Review Article Ankle Distraction Arthroplasty: … Article Ankle Distraction Arthroplasty:...

Review Article

Ankle Distraction Arthroplasty:Indications, Technique, andOutcomes

Abstract

Ankle distraction is an alternative to ankle arthrodesis or total anklearthroplasty in younger patients with arthritis. Ankle distractioninvolves the use of external fixation to mechanically unload the anklejoint, which allows for stable, congruent range of motion in the settingof decreased mechanical loading, potentially promoting cartilagerepair. Adjunct surgical procedures are frequently done to addresslower-extremity malalignment, ankle equinus contractures, andimpinging tibiotalar osteophytes. Patients can bear full weight duringthe treatment course. The distraction frame frequently uses a hinge,and patients are encouraged to do daily range-of-motion exercises.Although the initial goal of the procedure is to delay arthrodesis, manypatients achieve lasting clinical benefits, obviating the need for totalankle arthroplasty or fusion. Complications associated with externalfixation are common, and patients should be counseled that clinicalimprovements occur slowly and often are not achieved until at least 1year after frame removal.

Ankle osteoarthritis is generallya progressive condition, most

commonly the result of high-energytibial plafond fractures, bimalleolarankle fractures, recurrent ankle insta-bility, and neuropathy.1-3 Lower-extremity posttraumatic arthritis hasan estimated cost of $12 billionannually in the United States.4 It isoften disabling, predominantlyaffects young, active persons, and hasa negative effect on quality-of-lifemeasures that is comparable to thatof hip and knee arthritis.1 Themainstay of surgical treatment ofadvanced ankle arthritis has tradi-tionally included ankle arthrodesis ortotal ankle arthroplasty (TAA).Ankle arthrodesis reliably provides

pain relief. However, loss of anklemotion, increased stress at adjacentjoints that leads to degeneration, andincreased energy expenditure with

ambulation do occur.5 Unlike fusion,TAA does not affect range of motion(ROM); however, its use in youngactive patients may be contra-indicated because of wear, failure,and subsequent revisions.6

In contrast to the aims of TAA andankle fusion, the aim of distractionarthroplasty is to optimize the body’sregenerative capacity and the func-tion of the diseased joint.7,8 Anexternal fixator is used to mechan-ically unload the ankle to relievepain, preserve ROM, and potentiallydelay or even partially reverse theeffects of arthritis (Figure 1). Thesurgeon should be aware of thistreatment option, as well as itsindications, outcomes, and potentialadverse effects for ankle arthritis.Recent short- and intermediate-termevidence suggests that distractionarthroplasty may be a viable surgical

Mitchell Bernstein, MD, FRCSC

Jay Reidler, MD, MPH

Austin Fragomen, MD

S. Robert Rozbruch, MD

From the Department of OrthopaedicSurgery, Loyola University ChicagoStritch School of Medicine, Chicago,IL (Dr. Bernstein), the Department ofOrthopaedics, Johns HopkinsUniversity, Baltimore, MD(Dr. Reidler), and the Department ofOrthopaedic Surgery, Hospital forSpecial Surgery, Cornell University,New York, NY (Dr. Fragomen andDr. Rozbruch).

J Am Acad Orthop Surg 2016;0:1-11

DOI: 10.5435/JAAOS-D-14-00077

Copyright 2016 by the AmericanAcademy of Orthopaedic Surgeons.

Month 2016, Vol 0, No 0 1

Copyright � the American Academy of Orthopaedic Surgeons. Unauthorized reproduction of this article is prohibited.

http://dx.doi.org/10.5435/JAAOS-D-14-00077

option with the use of appropriatepatient selection criteria.9,10

Anatomy andBiomechanics of the AnkleJoint

The ankle is a highly constrained andcongruent weight-bearing joint. Its

ability to withstand up to 5.5 timesbody weight during ambulationdepends on a stable relationshipamong the bony and ligamentousstructures of the distal tibia, fibula,and talus.11 The tibial plafond has acentral ridge oriented in the sagittalplane that is complementary to aconcavity on the talar dome. Therelationship between the distal fib-ula, tibiotalar joint, and the medialmalleolus is maintained by the stoutligaments that make up the anklesyndesmosis.One study demonstrated that even

a 1-mm displacement of the talus inthe ankle mortise generates a 42%decrease in available joint contactarea.12 Consequently, the remainingcartilage is exposed to compressiveforces over a smaller surface area,potentially leading to degenerationand arthritis.The orientation of the ankle joint,

as described by a line perpendicularto the diaphysis of the tibia, is in slightvalgus. Named the lateral distal tibialangle, it measures on average 89�(range, 86º to 92�). The axis of theankle joint is created through therelationship between, and the geo-metric constraints of, the talar dome,tibial plafond, and the lateral andmedial malleoli.13 The traditionallyaccepted theory, proposed in the1950s, suggests that the ankle rotateswith one-degree of freedom about anaxis (ie, the Inman axis) that liesbetween the tips of the medial andlateral malleoli.14-16 This axis is thebasis for total ankle prosthesisdesigns17 and forms the basis onwhich the hinge is built when dis-traction arthroplasty is performed

with a circular fixator.18,19 Theconcept of the hinge is based on thetheory that the talar dome is a frus-tum of a cone with its apex mediallydirected.Articular cartilage, or hyaline car-

tilage, lacks blood supply, nerveinnervation, and lymphatic drainage.It receives nutrition and expels wastevia diffusion and imbibition. Its hightensile strength and elasticity func-tion to withstand high loads, protectsubchondral bone, and decreasefriction between the two opposingsurfaces. Hyaline cartilage is primar-ily composed of type II collagen,water, and chondrocytes. Chon-drocytes, which produce enzymes,proteins, and collagen, are responsi-ble for the normal and pathologicstate of the articular surface.20,21

Chondrocytes are metabolicallycontrolled via the surroundingmechanical environment and thus,can upregulate the synthesis of deg-radative cytokines, increasing cata-bolic enzymes in the local milieu—aconcept referred to as mechanoelec-trochemical events.22

