is common and can portend aninferior outcome. Neurologic inju-ries are generally neurapraxic, andresolution can be expected with de-compression.24 More severe nervecompression or stretch injuries re-sult in an inconsistent neurologic re-covery. Nyquist and Stern22 reportedon 10 cases of open radiocarpalfracture-dislocations in which all 10were complicated by an associatedinjury and 7 involved neurologiccompromise. At follow-up, all pa-tients had variable and inconsistentrecovery of sensibility. This is con-sistent with a study by Soong andRing,45 who reported on ulnar nervepalsies following distal radius frac-tures. These authors found that ul-nar nerve injuries are typicallyneurapraxic and that patients usual-ly experience normal or near-normalrecovery of function following de-compression.

Ulnar translocation and multidi-rectional instability may result fol-lowing complete radiocarpal liga-mentous injury in which there is nofracture of the distal radius, as in thecase of the group 1 injuries describedby Dumontier et al.25 Early reportshighlighted this instability patternas a late finding encountered duringclosed treatment.14,17,26 In a cadaver-ic study of radiocarpal instability,Rayhack et al31 sequentially sec-tioned the radiocarpal ligaments andfound that ulnar subluxation of thejoint required transection of both theradioscaphocapitate and the radiolu-nate ligaments. Viegas et al46 con-firmed this finding in their cadav-eric study and further studiedmultidirectional instability. Theseauthors found volar translation to beevident with less ligament disrup-tion than that needed for ulnar trans-lation. It was always evident in thepresence of ulnar translation of thecarpus. Such injury in conjunctionwith loss of the ulnolunate liga-ments led to progression of the inju-



ry from ulnar translocation to mul-tidirectional instability. The authorssuggested that the presence of ulnartranslation represents a much moreglobal ligament disruption.

Complications

The most common complication fol-lowing radiocarpal dislocation orfracture-dislocation is residual loss ofmotion and instability. On average,a patient can expect to lose 30% to40% of total arc of wrist flexion/extension.24,25 The other major com-plication is posttraumatic arthritis re-lated to residual articular step-off.22,24,25 Chronic radiocarpal anddistal radioulnar instability or ulnartranslation of the carpus are morecommon with group 1 injury pat-terns.14,15,17,25,26 Less commonly, sep-tic arthritis, tendon rupture, and hard-ware irritation have been reported.24,25

Summary

Radiocarpal fracture-dislocations arethe products of high-energy traumaand represent a shear and rotationalinjury to the wrist, with a variableamount of bone and soft-tissue in-jury. A high index of suspicion for as-sociated injuries must be main-tained, particularly for open wounds,neurologic compromise, and injuriesto other organ systems. Radiographsare adequate for diagnosis but mustbe carefully scrutinized for injury tothe normal carpal relationships. Al-though closed reduction has been de-scribed, we recommend open reduc-tion and internal fixation for thesecomplex and unstable injuries. Avolar and dorsal surgical approach isused. Surgical principles include con-centric reduction of the radiocarpaljoint, identification and treatment ofintercarpal injuries, and stable repairof the osseous-ligamentous avul-sions. Despite the complexity of theoriginal trauma, a satisfactory out-come is attainable. However, a resid-ual loss of motion of 30% to 40% isexpected.

References

Evidence-based Medicine: No levelI or II studies are cited. Level III/IV(case reports and case-control cohortstudies) references include 7-26, 34,35, 38, and 41-45. Level V (expertopinion) references include 1-6, 27,36, 37, 39, and 40.


1. Pouteau C: Mémoire ContenantQuelques Réflexions sur QuelquesFractures de L’avant-Bras sur les Lux-ations Incomplètes du Poignet et surLateral Epicondylitis Diastasis. Paris,France: Ph. Pierres, 1783.

2. Colles A: On the fracture of the carpalextremity of the radius. EdinburghMed Surg J 1814;10:182-186.

3. Fernandez DL, Jupiter JB: The fractureof the distal end of the radius: A his-torical perspective, in Fernandez DL,Jupiter JB (eds): Fractures of the DistalRadius: A Practical Approach toManagement (Chapter 1), ed 2. NewYork, NY: Springer, 2002, pp 1-22.

4. Malgaigne J: Treatise on Fractures.Philadelphia, PA: Translated by J.Packard Lippincott, 1859.

5. Dupuytren G: On the Injuries andDiseases of Bone. London, UK: Trans-lated by F. G. Clark Sydenham Soci-ety, 1847.

6. Voillemier M: Histoire d’une luxationcomplète et récente du poignet en ar-rière suivit de réflexions sur le mécan-isme de cette luxation. Arch GenMed 1839;6:401-417.

7. Dunn AW: Fractures and dislocationsof the carpus. Surg Clin North Am1972;52:1513-1538.

8. Destot E: Injuries of the Wrist: A Ra-diological Study. New York, NY: PaulB. Hoeber, Inc, 1926.

9. Weiss C, Laskin RS, Spinner M: Irre-ducible radiocarpal dislocation: Acase report. J Bone Joint Surg Am1970;52:562-564.

10. Freund LG, Ovesen J: Isolated dorsaldislocation of the radiocarpal joint: Acase report. J Bone Joint Surg Am1977;59:277.

11. Fehring TK, Milek MA: Isolated volardislocation of the radiocarpal joint: Acase report. J Bone Joint Surg Am1984;66:464-466.

12. Varodompun N, Limpivest P, Prin-yaroj P: Isolated dorsal radiocarpaldislocation: Case report and literature

review. J Hand Surg [Am] 1985;10:708-710.

13. Moore DP, McMahon BA: Anteriorradio-carpal dislocation: An isolatedinjury. J Hand Surg [Br] 1988;13:215-217.

14. Penny WH III, Green TL: Volar radio-carpal dislocation with ulnar translo-cation. J Orthop Trauma 1988;2:322-326.

15. Thomsen S, Falstie-Jensen S: Palmardislocation of the radiocarpal joint.J Hand Surg [Am] 1989;14:627-630.

16. Naranja RJ Jr, Bozentka DJ, PartingtonMT, Bora FW Jr: Radiocarpal disloca-tion: A report of two cases and a re-view of the literature. Am J Orthop1998;27:141-144.

17. Howard RF, Slawski DP, Gilula LA:Isolated palmar radiocarpal disloca-tion and ulnar translocation: A casereport and review of the literature.J Hand Surg [Am] 1997;22:78-82.

18. Watanabe K, Nishikimi J: Transsty-loid radiocarpal dislocation. HandSurg 2001;6:113-120.

19. Freeland AE, Ferguson CA, McCraneyWO: Palmar radiocarpal dislocationresulting in ulnar radiocarpal translo-cation and multidirectional instabili-ty. Orthopedics 2006;29:604-608.

20. Bilos ZJ, Pankovich AM, Yelda S:Fracture-dislocation of the radiocar-pal joint. J Bone Joint Surg Am 1977;59:198-203.

21. Moneim MS, Bolger JT, Omer GE: Ra-diocarpal dislocation: Classificationand rationale for management. ClinOrthop Relat Res 1985;192:199-209.

22. Nyquist SR, Stern PJ: Open radiocar-pal fracture-dislocations. J Hand Surg[Am] 1984;9:707-710.

23. Dumontier C, Lenoble E, Saffar P: Ra-diocarpal dislocations and fracture-dislocations, in Saffar P, Cooney WPIII (eds): Fractures of the Distal Radi-us. London, UK: Martin Dunitz, 1995,pp 267-279.

24. Mudgal CS, Psenica J, Jupiter JB: Radio-carpal fracture-dislocation. J HandSurg [Br] 1999;24:92-98.

25. Dumontier C, Meyer zu ReckendorfG, Sautet A, Lenoble E, Saffar P, AllieuY: Radiocarpal dislocations: Classifi-cation and proposal for treatment. Areview of twenty-seven cases. J BoneJoint Surg Am 2001;83:212-218.

26. Fennell CW, McMurtry RY, Fair-banks CJ: Multidirectional radiocar-pal dislocation without fracture: Acase report. J Hand Surg [Am] 1992;17:756-761.

27. Barton JR: Views and treatment of animportant injury of the wrist.Medical Examiner and Record of



Medical Science 1838;1:365-368.28. Ritt MJ, Stuart PR, Berglund LJ, Lin-

scheid RL, Cooney WP III, An KN: Ro-tational stability of the carpus relativeto the forearm. J Hand Surg [Am]1995;20:305-311.

29. Berger RA, Landsmeer JM: The pal-mar radiocarpal ligaments: A study ofadult and fetal human wrist joints.J Hand Surg [Am] 1990;15:847-854.

30. Siegel DB, Gelberman RH: Radial sty-loidectomy: An anatomical studywith special reference to radiocarpalintracapsular ligamentous morpholo-gy. J Hand Surg [Am] 1991;16:40-44.

31. Rayhack JM, Linscheid RL, DobynsJH, Smith JH: Posttraumatic ulnartranslation of the carpus. J HandSurg [Am] 1987;12:180-189.

32. Bohler L: Verrenkungen der handge-lenke. Acta Chir Scand 1930;67:154-177.

33. Mayfield JK, Johnson RP, KilcoyneRF: Carpal dislocations: Pathome-chanics and progressive perilunar in-stability. J Hand Surg [Am] 1980;5:226-241.

34. Ayekoloye CI, Shah N, Kumar A,Kurdy N: Irreducible dorsal radiocar-

pal fracture dislocation with dissocia-tion of the distal radioulnar joint: Acase report. Acta Orthop Belg 2002;68:171-174.

35. Fernandez DL: Irreducible radiocarpalfracture-dislocation and radioulnardissociation with entrapment of theulnar nerve, artery and flexor profun-dus II-V: Case report. J Hand Surg[Am] 1981;6:456-461.

36. Medoff RJ: Essential radiographicevaluation for distal radius fractures.Hand Clin 2005;21:279-288.

37. Gilula LA: Carpal injuries: Analyticapproach and case exercises. AJR AmJ Roentgenol 1979;133:503-517.

38. Mudgal C, Hastings H: Scapho-lunatediastasis in fractures of the distal radi-us: Pathomechanics and treatmentoptions. J Hand Surg [Br] 1993;18:725-729.

39. Navarro A: Cited by Scaramuzza RF:El moviomiento de rotacion en el car-po y su relacion con la fisiopatologicade sus lesiones traumaticas. BolTrabajos Soc Argent OrtopedTraumatol 1976;34:337.

40. Taleisnik J: The ligaments of the

wrist. J Hand Surg [Am] 1976;1:110-118.

41. Rikli DA, Regazzoni P: Fractures ofthe distal end of the radius treated byinternal fixation and early function: Apreliminary report of 20 cases. J BoneJoint Surg Br 1996;78:588-592.

42. Chin KR, Jupiter JB: Wire-loop fixa-tion of volar displaced osteochondralfractures of the distal radius. J HandSurg [Am] 1999;24:525-533.

43. Knirk JL, Jupiter JB: Intra-articularfractures of the distal end of the radiusin young adults. J Bone Joint SurgAm 1986;68:647-659.

44. Schoenecker PL, Gilula LA, ShivelyRA, Manske PR: Radiocarpal frac-ture–dislocation. Clin Orthop RelatRes 1985;197:237-244.

45. Soong M, Ring D: Ulnar nerve palsyassociated with fracture of the distalradius. J Orthop Trauma 2007;21:113-116.

46. Viegas SF, Patterson RM, Ward K: Ex-trinsic wrist ligaments in the patho-mechanics of ulnar translation insta-bility. J Hand Surg [Am] 1995;20:312-318.



Venous Thromboembolismin Spine Surgery

AbstractVenous thromboembolism is a life-threatening adverse event inspine patients and presents difficult decisions for the surgeon andpatient. Prophylactic protocols have been established to preventthe occurrence of venous thromboembolism and its sequelae,including venous occlusion, edema, postthrombotic syndrome, anddeath. Despite the known benefits of prophylaxis, some surgeonschoose not to use it because of concerns over increased bleedingcomplications and possible iatrogenic neurologic injury. Althoughmechanical prophylaxis remains an important element in venousthromboembolism prevention, low-molecular-weight heparin isbetter than other pharmacologic therapies in decreasing theincidence of major events.

Venous thromboembolism (VTE)causes significant morbidity

and mortality and has a notable eco-nomic impact. Deep vein thrombo-sis (DVT) and pulmonary embolism(PE) occur in 48 and 69 per 100,000patients per year, respectively. Theseevents were identified as fatal in 2and 17 per 100,000 patients per year,respectively.1 The cost per episode ofDVT and PE averages $9,337 and$12,795, respectively. These costsare compounded by the increasedrisk of future VTE events and theoccurrence of postthrombotic syn-drome. Recurrent DVT and fatal PEhave been reported to occur in21.5% and 2.6% of patients, respec-tively, after a single previous DVT.2

Orthopaedic patients are at signif-icant risk for thromboembolic dis-ease. Considerable research has beenconducted on the occurrence andtreatment of VTE in connectionwith traumatic orthopaedic injuries,total joint arthroplasties, and spinalsurgery. Increased age, intraoperativevenous stasis, and postoperative im-mobilization and convalescence put

these patients at risk for developingVTE. Numerous studies demon-strate that the prophylactic use ofpharmacologic anticoagulants, suchas low-molecular-weight heparin(LMWH) and low-dose heparin, andmechanical devices, such as inter-mittent pneumatic compressionsleeves, statistically reduces the in-cidence of VTE in the orthopaedicpatient population.3-5 However, con-cerns about bleeding complications,wound hemorrhage, and compres-sive spinal epidural hematoma aswell as the subsequent neurologicsequelae have resulted in diverseprophylactic and treatment proto-cols for VTE in the spine surgical pa-tient population.6

Pathophysiology

VTE, comprising DVT and PE, wasdescribed by Rudolph Virchow, aGerman physician and pathologist,in the 1850s. The Virchow triad ofhypercoagulability, venous stasis,and venous intimal injury that leadsto VTE was not described in the

Christopher A. Heck, MD

Christopher R. Brown, MD

William J. Richardson, MD

Dr. Heck is Orthopaedic Spine Surgeon,Southern Orthopaedic Surgeons LLC,Montgomery, AL. Dr. Brown is AssistantProfessor, Division of OrthopaedicSurgery, Duke University, Durham, NC.Dr. Richardson is Professor, Division ofOrthopaedic Surgery, Duke University.

