december1999 lr sinacore[1]

7/28/2019 December1999 LR Sinacore[1]

1/11

Journal of Orthopaedic & Sports Physical Therapyl999;29 (12) 73&746

Recognition and Management of AcuteNeuropathic (Charcot) Arthropathies of theFoot and AnkleDavid R. Sinacore, PhD, PT1Nina C. Withrington *

Study Design: Review of selected literature describing the outcomes related to themanagement of acute Charcot foot arthropathies in patients with diabetes mellitus.Objective: To familiarize the rehabilitationspecialist with the general principles ofnonsurgical management for patients with acute neuropathic arthropathies of the foot andankle.Background: Neuropathic (Charcot) arthropathy of the foot or ankle is the most destructiveand disabling chronic complication of all diabetic foot disease.Methods and Measures:We discuss the clinical presentation and the ro le that orthopaedicand sports physical therapists may have in identifying and preventing complications and thelong-term disabi lity associated with these arthropathies. We summarize the outcomes of 15published reports from 1985-1999 located using the MEDLlNE database from 1966-present.Studies were selected and included if the authors reported on (1) 2 or more patients withdiabetes mellitus and acute Charcot arthropathies; (2) the short-term or long-term outcomes,including the length of follow-up; and (3) the pattern or location of the arthropathy. Theshort-term outcomes (percentage of patients healed, average time to healing) and long-termoutcomes (percentage in whom treatment failed, amputation, disability) after treatment byimmobilization alone or immobilization after surgery were reviewed and summarized.Results: The prognosis for an indiv idual with severe neuropathic skeletal foot deformities ispoor. Eleven deaths (3.65%) in 301 patients were reported within the average follow-upper iod of 2.5 years after treatment for Charcot arthropathy. Pdrtial or complete footamputation occurred in 20 (6.6%) of 301, whereas 83 (28%) of 301 patients reviewed hadmobility limitations or required ankle-foot orthoses or permanent bracing or assistive devicesfor ambulation at the time of follow-up.Conclusion: Rehabilitation specialists can improve the short-term outcomes and limit thelong-term disabilities in patients with diabetes mellitus and peripheral neuropathy. Earlyrecognition and prompt immobilization are the basic principles of nonsurgical managementthat influence therapeutic outcome. ) Orthop Sports Phys Ther 1999;29:736-746.Key Words: diabetes mellitus, foot disease, immobilization, short-term and long-term outcomes

N uropathic (Char-cot) arthropathy ofthe foot and ankleis perhaps the mostdestructive and dis-abling chronic complication of alldiabetic foot disease. When Char-cot arthropathy is accompanied byulceration with infection, there isimminent danger of partial orcomplete foot amputation andeven death.

Roughly 16 million people inthe United States have signs orsymptoms of diabetes mellitus,with one third to nearly one halfof those not aware they have thedisease.lwThe incidence of acuteCharcot ankle or foot arthropa-thies among individuals with dia-betes and peripheral neuropathyis reportedly low, varying between0.15 and 2.5%,35.40 hereas theprevalence of Charcot jointchanges is somewhat higher, oc-curring in as many as 13-29% ofall people with diabetes who havefoot disease evaluated in specialtyclinics and regional medical cen-ters."'"

Assistant professor, Program in Physical Therapy and Research instructor, Division of Geriatrics andGerontology, Department of Medicine, Washington University School of Medicine, St Louis, Mo.Master's degree student, Program in Physical Therapy, Washington University School of Medicine ,St Louis, Mo.Funding in support of this work has been provided in part by grant ROI H D 3 680 242 .Send correspondence to David R. Sinacore, Program in Physical Therapy, Campus Box 8502, Wash-ington University School of Medicine, St Louis, M O 63 1 10. E-mail: [email protected]

OVERVIEWThe purpose of this literature

review is to familiarize the rehabil-itation specialist with the generalprinciples of nonsurgical manage-