Arthritis in the ankle is most oftensecondary, usually resulting fromtrauma.2,3 The energy dissipatedthrough the articular surface and theensuing inflammatory response isthought to be critical in post-traumatic arthritis.23,24 It has beendemonstrated that the inflammationresulting from the energy loss causesthe production of dysfunctionalcellular elements and chondrocyteapoptosis.25 In the subacute phase,articular cartilage changes its com-position, increasing its water con-tent and decreasing its proteoglycan

Figure 1

Ankle distraction frame mounted onan extremity model demonstrating ourcurrent assembly method. The red(circular) and blue (U-shaped) ringsare secured with 6-mmhydroxyapatite-coated half pins and1.8-mm bayonet Kirschner wires tothe tibia and foot, respectively. Thehinge (red arrow) is aligned with theInman axis. The acute and gradualdistraction is achieved with clockwiserotations of the square nuts (redarrowhead), which are attached to thetibial ring. Range-of-motion exercisesoccur with unlocking the connectionbetween the two rings (yellow arrow).Gradual dorsiflexion of the ankle, inthe setting of chronic ankle equinus, isachieved by gradually increasing thedistance between the square nut(yellow arrow) and the proximal end ofthe rod. (Copyright Mitchell Bernstein,MD, FRCSC, Chicago, IL.)

Dr. Bernstein or an immediate family member serves as a paid consultant to NuVasive, Smith & Nephew, and DuPuy Synthes and serves asa board member, owner, officer, or committee member of the Limb Lengthening Reconstruction Society. Dr. Fragomen or an immediatefamily member has received royalties from Small Bone Innovations; is a member of a speakers’ bureau, or has made paid presentations onbehalf of Smith & Nephew; and serves as a paid consultant to Smith & Nephew and DePuy Synthes. Dr. Rozbruch or an immediate familymember has received royalties from Small Bone Innovations and Smith & Nephew; is a member of a speakers’ bureau, or has made paidpresentations on behalf of Smith & Nephew; serves as a paid consultant to Small Bone Innovations and Smith & Nephew; and serves as aboard member, owner, officer, or committee member of the Limb Lengthening Reconstruction Society. Neither Dr. Reidler nor anyimmediate family member has received anything of value from or has stock or stock options held in a commercial company or institutionrelated directly or indirectly to the subject of this article.

Ankle Distraction Arthroplasty: Indications, Technique, and Outcomes

2 Journal of the American Academy of Orthopaedic Surgeons


concentration. Furthermore, type IIcollagen is weakened by a combina-tion of decreased production bychondrocytes and increased concen-tration of proinflammatory cytokines.To restore normal homeostasis,deeper layers containing “resting”chondrocytes proliferate to increaseanabolic activities. The exact cellu-lar mechanisms, signaling mole-cules, genetic factors, and the role ofmechanical influences currently arenot fully understood.As the catabolic processes over-

whelm resident chondrocytes, “full-thickness” chondral involvementensues, exposing subchondral bone.Healing is possible, albeit unpre-dictably. This spontaneous healingoccurs in part because of the releaseof growth factors from exposedmarrow spaces.26 The resultant localinflammatory response recruits plu-ripotent mesenchymal stem cells,which, depending on the local envi-ronment, can be manipulated todevelop fibrocartilage.27-30 This isone potential pathway that distrac-tion arthroplasty and adjunct pro-cedures may use to exploit theformation of hyaline cartilage.31

For younger patients withposttraumatic lesions, a durable,joint-sparing solution is desirable.Concurrently addressing all pathol-ogy, including equinus contracture, ti-biotalar osteophytes, supramalleolaror hindfoot malalignment, and insta-bility, is central to treatment. Cartilageregeneration is more reliable when itoccurs in the setting of a congruent,stable limb in anatomic alignment.19,32

Distraction Arthroplasty

The success of ankle distractiondepends on proper patient selectionand appropriate management ofexpectations. The ideal candidate forankle distraction arthroplasty is amotivated patient who seeks analternative to ankle fusion or TAA

andhas recalcitrant pain in the settingof a congruent joint with preservedmotion of .20�.33 Relative contra-indications include complex regionalpain syndrome, inflammatory ar-thritides, previous infection, neuro-pathic joint, and older age with lowfunctional demands. Patients with apainful stiff ankle (ie, ,20� ofmotion) are less likely to do well withdistraction because the proceduredoes not reliably increase ROM, andthus, these patients may be bettercandidates for arthrodesis or TAA.10

Extra-articular deformity, locatedin the hindfoot or distal tibia, is not acontraindication if the deformity isaddressed concurrently.34 Patientswith marked intra-articular defor-mity or a flat-top talus, however, arefelt to be poor candidates for ankledistraction. Asymmetric arthritis ofthe ankle is not a contraindicationfor ankle distraction. For example,patients with varus deformity at thedistal tibia and asymmetric jointwear on the medial side may benefitfrom a supramalleolar osteotomy,with correction of the varus defor-mity and ankle distraction to offloadthe diseased segment. Finally, ante-rior joint space narrowing associatedwith impinging anterior osteophytesshould be identified. In thesepatients, arthroscopic or open de-compression, possibly in conjunctionwith gastrocnemius recession,should be considered.The success of ankle distraction is

predicated on a thorough history,physical examination, and ancillarytests. Our evaluation includes a reviewof patients’ reasons for consultationand their perception of their disabilityin addition to the basic elements of athorough patient history: the locationand quality of the pain, aggravatingand alleviating factors, subjectivedescription of instability, previousnonsurgical and surgical treatments,and other musculoskeletal com-plaints. Because inflammatory arthri-tides are relative contraindications to

distraction, the etiology of anklearthritis is critical. A history of injuryor repetitive instability is thereforecarefully elucidated.The physical examination begins

with an assessment to identify anyipsilateral (or contralateral) extrem-ity malalignment, such as tibial mal-union, knee hyperextension, or tibiavara. The patient is screened for limblength inequality by comparing theheights of the iliac crests. Dynamicextremity instability, malalignment,and foot progression angle aredetermined by observing the patientambulate. A focused assessment ofthe ankle and foot, including stabilityand ROM testing, completes thephysical examination.No routine laboratory tests are

required. Infectious markers, such aswhite blood cell countand erythrocytesedimentation rate and C-reactiveprotein level, are ordered when activeinfection is suspected or needs to beruled out.Radiographic evaluation includes