None of the following authors or amember of their immediate families hasreceived anything of value from or ownsstock in a commercial company orinstitution related directly or indirectly tothe subject of this article: Dr. Heck,Dr. Brown, and Dr. Richardson.

Reprint requests: Dr. Heck, SouthernOrthopaedic Surgeons LLC, 2119 EastSouth Boulevard, Montgomery, AL36116.




medical literature until approxi-mately 100 years after his originalwork.7 Surgery patients with VTE of-ten demonstrate many, if not all, ofthe characteristics of the Virchowtriad.

More recently, patient popula-tions that are at increased risk forVTE from an inherited or acquiredhypercoagulable state have beenidentified. Inheritable abnormalitiesinclude the factor V Leiden muta-tion; deficiencies of proteins C, S,and antithrombin III; mutations of

the prothrombin promoter G20210Aand methylenetetrahydrofolate re-ductase C677T genes; and polymor-phism of the plasminogen activatorinhibitor-1 4G/4G and platelet gly-coprotein IIb/IIIa A1/A2 and A2/A2proteins. Factor V Leiden mutation,which accounts for 25% of all inher-ited thrombophilic disorders, occursas a result of a point mutation in thefactor V gene, which makes the fac-tor resistant to degradation by acti-vated protein C and, thus, propelsboth the intrinsic and extrinsic coag-

ulation cascades forward8 (Figure 1).However, in a matched cohort of to-tal hip arthroplasty (THA) patientswith and without postoperativeDVT, the incidence of factor VLeiden mutations (8% versus 4%,respectively) was not different.9 Infact, antithrombin III deficiency andprothrombin G20210A mutationwere statistically correlated withpostoperative PE in another matchedcohort study of THA patients receiv-ing postoperative DVT prophylax-is.10 Common acquired thrombo-

Figure 1

The coagulation cascade. Subendothelial collagen and tissue factor, both released during vessel intimal injury, initiate theintrinsic and extrinsic pathways, respectively, of the cascade. Ca++ = calcium, HMWK = high-molecular-weight kininogen,PK = prekallikrein

Christopher A. Heck, MD, et al


philic states include malignancy,elevated hormone conditions, (eg,hormone replacement, oral contra-ceptive therapy, late pregnancy), andantiphospholipid antibodies (includ-ing lupus anticoagulant and anticar-diolipin antibody). Malignancy hasrecently been associated with up to20% of all new diagnoses of VTE.11

Venous stasis occurs during surgi-cal positioning and retraction ofvenous structures. The condition iscreated by the absence of activemuscle contraction, vessel occlusionfrom both external and internal (sur-gical retraction) sources, and postop-erative bed rest and immobiliza-tion.

Venous intimal injury during sur-gery can initiate both the intrinsicand extrinsic coagulation cascades.The intrinsic cascade begins whensubendothelial collagen binds toplasma proteins, which, in turn, ac-tivate factor XII. The extrinsic cas-cade starts with activation of factorVII by calcium and tissue factor,a cellular membrane lipoprotein,which is exposed during intimal in-jury (Figure 1). With the exception ofanterior lumbar interbody fusionprocedures, during which the venacava and common iliac veins aresubject to injury, this mechanism ofthrombosis initiation is not highlygermane to the spine surgery popula-tion.

Elective SpinalReconstruction andDecompression

IncidenceThe natural history of VTE as a

result of elective spinal surgerywithout prophylaxis is poorly report-ed. In 1966, Prothero et al12 reportedon two series, each of 500 lumbarand lumbosacral fusion patients,compared one decade apart. Theyfound that the initial incidence ofVTE in spine patients was 4.2% andhad decreased to 2.2% at the secondevaluation. The method of surveil-lance and the use or nonuse of pro-

phylaxis were not noted. More re-cent studies have provided greaterdetail regarding prophylactic regi-mens and the subsequent incidenceof VTE at the expense of providingan untreated control group. In astudy employing only intermittentpneumatic compression prophylaxisand prospectively using duplex ul-trasound screening, Epstein13 dem-onstrated a 2.8% incidence of VTEin patients undergoing multilevellaminectomy with instrumented fu-sion.

In eastern Asia, extremely lowrates of VTE have been described inpatients undergoing THA withoutprophylaxis.14 Similarly, the onlystudies evaluating spine surgery co-horts without the use of mechanicalor pharmacologic prophylaxis havecome from that region. Using con-trast venography, one study reporteda 15.5% incidence of DVT, althoughonly 0.9% of occurrences were prox-imal to or inclusive of the poplitealvein.15 No patients were clinicallysymptomatic.

Pharmacologic prophylaxis toprevent VTE is not universally usedin elective spine surgery. Many sur-geons fear resulting complications,including wound hemorrhage andspinal epidural hematoma. In a pro-spective randomized double-blindinvestigation, a once-daily dose ofLMWH and twice-daily low-doseheparin demonstrated similar DVTrates (1.1% and 2.2%, respectively).Venography confirmed this result.No bleeding complications or neuro-logic deficits were identified.16

More recently, in a retrospectivestudy of 1,954 patients undergoingelective spine surgery in the cervi-cal, thoracic, and lumbar spine,Gerlach et al17 reported a 0.05% riskof VTE and a 0% risk of PE whencompression stockings were usedin conjunction with pharmacologicprophylaxis consisting of LMWH (ie,nadroparin, 2,850 IU/day) initiatedwithin 24 hours after surgery. Ofconcern, however, was a 0.4% inci-dence of spinal epidural hematoma.

A progressive postoperative neuro-logic deficit was present in 77% ofpatients with a spinal epidural he-matoma. Only 60% of the patientswho developed a progressive deficitwere discharged with a normalneurologic examination, even afterprompt surgical decompression.This rate of spinal epidural hemato-ma is slightly higher than other re-ports of 0.1% to 0.19%.18,19

Few data exist regarding warfarinutilization for VTE prophylaxis inspine surgery. In a prospective studygroup of 110 patients, 35 were ran-domized to receive warfarin andelastic stockings.20 No VTE was ob-served with prospective ultrasoundscreening. However, two patientsexperienced excessive postoperativeblood loss. This did not occur in theother two randomized groups inwhich patients received mechanicalprophylaxis alone.

Risk FactorsIdentification of risk factors for

VTE associated with elective spinalreconstruction is difficult because ofa lack of specifics in the literature re-garding surgical approaches in pa-tients who sustain a VTE. In addi-tion to the general surgical riskfactors discussed above, a history ofVTE, decreased postoperative mobil-ity, and increased surgical time havebeen reported as risk factors.21

Although statistical support forspine-specific risk factors is lacking,several studies note common obser-vations for patients who sustained aVTE. The most commonly cited as-sociation was an anterior or com-bined anterior/posterior approach tothe thoracolumbar spine. Dearbornet al22 cited a 6.1% incidence of PEin 97 patients undergoing combinedanterior/posterior approaches. How-ever, in 201 patients undergoing aposterior-only approach, there wasjust one occurrence of PE (0.5%).

Studying spine patients treatedwithout pharmacologic or mechani-cal prophylaxis, Oda et al15 demon-strated a statistically significant risk

Venous Thromboembolism in Spine Surgery


of VTE associated with age and ana-tomic location of surgery. The aver-age age of the 17 patients who sus-tained a VTE was 66 years, whereasthose who did not had an average ageof 58 years (P = 0.04). Also, there wasa significantly higher rate (P = 0.003)of VTE among patients undergoinglumbar surgery (26.5%) versus thosetreated with cervical surgery (5.6%).No statistically significant differ-ence was noted for thoracic spinesurgery (incidence, 14.3%) comparedwith cervical or lumbar surgery. Theinvestigators concluded that therewas no statistically increased riskassociated with sex, surgical time,surgical blood loss, duration of post-operative recumbency, or intervalfrom surgery to venography. Otherinvestigators have also noted thepoor positive predictive value ofDVT screening, with venography orultrasound, for predicting the inci-dence of PE.22

TraumaticFracture/Dislocation

IncidenceIn a prospective study evaluating

hypercoagulability in 101 trauma pa-tients with a minimal and averageInjury Severity Score of 15 and 27,respectively, the overall incidence ofVTE was 30%.23 The authors did notfind a statistically significant differ-ence in the rate of VTE betweenthose receiving prophylaxis andthose not receiving it. They reportedan increased rate of VTE in obese pa-tients, in patients older than age 40years, in patients treated with >3days immobilization, and in patientswith spine or lower extremity frac-tures. However, when a logistic re-gression model was used, only obe-sity (P = 0.004) and prolongedimmobilization (P = 0.05) remainedstatistically significant risk factors.The authors concluded that traumainduces a hypercoagulable state, asevidenced by elevated coagulationmarkers that may remain ≤1 monthafter injury. One year later, however,

in a retrospective databank analysisof more than 450,000 patients, dif-ferent investigators noted only a0.36% incidence of clinically evi-dent VTE.24 Thus, the reported ratesof VTE in the trauma population arelargely dependent on the studymethodology.

In more clinically oriented retro-spective investigations, symptomat-ic VTE in spine trauma patients withno or minimal neurologic injury (ie,Frankel scale E or D) is low. Tropi-ano et al25 reported an absence ofVTE events in 45 patients with tho-racolumbar or lumbar burst frac-tures treated with closed reductionand casting. A specific prophylacticprotocol was not described. The pa-tients were mobilized after only 12hours postoperatively. In a retrospec-tive review of 143 surviving traumapatients with thoracolumbar burstfractures, the rate of VTE was 2.1%,with no statistically significant dif-ference between those treated surgi-cally versus nonsurgically.26

Only one study prospectively com-pared alternative types of prophylaxisin patients with spinal trauma. Kur-toglu et al27 used duplex ultrasoundto screen 120 patients with severehead and spinal trauma, including 11patients with spinal fractures. Pa-tients were randomized to receive ei-ther intermittent pneumatic com-pression or intermittent pneumaticcompression and LMWH (enoxaparin,40 mg/day). The LMWH was initiatedapproximately 24 hours after admis-sion. Overall, there was a 5.8% inci-dence of DVT and a 5% incidence ofPE. The latter occurred more fre-quently in patients with known DVT.There was no difference betweentreatment groups. More important,the investigators found no increasedrate of cranial epidural hematoma ex-acerbation in the group treated withLMWH, as evidenced on follow-upcomputerized tomography scans ofthe head. Although the authors didnot objectively evaluate for spinal epi-dural hematoma, none was foundsubjectively in either group.

Risk FactorsBecause of the aggressive moni-

toring and resuscitation of traumapatients, the reported rates of symp-tomatic VTE in such patients arevery low. Thus, there are few identi-fiable risk factors for VTE in spinetrauma patients. Risk factors forthese patients, therefore, must beobtained from the general traumaliterature. In a review of 1,602 trau-ma patients with VTE, Knudson etal24 identified the strongest factors toinclude ventilator dependency >3days, age >40 years, lower extremityfracture, major head injury, venousinjury, and major surgical procedure.Other studies suggest that indepen-dent risk factors also include obesi-ty, blood transfusion, and spinal cordinjury.23,28

Spinal Cord Injury

IncidenceWhen prophylaxis is not used,

VTE in the patient with spinal cordinjury is common, with a minimumincidence of 80%.28,29 However,when mechanical prophylaxis is ini-tiated, the rates of VTE statisticallydecrease. In a small, prospectivestudy, Green et al30 reported a 40%DVT rate in 15 patients who re-ceived intermittent pneumatic com-pression alone.

Both low-dose heparin andLMWH have been shown to decreasethe incidence of VTE in patientswith spinal cord injury. In a multi-center randomized trial, the efficacyof low-dose heparin was comparedwith enoxaparin in 476 patientswith spinal cord injury.31 Althoughall patients were evaluated for thesafety arm of the study, only 107were evaluated by prospective du-plex ultrasound examination of thelower extremities and monitored forsymptomatic PE, which was con-firmed with imaging studies. VTEwas found in 26 patients. No statis-tically significant difference was ob-served between treatment groups(12.4% for patients treated with



LMWH [enoxaparin] versus 16.3%for those treated with low-dose hep-arin).

From this group of patients, 119underwent 6 weeks of inpatient reha-bilitation.32 The group treated withlow-dose heparin had a 2.5 timesgreater risk of VTE during the reha-bilitation phase than did the patientstreated with LMWH (21.7% versus8.5%, respectively; P = 0.052).