2/11

ment and the short-term and long-term outcomes ofneuropathic (Charcot) arthropathies of the foot andankle. It is imperative that all physical therapists, par-ticularly those orthopaedic and sports physical thera-pists in states with direct patient access, be able torecognize the clinical presentation of acute Charcotfoot arthropathies and to begin early, protective im-mobilization. Early recognition and prompt immobi-lization may prevent or limit foot deformity and thesubsequent disability, including lower extremity am-putation. This review will summarize several perti-nent short-term and long-term outcomes from pub-lished studies related to the management of Charcotfoot arthropathies and discuss factors that appear toinfluence those outcomes. Finally, strategies for theidentification and prevention of acute Charcot ar-thropathies and the role that physical therapists mayhave in preventing long-term complications and dis-ability associated with these devastating neuropathiccomplications are discussed.What Is Neuropathic (Charcot) Arthropathy and WhatCauses It?

Neuropathic (osteo)arthropathy is an insidious,noninfective, chronic destruction of bones andjoints, resulting in pathological fractures, disloca-tions, or subluxations. There have been 2 dozen ormore chronic diseases in which neuropathic fracturesand joint changes have been described."." Histori-cally, the famous French neurologist Jean-MarieCharcot meticulously detailed the description of jointdestructions that occur in tabes dorsalis." His thor-ough clinical descriptions are the reason the diseaseis commonly referred to as Charcot arthropathy.

Since a report by Jordan in 1936, diabetes mellitushas been recognized as the major chronic disease inwhich Charcot foot and ankle arthropathies most of-ten o~cur.~%urrently,t is believed that Charcot ar-thropathy is a multisystem complication that is exac-erbated by both mechanical and metabolic stresses(ie, persistent or fluctuating hyperglycemia). Thebony and joint destruction has been commonly ass-ciated with peripheral neuropathy of both somaticand autonomic nerve axons. Peripheral neuropathyoccurs in nearly 50% of individuals with chronic dia-betes mellitus lasting 25 years or longer and is be-lieved to be the most critical and necessary factorthat contributes to the early onset of foot degenera-tion and subsequent def~rrnity.~"

The etiogenesis of acu tmnset Charcot arthropathyof the foot in diabetes mellitus remains poorly un-derstood. At least 2 different theories have been pro-posed.YThe "neurotraumatic" theory stated thatCharcot arthropathies are due to unperceived trau-ma or forceful injuries in an insensitive, anestheticfoot. This view held that patients with diabetes whoexhibit loss of protective sensation and propriocep

tion experience no symptoms when stress fracturesoccur, so they continue to use the fo~t.~"his resultsin progressive and permanent damage to the boneand joint structures. Secondary bacterial infectionmay follow, leading to further tissue destructionthrough cellulitis, abscess, or osteomyelitis.

A second mechanism, known as the "neurovascu-lar" theory, involves autonomic neuropathy that re-sults in the loss of sympathetic regulation of bloodflow to pedal bones.YThe loss of sympathetic vaso-constriction results in an enhanced pedal blood flowand causes a mismatch between bony resorption (anincreased osteoclastic activity) and bony synthesis (re-duced osteoblastic activity), leading to a "hyperemicdemineralization" and bone weaknessg Studies byCundy et all%nd Young et a14i suggest that the lossof bone density may predispose the foot to injuryand fractures. Hyperemic decalcification that resultsin osteopenia, particularly of the metatarsal bonesand forefoot, may be the earliest hallmark of disa-bling neuropathic arthropathies.'%ether there isunderlying predisposing bone disease present beforean onset of overt fracture, dislocation, or subluxationis still not entirely clear or uniformly accepted.

Currently, acute Charcot arthropathies are believedto be an amalgam of peripheral neuropathy (bothsomatic and autonomic), unperceived and repetitive(minor) trauma, and an underlying propensity forbone weakness. It now appears these are the 3 mostcommon factors that contribute to the onset of acuteCharcot foot arthropathies in diabetes mellitus.Clinical Presentation of Acute-Onset CharcotArthropathies