AP, lateral, and mortise weight-bearing views of the ankle (Figure 2).Radiographs of the tibia and/ora standing hindfoot alignment(Saltzman) view of the foot are ob-tained in the case of pathology and/ormalalignment. CT scans of the ankleare not routinely ordered. MRI is usedwhen osteochondral lesions need to bedelineated or in the case of nonosseouspathology (eg, lateral ankle ligamentpathology, posterior tibial tendinosis).The ankle radiographs are also

used to measure the weight-bearingpreoperative joint space in prepara-tion for the required increase of 5mmduring distraction to effectivelyunload the ankle joint.10,35 Thepresence of subchondral sclerosisand subchondral cysts are noted. Inaddition, the ankle joint should bescrutinized for asymmetric wear.This wear should be correlated to anassociated deformity. For example, ifthe medial aspect of the ankle jointdemonstrates arthritis with relative

Mitchell Bernstein, MD, FRCSC, et al



preservation of the lateral joint andthe patient has a posttraumatic varusdeformity with an apex at the distaltibial metaphysis, joint distraction inaddition to a supramalleolar osteot-omy may be indicated.

History of the Procedure

In 1978, seeking alternatives toTAA for joint arthritis, Judet and

Judet36 described a technique usingexternal fixation to mechanicallyseparate opposing joint surfaces toallow “for fibrous tissue betweenthe bone ends.” Their histologicanalysis of regenerated tissue indogs was done after the tibiotarsaljoints were devoid of articularcartilage and distracted for 30 dayswith a 4- to 8-mm gap. At 1 year,they reported metaplastic changesin the joint surface resembling

those of normal hyaline cartilage.36

Aldegheri et al37 reported on theuse of a hinged distractor for hiparthritis in 80 patients. Based ongood results achieved in 46 patientsat a minimum follow-up of 5 years,the authors concluded that radio-graphic results do not always cor-relate with clinical outcomes.Patients aged .45 years with orwithout inflammatory arthritis haduniformly poor results.

Figure 2

Preoperative AP (A) and lateral (B) standing radiographs of the ankle joint in a 53-year-old woman with posttraumatic arthritis.Subchondral sclerosis and cysts, as well as decreased joint space are noted. Mild flattening of the talar dome is evidentanteriorly on the lateral radiograph. AP (C) and lateral (D) standing radiographs obtained immediately postoperatively afterdistraction arthroplasty and osteophyte excision. Note the 1-inch calibration ball to measure the joint space E, Photograph ofthe patient wearing the frame. AP (F) and lateral (G) standing radiographs of the ankle joint demonstrating joint space andremodeling of the joint at an 18-month follow-up visit. (Copyright Mitchell Bernstein, MD, FRCSC, Chicago, IL.)




Mechanism of Action

The exact mechanism of action fordistraction arthroplasty is unknown;however, it is based on the belief thatthe biologic aspects of cartilageregeneration are most likely to occurin a mechanically unloaded, well-aligned limb.38,39 The stiffness ofthe circular ring fixator may allowfor sufficient stress shielding at theankle joint to allow subchondralbone remodeling, which has beenshown to be of clinical benefit.40

However, this phenomenon, and itsclinical correlation to pain relief, isstill controversial.In addition, weight bearing, sta-

bility, and motion—crucial for pro-moting durable articular cartilage—are possible for an extended periodin the external fixator.41-43

van Valburg et al43 measured intra-articular hydrostatic pressure dur-ing distraction by inserting apressure-sensitive catheter into theankle. Recordings demonstratedintra-articular pressure fluctuations,and the authors suggested that thesefluctuations combined with theabsence of mechanical stress wereinstrumental in articular cartilagerepair.43-45 They theorized that theintermittent fluid flow inside thejoint caused by pressure fluctuationseffectively mimics normal physio-logic processes in the absence ofload and shear and that this pro-motes cartilage repair. In a study ofknee arthritis (induced by anteriorcruciate ligament resection) in acanine model, van Valburg et al31

observed that, in addition to theintermittent fluid flow caused byan articulated knee hinge on theIlizarov device, a change occurred inthe proteoglycan metabolism thatresembled the nonarthritic controlknee.Factors that have been associated

with cartilage repair include a decreasein joint reactive forces (shear), an

increase in protoeglycan synthesis,recruitment of mesenchymal stemscells, and optimization of themechanical environment.22,31,45 Fur-thermore, an improvement is seen inthe density of the subchondral bonewith a decrease in sclerosis.40 Thissubchondral change may improve thebiomechanical environment of thearthritic joint. In addition, resorptionof subchondral cysts and improve-ment in edema may contribute to theclinical improvement that we haveobserved.

Outcomes

In 1995, van Valburg et al44 pub-lished a preliminary retrospectivestudy on the use of ankle distractionin 11 patients with severe post-traumatic arthritis scheduled forankle arthrodesis. The patientcohort was relatively young (meanage, 35 years). An Ilizarov distrac-tion apparatus was used for a meanof 15 weeks. Their technique, whichhas since been slightly modified,consisted of ensuring a post-operative distraction gap of 5 mmand the addition of hinges between6 and 12 weeks postoperatively. Ata mean follow-up of 20 months(range, 10 to 60 months), allpatients reported a decrease in pain,and five patients (45%) reported acomplete resolution of pain.44 Sixpatients (55%) had increased ROM,and five (50%) had radiographicevidence of increased tibiotalar jointspace. Encouraged by their results,the authors published a prospectivestudy with a minimum 2-yearfollow-up.46 The study included 17patients (mean age, 39.6 years) whounderwent annual evaluations forfunctional ability, mobility, andpain scores. Ankle distraction failedin four patients, who ultimatelyproceeded with arthrodesis. Of theremaining 13 patients, clinical out-come scores at 2-year follow-up

were better (P , 0.004) than thoseobtained at 1-year follow-up.Although poorly understood, sev-eral studies support the process ofankle joint remodeling after frameremoval.10,40,46,47 The restorationof preinjury ankle architecture andmechanics is thought to be instru-mental for prolonged benefit. Thenotion that improvements mayoccur over 12 to 24 months sub-sequent to frame removal shouldbe considered when counselingpatients in their recovery period.In 2008, Paley et al48 published