Clinically symptomatic VTEevents have been identified less fre-quently. In a retrospective review ofmedical records from 16,240 pa-tients with spinal cord injury, Joneset al33 found a 5.4% incidence ofVTE in the first 3 months after inju-ry. This increased to 6.0% at 1 year.

Risk FactorsSpinal cord injury is a significant

independent risk factor for VTE.28

Paradoxically, paraplegia has beenassociated with a higher incidence ofthromboembolic disease than hasquadriplegia. Jones et al33 found astatistically higher rate (P = 0.009) of

VTE at 3 months in complete para-plegics (11%) versus complete quad-riplegics (7.8%). It is hypothesizedthat persons with lower-level inju-ries, such as at the thoracolumbarjunction, develop flaccid paralysis,and consequently increased venouscapacitance and stasis, which mayincrease the incidence of VTE com-pared with quadraplegic patientswith spastic paralysis.34

Increased age in persons with spi-nal cord injury is a significant riskfactor. Jones et al33 reported that pa-tients older than age 30 years had astatistically higher rate of VTE; therate then statistically diminished inpatients older than age 80 years.

Another commonly reported riskfactor is the temporal relationship toinjury. Several authors have docu-mented that 79% to 95% of VTE di-agnoses in patients with spinal cordinjury occur within the first 3months after injury.35,36 DeVivo etal37 further noted a higher risk forfatal PE for spinal cord injury—500times that of uninjured matched

control patients—within 1 monthafter injury; this decreased to 20times higher risk at 6 months. De-spite the decreased risk with time, abimodal distribution for VTE hasbeen reported, peaking approximate-ly at days 30 and 100 after injury.38

The first peak coincided with thepreviously reported early risk. Thesecond peak, however, coincidedwith rehabilitation and was also as-sociated with discontinuation ofprophylaxis at approximately 55days after injury.

Aito et al39 recently evaluated therelationship between delayed pro-phylaxis and risk assessment. In pa-tients admitted to their rehabilita-tion hospital, the authors identifieda 2% rate of DVT in patients receiv-ing prophylaxis <72 hours after inju-ry versus a 26% rate in patients forwhom prophylaxis was delayedlonger than 7 days after injury. How-ever, this study does not addresswhat, if any, prophylaxis had beenused at the primary hospital. Fur-ther, the severity of the injury mayhave increased the risk of DVT inpatients who could not be trans-ferred during the early (<72 hours)treatment period, thus potentiallyaffecting the reported rate of DVT.Nevertheless, it is evident that pa-tients are at an increased risk forVTE during the early stages after in-jury and that early prophylaxis isprudent.

Prophylaxis

Guidelines issued by the SeventhAmerican College of Chest Physi-cians (ACCP) Conference recom-mend no prophylaxis for patientswithout additional risk factors whoreceive elective spine surgery21 (Ta-ble 1). For the patient with a historyof VTE or any previously listed riskfactor, investigators recommendLMWH and intermittent pneumaticcompression alone or in combina-tion. For this patient population,data support the initiation of phar-macologic prophylaxis within 24

Table 1

Venous Thromboembolism Prophylactic Guidelines Recommended by theSeventh American College of Chest Physicians Conference21

Type ofSpine Surgery Risk Factors

ProphylacticRecommendations

Electivereconstructive/decompression

No risk factorsOne or more risk factors:

age ≥60 years, BMI≥30kg/m2, geneticthrombophilia, historyof VTE, anterior orcombined procedure,thoracic/lumbar/sacralprocedure

NoneEarly (<24 hrs

postoperatively) LMWHand IPC

Traumaticfracture/dislocation

No spinal cord injuryEvidence of systemic

hemorrhage

Early LMWH and IPCDelayed LMWH and

early IPCSpinal cord injury — Early (24-72 hrs after

injury) LMWH and IPC— Conversion to warfarin

(INR 2.5) until 4months after injury

BMI = body mass index, INR = international normalized ratio, IPC = intermittentpneumatic compression, LMWH = low-molecular-weight heparin, VTE = venousthromboembolism



hours after surgery.16,17 However, thetreating surgeon should monitor forprogressive neurologic deficits sug-gesting a spinal epidural hematoma.Data for warfarin use in this patientpopulation are limited and discour-aging; thus, warfarin is not recom-mended.

For a patient who sustains a trau-matic fracture or dislocation of thespine without spinal cord injury, theSeventh ACCP Conference recom-mends immediate mechanical com-pression devices and LMWH as soonas safety permits.21 Current data sug-gest that simultaneous utilization ofintermittent pneumatic compres-sion and LMWH may decrease bothlocal occurrence and proximal prop-agation of DVT.31 Concern exists,however, that early pharmacologicprophylaxis may exacerbate localtraumatic spinal epidural hematomaand iatrogenic neurologic deteriora-tion. Recent prospective data indi-cate no increased incidence of spinalor cranial epidural hematoma whenLMWH is initiated approximately24 hours after injury in patients withno evidence of continuing hemor-rhage.27

Recommendations by the Sev-enth ACCP Conference are the samefor trauma patients with spinal cordinjury.21 Although most reports de-scribe the safety of LMWH when itis initiated within 72 hours of inju-ry,31,39 there are data to mandate ear-lier initiation, within 24 hours afterinjury, in attempts to diminish therisk of VTE.40 However, the effect ofpharmacologic anticoagulation onthe injured spinal cord is unknown;thus, such therapy should be usedwith caution in patients with in-complete neurologic deficits. TheSeventh ACCP Conference also rec-ommends continuation of LMWH orconversion to warfarin with a targetinternational normalized ratio of 2.5during the rehabilitation phase.21 Be-cause most VTE events do occurwithin 3 months after injury, currentrecommendations are to continueprophylaxis for 3 to 4 months, de-

pending on the activity level of thepatient.33,35,36 Suggested rationalesfor this delayed decreased risk in-clude acquired spasticity of the in-volved extremity muscle groups andintrinsic vascular changes in para-lyzed limbs.34,41

Treatment

Treatment of existing VTE in thepostoperative spine surgery patientpopulation remains controversial.Bleeding complications and subse-quent neurologic deterioration havebeen associated with pharmacologicanticoagulation.42,43 Similar poor re-sults and high mortality with obser-vation alone are reported.42 For pa-tients with a known VTE or who areconsidered to be at high risk for VTEbut are not considered candidates forpharmacologic anticoagulation ther-apy, inferior vena cava filters haveshown good results in the electivereconstruction and trauma patientpopulation.44-47 Although a high rateof DVT occurs, the symptomatic PErate is 0% to 1.3%, with no fatalitiescaused by filter insertion. This is incontrast with a 13% PE incidence ina retrospective, matched cohort ofhigh-risk spine surgery patients.Aside from efficacy, these studies

also support the safety of the inferi-or vena cava filters. In appropriatepatients, a retrievable inferior venacava filter has a 95% success rate bythe recommended 2-week retrievaldate. Current multicenter investiga-tions are evaluating whether thissuccess rate can be achieved whenretrieval is delayed to 4 weeks afterinsertion.

No data exist regarding timing ofpharmacologic therapy initiation forknown VTE after elective surgery ortraumatic injury. In the report byCain et al43 on therapeutic anticoag-ulation for PE after spine surgery,five of six patients who sustainedmajor bleeding complications didnot receive therapeutic heparin until≥6 days after surgery. Therefore,treatment is left to the discretion ofthe surgeon and should be based onclinical judgment.

Complications WithAnticoagulation

The most significant complication as-sociated with perioperative pharma-cologic prophylaxis in spine surgeryis spinal epidural hematoma (Figure2) and its neurologic sequelae. Inci-dence rates ranging from 0.1% to0.7% have been reported.17-19 How-

Figure 2

T1-weighted axial (A) and T2-weighted sagittal (B) emergent magnetic resonanceimaging scans demonstrating a dissecting, compressive epidural hematoma in a79-year-old man who developed acute onset writhing low back pain after an L2/3and L3/4 laminectomy, followed by lower extremity weakness over the next 24hours. The patient underwent emergent decompression with eventual completereturn of neurologic function.



ever, this occurrence has also beenstudied in the absence of anticoagu-lation.48 Although pharmacologic pro-phylaxis and management of VTEmay increase the risk of spinal epidu-ral hematoma, because it is a rare out-come, minimal data exist regardingwhich patients are more susceptibleto developing this complication. Us-ing logistic regression to evaluate riskfactors, Kou et al18 suggested that onlypatients having multilevel proceduresand with coagulopathy were at a sta-tistically higher risk of developing aspinal epidural hematoma.

Typical signs and symptoms ofspinal epidural hematoma includeprogressive neurologic deficit; how-ever, they may be less specific, withback pain as the only symptom.17,49

In their report of 13 patients withpostoperative spinal epidural he-matoma, Gerlach et al17 found a newneurologic deficit in 77%. Once thediagnosis is made, early surgical de-compression provides the best sce-nario for recovery of neurologicfunction. Vandermeulen et al50 de-scribed good or partial neurologic re-covery in patients who had a surgicaldecompression within 8 hours aftersymptom onset. Gerlach et al17 re-ported a 60% complete neurologicrecovery after early surgical decom-pression.

Other sites of postoperative bleed-ing have been associated with phar-macologic prophylaxis in the spinaltrauma population. In an evaluationof 476 patients with spinal cord in-jury, upper gastrointestinal and sur-gical site bleeding accounted for 16 of19 major bleeding complications.31

No spinal epidural hematomas werereported.

Summary

VTE in patients undergoing spinesurgery remains difficult to prevent,diagnose, and treat. Aito et al39 doc-umented symptoms in only 35% ofpatients with confirmed DVT de-spite prophylaxis. The variability ofpatients undergoing spine surgery

contributes to the complexity of itsmanagement. The utilization of me-chanical and pharmacologic prophy-laxis has led to decreased incidenceof VTE in patients undergoing spinesurgery. With current treatmentprotocols, including immediatemechanical and early (<24 hours)pharmacologic prophylaxis, sympto-matic VTE rates are reported at ap-proximately 0.05%, 2.1%, and 6.0%for patients who underwent electivespine surgery, sustained spinal frac-tures, and had spinal cord injury, re-spectively.17,26,33

Data exist to support prophylacticand treatment protocols for spine pa-tients. After elective spine surgery,patients with no concurrent risk fac-tors do not require prophylaxis. How-ever, because most elective spine pa-tients are elderly or have other riskfactors, most of these patients shouldreceive LMWH and intermittentpneumatic compression in the earlypostoperative period. Prophylaxis forthe patient with spinal fracture oftenrequires a multidisciplinary approachby the spine surgeon, trauma surgeon,and additional treating physicians.When no internal or external sourcesof bleeding are apparent, prophylaxisinitiation should begin early withLMWH and intermittent pneumaticcompression. Similar prophylaxis pro-tocols should be followed for the pa-tient with spinal cord injury. Datasuggest that initiation of pharmaco-logic prophylaxis within 24 to 72hours safely decreases the incidenceof VTE. Conversion to warfarinshould be continued for up to 4months, with a target internationalnormalized ratio of 2.0 to 3.0.

When a VTE event occurs in theearly postoperative or postinjury pe-riod, observation alone can result inpoor outcomes. Inferior vena cavafilter placement has been demon-strated to be safe, with the potentialfor removal in select patients. Oncedeemed safe by the treating surgeon,pharmacologic therapy should beinitiated regardless whether the infe-rior vena cava filter is removed.

An increased risk of bleedingcomplications, including spinal epi-dural hematoma, exists with earlyinitiation of pharmacologic prophy-laxis. However, the risk/benefit ratiosupports early prophylaxis ratherthan observation. Thus, the treatingsurgeon should be mindful of signsand symptoms of bleeding in thesepatients. Once a spinal epidural he-matoma is diagnosed, prompt surgi-cal decompression provides the bestchance for neurologic recovery.

References

Evidence-based Medicine: There areseveral level I/II randomized pro-spective studies (references 4, 5, 10,16, 28, 32, and 38). The remainingreferences are level III/IV case con-trol and cohort studies, case reports,or level V expert opinion.


1. Silverstein MD, Heit JA, Mohr DN,Petterson TM, O’Fallon WM, MeltonLJ III: Trends in the incidence of deepvein thrombosis and pulmonary em-bolism: A 25-year population-basedstudy. Arch Intern Med 1998;158:585-593.

2. Hansson PO, Sörbo J, Eriksson H: Re-current venous thromboembolism af-ter deep vein thrombosis: Incidenceand risk factors. Arch Intern Med2000;160:769-774.

3. Demers C, Ginsberg JS, Brill-EdwardsP, Panju A, McGinnis J: Heparin andgraduated compression stockings inpatients undergoing fractured hip sur-gery. J Orthop Trauma 1991;5:387-391.

4. Planes A, Vochelle N, Darmon JY,Fagola M, Ballaud M, Huet Y: Risk ofdeep-venous thrombosis after hospi-tal discharge in patients having under-gone total hip replacement: Double-blinded randomised comparison ofenoxaparin versus placebo. Lancet1996;348:224-228.

5. Freedman KB, Brookenthal KR,Fitzgerald RH Jr, Williams S, LonnerJH: A meta-analysis of thromboem-bolic prophylaxis following electivetotal hip arthroplasty. J Bone JointSurg Am 2000;82:929-938.



6. Janku GV, Paiement GD, Green HD:Prevention of venous thromboembo-lism in orthopaedics in the UnitedStates. Clin Orthop Relat Res 1996;325:313-321.