Acute Charcot arthropathies of the ankle or footalways show evidence of acute inflammation, that is,swelling, increased local skin temperature, and red-ness of the skin (Figure 1) .Many individuals com-plain of foot pain at their initial presentation, andsome retain pain sensation, although their pain re-sponses may be disproportionately low comparedwith their involvement. Usually there is instabilityand loss of joint function. Passive movement of all orparts of the foot may exhibit a "loose bag of bones,"since joint alignment and bony congruency may belost. Often the skin is intact without ulceration or ev-idence of bacterial infection. Plantar-surface ulcera-tion, particularly in the forefoot and midfoot, fre-quently occurs with acute arthropathies of the toes,metatarsus, and lesser tarsi (Sanders' patterns I and11). This often makes Charcot arthropathies difficultto diagnose and very often confused with osteomyeli-ti^.'^

Typically, the individual remains afebrile, althoughlocal dermal foot temperature elevations may exceed3-7C compared with the contralateral (uninvolved)extremity.' The use of handheld infrared thermome-

J Orthop Sports P hy Ther .Volume 29. Number 12. December 1999


3/11

FIGURE 1. An acute neuropathic arthropathy of the right ankle in a 34-year-old woman w ith type 2 diabetes mellitus wh o sustained a m ild twist-ing injury. She was examined in an emergency department, where radio-graphs were negative for evidence of ankle fracture. Several weeks laterher ankle remained painful and swollen, prompting further medical careand subsequent imm obilization in a total-contact cast. She has "protectivesensation" of both feet.

try for measuring local dermal temperatures and as-sessing the presence and healing progression ofacute Charcot arthropathies has aided the identifica-tion of patients at risk for these complications.'

Unperceived trauma has been implicated as a com-mon clinical feature, although most individuals donot recall a specific major trauma that precipitatedthe inflammatory arthr~ pathy.~ ." n thorough ques-tioning, some individuals may admit to minor trauma(a mild sprain or bruise) or to wearing too flexiblefootwear, which may have precipitated their acute on-set of symptoms.

Most individuals who develop an acute Charcotfoot arthropathy have had diabetes mellitus for 10-15 years or longer; therefore, they have some clinicalevidence of peripheral neuropathy, such as symptomsof foot numbness or tingling paresthesia. Severe sen-sory neuropathy of the foot is not always present noris it always uniform in its distribution, although somevariable loss of protective sensation and vibrationsense are common. Most individuals retain sometemperature awareness, kinesthetic sense, and jointproprioception of both feet.

The most common feature of first occurrence

Charcot arthropathies is the presence of an exuber-ant pedal blood flow. Foot pulses are almost invari-ably present and readily palpable, giving the clinicalimpression of an enhanced pedal flow. The feet aretypically warm to touch, and the veins may appeardistended or engorged. Arteriovenous shunting anda dysregulation of vasoconstrictive responses to evenminor trauma occur in acute Charcot arthropathiesand are an indication of the presence of autonomic(sympathetic) dy sf ~n cti on. ~V heoss of normal footsweating and hair growth on the dorsal distal aspectsof the toes are further clinical signs of autonomicneuropathy.

Like stress fractures of the foot, early radiographicevidence of fracture or subluxation may not be evi-dent. However, repeat radiographs or bone scans of-ten confirm the clinical presentation and help ruleout underlying bone infection, if there is ulcerationpresent. Most commonly, there is unilateral involve-ment, although in some cases where individuals werejumping from heights, both feet may be involved.Acute Charcot arthropathy appears to affect bothmen and women equally, and ethnic groups appearequally susceptible to this c~m pl ic at io n. '~ .~ ~ighplantar pressures, peripheral neuropathy, and func-tional equinus deformities (ie, limited dorsiflexionrange of motion) due to Achilles tendon shortnessappear to be strong clinical factors associated with atleast some patterns of Charcot ar th ropa thi e~ .~