their results on a modified distractionframe that was being used in Europe.Motion during distraction wasstressed, and an anatomically locatedhinge based on the Inman axisbecame the foundation of the frame.Adjunctive procedures, such as os-teophyte resection, gastrocnemius-soleus complex recession, andextremity realignment with osteot-omy, were also done to maximizeclinical outcomes.48 Nine males and14 females were available for ret-rospective chart review after a mean17 weeks with external fixation.Preoperative and postoperativeankle motion was 28� and 27�,respectively. At a mean follow-up of64 months (range, 24 to 157months), 77% patients reportedambulation for pleasure, and 33%reported an ability to run. Radio-graphic analysis revealed that thejoint space after frame removal didnot remain distracted, although thisdid not affect clinical results. Theauthors claimed that the purposeof articular cartilage repair wasto “seal cartilage cracks anddefects,”48 which was supported byan MRI study of three patients whowere treated with ankle distrac-tion.49 Of note, 33% were not sat-isfied with the outcome of theprocedure. Treatment failed in twopatients (11%); one patient under-went ankle fusion and anotherunderwent TAA.48




A hinge allows for ROM duringrehabilitation, but ROMwill not likelyincrease after frame removal.10,32 In

addition, while hinge distraction inanimals supports robust and dura-ble articular-like cartilage regener-

ation, histologic proof of hyalinecartilage regeneration in humansis lacking.8,31 In a prospective

Table 1

Summary of Outcomes and Adverse Events From Studies on Distraction Arthroplasty

StudyNo. of

Patients Follow-upa Agea (yr) Outcomes Adverse Events

Marijnissenet al9

111 2 yr minimum 42.76 9.8 Pain and disability scoresdecreased from 67% and68% to 38% and 36%,respectively, at 2 yr

48 patients (44%) hadsubsequent arthrodesis

Tellisi et al10 23 30.5 mo(12–60 mo)

43 (16–73) Decrease in pain in 91% ofpatients

Pin-site infection in all patients; 2of 23 patients (9%) went on toarthrodesis

Saltzmanet al32

29 2 yr Fixed: 42.4(18–53)

Motion: 42.7(27–59)

Better pain improvement inmotion group at 2 yr; bothgroups better at 2 yr thanbaseline

19 of 29 patients (66%) hadrecurrent pin-site infections; 2 of29 patients (7%) hadosteomyelitis. 8 of 29 patients(28%) had nerve injury of medialcalcaneal branch of the tibialnerve and deep peroneal nerve;1 of 29 patients (3%) had deepvein thrombosis

Intema et al40 26 2 yr 416 9 Decrease in AOS pain anddisability scores; correlationwith subchondral boneremodeling and clinicalimprovement

Not reported

Ploegmakerset al47

22 10 yr(7–15 yr)

376 11 Decrease in pain scores from78% to 30%; increase infunction scores from 20% to73%

6 of 22 patients (27%) hadarthrodesis; 1 of 22 patients(5%) had complex regional painsyndrome

van Valburget al44

11 20 wk(10–60 wk)

35 (20–70) Pain decreased in all patients5 patients pain free

Not reported

van Valburget al46

17 2 yr 40 (17–55) Decrease in physical,functional, and pain disabilityscores at 2 yrs (P , 0.003)

4 of 17 patients (24%) hadarthrodesis; 4 of 17 patients (24%)had broken Kirschner wires

Paley et al48 23 64 mo (24–157 mo)

45 (17–62) 71% of patients ambulating forpleasure; 33% can run, 22%using assistive devices; 11%with severe limitations

17 of 23 patients (74%) had pin-site infection; 1 of 23 patients(4%) had arthrodesis; 1 of 23patients (4%) had total anklearthroplasty; 10 of 23 patients(43%) returned to operatingroom for unplanned procedure

Nguyenet al50

36 8.3 yr (6.1–10.5 yr)

Fixed: 42.4(18–53)

Motion: 42.7(27–59)

AOS score,43; age at time ofdistraction, and fixed versusmotion ankle distractionpredictive of failure at 2 yrpostoperatively

16 of 36 patients (45%) failedtreatment; 8 of 16 patients (50%)hadankle fusion, 5 of 16 patients(31%) had total anklearthroplasty

Marijnissenet al51

57 2.8 yr (2.5–3.1 yr)

44 (18–65) Decrease in pain scores by 38%(P , 0.0001); 69% increase infunction (P, 0.0001); increasein clinical condition by 120%(P , 0.0001)

16 of 57 patients (28%) had pin-site infections; 8 of 57 patients(14%) had broken Kirschnerwires

AOS = Ankle Osteoarthritis Scalea The values are given as the mean with the range in parentheses.Copyright Mitchell Bernstein, MD, FRCSC, Chicago, IL.




randomized controlled trial,Saltzman et al32 compared 36patients who underwent distractionarthroplasty with or without ahinge. Two years after frameremoval, clinical scores were betterin the hinge group, although anklemotion was similar in both groups.However, in a subsequent report onthe same cohort with longer follow-up, the authors reported thatpatients without a hinge hadimproved outcomes.50 The authorscould only speculate on the reasonfor the contradictory results, andfurther research on the benefit of ahinge is necessary. In a retrospec-tive study of 23 patients withhinged distraction arthroplasty,Tellisi et al10 reported that allpatients in the hinge group hadsevere posttraumatic arthritis andwere being considered for anklearthrodesis. At a mean follow-up of30.5months (range, 12 to 60months),no patient demonstrated a change inankle motion. At the latest follow-up,21 of 23 patients (91%) reportedimproved pain, and 17 patients (74%)had notable improvement in Ameri-can Foot and Ankle Society scores.The initial enthusiasm for ankle

distraction focused on the ability todelay arthrodesis or TAA. Propo-nents of distraction arthroplastycite several advantages, includingthe minimally invasive nature ofthe procedure, no required internalfixation, and no interference withfuture reconstructive efforts.10 How-ever, studies of ankle arthrodesis andTAA after distraction arthroplastyare lacking.Nevertheless, clinical failures in the