7. Dickson B: Virchow’s triad? SouthMed J 2004;97:915-916.

8. Handin RI: Disorders of coagulationand thrombosis, in Fauci AS (ed): Har-rison’s Principles of Internal Medi-cine, ed 14. New York, NY: McGraw-Hill, 1998, pp 736-743.

9. Woolson ST, Zehnder JL, Maloney WJ:Factor V Leiden and the risk of proxi-mal venous thrombosis after total hiparthroplasty. J Arthroplasty 1998;13:207-210.

10. Westrich GH, Weksler BB, Glueck CJ,Blumenthal BF, Salvati EA: Correla-tion of thrombophilia and hypofibri-nolysis with pulmonary embolismfollowing total hip arthroplasty: Ananalysis of genetic factors. J BoneJoint Surg Am 2002;84:2161-2167.

11. Lee AY, Levine MN: Venous throm-boembolism and cancer: Risks andoutcomes. Circulation 2003;107:I17-I21.

12. Prothero SR, Parkes JC, StinchfieldFE: Complications after low-back fu-sion in 1000 patients: A comparison oftwo series one decade apart.1966.Clin Orthop Relat Res 1994;306:5-11.

13. Epstein NE: Efficacy of pneumaticcompression stocking prophylaxis inthe prevention of deep venous throm-bosis and pulmonary embolism fol-lowing 139 lumbar laminectomieswith instrumented fusions. J SpinalDisord Tech 2006;19:28-31.

14. Kim YH, Suh JS: Low incidence ofdeep vein thrombosis after cement-less total hip replacement. J BoneJoint Surg Am 1988;70:878-882.

15. Oda T, Fuji T, Kato Y, Fujita S, Kane-mitsu N: Deep venous thrombosis af-ter posterior spinal surgery. Spine2000;25:2962-2967.

16. Voth D, Schwarz M, Hahn K, Dei-Anang K, al Butmeh S, Wolf H: Pre-vention of deep vein thrombosis inneurosurgical patients: A prospectivedouble-blind comparison of two pro-phylactic regimen. Neurosurg Rev1992;15:289-294.

17. Gerlach R, Raabe A, Beck J, WoszczykA, Seifert V: Postoperative nadroparinadministration for prophylaxis ofthromboembolic events is not associ-ated with an increased risk of hemor-rhage after spinal surgery. Eur Spine J2004;13:9-13.

18. Kou J, Fischgrund J, Biddinger A,Herkowitz H: Risk factors for spinal

epidural hematoma after spinal sur-gery. Spine 2002;27:1670-1673.

19. Yi S, Yoon do H, Kim KN, Kim SH,Shin HC: Postoperative spinal epidu-ral hematoma: Risk factor and clinicaloutcome. Yonsei Med J 2006;47:326-332.

20. Rokito SE, Schwartz MC, NeuwirthMG: Deep vein thrombosis aftermajor reconstructive spinal surgery.Spine 1996;21:853-858.

21. Geerts WH, Pineo GF, Heit JA, et al:Prevention of venous thromboembo-lism: The Seventh ACCP Conferenceon Antithrombotic and ThrombolyticTherapy. Chest 2004;126:338S-400S.

22. Dearborn JT, Hu SS, Tribus CB, Brad-ford DS: Thromboembolic complica-tions after major thoracolumbar spinesurgery. Spine 1999;24:1471-1476.

23. Meissner MH, Chandler WL, ElliottJS: Venous thromboembolism in trau-ma: A local manifestation of systemichypercoagulability? J Trauma 2003;54:224-231.

24. Knudson MM, Ikossi DG, Khaw L,Morabito D, Speetzen LS: Throm-boembolism after trauma: An analy-sis of 1602 episodes from the Ameri-can College of Surgeons NationalTrauma Data Bank. Ann Surg 2004;240:490-498.

25. Tropiano P, Huang RC, Louis CA,Poitout DG, Louis RP: Functional andradiographic outcome of thoracolum-bar and lumbar burst fractures man-aged by closed orthopaedic reductionand casting. Spine 2003;28:2459-2465.

26. Dai LY, Yao WF, Cui YM, Zhou Q:Thoracolumbar fractures in patientswith multiple injuries: Diagnosis andtreatment. A review of 147 cases.J Trauma 2004;56:348-355.

27. Kurtoglu M, Yanar H, Bilsel Y, et al:Venous thromboembolism prophy-laxis after head and spinal trauma: In-termittent pneumatic compressiondevices versus low molecular weightheparin. World J Surg 2004;28:807-811.

28. Geerts WH, Code KI, Jay RM, Chen E,Szalai JP: A prospective study ofvenous thromboembolism after ma-jor trauma. N Engl J Med 1994;331:1601-1606.

29. Brach BB, Moser KM, Cedar L, Mint-eer M, Convery R: Venous thrombosisin acute spinal cord paralysis.J Trauma 1977;17:289-292.

30. Green D, Rossi EC, Yao JS, Flinn WR,Spies SM: Deep vein thrombosis inspinal cord injury: Effect of prophylax-

is with calf compression, aspirin, anddipyridamole. Paraplegia 1982;20:227-234.

31. Spinal Cord Injury Thromboprophy-laxis Investigators: Prevention ofvenous thromboembolism in theacute treatment phase after spinalcord injury: A randomized, multi-center trial comparing low-dose hep-arin plus intermittent pneumaticcompression with enoxaparin. J Trau-ma 2003;54:1116-1126.

32. Spinal Cord Injury Thromboprophy-laxis Investigators: Prevention ofvenous thromboembolism in the re-habilitation phase after spinal cord in-jury: Prophylaxis with low-dose hep-arin or enoxaparin. J Trauma 2003;54:1111-1115.

33. Jones T, Ugalde V, Franks P, Zhou H,White RH: Venous thromboembo-lism after spinal cord injury: Inci-dence, time course, and associatedrisk factors in 16,240 adults and chil-dren. Arch Phys Med Rehabil 2005;86:2240-2247.

34. Green D, Hartwig D, Chen D, Sol-tysik RC, Yarnold PR: Spinal cord in-jury risk assessment for thromboem-bolism (SPIRATE study). Am J PhysMed Rehabil 2003;82:950-956.

35. Lamb GC, Tomski MA, Kaufman J,Maiman DJ: Is chronic spinal cord in-jury associated with increased risk ofvenous thromboembolism? J AmParaplegia Soc 1993;16:153-156.

36. El Masri WS, Silver JR: Prophylacticanticoagulant therapy in patientswith spinal cord injury. Paraplegia1981;19:334-342.

37. DeVivo MJ, Kartus PL, Stover SL, RuttRD, Fine PR: Cause of death for pa-tients with spinal cord injuries. ArchIntern Med 1989;149:1761-1766.

38. Thumbikat P, Poonnoose PM, Bala-subrahmaniam P, Ravichandran G,McClelland MR: A comparison ofheparin/warfarin and enoxaparinthromboprophylaxis in spinal cordinjury: The Sheffield experience.Spinal Cord 2002;40:416-420.

39. Aito S, Pieri A, D’Andrea M, MarcelliF, Cominelli E: Primary prevention ofdeep venous thrombosis and pulmo-nary embolism in acute spinal cord in-jured patients. Spinal Cord 2002;40:300-303.

40. Harris S, Chen D, Green D: Enoxaparinfor thromboembolism prophylaxis inspinal injury: Preliminary report on ex-perience with 105 patients. Am JPhys Med Rehabil 1996;75:326-327.

41. Gaber TA: Significant reduction of therisk of venous thromboembolism in



all long-term immobile patients a fewmonths after the onset of immobility.Med Hypotheses 2005;64:1173-1176.

42. Swann KW, Black PM, Baker MF:Management of symptomatic deepvenous thrombosis and pulmonaryembolism on a neurosurgical service.J Neurosurg 1986;64:563-567.

43. Cain JE Jr, Major MR, Lauerman WC,West JL, Wood KB, Fueredi GA: Themorbidity of heparin therapy after de-velopment of pulmonary embolus inpatients undergoing thoracolumbaror lumbar spinal fusion. Spine 1995;20:1600-1603.

44. Leon L, Rodriguez H, Tawk RG, On-dra SL, Labropoulos N, Morasch MD:The prophylactic use of inferior vena

cava filters in patients undergoinghigh-risk spinal surgery. Ann VascSurg 2005;19:1-6.

45. Rosner MK, Kuklo TR, Tawk R, Mo-quin R, Ondra SL: Prophylactic place-ment of an inferior vena cava filterin high-risk patients undergoing spi-nal reconstruction. Neurosurg Focus2004;17:E6.

46. Duperier T, Mosenthal A, Swan KG,Kaul S: Acute complications associat-ed with greenfield filter insertion inhigh-risk trauma patients. J Trauma2003;54:545-549.

47. Allen TL, Carter JL, Morris BJ, HarkerCP, Stevens MH: Retrievable venacava filters in trauma patients forhigh-risk prophylaxis and prevention

of pulmonary embolism. Am J Surg2005;189:656-661.

48. Cuenca PJ, Tulley EB, Devita D, StoneA: Delayed traumatic spinal epiduralhematoma with spontaneous resolu-tion of symptoms. J Emerg Med2004;27:37-41.

49. Han YM, Kwak HS, Jin GY, ChungGH, Song KJ: Spinal epidural hemato-ma after thrombolysis for deep veinthrombosis with subsequent pulmo-nary thromboembolism: A case re-port. Cardiovasc Intervent Radiol2006;29:450-453.

50. Vandermeulen EP, Van Aken H, Ver-mylen J: Anticoagulants and spinal-epidural anesthesia. Anesth Analg1994;79:1165-1177.



Surgical Management ofHip Fractures: AnEvidence-based Reviewof the Literature. II:Intertrochanteric Fractures

AbstractTreatment of intertrochanteric hip fracture is based on patientmedical condition, preexisting degenerative arthritis, bone quality,and the biomechanics of the fracture configuration. A criticalreview of the evidence-based literature demonstrates a preferencefor surgical fixation in patients who are medically stable. Stablefractures can be successfully treated with plate-and-screw implantsand with intramedullary devices. Although unstable fractures maytheoretically benefit from load-sharing intramedullary implants,this result has not been demonstrated in the current evidence-based literature.

Intertrochanteric hip fractures areextracapsular fractures of the

proximal femur involving the areabetween the greater and lesser tro-chanter. Such fractures that extendinto the area distal to the lesser tro-chanter are described as having asubtrochanteric component. Theintertrochanteric region has anabundant blood supply, whichmakes fractures in this area muchless susceptible to osteonecrosis andnonunion than are femoral neckfractures. Fractures just proximal tothe intertrochanteric line, so-calledbasicervical fractures, are at greaterrisk for osteonecrosis (secondary topossibly being intracapsular) andmalunion (as a result of head rota-tion during implant insertion). How-ever, they may be treated with thesame implants that are used forintertrochanteric fractures.

Internal fixation of intertrochan-teric fractures is the mainstay of

treatment, although prosthetic re-placement is occasionally indicated.The major difficulty stems from thecombination of the presence of oftenosteopenic bone and the adverse bio-mechanics of many intertrochan-teric fracture patterns. Other factorsaffecting the choice of fixation in-clude preexisting hip symptoms, thepresence of osteoarthritis, bone qual-ity, degree of comminution, and thepatient’s medical condition.

Most of the classification systemsfor intertrochanteric fractures havepoor reliability and reproducibility.A simplified system to aid in evalu-ating treatment algorithms when as-sessing the literature is based onfracture stability, which is related tothe condition of the posteromedialcortex. Fractures are considered sta-ble in the absence of a comminutedposteromedial cortex, reverse obliq-uity, and subtrochanteric extension(Figure 1).

Kevin Kaplan, MD

Ryan Miyamoto, MD

Brett R. Levine, MD

Kenneth A. Egol, MD

Joseph D. Zuckerman, MD

Dr. Kaplan is Sports Medicine Fellow,Kerlan Jobe Orthopaedic Clinic, LosAngeles, CA. Dr. Miyamoto is SportsMedicine Fellow, Steadman HawkinsClinic, Vail, CO. Dr. Levine is AdultReconstructive Surgeon, Department ofOrthopaedic Surgery, Rush UniversityMedical Center, Chicago, IL. Dr. Egol isChief of Fracture Service, Department ofOrthopaedic Surgery, NYU–Hospital forJoint Diseases, New York, NY. Dr.Zuckerman is Professor and Chairman,Department of Orthopaedic Surgery,NYU–Hospital for Joint Diseases.

Reprint requests: Dr. Zuckerman,Department of Orthopaedic Surgery,NYU–Hospital for Joint Diseases, 301East 17th Street, New York, NY 10003.




The literature regarding intertro-chanteric fractures points to the dif-ficulty in applying evidence-basedtreatment algorithms. The currentevidence is conflicting and does notalways support the treatment mo-dalities that are widely used in prac-tice. Techniques and implants con-tinue to be modified, making theolder literature less relevant to cur-rent practice. Varying fracture pat-terns may not be distinguished inclinical studies. The ability to makeabsolute recommendations based onclear evidence is limited by theseproblems.

The Centre for Evidence-BasedMedicine created criteria for assign-ing levels of evidence (Table 1). Weperformed a thorough literature re-view to determine the most pertinent

and highest-level studies related tothe treatment of intertrochanteric hipfracture. Although level IV case stud-ies contribute general recommenda-tions for the management of thesefractures, we have focused on level I,II, and III studies.