Patterns of Acute Charcot Foot and AnkleArthropathiesSeveral investigators have characterized the loca-

tions or patterns where neuropathic fractures, sub-luxation~, nd dislocations most commonly occur inthe f o ~ t . " ~ " . ~ ~ . ~ ~arris and Brand were the first to de-scribe 5 distinct patterns of foot destruction that oc-cur in Hansen di~ease .~"heir patterns were basedlargely on the habitual posture of the foot, structuralintegrity of the bones, and the weight-bearing forcesencountered throughout the foot. Pattern 1 involvedthe posterior pillar and calcaneus of the hindfoot.Pattern 2 involved the body of the talus and the dis-integration pattern frequently observed on the un-dersurface of the talus primarily due to subtalar in-congruence. Pattern 3 involved the anterior pillar-medial arch, often the navicular or head of the talus.With this pattern, the medial longitudinal arch tendsto progressively collapse and flatten, causing the skinto ulcerate in the medial midfoot. Pattern 4 involvedthe anterior pillar-lateral arch, particularly the cuboidand base of the fifth metatarsal bone. Harris andBrand point out that ulceration and sepsis usually ac-company this pattern, and the severe varus positionof the foot may be more common in leprosy and lessso in diabetes mellitus. Pattern 5 involves the cunei-

1 Orthoo S m ru Phvs Ther-Vo lume 29. Number 12. December 1999


4/11

form and metatarsal bases, which is uncommon inleprosy but prevalent in diabetes.'O

Brodsky and Rouse described 3 anatomic classifica-tions of Charcot joints (types 1, 2, and 3a and b).8Type 1 primarily involves the midfoot, that is, the tar-sometatarsal or naviculocuneiform joints. This typeof involvement frequently leads to symptomatic bonyprominence and midfoot ulcerations. Type 2 is thehindfoot type, primarily involving the subtalar jointcomplex, including the talonavicular, calcaneocu-boid, or talocalcaneal joints. Type 3a involves the an-kle, that is, the tibiotalar joint, whereas 3b is a patho-logical fracture of the tubercle of the os calci~.~hisclassification scheme appears to be most often usedby orthopaedic surgeons with a foot and ankle sub-specialty.

Sanders and Frykberg characterized 5 patternsmost often seen in the feet of individuals with diabe-tes me ll i tu~ .~~attern I involved the forefoot, includ-ing the interphalangeal joints, phalanges, metatarso-phalangeal joint, or distal metatarsal bones. This pat-tern occurs in 2 647% of all the affected sites and isroutinely accompanied by plantar ul c e r a t i ~ n . ~ ~c+field et a1 reported that 91% of their patients withradiographic evidence of metatarsophalangeal ointinvolvement had concomitant ulcers.1sPattern I1 af-fects the tarsometatarsal (Lisfranc) joints and occursin 1543% of the arthropathy sites.JgA most com-mon site of pattern I1 is the second metatarsal baseand middle cuneiform articulation, which may causeprogressive collapse of the longitudinal and trans-verse arches of the foot. Pattern 111 involves the navi-culocuneiform, talonavicular, or calcaneocuboid ar-ticulations, which occur in 32% of the foot site^.^"Midtarsal joint dislocations and medial midfoot ulcer-ations below the navicular and first cuneiform bonesresult when structural stability of the medial longitu-dinal arch is lost. Sanders' pattern IV occurs at theankle joint, that is, the tibiotalar or tibiofibulotalararticulations. Although occurring in only 3-10% ofall sites, this pattern results in the most severe struc-tural deformity with functional instability. This pat-tern may be associated with very minor trauma (amild ankle sprain) yet progress rapidly to a severelydisintegrating ankle joint that cannot properly bearweight, therefore requiring fusion or amputation.Pattern V involve fractures of the posterior third ofthe calcaneus or an avulsion type of the posterior tu-bercle of the calcaneus. A tight, short Achilles ten-don coupled with weakening of the calcaneal trabec-ulae may precipitate this pattern of involvement. Al-though these patterns are described individually,their occurrence may not be mutually exclusive,since more than one type or location may occur si-multaneously in one o r both feet.

Based on their clinical experiences with 9000 indi-viduals with diabetes mellitus seen during a 7-year in-terval, Schon et a1M337 lassified the 4 most common