formof ankle fusion or TAAdo occurfollowing distraction.9,10,46,50 Asnoted, van Valburg et al46 reportedon 17 patients with a mean age of39.6 years who were treated withfixed ankle joint distraction. Fourpatients (24%) required ankle fusionwithin 1 year postoperativelybecause of the recurrence of severe

pain.46 Similar rates of failure werereported by Marijnissen et al51 andPloegmakers et al47 (24% and 27%,respectively), who reported thatclinical recurrence of pain 1 yearafter frame removal was the reasonfor ankle arthrodesis.51 Marijnissenet al9 recently updated their clinicalresults with data from a 12-yearfollow-up, noting a 44% rate ofconversion to ankle arthrodesis. Inthe same study, Cox regressionanalysis revealed that female sex waspredictive of failure, whereas pre-operative ankle motion permittingdistraction was protective.9 Nguyenet al50 reported on their cohort of 36patients who underwent ankle dis-traction for end-stage osteoarthritis.At a mean follow-up of 8.3 years, 29patients (81%) were available forfollow-up. Treatment failed in 13patients (45%), requiring eitherankle fusion or TAA. The authorsreported that age, Ankle Osteoar-thritis Scale score, and the presenceof a hinge to allow ankle ROM werepredictors of failure at 2 years.50

Finally, it is important to notethat published results primarilyanalyze patients with severe anklearthritis who otherwise would beconsidered candidates for arthrod-esis (Table 1). Selection of patientswith moderate arthritis could leadto improved long-term outcomes.Further research is necessary.

Authors’ Preferred SurgicalTechnique

The patient’s history and physicalexamination, as well as the results ofappropriate imaging studies willdictate which, if any, adjunct surgi-cal treatments are required beforethe circular frame is mounted to theleg. Ankle equinus contracture witha positive Silfverskiöld test is treatedwith a gastrocnemius Strayer orVulpius recession through a postero-medial or direct posterior surgical

approach, respectively. An arthrot-omy is used to remove impinginganterior osteophytes. A supra-malleolar osteotomy is added tocorrect concomitant coronal or sag-ittal malalignment.19,34 This pro-cedure requires placement of anadditional circular ring at theproximal tibia, with the distractiontibial ring closer to the ankle joint(Figure 3). The supramalleolar os-teotomy begins with the patient in asupine position on a radiolucenttable. A bump is placed under theipsilateral buttock to ensure the limbis in neutral rotation (ie, patellafacing upward).In addition to mechanical distrac-

tion, we prefer to inject bone marrowautograft concentrate (BMAC) fromthe ipsilateral iliac crest as describedby Hernigou et al.52 An aspirate of60 mL of marrow yields approxi-mately 7 mL of BMAC. This aspiratecontains pluripotent stem cells,which are injected into the anklejoint. We inject this percutaneouslyat the end of the case, after the acutedistraction has been performed. Weroutinely administer this aspirate aspart of the ankle distraction pro-cedure. Although clinical evidence islacking, compelling basic science andanimal studies support the use ofBMAC to augment the cartilageregeneration.8,29,53-55

The application of the ankle dis-traction frame begins by choosing atibial ring that allows for two finger-breadths of space circumferentiallybetween the skin and the ring. Themedial malleolus and the anteriorand posteromedial border of thedistal tibia are marked. The prox-imal ring is secured with two 6-mmhydroxyapatite-coated pins.56,57 Thefirst pin is placed approximately 6cm proximal to the medial malleolusdirectly anterior in the tibial crestusing a 4.8-mm drill bit. The pin isplaced perpendicular to the shaftof the tibia and secured to thering with a three-hole cube. Before




final tightening, intraoperative fluo-roscopy is used to confirm that thering is positioned perpendicular to

the axis of the tibial shaft. Universalhinges are then applied in line with aKirschner wire, inserted from the tip

of the lateral malleolus and exiting atthe tip of the medial malleolus, in aposterolateral-to-anteromedial

Figure 3

Preoperative AP (A) and lateral (B) standing radiographs of the ankle joint in a 62-year-old woman who sustained a closed rightdistal tibia fracture in a skiing accident. Note the decreased joint space, subchondral sclerosis, cysts, and anterior ankleosteophytes. Lateral translation, recurvatum, and anterior subluxation of the talus also are evident. Supramalleolar osteotomywasperformed to correct extremity malalignment, and ankle distraction arthroplasty was done in conjunction with arthrotomy, excisionof tibiotalar osteophytes, microfracture of the talar dome and tibial plafond, and gastrocnemius-soleus complex recession. C,Intraoperative fluoroscopic image demonstrating bone marrow aspirate injection into the ankle joint.D, Clinical photograph of thelower extremity after frame application. Note the additional ring and struts used to correct supramalleolar malalignment. Lateral (E)and AP (F) radiographs of the ankle joint at 6 months postoperatively. Reduction of the tibiotalar joint and restoration of coronaland sagittal alignment have been achieved. Joint space is increased on weight-bearing radiographs. Clinical examinationdemonstrated 10� of dorsiflexion and 25� of plantar flexion. (Copyright Mitchell Bernstein, MD, FRCSC, Chicago, IL.)




direction18 (Figure 4). This approx-imates the Inman axis. A footplate issecured 1 inch proximal and parallelto the plantar aspect of the foot. Alocking rod connects the footplate tothe proximal adjustable ring, whichallows for gradual dorsiflexion tocorrect equinus contractures. Thering is unlocked four times dailyfor ROM exercises (15 repetitions/session). Typically, the ankle isacutely distracted 3 mm in theoperating room by turning thesquare nuts on the proximal ring.Acute distraction beyond that isdiscouraged to avoid neurologictraction injury; acute correction ofequinus contracture is avoided forthe same reason. Once normal post-operative plantar sensation is con-firmed, an additional 2 mm ofdistraction is usually done on post-operative day 1 and another 1 mm onpostoperative day 2. At the 2-weekclinic visit, another 1 to 2 mm of dis-traction is done. Fluoroscopy is used toconfirm that a congruent distractiongap exists on the AP and lateral views.Postoperatively, the patient is al-