Nonsurgical VersusSurgical Treatment

Nonsurgical treatment of intertro-chanteric hip fractures is usually re-served for patients with comorbidi-ties that place these patients atunacceptable risk from anesthesia,the surgical procedure, or both. Mor-tality from surgical treatment typi-cally results from cardiopulmonarycomplications, thromboembolism,and sepsis.1

There is a paucity of level I evi-dence concerning whether nonsurgi-cal treatment can provide a compa-rable outcome to that of surgicalfixation for intertrochanteric hipfractures (Table 2). In 1989, Hornbyet al2 performed a randomized pro-spective study comparing nonsurgi-cal treatment (ie, traction) with asliding hip screw (SHS) in 106 pa-tients with intertrochanteric hipfracture. Complications were low inboth groups, with no significant dif-ference in 6-month mortality, pain,leg swelling, or pressure sores. Ana-tomic reduction was achieved morecommonly with surgical treatment,and these patients had shorter hospi-tal stays. Patients treated with trac-tion had greater loss of independenceat 6-month follow-up. The authorsrecommended surgical treatment formedically stable patients.

A 1981 prospective (level II) trialof 150 patients compared nonsurgi-cal treatment (ie, skeletal tractionwith a tibial pin) with surgical treat-ment (eg, medial displacement os-teotomy, valgus osteotomy).3 Theauthors concluded that excellent re-sults were feasible with tractionalone provided that a high standardof nursing care was maintained.Careful attention to bedside physicaltherapy, respiratory care, deep veinthrombosis prophylaxis, and preven-tion of ulcers were vital to satisfac-tory outcomes in nonsurgicallytreated patients.

A 2003 retrospective level III studyreviewed a population database tocompare mortality rates in patientswith severe comorbidities who weretreated either nonsurgically or surgi-cally for intertrochanteric hip frac-ture.4 The 30-day mortality rate waslower in patients treated surgically.

Dr. Egol or a member of his immediate family has participated in a speakers bureau or given paid presentations for Biomet; is an unpaidconsultant for Biomet; has received research or institutional support from Biomet, Smith & Nephew, Stryker, and Synthes; and holds stock orstock options in Johnson & Johnson. Dr. Zuckerman or a member of his immediate family is affiliated with Neostem and Starmed as a boardmember, owner, officer, or committee member; has received royalties from Exactech; and has received research or institutional support fromExactech and Stryker. None of the following authors or a member of their immediate families has received anything of value from or owns stockin a commercial company or institution related directly or indirectly to the subject of this article: Dr. Kaplan, Dr. Miyamoto, and Dr. Levine.

Figure 1

Intertrochanteric hip fracture. A, Standard oblique fracture (type I). B, Reverseoblique fracture (type II).

Surgical Management of Hip Fractures: An Evidence-based Review of the Literature. II: Intertrochanteric Fractures


Table 1

Levels of Evidence for Primary Research Question*

Type of Study

Level

Therapeutic Studies—Investigating the resultsof treatment

Prognostic Studies—Investigating theeffect of a patientcharacteristic on theoutcome of disease

Diagnostic Studies—Investigating adiagnostic test

Economic and DecisionAnalyses—Developingan economic ordecision model

I High-quality RCT withstatistically significantdifference or nostatistically significantdifference but narrowconfidence intervals

Systematic review† oflevel I RCTs (andstudy results werehomogeneous‡)

High-quality prospectivestudy§ (all patients wereenrolled at the samepoint in their diseasewith ≥80% follow-up ofenrolled patients)

Systematic review† oflevel I studies

Testing of previouslydeveloped diagnosticcriteria on consecutivepatients (with universallyapplied reference “goldstandard”)


Sensible costs andalternatives; valuesobtained from manystudies; withmultiway sensitivityanalyses


II Lesser quality RCT (eg,<80% follow-up, noblinding, or improperrandomization)

Prospective§

comparative study¶

Systematic review† oflevel II studies or levelI studies withinconsistent results

Retrospective# studyUntreated controls

from an RCTLesser quality prospective

study (eg, patientsenrolled at differentpoints in their disease or<80% follow-up)

Systematic review† oflevel II studies

Development of diagnosticcriteria on consecutivepatients (with universallyapplied reference “goldstandard”)


Sensible costs andalternatives; valuesobtained from limitedstudies; withmultiway sensitivityanalyses


III Case-control study**

Retrospective#

comparative study¶

Systematic review† oflevel III studies

Case-control study** Study of nonconsecutivepatients (withoutconsistently appliedreference “gold standard”)


Analyses based onlimited alternativesand costs; and poorestimates


IV Case series†† Case series Case-control studyPoor reference standard

Analyses with nosensitivity analyses

V Expert opinion Expert opinion Expert opinion Expert opinion

RCT = randomized clinical trial

*A complete assessment of quality of individual studies requires critical appraisal of all aspects of the study design

†A combination of results from two or more prior studies

‡Studies provided consistent results

§The study was started before the first patient enrolled

¶Patients treated one way (eg, cemented hip arthroplasty) compared with a group of patients treated in another way (eg,uncemented hip arthroplasty) at the same institution

#The study was started after the first patient enrolled

**Patients identified for the study based on their outcome, called “cases” (eg, failed total arthroplasty), are compared to those whodid not have that outcome, called “controls” (eg, successful total hip arthroplasty)

††Patients treated one way with no comparison group of patients treated in another way

Data for this table are from http://www.ejbjs.org/misc/public/instrux.shtml and http://www.cebm.net/levels_of_evidence.asp

Reproduced from Spindler KP, Kuhn JE, Dunn W, Matthew LE, Harrell FE Jr, Dittus RS: Reading and reviewing the orthopaedicliterature: A systematic, evidence-based medicine approach. J Am Acad Orthop Surg 2005;13:220-229.

Kevin Kaplan, MD, et al


However, when the authors com-pared surgical fixation with nonsur-gical treatment with early mobiliza-tion (ie, out of bed to chair), theyfound no significant difference inmortality rate. Thus, when feasible,

the authors recommend early mobi-lization out of bed to chair in patientswith nonsurgically managed hip frac-ture.

The evidence-based literaturesupports surgical fixation2 while also

providing valuable information inregard to medically unstable pa-tients who must be treated nonsurgi-cally.3,4

Intramedullary VersusExtramedullary Fixation

The mechanical environment andblood supply to the peritrochantericregion of the hip is more robust, mak-ing surgical treatment of intertro-chanteric hip fractures different fromthat of femoral neck fractures. Be-cause the risk of osteonecrosis is min-imal, the need for prosthetic replace-ment is reduced. Experience withfixed-angle screw-plate constructsindicates that uncontrolled fractureimpaction is a problem, with compli-cations including implant joint pen-etration and implant failure.5

Two types of implant are used inthe treatment of patients with inter-trochanteric hip fracture: an SHS witha side plate, and an intramedullary(IM) nail with an SHS component.The latter may have several advan-tages over the SHS and side plate. TheIM component helps to buttressagainst fracture collapse and medial-ization of the distal fracture fragment,particularly in unstable (ie, reverseobliquity) intertrochanteric fractures.Furthermore, the percutaneous inser-tion of the IM device may reduce theamount of surgical trauma. Numer-ous studies have compared thesetypes of implant.5-14 The correct inter-pretation of these data to guide cur-rent practice is one of the majorcontroversies in the treatment ofintertrochanteric fractures (Table 3).

In 1991, Bridle et al6 reported on100 patients with 41 stable intertro-chanteric fractures who were ran-domized to receive either a Gammanail (Stryker, Mahwah, NJ) or a dy-namic hip screw (DHS). In this levelI study, no differences were demon-strated in surgical time, blood loss,wound complications, length of stay,or patient mobility at a minimumfollow-up of 6 months. Loss of re-duction (lag screw, nail cutout) was

Table 2

Nonsurgical Versus Surgical Treatment of Intertrochanteric Hip Fracture

Evidence Treatment Results/Recommendations

Level I2 Traction vs slidinghip screw

No significant difference in6-month mortality

With surgical treatment,better anatomic reduction,decreased hospital stay,increased independence

Level II3 and III4 Traction with tibialpin vs medialdisplacementosteotomy or valgusosteotomy

Nonsurgical treatment canbe as successful as surgicaltreatment, provided a highstandard of nursing care ismaintained

Authors’ experience Tibial traction withearly mobilizationvs surgicaltreatment(dependent onevaluation of thefracture)

Surgical treatment results inearlier mobilization andlower perioperativemorbidity

Nonsurgical treatment ispreferred for the patientwhose medical condition isnot stable

Table 3

Intramedullary Versus Extramedullary Fixation for Intertrochanteric HipFracture


Level I, II,and III6-23

Gamma nail (Stryker,Mahwah, NJ) vs DHS;6,17 IMnail vs SHS;7,8,18 IM nail vsDHS and side plate;9 Gammanail vs compression hipscrew;10-12 DHS vs PFN;13,19

IM hip screw vs SHS;14 IMdevice vs fixed-anglescrew-plate;15 SHS, Gammanail, PFN;16 SHS vs shorttrochanteric nail23

No significant difference inwound complications,fracture union, mortality,or functional outcomes

Authors’experience

DHS or IM implant (stablefracture), IM device (unstablefracture)

DHS or IM implant forstable fractures (based onclinical experience andfinancial considerations)

IM device (unstablefractures) aids in earlymobilization and results indecreased blood loss andreduced surgical time

DHS = dynamic hip screw, IM = intramedullary, PFN = proximal femoral nail,SHS = sliding hip screw



similar between the two groups; ofthe patients treated with the Gam-ma nail, four experienced femoralshaft fracture requiring revision sur-gery. For both groups, union oc-curred at an average of 6 months.Radford et al7 and Saudan et al8

found nearly identical results intheir level I studies of 200 and 206patients, respectively, who were ran-domized to receive either an IM nailor SHS fixation.

In 2001, Adams et al9 published aprospective, randomized controlledtrial assessing IM nailing versus aDHS and side plate in 400 patients.Revision rates, femoral shaft frac-tures, and lag screw cutout wereslightly higher in patients treatedwith IM nailing but did not differsignificantly from the cohort treatedwith a DHS. There was no differencein early or 1-year functional out-comes.

Ahrengart et al10 randomized 426intertrochanteric fractures to treat-ment with either the Gamma nail ora compression hip screw. The lattercohort required significantly lesssurgical time, and patients experi-enced less blood loss (P < 0.05). How-ever, in unstable intertrochantericfractures, surgical time was not sig-nificantly different between the twogroups. In patients treated with theGamma nail, difficulty was encoun-tered with the distal locking tech-nique. There was also a higher inci-dence of cephalic position of thecompression screw within the fem-oral head, screw cutout, and intraop-erative fracture in the Gamma nailgroup. Walking ability was the samein both groups. The authors recom-mended compression hip screws forless comminuted fractures, reserv-ing Gamma nails for comminutedpatterns. In 1995, O’Brien et al11

found no significant difference be-tween Gamma nail and compressionhip screw fixation in terms of bloodloss, days in the hospital, time tounion, and functional outcome.

Utrilla et al12 found no differencein total surgical time in their level I

study comparing the Gamma nailwith a compression hip screw in 210stable and unstable fractures. How-ever, the Gamma nail group had asignificantly lower postoperativetransfusion requirement (P = 0.013).Mortality, fracture healing, andintra- and postoperative complica-tion rates were not significantly dif-ferent between the two groups. Inpatients with unstable fracture pat-terns, postoperative ambulation wassignificantly improved in the Gam-ma nail group (P = 0.017).

Recovery of ambulation was a fo-cus of the study by Pajarinen et al,13

who compared a DHS with a proxi-mal femoral nail (PFN) (Synthes,Oberdorf, Switzerland) in 108 pa-tients. Although the immediatepostoperative outcomes did not dif-fer between the two groups, patientstreated with IM devices had a signif-icantly faster return to preoperativeambulation levels (P = 0.04). Frac-ture healing was similar between thetwo groups at 4 months, with twopatients in each group requiring revi-sion. This study also suggested thatthe PFN provided faster restorationof walking ability than did the DHSin patients with unstable fracturepatterns.

Baumgaertner et al14 randomized135 unstable intertrochanteric frac-tures to either an IM hip screw (In-tramedullary Hip Screw [IMHS];Smith & Nephew, Memphis, TN) oran SHS. Patients with unstable frac-tures treated with the IMHS re-quired 23% less time in the operat-ing room and experienced 44% lessblood loss than did the SHS cohort.Functional outcome was not signif-icantly different between the twogroups.

Sadowski et al15 reported the re-sults of 39 unstable reverse obliqui-ty intertrochanteric fractures man-aged with either an IM device or afixed-angle screw-plate device (Dy-namic Condylar Screw; Synthes).Clinical and radiographic follow-updemonstrated a shorter mean surgi-cal time for patients treated with IM

nailing and a significantly higherrate of implant failure and nonunionin the group treated with the Dy-namic Condylar Screw (P = 0.008and P = 0.007, respectively). Exclud-ing patients with nonunion or fail-ure, there was no significant differ-ence in postoperative walkingability or level of independence.

In 2005, Papasimos et al16 per-formed a randomized, prospectivestudy of 120 patients with unstableintertrochanteric fractures compar-ing an SHS, Gamma nail, and PFN.Mean blood loss, length of hospitalstay, screw cutout, and fracture re-duction were not statistically differ-ent between the three groups. Pa-tients treated with PFN had asignificantly longer surgical time(P < 0.05), which the investigatorssuggested was due to lack of surgeonexperience with that device.