midfoot sites of involvement and staged the midfootinvolvement into 3 different levels based on the se-verity of the arch collapse. In their experience, themore proximal the involvement and the more severethe collapse of the medial or lateral longitudinalarches, the worse the outcome when treated eitherby nonsurgical (immobilization) or surgical interven-tions.%eJ7The value of describing the specific patternand locations of involvement is being able to esti-mate the length of time required for immobilization,to predict the eventual outcome, and to judge the ef-fectiveness of new therapies compared with conven-tional treatment regimens. Sinacore reported thathealing times of acute Charcot arthropathies of theforefoot healed faster than midfoot, hindfoot, or an-kle arthropathie~.~~n addition, the pattern of ar-thropathy determines the amount of weight-bearingrelief necessary to achieve healing and the predilec-tion for ul ce ra t i ~n .~ ~hysical therapists should famil-iarize themselves with the arthropathy patterns andneuropathic fracture locations, because treatmentmethod, severity of involvement, and short-term heal-ing outcomes vary for each pattern.J.JgNonsurgical Management of Acute-Onset CharcotArthropathies

The basic principles of nonsurgical managementinclude early recognition, immediate and adequateimmobilization with protected weight-bearing, andvigilant use of therapeutic footwear to prevent orlimit permanent foot deformity, subsequent disability,and potential lower extremity a m p u t a t i ~ n . ~ ~

The choice of management strategies of acute-on-set foot arthropathies or fractures depends mainly on2 factors, namely, the stage of development and thelocation (pattern) of fracture or arthr~pathy.~'i-chenholtz described 3 temporal stages throughoutthe fracture or joint healing pr oce ~s .' ~he first stageis the dissolution phase characterized by local signsof acute inflammation, including swelling, redness,increased local skin temperature, and loss of footfunction due to bony fracture, joint subluxation, ordislocation. The duration of this inflammatory stageis somewhat variable and depends on the speed withwhich immobilization is initiated and the effective-ness of the weight-bearing reduction. Typically, strictnon-weight-bearing of the involved foot is advisedthroughout this stage to limit the persistent inflam-matory responses, including hyperemia and bony de-struction. Several authorities suggest 6-10 weeks ofstrict non-weight-bearing is required to sufficientlyreduce the inflammatory p h a ~ e . ~ ~ , ~ '

The second Eichenholtz stage is the coalescence(quiescent or healing) phase and is characterized byclinically evident reductions in local swelling, red-ness, skin temperature, and the return of joint integ-rity.I9 The foot becomes more stable with passive

J Orthop Sports Phys Ther.Volum e 29. Number 12. December 1999 739


5/11

movements, and there is usually radiographic evi-dence of the consolidation of bony fragments. Pro-tection and continued immobilization are still re-quired so as not to refracture or relapse the inflam-matory processes. Some authors use braces adaptedto custom-fitted shoes or ankle-foot orthoses (AFOs)in lieu of plaster casts during this stage"; however,partial or protected weight-bearing using assistive de-vices should continually be advocated during ambula-tion until consecutive skin temperatures have re-turned to within 1C of the uninvolved foot's temper-a t ~ r e . ~ . ~ 'efining healing has historically been prob-lematic, because there has been no single universallyaccepted criterion; however, the definition is most of-ten based on clinical indicators of reduced signs ofinflammation and the lack of temperature differ-ences between feet.J The decision to terminate im-mobilization with weight-bearing restrictions also var-ies considerably among individual clinicians, but, ingeneral, adequate healing must be evident. Sinacoreshowed that healing times of acute Charcot arthropa-thies judged by 3 different orthopaedic physicians atthe same medical center did not differ significantlyfrom one another.Jy

The final stage (Eichenholtz stage 3) is the resolu-tion (remodeling) phase and is characterized by ra-diographic evidence of bony consolidation, reducedbony fragmentation, reduced or absent temperaturedifferences between feet, reduced swelling and red-ness, and improved bony stability.ly Bony deformityand joint subluxation may remain, although fewsigns of active inflammation remain in the local tis-sues. The normal bony architecture and alignmentmay be completely restored, although commonly itremains lost, leading to permanent, nonreduciblefoot deformity, which may require custom-made or-thoses and footwear to be fabricated to prevent fu-ture skin ulceration. If permanent bony malalign-ments result, particularly in the midfoot or hindfoot,the longitudinal arches may collapse and sublux tothe point where the foot begins to take on a "rock-er" appearance, leading to midfoot ulcerations over-lying bony prominence (Figure 2a through c). Al-though stable without evidence of acute inflamma-tion, the severely chronically deformed neuropathicfoot is often referred to as a Charcot foot or Charcotdeformity. Relapses of joint inflammation and pro-gressive bony destruction with new or recurrent plan-tar ulcerations make the neuro(osteo)arthropathicfoot at extreme risk for lower extremity amputat ion.Immobilization and Weight-Bearing Status