lowed full weight bearing as toler-ated, with crutches. The neutralposition (ie, ankle dorsiflexion) ismarked on the hinge, and the physicaltherapist teaches the patient how tounlock the hinge and do active-assisted ROM exercises with a footstrap. Once patients are comfortable,they are encouraged to ambulatewiththe frame’s hinge unlocked. Anyresidual distraction beyond whatwas done in the operating room isundertaken by the physician in thehospital or 2 weeks later at the firstclinical visit. We typically do notdistract .3 mm acutely. Based on arecent biomechanical study, a rela-tive increase of 5 mm of joint spaceshould be obtained relative to thepreoperative standing radiograph toensure that the articular surfaces ofthe tibial plafond and talar body donot come in contact during weightbearing.35 Animal models support

the use of the frame in distraction forat least 8 weeks, and no addedbenefit has been seen beyond 12weeks.8,31 We prefer to use the framefor 12 weeks.In addition to distraction, the

senior authors (A.T.F., S.R.R.) cur-rently inject autologous bone mar-row aspirate into the ankle joint androutinely affect microfracture.Although the mechanism of actionof hyaline cartilage regenerationremains elusive, and clinical dataare lacking, we feel that theseadjunctive procedures may opti-mize the local healing environment.

Complications

The most common complicationassociated with ankle distraction ar-throplasty is a superficial pin-siteinfection, which typically resolveswith a course of oral antibiotics. Thereported incidence ranges from 14%to 100%.10,32,48,50,51,58 Osteomyeli-tis that requires hospital admissionand intravenous antibiotics is lesscommon, with a reported incidenceof 1.2% to 5.5%.32,48,51 Pin break-age does occur, usually in the mid-foot because of the motion-inducedcyclic fatigue of the Kirschner wire.Likely underreported, the estimatedincidence is 14% to 24% in twostudies of 74 patients.51,58 Typically,this breakage occurs at the junctionof the wire connection onto the ringand therefore is rectified by modi-fying the connection of the wire-fixation bolt closer to the skin.Complications are best avoided

with stringent and consistent patientselection, meticulous surgical tech-nique, and close clinical follow-up.Patients should be screened at theinitial clinical visit for the inability tocomply with postoperative regimens.Educational level, ability to take timeoff work, living situation, and theavailability of supportive family and/or friends are determined. Patients

are educated preoperatively and inthe hospital before discharge regard-ing the appropriate use of theirexternal fixation device. We recom-mend daily pin-site care using asolution of 50% normal saline and50%hydrogen peroxide appliedwithsterile cotton-tipped swabs. To pro-tect the soft tissues, each group ofpins should be wrapped with 2-inchcotton gauze.A thorough knowledge of cross-

sectional anatomy in the lowerextremity is required to avoid inad-vertent perforation or incarcerationof neurovascular structures. Specifi-cally, when placing the tibial ring,the tibialis anterior tendon andanterior neurovascular bundle are atrisk of injury. The surgeon shouldhave access to new, sharp 4.8-mm

Figure 4

AP fluoroscopic image of the ankledemonstrating insertion of aKirschner wire to approximate theInman axis during application of anankle distraction frame. The wire isinserted from the tip of the lateralmalleolus aiming toward the distalaspect of the medial malleolus in aposterolateral-to-anteromedialdirection. (Copyright MitchellBernstein, MD, FRCSC, Chicago, IL.)




bits to drill pilot holes for the 6-mmhalf pins in each case. This pre-caution avoids thermal damage tobone, premature loosening, and pin-site infection. The foot ring is appliedwith care taken to avoid the medialneurovascular structures. However,in our experience, patients reportheel numbness, which can be attrib-uted to medial calcaneal branchnerve irritation from the crossedhindfoot wires. These symptomsshould not be mistaken for plantarnumbness at the forefoot consistentwith tibial nerve injury because theformer should resolve, whereas thelatter requires urgent release of dis-traction and possible tarsal tunneldecompression. Posterior tibialnerve neurapraxia often occurs withlarger, acute distraction (.5 mm) inpatients with contracted postero-medial soft tissues.

Summary

Ankle distraction arthroplasty is asurgical procedure whereby theankle joint is temporarily mechan-ically unloaded with an externalfixator and is performed in con-junction with osteophyte removal,microfracture, soft-tissue release,and deformity correction, if neces-sary. Although robust clinical evi-dence is lacking, studies usinganimal models support the theorythat the addition of mechanical un-loading and alignment correctionimproves the inherent ability ofhuman cartilage repair to occur.The goals of the procedure are toprovide pain relief, preservemotion, and to generate hyalinecartilage or a durable hyaline-likecartilaginous substance. Althoughclinical studies have demonstratedgood short- to intermediate-termclinical outcomes, the mechanismsfor success and failure remainunknown. Further clinical researchon this procedure and the histologic

composition of the resultant gener-ated tissue is needed.

References

Evidence-based Medicine: Levels ofevidence are described in the table ofcontents. In this article, references 1,9, 32, and 57 are level I studies.Reference 51 and 55 are level IIstudies. References 5, 6, 11, and 33are level III studies. References 2-4,7, 10, 17, 18, 24, 34, 37, 39, 40, 44,46-50, and 53 are level IV studies.References 15, 19, and 38 are level Vexpert opinion.

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

1. GlazebrookM, Daniels T, Younger A, et al:Comparison of health-related quality of lifebetween patients with end-stage ankle andhip arthrosis. J Bone Joint Surg Am 2008;90(3):499-505.

2. Saltzman CL, Salamon ML, Blanchard GM,et al: Epidemiology of ankle arthritis: Reportof a consecutive series of 639 patients from atertiary orthopaedic center. Iowa Orthop J2005;25:44-46.

3. Valderrabano V, Horisberger M, Russell I,Dougall H, Hintermann B: Etiology ofankle osteoarthritis. Clin Orthop Relat Res2009;467(7):1800-1806.

4. Brown TD, Johnston RC, Saltzman CL,Marsh JL, Buckwalter JA: Posttraumaticosteoarthritis: A first estimate of incidence,prevalence, and burden of disease. J OrthopTrauma 2006;20(10):739-744.