Several level II studies have beenpublished on this topic. In 1992,Leung et al17 reported the results ofa prospective trial comparing Gam-ma nails with DHSs and found thatpatients treated with Gamma nailshad smaller incisions, less intraoper-ative blood loss, and earlier fullweight bearing. No significant differ-ence was found in mortality andpostoperative mobility (both groupslost one level of mobility). Of note,the investigators cited a higher inci-dence of fractures of the lateral cor-tex (three in the nail group and twoin the DHS group) during insertionand noted two femoral shaft frac-tures within 3 months of surgery inthe Gamma nail group.

Guyer et al18 reviewed 100 pa-tients treated with either an IM de-vice or an SHS. There was no signif-icant difference in intraoperativeblood loss or perioperative complica-tions between the two devices. Theauthors suggested that the Gammanail was preferable to DHSs for un-stable fracture patterns becausethree patients in the DHS group ex-perienced proximal screw perfora-tion during attempted mobiliza-tion.



Nuber et al19 evaluated 129 pa-tients with unstable intertrochantericfractures treated with either a DHS ora PFN. Revision rates were similar be-tween the two groups. However,there was a significantly shorter sur-gical time (44.3 versus 57.3 min) andhospital stay (18.6 versus 21.3 days)in the PFN cohort. Full weight bear-ing was possible immediately postop-eratively in 97% of the proximal nailcohort, compared with 88% of the pa-tients treated with a DHS. At6-month follow-up, considerablylower pain intensity scores werefound in the PFN cohort.

In several level II trials comparingextramedullary and IM devices, theuse of a nail was shown to have anincreased risk of intraoperative andpostoperative fracture, with an in-creased rate of revision.14,20-22 No sig-nificant differences were reported inregard to wound infection, medicalcomplications, mortality, functionaloutcomes, postoperative complica-tions, hip function, quality of life,and activities of daily living at 1 yearpostoperatively. Complications as-sociated with the Gamma nail, par-ticularly the intraoperative fracturerate, resulted in specific designand technique modifications. Thesechanges, combined with increasedsurgeon experience, contributed to alower rate of intraoperative compli-cations in subsequent studies.

One retrospective (level III) studyreviewed 93 patients who were treat-ed with either an SHS or a short tro-chanteric nail.23 Fracture healingwas uneventful in 94% of the pa-tients treated with SHS and in 89%of the patients treated with trochan-teric nailing. Complications includ-ed one lag screw cutout in the SHScohort compared with three in thetrochanteric nail cohort. Other out-come measures were similar be-tween the two groups, and the au-thors concluded that both methodsresulted in successful treatment ofintertrochanteric fractures.

Analysis of level I studies pro-vides insight regarding the two most

commonly used methods of intertro-chanteric fracture fixation: IM nail-ing and DHS fixation. Most level Istudies indicate that there are no sig-nificant differences in operatingroom time, blood loss, wound com-plications, length of stay, mobility,functional outcomes, loss of reduc-tion, union rate, mortality, and com-plication rates when comparing IMdevices with SHS constructs. How-ever, several studies report a fasterreturn to preoperative ambulation,reduced operating room time, andless blood loss when an IM device isused, especially in patients with un-stable fracture patterns.6-16 Analysisof level II studies demonstrates apreference for IM devices.

Surgical OutcomesUnfortunately, the 18 studies dis-

cussed herein provide inconsistentevidence for treatment recommen-dations. A well-defined outcomemeasure such as surgical time is agood example. Two level I studies in-dicated no significant difference insurgical time between IM and ex-tramedullary implants.6,12 However,two level I studies and one level IIstudy found a significantly highersurgical time when an SHS is used(P < 0.05), with longer times beingassociated with unstable pat-terns.14,15,19 Two level I studies dem-onstrated a longer surgical time withthe use of an IM implant.10,16

Femoral Shaft FractureThree level I studies and one lev-

el II study found an increased inci-dence of femoral shaft fracture at thetip of the implant when using IMnails.9,10,12,17 Most authors concludedthat this increase was in part dueto a lack of experience and to sub-optimal hardware design. Newer-generation nails have a radius of cur-vature that better conforms to theanatomic shape of the femur. Al-though this statement is not sup-ported by evidence-based literature,this feature may potentially reducethe rate of intraoperative frac-

ture.9,10,12,17 In contrast to earlier re-ports, recent studies show no signif-icant difference in complications orrevision rates between the two typesof implants, which may be attribut-ed to improved nail design and in-creasing surgeon experience.5,24

Blood LossSix level I studies and one level II

study found no significant differencein blood loss or transfusion ratesbetween IM and extramedullaryimplants.6-8,11,16,18 However, two lev-el I studies (P < 0.05,12 P < 0.01314)and one level II study (P < 0.05)17

found significantly less blood losswith IM implants, while one level Istudy states that there was signifi-cantly less blood loss with the use ofa DHS (P < 0.05).10

Patient Ambulation andComplications

Five level I studies suggested thatpatients regain equal ambulatory sta-tus regardless of fixation type.6-8,10,15

However, two level I and two level IIstudies concluded that IM devices ex-pedite return to pretreatment ambu-latory function.12,13,17,19 It is importantto note that many current studieshave not separated stable from unsta-ble patterns when assessing ambula-tory status. The literature is consis-tent, however, in regard to woundcomplications, fracture nonunion,mortality rates, and functional out-comes and overall incidence of com-plications, with no significant differ-ence between IM and extramedullaryimplants.6-24

Authors’ RecommendationThere is no consensus regarding

the ideal implant for treating intertro-chanteric fractures. However, basedon the available evidence-based data,we recommend either a DHS or an IMdevice for stable intertrochantericfractures. For unstable fractures, werecommend an IM device. Althoughthis has not been proved in the cur-rent evidence-based literature, we be-lieve that an IM device is a biome-



chanically stronger construct and isbetter suited to preventing increasedfracture collapse in unstable fractures.In addition, evidence suggests that IMdevices aid in early mobilization, re-turn of ambulatory function, de-creased blood loss, and less surgicaltime.12,13,17,19 However, there seems tobe a higher cost associated with theuse of IM devices.5

Open Reduction andInternal Fixation VersusArthroplasty

Prosthetic hip replacement generallyhas not been considered a primarytreatment option for intertrochantericfractures. Unlike femoral neck frac-tures, which retain some of the fem-oral neck in addition to the abductormechanism, intertrochanteric frac-tures involve more distal femoralbone, and often the greater trochanterand the abductor are not attached tothe proximal femur. In this setting,prosthetic replacement for intertro-chanteric fractures typically requiresa more complex surgical procedurewith potentially higher morbidity. Inthe patient with preexisting sympto-matic degenerative arthritis, primaryprosthetic replacement may be thebest option. It can also be consideredfor intertrochanteric fractures withextreme comminution in severely os-teoporotic bone in which internal fix-ation methods are unlikely to be suc-cessful.25

In 2005, Kim et al26 performed aprospective randomized (level I) studyof unstable intertrochanteric fracturesin elderly patients in which long-stemcementless calcar-replacement hemi-arthroplasty was compared with aPFN. No significant differences werefound between the two groups interms of functional outcomes, hospi-tal stay, time to weight bearing, andrisk of complications. However, sur-gical time (P < 0.001), blood loss (P <0.001), need for blood transfusions(P < 0.001), and mortality rates (P <0.006) were all significantly lower inthe PFN group.

In another level I study, Stap-paerts et al27 treated 47 patients withcompression hip screws and 43 withhemiarthroplasty. No significant dif-ference was found between surgicaltime, wound complications, or mor-tality rates. However, the hemiar-throplasty group was reported tohave higher transfusion rates.

Haentjens et al28 reported on aprospective (level II) study compar-ing the results of 79 consecutive pa-tients aged 75 years and older whowere treated with either bipolarhemiarthroplasty (37 patients) or in-ternal fixation (42 patients). The bi-polar group experienced easier andfaster rehabilitation, with a lower in-cidence of decubiti and pulmonarycomplications. The decrease in com-plications was attributed to an earli-er return to full weight bearing.

The remainder of evidence re-garding arthroplasty to treat inter-trochanteric fractures comes fromlevel III and IV studies. These stud-ies suggest that a cemented hemiar-throplasty with standard implants isa reasonable alternative to open re-duction and internal fixation. In ad-dition, they indicate that arthroplas-ty has the advantage of early weightbearing and avoids the potential offixation failure and the need for sub-sequent revision.29-33

There is no overwhelming evi-dence from randomized clinical tri-als to indicate that arthroplasty ismore effective than IM and ex-tramedullary fixation of intertro-chanteric hip fractures (Table 4). Nosignificant differences in complica-tions have been reported betweenhemiarthroplasty or THA versus IMfixation.26,28 However, the incidenceof decubiti and pulmonary compli-cations may be higher with internalfixation.27 Two level I studies founda significantly lower transfusion ratewhen a PFN was used (P < 0.001,26

P < 0.0527). No significant differencein functional outcomes or rehabilita-tion was shown between unstablefractures treated with hemiarthro-plasty or with a PFN.26 However,

one level II study concluded thatpatients treated with bipolar in-strumentation had a faster rate ofrehabilitation, although the time dif-ferences were not statistically signif-icant.28

Based on the available evidenceon, and our clinical experience with,intertrochanteric hip fractures, ar-throplasty should be reserved for pa-tients with preexisting symptomat-ic degenerative arthritis, those inwhom internal fixation is not ex-pected to be successful because offracture comminution or bone qual-ity, and in patients who require sal-vage for failed internal fixation.

Summary

With ongoing improvements in en-doprostheses and total hip replace-ments, increased surgeon experi-ence, and the need to separate stablefrom unstable fractures, it is difficultto recommend one optimum treat-ment of intertrochanteric fracturesfrom a purely evidence-based per-spective. Even so, we believe thatcombining current evidence-basedliterature with clinical experiencecan guide clinical decision making.

Surgical intervention is preferableto nonsurgical treatment of intertro-chanteric fractures in the medicallystable patient. This is the case de-spite evidence demonstrating thatpatients can have equivalent out-comes with nonsurgical treatmentwhen nursing care is excellent. Pa-tients treated nonsurgically mayhave a higher mortality rate if theyare not mobilized early. Althoughthere is no evidence-based literatureto support these findings, nonsurgi-cally treated patients appear to beat higher risk for complications suchas decubiti, pneumonia, and deepvein thrombosis.

When considering surgical inter-vention, it is important to considerthe character of each fracture pattern,surgeon clinical experience, and thereported evidence regarding the var-ious internal fixation implants. Pa-



tient outcome has not been shown todiffer significantly between fixationof stable intertrochanteric fractureswith plate-and-screw implants versusIM devices. Thus, factors in thedecision-making process should in-clude surgeon experience with the de-vices and cost-effectiveness of theprocedure. Unstable intertrochantericfractures are a distinct subset that bio-mechanically should benefit from anIM device; however, there is no over-whelming evidence to prove this rec-ommendation. Studies on functionaloutcome have not yet been performedin sufficient detail to demonstrate sig-nificant differences between devices.Comparisons between specific typesof IM implants have not been re-ported in sufficient numbers or detail

to determine whether nail design hasan effect on outcome.

Patients with severe degenera-tive disease or with comminutedfracture in osteoporotic bone can besuccessfully treated with an en-doprosthetic replacement or a THA.This surgery is more complex thaninternal fixation and is associatedwith a higher rate of postoperativecomplications. The evidence-basedliterature does not show a signifi-cant difference in terms of time toambulation and length of hospitalstay between arthroplasty and inter-nal fixation. However, given thevariety of clinical presentations andfracture patterns, such treatmentmay be considered in select pa-tients.

References

Evidence-based Medicine: Referenc-es 1-3, 5-22, and 26-28 are level I/IIprospective, randomized studies orsystematic reviews of level I studies.The remainder are level III/IV casereports and case-control cohort stud-ies.


1. Parker MJ, Handoll HH: Conserva-tive versus operative treatment forextracapsular hip fractures. Co-chrane Database Syst Rev 2000;2:CD000337.

2. Hornby R, Evans JG, Vardon V: Oper-ative or conservative treatment fortrochanteric fractures of the femur: Arandomized epidemiological trial inelderly patients. J Bone Joint Surg Br1989;71:619-623.

3. Bong SC, Lau HK, Leong JC, Fang D,Lau MT: The treatment of unstableintertrochanteric fractures of the hip:A prospective trial of 150 cases.Injury 1981;13:139-146.

4. Jain R, Basinski A, Kreder HJ: Nonop-erative treatment of hip fractures. IntOrthop 2003;27:11-17.

5. Parker MJ, Handoll HH: Gamma andother cephalocondylic intramedul-lary nails versus extramedullary im-plants for extracapsular hip fracturesin adults. Cochrane Database SystRev 2005;4:CD000093.

6. Bridle SH, Patel AD, Bircher M, Cal-vert PT: Fixation of intertrochantericfractures of the femur: A randomisedprospective comparison of the gammanail and the dynamic hip screw.J Bone Joint Surg Br 1991;73:330-334.

7. Radford PJ, Needoff M, Webb JK: Aprospective randomised comparisonof the dynamic hip screw and the gam-ma locking nail. J Bone Joint Surg Br1993;75:789-793.