The traditional management recommendation foracute (nondisplaced) Charcot arthropathies of thefoot or ankle is strict non-weight-bearing cast immo-bilization for at least 3 month^.^ Most individualswith diabetes mellitus find it difficult, if not impossi-

ble, to fully comply with this recommendati~n.~~Many individuals have difficulty walking, which is at-tributable to their diabetes and sensory neuropathy."'These difficulties become more pronounced (if notimpossible) when immobilized in a cast and instruct-ed to be strictly non-weight-bearing on their affectedextremity. For that reason, a well-fitted totalcontactcast (TCC) and the use of assistive devices whenwalking have been recommended to enhance stabili-ty, reduce the chance of injurious falls, and improvehealing times by reducing weight-bearing forces onthe fracture site^.^'.^^.^' Schon et a1 observed that im-mobilization with serial contact casts and compliancewith restricted weight-bearing instructions, particular-ly in the early phases (Eichenholtz stage I and 11) forankle and hindfoot arthropathies, have improvedboth short-term and long-term outcomes in a largeseries of patients with diabetes and n eu r~ pa th y. ~"

Studies that report full weight-bearing while immo-bilized in plaster casts, braces, or AFOs appear tohave longer healing times than when strict non-weight-bearing or partial (protected) weight-bearingis required."-%'Adherence to partial weight-bearingusing an assistive device while walking in a TCC re-portedly reduces the healing time compared withnoncompliance with partial weight-bearing with anassistive device.Jy

The type of immobilization may affect the healingtimes and perhaps the long-term disability in Charcotarthropathies. The use of metal braces, custom-mold-ed AFOs, and commercial AFOs may prolong thehealing times compared with a well-fitted TCC. Total-contact casts have the advantages of quickly reducingswelling and inflammation and forcing compliance,since they are no t readily removed by the patient.41All braces and AFOs are removable and are not as ef-fective in controlling persistent edema as a TCC, par-ticularly in the acute (Eichenholtz stage 1) phase.These factors may account for the prolonged healingtimes observed with these alternative types of immo-bilization.kX' Additional studies are needed to deter-mine if TCC or alternative forms of immobilizationare equivalent in both short-term and long-term out-comes of acute Charcot arthropathies.Surgical Interventions

The goal of most operative interventions for Char-cot arthropathies is to achieve a stable, well-alignedfoot and ankle that does not ulcerate and can be fit-ted with protective footwear to allow for plantigradearnbulati~n.~'urgical interventions are indicated formalaligned, unstable, or nonreducible neuropathicfractures or dislocations when nonsurgical manage-ment by casting, bracing, or custom-made footwearfails to achieve or maintain foot integrity. In the ex-tensive series of feet in patients with diabetes andneuropathy reported by Schon et al, 85 (38.5%) of

740 J Orthop Sports Phys Ther Volume 29. Number 12 December 1999


6/11

FIGURE 2. A foot with chronic Charcot arthropathy in a 69-year-old woman with type 2 diabetes mellitus. These Charcot deformities are rigid, whichmade the fitting of custom-made, protective footwear a challenge for the rehabilitation specialist. a, Side view of left foot; note the complete collapseof the medial, lateral, and transverse arches. b, Radiograph of left foot showing bony "rocker" appearance. c, Front view of same foot; note multiple-plane deformities at the ankle, midfoot, hallux, and lesser toes.

221 required some type of surgery or reconstructionto prevent future complications or lower extremitya m p ~ t a t i o n . ~

Limbsalvage surgical techniques and proceduresappear to vary by institution, geographical region,and anatomic location of the arthropathy. Surgicalstabilization procedures vary from screw fixation toarthrodesis, osteotomies, and reconstructive tech-niques to partial or complete foot amputation. Totaljoint arthroplasty is rarely advised or performed formidfoot and hindfoot involvements, since most oftenthe bone stock is compromised and inflammatory re-currences are common.