5. Thomas R, Daniels TR, Parker K: Gaitanalysis and functional outcomes followingankle arthrodesis for isolated anklearthritis. J Bone Joint Surg Am 2006;88(3):526-535.

6. Brunner S, Barg A, Knupp M, et al: TheScandinavian total ankle replacement:Long-term, eleven to fifteen-year,survivorship analysis of the prosthesis inseventy-two consecutive patients. J BoneJoint Surg Am 2013;95(8):711-718.

7. Wiegant K, van Roermund PM, Intema F,et al: Sustained clinical and structuralbenefit after joint distraction in thetreatment of severe knee osteoarthritis.Osteoarthritis Cartilage 2013;21(11):1660-1667.

8. Yanai T, Ishii T, Chang F, Ochiai N: Repairof large full-thickness articular cartilagedefects in the rabbit: The effects of jointdistraction and autologous bone-marrow-

derived mesenchymal cell transplantation. JBone Joint Surg Br 2005;87(5):721-729.

9. Marijnissen AC, Hoekstra MC, Pré BC,et al: Patient characteristics as predictors ofclinical outcome of distraction in treatmentof severe ankle osteoarthritis. J Orthop Res2014;32(1):96-101.

10. Tellisi N, Fragomen AT, Kleinman D,O’Malley MJ, Rozbruch SR: Jointpreservation of the osteoarthritic ankleusing distraction arthroplasty. Foot AnkleInt 2009;30(4):318-325.

11. Stauffer RN, Chao EY, Brewster RC: Forceand motion analysis of the normal,diseased, and prosthetic ankle joint. ClinOrthop Relat Res 1977;127:189-196.

12. Ringleb SI, Udupa JK, Siegler S, et al: Theeffect of ankle ligament damage andsurgical reconstructions on the mechanicsof the ankle and subtalar joints revealed bythree-dimensional stress MRI. J Orthop Res2005;23(4):743-749.

13. Siegler S, Toy J, Seale D, Pedowitz D: TheClinical Biomechanics Award 2013presented by the International Society ofBiomechanics: New observations on themorphology of the talar dome and itsrelationship to ankle kinematics. ClinBiomech (Bristol, Avon) 2014;29(1):1-6.

14. Barnett CH, Napier JR: The axis of rotationat the ankle joint in man: Its influence uponthe form of the talus and the mobility of thefibula. J Anat 1952;86(1):1-9.

15. Close JR: Some applications of thefunctional anatomy of the ankle joint. JBone Joint Surg Am 1956;38(4):761-781.

16. Inman VT: Inman’s Joints of the Ankle, ed2. Baltimore, MD, William and Wilkins,1991.

17. Hintermann B, Valderrabano V,Dereymaeker G, Dick W: The HINTEGRAankle: Rationale and short-term results of122 consecutive ankles. Clin Orthop RelatRes 2004;424:57-68.

18. Inda DA, Blyakher A, O’Malley MJ,Rozbruch SR: Distraction arthroplasty forthe ankle using the Ilizarov frame. TechFoot Ankle Surg 2003;2(4)249-253.

19. Rozbruch SR: Posttraumatic reconstructionof the ankle using the Ilizarov method. HSSJ 2005;1(1):68-88.

20. Millward-Sadler SJ, Salter DM: Integrin-dependent signal cascades in chondrocytemechanotransduction. Ann Biomed Eng2004;32(3):435-446.

21. Pulai JI, Chen H, Im HJ, et al: NF-kappa Bmediates the stimulation of cytokine andchemokine expression by human articularchondrocytes in response to fibronectinfragments. J Immunol 2005;174(9):5781-5788.

22. Mow VC, Wang CC, Hung CT: Theextracellular matrix, interstitial fluid andions as a mechanical signal transducer in




articular cartilage. Osteoarthritis Cartilage1999;7(1):41-58.

23. McKinley TO, Borrelli J Jr, D’Lima DD,Furman BD, Giannoudis PV: Basic scienceof intra-articular fractures andposttraumatic osteoarthritis. J OrthopTrauma 2010;24(9):567-570.

24. Haller JM, McFadden M, Kubiak EN,Higgins TF: Inflammatory cytokineresponse following acute tibial plateaufracture. J Bone Joint Surg Am 2015;97(6):478-483.

25. D’Lima DD, Hashimoto S, Chen PC,Colwell CW Jr, Lotz MK: Humanchondrocyte apoptosis in response tomechanical injury. Osteoarthritis Cartilage2001;9(8):712-719.

26. Shapiro F, Koide S, Glimcher MJ: Cellorigin and differentiation in the repair offull-thickness defects of articular cartilage.J Bone Joint Surg Am 1993;75(4):532-553.

27. BreinanHA,Martin SD,HsuHP, SpectorM:Healing of canine articular cartilage defectstreated with microfracture, a type-II collagenmatrix, or cultured autologous chondrocytes.J Orthop Res 2000;18(5):781-789.

28. FurukawaT, EyreDR, Koide S, GlimcherMJ:Biochemical studies on repair cartilageresurfacing experimental defects in the rabbitknee. J Bone Joint Surg Am 1980;62(1):79-89.

29. Im GI, Kim DY, Shin JH, Hyun CW,Cho WH: Repair of cartilage defect in therabbit with cultured mesenchymal stemcells from bone marrow. J Bone JointSurg Br 2001;83(2):289-294.

30. Kadiyala S, YoungRG,ThiedeMA, Bruder SP:Culture expanded canine mesenchymal stemcells possess osteochondrogenic potentialin vivo and in vitro.Cell Transplant 1997;6(2):125-134.

31. van Valburg AA, van Roermund PM,Marijnissen AC, et al: Joint distraction intreatment of osteoarthritis (II): Effects oncartilage in a canine model. OsteoarthritisCartilage 2000;8(1):1-8.

32. Saltzman CL, Hillis SL, Stolley MP,Anderson DD, Amendola A: Motionversus fixed distraction of the joint in thetreatment of ankle osteoarthritis: Aprospective randomized controlled trial. JBone Joint Surg Am 2012;94(11):961-970.

33. Smith NC, Beaman D, Rozbruch SR,Glazebrook MA: Evidence-basedindications for distraction anklearthroplasty. Foot Ankle Int 2012;33(8):632-636.