8. Saudan M, Lübbeke A, Sadowski C,Riand N, Stern R, Hoffmeyer P: Per-

“Surgical Management of HipFractures: An Evidence-based Re-view of the Literature. I: FemoralNeck Fractures” appeared in theOctober 2008 issue of the Jour-nal of the American Academy ofOrthopaedic Surgeons.

Table 4

ORIF Versus Hemiarthroplasty and THA for Intertrochanteric Hip Fracture


Level I26,27 andII28

Hemiarthroplasty vsPFN,26

compression hipscrews vshemiarthroplasty,27

hemiarthroplastyvs ORIF28

Shorter surgical time with PFNcompared with hemiarthroplasty,shorter surgical time with THAversus IM and extramedullaryimplants

No significant differences incomplications

Higher incidence of decubiti andpulmonary complications withbipolar hemiarthroplasty

Significantly lower transfusionrate with PFN

No significant differences infunctional outcomes

Faster rehabilitation in patientstreated with bipolarinstrumentation

Level III andIV29-33

Hemiarthroplasty vsORIF

Cemented hemiarthroplasty areasonable alternative to ORIF

With arthroplasty, earlier weightbearing and lack of fixationfailure, so no need for revision

Authors’experience

Hemiarthroplasty vsORIF

Arthroplasty reserved for patientswith preexisting symptomaticdegenerative arthritis and thosein whom internal fixation is notexpected to be successfulbecause of comminution or bonequality, and as a salvageprocedure for failed internalfixation

IM = intramedullary, ORIF = open reduction and internal fixation, PFN = proximalfemoral nail, THA = total hip arthroplasty



trochanteric fractures: Is there an ad-vantage to an intramedullary nail? Arandomized, prospective study of 206patients comparing the dynamic hipscrew and proximal femoral nail.J Orthop Trauma 2002;16:386-393.

9. Adams CI, Robinson CM, Court-Brown CM, McQueen MM: Prospec-tive randomized controlled trial of anintramedullary nail versus dynamicscrew and plate for intertrochantericfractures of the femur. J OrthopTrauma 2001;15:394-400.

10. Ahrengart L, Törnkvist H, FornanderP, et al: A randomized study of thecompression hip screw and Gammanail in 426 fractures. Clin OrthopRelat Res 2002;401:209-222.

11. O’Brien PJ, Meek RN, Blachut PA,Broekhuyse HM, Sabharwal S: Fixa-tion of intertrochanteric hip fractures:Gamma nail versus dynamic hipscrew. A randomized, prospectivestudy. Can J Surg 1995;38:516-520.

12. Utrilla AL, Reig JS, Muñoz FM, Tufa-nisco CB: Trochanteric gamma nailand compression hip screw for tro-chanteric fractures: A randomized,prospective, comparative study in 210elderly patients with a new design ofthe gamma nail. J Orthop Trauma2005;19:229-233.

13. Pajarinen J, Lindahl J, Michelsson O,Savolainen V, Hirvensalo E: Pertro-chanteric femoral fractures treatedwith a dynamic hip screw or a proxi-mal femoral nail: A randomised studycomparing post-operative rehabilita-tion. J Bone Joint Surg Br 2005;87:76-81.

14. Baumgaertner MR, Curtin SL, Lind-skog DM: Intramedullary versus ex-tramedullary fixation for the treat-ment of intertrochanteric hipfractures. Clin Orthop Relat Res1998;348:87-94.

15. Sadowski C, Lübbeke A, Saudan M,Riand N, Stern R, Hoffmeyer P: Treat-ment of reverse oblique and trans-verse intertrochanteric fractures withuse of an intramedullary nail or a 95degrees screw-plate: A prospective,randomized study. J Bone Joint SurgAm 2002;84:372-381.

16. Papasimos S, Koutsojannis CM, Pana-

gopoulos A, Megas P, Lambiris E: Arandomised comparison of AMBI,TGN and PFN for treatment of unsta-ble trochanteric fractures. Arch Or-thop Trauma Surg 2005;125:462-468.

17. Leung KS, So WS, Shen WY, Hui PW:Gamma nails and dynamic hip screwsfor peritrochanteric fractures: A ran-domised prospective study in elderlypatients. J Bone Joint Surg Br 1992;74:345-351.

18. Guyer P, Landolt M, Eberle C, KellerH: The gamma-nail as a resilient alter-native to the dynamic hip screw in un-stable proximal femoral fractures inthe elderly [German]. Helv Chir Acta1992;58:697-703.

19. Nuber S, Schönweiss T, Rüter A: Sta-bilisation of unstable trochantericfemoral fractures: Dynamic hip screw(DHS) with trochanteric stabilisationplate vs. proximal femur nail (PFN)[German]. Unfallchirurg 2003;106:39-47.

20. Hardy DC, Descamps PY, Krallis P, etal: Use of an intramedullary hip-screwcompared with a compression hip-screw with a plate for intertrochan-teric femoral fractures: A prospective,randomized study of one hundred pa-tients. J Bone Joint Surg Am 1998;80:618-630.

21. Harrington P, Nihal A, Singhania AK,Howell FR: Intramedullary hip screwversus sliding hip screw for unstableintertrochanteric femoral fractures inthe elderly. Injury 2002;33:23-28.

22. Hoffmann R, Schmidmaier G, SchulzR, Schütz M, Südkamp NP: Classicnail versus DHS: A prospective ran-domised study of fixation of trochan-teric femur fractures [German].Unfallchirurg 1999;102:182-190.

23. Crawford CH, Malkani AL, Cordray S,Roberts CS, Sligar W: The trochanter-ic nail versus the sliding hip screw forintertrochanteric hip fractures: A re-view of 93 cases. J Trauma 2006;60:325-328.

24. Egol KA, Chang EY, Cvitkovic J,Kummer FJ, Koval KJ: Mismatch ofcurrent intramedullary nails with theanterior bow of the femur. J OrthopTrauma 2004;18:410-415.

25. Parker MJ, Handoll HH: Replacement

arthroplasty versus internal fixationfor extracapsular hip fractures.Cochrane Database Syst Rev 2000;1:CD000086.

26. Kim SY, Kim YG, Hwang JK: Cement-less calcar-replacement hemiarthro-plasty compared with intramedullaryfixation of unstable intertrochantericfractures: A prospective, randomizedstudy. J Bone Joint Surg Am 2005;87:2186-2192.

27. Stappaerts KH, Deldycke J, Broos PL,Staes FF, Rommens PM, Claes P:Treatment of unstable peritrochant-eric fractures in elderly patients witha compression hip screw or with theVandeputte (VDP) endoprosthesis: Aprospective randomized study. J Or-thop Trauma 1995;9:292-297.

28. Haentjens P, Casteleyn PP, De BoeckH, Handelberg F, Opdecam P: Treat-ment of unstable intertrochantericand subtrochanteric fractures in el-derly patients: Primary bipolar arthro-plasty compared with internal fixa-tion. J Bone Joint Surg Am 1989;71:1214-1225.

29. Chan KC, Gill GS: Cemented hemiar-throplasties for elderly patients withintertrochanteric fractures. Clin Or-thop Relat Res 2000;371:206-215.

30. Rodop O, Kiral A, Kaplan H, Akmaz I:Primary bipolar hemiprosthesis forunstable intertrochanteric fractures.Int Orthop 2002;26:233-237.

31. Harwin SF, Stern RE, Kulick RG: Pri-mary Bateman-Leinbach bipolar pros-thetic replacement of the hip in thetreatment of unstable intertrochan-teric fractures in the elderly. Ortho-pedics 1990;13:1131-1136.

32. Berend KR, Hanna J, Smith TM, Mal-lory TH, Lombardi AV: Acute hip ar-throplasty for the treatment of inter-trochanteric fractures in the elderly.J Surg Orthop Adv 2005;14:185-189.

33. Haentjens P, Casteleyn PP, OpdecamP: Primary bipolar arthroplasty or to-tal hip arthroplasty for the treatmentof unstable intertrochanteric and sub-trochanteric fractures in elderly pa-tients. Acta Orthop Belg 1994;60(suppl 1):124-128.



Antibiotic Beads

Bone infection, or osteomyelitis,can be one of the most difficult

problems confronted by the ortho-paedic surgeon. Common causes ofosteomyelitis include open frac-tures, hematogenous spread of bacte-ria to bone, and orthopaedic surgicalprocedures complicated by infec-tion. Assessment involves identifi-cation of the offending organism bytissue culture and sensitivity to an-tibiotics; radiographic assessment ofthe extent of the infection; clinicalevaluation of the patient’s generalhealth and ability to fight infection;and determination of the local ana-tomic condition of the bone and softtissue.

Treatment is individualized butgenerally involves prolonged use ofsystemic antibiotics, surgical dé-bridement, and support of the pa-tient’s overall health.1,2 Results aregenerally good but are not universal-ly successful. Difficulties includerecurrence or lack of control of in-fection, systemic toxicity to antibi-otics, scar formation, and persistentnonunion of fractures.2 Other prob-lems include expense and practicaldifficulties for the patient and thesurgeon, including multiple surgicaltreatments, lengthy and frequenthospitalizations, and prolonged limbdysfunction. Control of infectioneventually may be obtained, but pa-tient function still may be limitedby scar, stiffness, and weakness, aswell as nonunion or malunion of thefracture. Late recurrence of infectionis also a problem.

To deliver therapeutic tissue lev-els of parenteral antibiotics to thetarget area, high serum levels of an-tibiotics must be achieved.3 Thesehigh serum levels, however, mayresult in an increased incidence ofsystemic side effects such as nephro-toxicity and ototoxicity.4 As an alter-

native, the use of antibiotic-impregnated beads as an adjunct toother treatment offers advantagescompared with systemic aminogly-cosides. With the bead pouch tech-nique, the systemic levels are low,and the systemic complications arevirtually eliminated, while the localconcentration, where it is needed, isextremely high.3 Antibiotic beadsalso offer the benefit of managementof dead space. They are relatively in-expensive and are easy for the sur-geon to insert and the patient to tol-erate.5

Indications andContraindications

Antibiotic beads can be used in mul-tiple different applications. Typicalindications include prevention of in-fection (eg, open fracture antibioticbead pouch prophylaxis), treatmentof established bone infection (ie,acute and chronic osteomyelitis),treatment of infected joint arthro-plasties, dead space management inpatients with large soft-tissue inju-ries, and chronic infected non-unions.

Contraindications to the use ofantibiotic bead pouches in the treat-ment of open fractures include pa-tient hypersensitivity to specific an-tibiotics, small wounds (for whichbeads are not necessary), and unsal-vageable limbs (because beads do notovercome massive tissue injuries).Contraindication to the use of anti-biotic beads in the treatment of os-teomyelitis include patient hyper-sensitivity to a specific antibioticand the presence of resistant andslime-forming organisms such asEnterococcus. The foreign-body sur-face of the methacrylate beads them-selves is conductive to slime-producing organisms, and the slime

Thomas A. DeCoster, MD

Shahram Bozorgnia, MD

Dr. DeCoster is Professor and ViceChair, Department of Orthopaedics andRehabilitation, University of New MexicoSchool of Medicine, Albuquerque, NM.Dr. Bozorgnia is Trauma Fellow,Department of Orthopaedics andRehabilitation, University of New MexicoSchool of Medicine.

Dr. DeCoster or a member of hisimmediate family has received researchor institutional support from Biomet, EBI,Orthofix, Smith & Nephew, Stryker, andZimmer; has stock or stock options heldin Merck and Wyeth; and has receivedroyalties from Innomed. NeitherDr. Bozorgnia nor a member of hisimmediate family has received anythingof value from or owns stock in acommercial company or institutionrelated directly or indirectly to thesubject of this article.

Reprint requests: Dr. DeCoster,Department of Orthopaedics andRehabilitation, University of New MexicoSchool of Medicine, 1127 UniversityBoulevard NE, Albuquerque, NM87131.



The video that accompaniesthis article is “Preparation andUse of Antibiotic-ImpregnatedBeads for Orthopaedic Infec-

tions,” available on the Orthopaedic Knowl-edge Online Website, at http://www5.aaos.org/oko/jaaos/surgical.cfm

Surgical Techniques


http://www5.aaos.org/oko/jaaos/surgical.cfm

barrier severely limits the efficacy ofthe antibiotics in controlling the in-fection.6,7

Specific Characteristicsof Antibiotic Beads

Antibiotic-impregnated polymethyl-methacrylate (PMMA) cement beadsare a popular modality used in con-junction with surgical débridementand intravenous antibiotic therapyfor the treatment or prophylaxis oforthopaedic infections. The beadsvary in size, type and amount of an-tibiotic used, and type of bone ce-ment used. The beads can be pre-pared in advance by molding or byrolling by hand in the operatingroom.

Nonbiodegradable PMMA ce-ment is the most common carrierused. To incorporate antibiotics, an-tibiotic powder is mixed with thepowdered cement polymer beforeadding the methylmethacrylate liq-uid monomer. The antibiotic mustbe water soluble, available as pow-der, able to remain stable despite theheat generated during the polymer-ization reaction, and hypoallergenic,as well as have a broad spectrum ofactivity. Antibiotic release is highestin the first 4 days following implan-tation; the remaining elution at ther-apeutic concentration persists forweeks to months.8 Wahlig et al3

showed that, if gentamicin-PMMAchains are implanted and the woundis closed, then the local concen-trations of antibiotic achieved are200 times the levels achieved withsystemic antibiotic administration.However, the use of the beads in anopen system or in combination withsuction irrigation rapidly lowers lo-cal antibiotic concentrations, andthe therapeutic advantage is dimin-ished. Therefore, this technique isnot recommended.

Aminoglycosides are the mostcommonly used antibiotics in thiscontext. They are effective againstaerobic gram-negative bacilli andstaphylococci as well as streptococ-

ci, enterococci, and anaerobes.9

Among the aminoglycosides are to-bramycin and gentamicin. Tobramy-cin has been substituted for genta-micin in the United States because itis available as a pharmaceutical-grade powder, whereas gentamicin isnot. There is extensive informationon the elution patterns of aminogly-cosides from beads in a variety ofclinical scenarios.4,10

Vancomycin should be consid-ered when there is a risk of resistantstaphylococcal organisms, that is,methicillin-resistant Staphylococ-cus aureus. Vancomycin is availablein powder form and is not neutral-ized by the heat of methacrylate po-lymerization. Effective elution ofvancomycin has also been report-ed.11

Bead molds are available in a va-riety of sizes. A diameter of 8 mm isthe largest and 2 mm, the smallest.Small-diameter beads are used inwounds of the hand and in othersmall wounds to maximize the sur-face area and antibiotic elution.Consistency in size and shape of thebeads also facilitates their passageinto tight spaces, including the med-ullary canal.

Surgical Technique

The technique of application of anti-biotic beads involves the production

of beads, followed by surgical place-ment within the débrided woundand in place of débrided bone (Fig-ures 1 and 2). Following bead im-plantation, the soft tissue is closed.The beads allow for very high levelsof antibiotic bathing of the woundand also assist in fighting infection.The beads occupy space, preventingthe accumulation of hematoma thatotherwise would be a potential sitefor infection. Antibiotic beads alsoprovide management of dead spaceby preventing the formation of scartissue in the bone-defect site. If in-fection persists, a repeat débride-ment, culture, and antibiotic beadexchange may be performed afterseveral days or weeks. Once the in-fection is well controlled, the beadsare surgically removed. For non-union or a large bone void, bone graftmay be placed in the area that thebeads occupied.

Although the antibiotic withinthe beads produces a high local con-centration in the surrounding tis-sues, the systemic level of antibiot-ic resulting from beads is low, whichminimizes complications, includingrenal toxicity and ototoxicity. Thehigh local concentration of antibiot-ic reduces and often eliminates theneed for intravenous antibiotics;therefore, intravenous access andcompliance are not required. Osteo-myelitis typically involves seques-

Figure 1

A string of antibiotic beads is placedinto the wound for dead spacemanagement and to provide a highconcentration of local antibiotic to thewound.

Figure 2

In open fractures with significant soft-tissue injuries, antibiotic beads can beplaced in the open wound. The soft-tissue defect will be covered with anadhesive, porous, polyethylene woundfilm.

Thomas A. DeCoster, MD, and Shahram Bozorgnia, MD


tration of dead infected bone withlittle to no blood supply. Many sys-temic antibiotics exhibit poor pene-tration of bone, even when the boneis vascularized. Hence, systemic an-tibiotics and blood-borne antimicro-bial cells may not reach the areaof greatest need. Antibiotic beads,however, provide high local concen-trations of antibiotic not dependenton blood supply to the bone. Neitherare the antibiotics dependent onbone penetration.5

Antibiotic BeadPreparation Technique

Bead PreparationThe bead preparation technique

described here results in spherical,uniform tobramycin-impregnatedPMMA beads ( video, “Ingredi-ents”). These beads can be made inadvance, sterilely packaged, andstored in the operating room readyfor use. Alternatively, beads can beprepared in the operating room atthe time of implantation, with orwithout the use of bead molds.When this technique is used, suffi-cient time (at least 30 minutes) mustbe allowed for full curing of the ce-ment to minimize toxic monomers.

MaterialsThe materials needed for the pro-

duction of tobramycin-impregnatedPMMA beads are the following:

• One package (40 g) of PMMA

bone cement• 20 mL of liquid monomer• Two 1.2-g vials of tobramycin

powder• Two twisted strands of 26-gauge

wire• A cast-metal, Teflon-coated bead

mold• One medium-size (approximate-

ly 200 mL) plastic bowl• Tongue depressors• Two 10-mL syringes• Tweezers from a suture-removal

kit• Gloves

TechniqueTwist together two strands of 26-

gauge wire (50 turns using a handdrill) ( video, “Wire-Twisting Pro-cedure”). Carefully place the wireinto the grooves of the bead moldand tighten the mold ( video,“Preparation Technique A [Spatu-la]”). A cool working environmentand precooling of the mold andPMMA prolong the polymerizationtime. In a medium-size plastic bowl,mix most of one package of bone ce-ment (ie, 35 of 40 g) with two 1.2-gvials of tobramycin powder. This al-lows a prolonged liquid phase, whichfacilitates better bead production.Stir thoroughly to ensure homogene-ity of the mixture. Pour 20 mL of liq-uid monomer into the powder andstir vigorously for about 30 secondsuntil the mixture liquefies. Contin-ue to stir the cement mixture, mov-

ing the cement away from the bot-tom and edges of the bowl to preventhardening.

With a tongue depressor, forceful-ly press the liquid cement mixtureinto the bead mold (Figure 3). Quick-ly and carefully fill all of the holesuntil the entire mold is filled withthe cement mixture. It is imperativethat this step is done in less than 8minutes, before the mixture hard-ens. Smooth the surface of the moldby scraping off the remaining ce-ment with a tongue depressor. Placethe mold on its side and let it sit for15 minutes as the cement hardens.

An alternative technique for plac-ing the cement into the mold holesis to pour 3 mL of the liquid cementmixture into a 10-mL syringe withthe plunger removed and the tipcapped, then replace the plunger andremove the syringe cap ( video,“Preparation Technique B [Sy-ringe]”). While applying pressure,use the syringe to fill each hole inthe mold with the cement mixture.Keep the syringe at a 90° angle aseach hole is filled and apply pressureto maintain a tight seal between themold and the syringe (Figure 4). Ce-ment should bulge into the adjacenthole. Proceed rapidly because the sy-ringe technique requires the cementto be in the liquid state. Four sy-ringes typically are needed for two20-bead chains. The tip of each sy-ringe is cut off so that the lumenmatches the diameter of the mold.Smooth the surface of the mold byscraping off the remaining cementwith a tongue depressor. Place themold on its side and let it sit for 15minutes until the cement is com-pletely hardened.

After 15 minutes, take the moldapart and gently remove the beadsby pulling the wire (Figure 5). Re-move the excess cement (ie, flash-ing) between the beads using thetweezers from the suture kit. Excesscement in the mold also should beremoved to facilitate the next useof the mold. A pair of scissors froma suture-removal set is the exact

Figure 3

The liquid cement mixture is pressedforcefully with a tongue depressor intothe bead mold.

Figure 4

The liquid cement mixture is pressedforcefully with a cement-filled syringeinto the bead mold.

Antibiotic Beads






width of the slots in the mold andthus can be used to remove residualcement from the mold.

Transfer the beads into gas-sterilization packages (Figure 6) andhave them sterilized. Wash yourhands and all work surfaces thor-oughly after working with the ce-ment. This same process can be per-formed in the operating room, usingsterile technique, for immediateuse.

Complications

Complications caused by the use ofantibiotic beads are uncommon;however, difficulty with bead re-moval can occur when the beads areleft in place too long, especially inthe medullary canal of long bones.

Antibiotic beads reduce the rate ofinfection in open fractures, but theydo not eliminate the risk. Beads helpto control established infections, butnot all infections will completely re-solve. A major complication is persis-tent or recurrent infection. Also, spac-ers can fragment or dislodge when

subjected to excessive or chronicloads. Infection resulting from inad-equate débridement will not be over-come by the use of antibiotic beads.

Knowledge of the effectiveness ofantibiotic beads is limited. The opti-mal dose, duration of treatment, andrelative efficacy of various antibioticclasses are not known. Differentialelution of antibiotics from variousforms of PMMA has been report-ed,12,13 but the ideal carrier mediumis still a matter of debate. The roleand efficacy of absorbable beads is notyet known. Local antibiotic deliveryby pump or other mechanism is an al-ternative route of local antibiotictherapy. In addition, the efficacy ofbeads compared with the efficacy ofother delivery methods is not yetknown.

Summary

Antibiotic beads are an attractivemethod of treatment in the manage-ment and prevention of osteomyeli-tis. Antibiotic beads provide high lo-cal concentrations of antibiotic at

Pearls• Remove all avascular, necrotic, and contaminated tissue before applying antibiotic beads. Infection re-

sulting from inadequate débridement will not be overcome by the use of antibiotic beads.

• Beads can be made in advance and stored in the operating room.

• Thoroughly mix the bone cement with the antibiotic powder before adding the liquid monomer.

• Fill all of the holes of the mold as quickly as possible before the mixture hardens.

• An assistant is helpful.

Pitfalls• Do not use the beads in an open system or in combination with suction irrigation because doing so pre-

vents development of effective antibiotic levels in the wound.

• Do not use the beads as an initial measure in inflamed and suppurating wounds.

• Do not use the beads in the treatment of osteomyelitis in the presence of resistant organisms or slime-forming organisms (eg, Enterococcus) because effective elution of antibiotic is not achieved.

• Do not substitute antibiotic beads for thorough wound débridement.

• Do not insert handmade beads that are too large to fit into the medullary canal or too big to completelyfill the wound.

• Use care to produce bead chains in a timely fashion during the short available working time (8 minutes)as the methacrylate polymerizes. Failure to do so results in wasted material and surgeon frustration. Thesecan be avoided by planning and relying on an assistant.

• Do not leave beads in the patient so long (>3 weeks) that removal is difficult.

Figure 5

After 15 minutes, the mold is takenapart, and the beads are gentlyremoved by pulling the wire.

Figure 6

The beads are transferred into gas-sterilization packages.

Thomas A. DeCoster, MD, and Shahram Bozorgnia, MD


the site of infection without signifi-cant systemic toxicity. A variety oftechniques to provide local antibiot-ics has been reported, including ab-sorbable beads with various types ofantibiotics, antibiotic sticks, coatednails, and coated joint spacers. Beadscan be prepared in the operatingroom or in advance. Antibiotic beadsassist in dead space managementand help facilitate the filling of bonevoids and healing of infected non-unions. Results demonstrate im-proved efficacy in the control of in-fection and enhanced outcomes,with financial and practical sav-ings.14

References


1. Gustilo RB, Anderson JT: Preventionof infection in the treatment of onethousand and twenty-five open frac-tures of long bones: Retrospective andprospective analyses. J Bone JointSurg Am 1976;58:453-458.

2. Patzakis MJ, Harvey JP Jr, Ivler D: Therole of antibiotics in the managementof open fractures. J Bone Joint SurgAm 1974;56:532-541.

3. Wahlig H, Dingeldein E, Bergmann R,Reuss K: The release of gentamicinfrom polymethylmethacrylate beads.J Bone Joint Surg Br 1978;60:270-275.

4. Schentag JJ, Lasezkay G, Plant ME,Jusko WJ, Cumbo TJ: Comparativetissue accumulation of gentamicinand tobramycin in patients.J Antimicrob Chemother 1978;4:23-30.

5. Cunningham A, Demarest G, RosenP, DeCoster TA: Antibiotic bead pro-duction. Iowa Orthop J 2000;20:31-35.

6. van de Belt H, Neut D, Schenk W, vanHorn JR, van Der Mei HC, BusscherHJ: Staphylococcus aureus biofilmformation on different gentamicin-loaded polymethylmethacrylate bonecements. Biomaterials 2001;22:1607-1611.

7. Ensing GT, van Horn JR, van der MeiHC, Busscher HJ, Neut D: Copal bonecement is more effective in prevent-ing biofilm formation than PalacosR-G. Clin Orthop Relat Res 2008;466:1492-1498.

8. Anagnostakos K, Kelm J, Regitz T,Schmitt E, Jung W: In vitro evaluationof antibiotic release from and bacteriagrowth inhibition by antibiotic-

loaded acrylic bone cement spacers.J Biomed Mater Res B Appl Biomater2005;72:373-378.

9. Popham GJ, Mangino P, Seligson D,Henry SL: Antibiotic-impregnatedbeads: Part II. Factors in antibioticselection. Orthop Rev 1991;20:331-337.

10. Walenkamp GH, Vree TB, van RensTJ: Gentamicin-PMMA beads: Phar-macokinetic and nephrotoxicologicalstudy. Clin Orthop Relat Res 1986;205:171-183.

11. Sasaki T, Ishibashi Y, Katano H,Nagumo A, Toh S: In vitro elution ofvancomycin from calcium phosphatecement. J Arthroplasty 2005;20:1055-1059.

12. Greene N, Holtom PD, Warren CA, etal: In vitro elution of tobramycin andvancomycin polymethylmethacryl-ate beads and spacers from Simplexand Palacos. Am J Orthop 1998;27:201-205.

13. Nelson CL, Griffin FM, Harrison BH,Cooper RE: In vitro elution character-istics of commercially and noncom-mercially prepared antibiotic PMMAbeads. Clin Orthop Relat Res 1992;284:303-309.

14. Henry SL, Seligson D, Mangino P,Popham GJ: Antibiotic-impregnatedbeads: Part I. Bead implantation ver-sus systemic therapy. Orthop Rev1991;20:242-247.

Antibiotic Beads


j aaos -2008

Documents