34. Horn DM, Fragomen AT, Rozbruch SR:Supramalleolar osteotomy using circularexternal fixation with six-axis deformitycorrection of the distal tibia. Foot Ankle Int2011;32(10):986-993.

35. Fragomen AT, McCoy TH, Meyers KN,Rozbruch SR: Minimum distraction gap:How much ankle joint space is enough inankle distraction arthroplasty? HSS J 2014;10(1):6-12.

36. Judet R, Judet T: The use of a hingedistraction apparatus after arthrolysis andarthroplasty (author’s transl) [French]. RevChir Orthop Reparatrice Appar Mot 1978;64(5):353-365.

37. Aldegheri R, Trivella G, Saleh M: Articulateddistraction of the hip: Conservative surgeryfor arthritis in young patients. Clin OrthopRelat Res 1994;301:94-101.

38. Felson DT, Kim YJ: The futility of currentapproaches to chondroprotection. ArthritisRheum 2007;56(5):1378-1383.

39. Kanamiya T, Naito M, Hara M,Yoshimura I: The influences ofbiomechanical factors on cartilageregeneration after high tibial osteotomy forknees with medial compartmentosteoarthritis: Clinical and arthroscopicobservations. Arthroscopy 2002;18(7):725-729.

40. Intema F, Thomas TP, Anderson DD, et al:Subchondral bone remodeling is related toclinical improvement after joint distractionin the treatment of ankle osteoarthritis.Osteoarthritis Cartilage 2011;19(6):668-675.

41. Hung SC, Nakamura K, Shiro R, Tanaka K,Kawahara H, Kurokawa T: Effects ofcontinuous distraction on cartilage in amoving joint: An investigation on adultrabbits. J Orthop Res 1997;15(3):381-390.

42. Nishino T, Chang F, Ishii T, Yanai T,Mishima H, Ochiai N: Joint distraction andmovement for repair of articular cartilage ina rabbit model with subsequent weight-bearing. J Bone Joint Surg Br 2010;92(7):1033-1040.

43. van Valburg AA, van Roy HL, Lafeber FP,Bijlsma JW: Beneficial effects ofintermittent fluid pressure of lowphysiological magnitude on cartilage andinflammation in osteoarthritis: An in vitrostudy. J Rheumatol 1998;25(3):515-520.

44. van Valburg AA, van Roermund PM,Lammens J, et al: Can Ilizarov jointdistraction delay the need for an arthrodesisof the ankle? A preliminary report. J BoneJoint Surg Br 1995;77(5):720-725.

45. Lafeber F, Veldhuijzen JP, Vanroy JL,Huber-Bruning O, Bijlsma JW: Intermittenthydrostatic compressive force stimulatesexclusively the proteoglycan synthesis ofosteoarthritic human cartilage. Br JRheumatol 1992;31(7):437-442.

46. van Valburg AA, van Roermund PM,Marijnissen AC, et al: Joint distraction intreatment of osteoarthritis: A two-yearfollow-up of the ankle. OsteoarthritisCartilage 1999;7(5):474-479.

47. Ploegmakers JJ, van Roermund PM, vanMelkebeek J, et al: Prolonged clinical benefitfrom joint distraction in the treatment ofankle osteoarthritis. Osteoarthritis Cartilage2005;13(7):582-588.

48. Paley D, LammBM, Purohit RM, Specht SC:Distraction arthroplasty of the ankle: Howfar can you stretch the indications? FootAnkle Clin 2008;13(3):471-484, ix.

49. Lamm BM, Gourdine-Shaw M: MRIevaluation of ankle distraction: Apreliminary report. Clin Podiatr Med Surg2009;26(2):185-191.

50. Nguyen MP, Pedersen DR, Gao Y,Saltzman CL, Amendola A: Intermediate-term follow-up after ankle distraction fortreatment of end-stage osteoarthritis. JBone Joint Surg Am 2015;97(7):590-596.

51. Marijnissen AC, Van Roermund PM, VanMelkebeek J, et al: Clinical benefit of jointdistraction in the treatment of severeosteoarthritis of the ankle: Proof of conceptin an open prospective study and in arandomized controlled study. ArthritisRheum 2002;46(11):2893-2902.

52. Hernigou P, Mathieu G, Poignard A,Manicom O, Beaujean F, Rouard H:Percutaneous autologous bone-marrowgrafting for nonunions: Surgical technique.J Bone Joint Surg Am 2006;88(suppl 1 pt2):322-327.

53. Kim JD, Lee GW, Jung GH, et al: Clinicaloutcome of autologous bone marrowaspirates concentrate (BMAC) injection indegenerative arthritis of the knee. Eur JOrthop Surg Traumatol 2014;24(8):1505-1511.

54. Betsch M, Thelen S, Santak L, et al: Therole of erythropoietin and bone marrowconcentrate in the treatment ofosteochondral defects in mini-pigs. PLoSOne 2014;9(3):e92766.

55. Hegde V, Shonuga O, Ellis S, et al: Aprospective comparison of 3 approvedsystems for autologous bone marrowconcentration demonstratednonequivalency in progenitor cell numberand concentration. J Orthop Trauma 2014;28(10):591-598.

56. Moroni A, Cadossi M, Romagnoli M,Faldini C, Giannini S: A biomechanical andhistological analysis of standard versushydroxyapatite-coated pins for externalfixation. J Biomed Mater Res B ApplBiomater 2008;86(2):417-421.

57. Pizà G, Caja VL, González-Viejo MA,Navarro A: Hydroxyapatite-coatedexternal-fixation pins: The effect on pinloosening and pin-track infection in leglengthening for short stature. J Bone JointSurg Br 2004;86(6):892-897.

58. van Roermund PM, Lafeber FP: Jointdistraction as treatment for ankleosteoarthritis. Instr Course Lect 1999;48:249-254.


Month 2016, Vol 0, No 0 11


Review Article Ankle Distraction Arthroplasty: … Article Ankle Distraction Arthroplasty:...

Documents

Transcript of Review Article Ankle Distraction Arthroplasty: … Article Ankle Distraction Arthroplasty:...