Although it is not the intent of this article to re-view the outcomes related to each type of surgical in-tervention, some general conclusions may be drawnfrom published reports2 Surgical stabilization, wheth-er by internal fixation o r bony arthrodesis, is typicallyadvocated for displaced fractures that cannot bemanaged by closed reduction o r after failed nonsurg-ical management by casting and imm~bil izat ion.~'t

is rarely performed during the acute inflammatoryphase (Eichenholtz stage 1 or 2) because of the highproportion of failures, pseudoarthrosis, delayed un-ions, and nonunions. Displaced fractures of the an-kle or hindfoot or fractures with dislocations that re-quire surgical fixation usually will result in prolongedperiods of healing and immobilization ( 6 1 2 monthsof protected ~eight-bearing)~~.~~nd variableamounts of permanent deformity, which may have tobe managed with a combination of bracing, custom-fitted footwear, or permanent AFOs. Few studies thatdescribe the long-term disability ratings of individualswith chronic Charcot foot arthropathies have beenpublished; however, the prognosis for an individualwith severe neuropathic skeletal foot deformities ispoor, since they often lead to mobility limitationsand lower extremity amputation. The progressive lossof protective sensation and pain (anesthesia and an-algesia) and fixed foot deformities that result in theloss of structural architecture pose a formidable chal-lenge for orthopaedic physicians. Similarly, such

J Onhop Sports Phys There Volume 29. Number 12 December 1999 74 1


7/11

TABLE 1. Patient characteristics and interventions of selected studies on management of acute Charcot arthropathies since 1985.Length of Interven tion typeDiabetesNo. of Age Mel litus Sanders' pattern Imm obil izat ion SurgerySource, year patients (years) (years) (No. of each pattern)* (No.) (No.)

Brindley and Cofield,' 1985 19 (of 42 feet) 49 20 1 and 11 (7), 111 and IV (91, V (3) 19 5Clohisy and Tho mp~on ,'~ 988 18 37 2 4 15 2 7 I It), IV (51, V (7) 18 0Lesko and Ma ~rer: ~ 989 4 50 ?1 2 10 ? 7 111(4) 3 1Stuart and M ~ r r e y , ~990 13 48 NR IV( 13) After surgery 13Pappa et al,l2 1993 29 57 2 10 NR Ill (81, IV (81, V (18) After surgery 29Thompson and Clohi~y, '~ 993 14 (of 15 feet) 41 18 11 (11 Ill (6), IV (71, V (1) After surgery 15Biehl et 1993 2( of 4f ee t) 5 4 2 1 1 0 2 1 3 V ( 4 ) 4 1Morgan et aI,l0 1993 18 54 19 11 (71, Ill (71, IV (I ), V (3) 8 10Myenon et al," 1994 68 (of 89 feet) 54 18 IlAl l (all) 80 42Holmes and 1994 18 55 NR 1 (21, 111 (81, IV (61, V (4) 14 4Boninger and Leonard: 1996 6 52 NR NR 6 0Armstrong et al,' 1997 55 59 2 8 16 2 6 I 21, I1(261, Ill (19), IV (71, V (1) 55 14Sinacore,j9 1998 30 (of 35 feet) 55 2 9 21 2 12 1 (71, llnll (16), IV (81, v (4) 35 0Connolly and Csencsitz,14 1998 5 4 8 2 1 2 >15 V(5) 5 4Young461999 2 5 0 2 1 1 2 1 2 1 3 I V ( l ) , V ( l ) 2 1NR indicates not reported; Sanders' pattern I, forefoot, metatarsal bones, metatarsal phalangeal joints, phalanges, or interphalangeal joints; Sanders'pattern II, tarsometatarsal joints; Sanders' pattern Ill, naviculocuneiform, talonavicular, or calcaneocuboid oints; Sanders' pattern IV, ankle joint; andSanders' pattern V calcaneus.* Number of each pattern is the number of feet or fractures reported by each author.

complications make fitting therapeutic footwear a se-rious challenge for physical therapists and other spe-cialists involved in the care and rehabilitation ofthese individuals.Short-Term and Long-Term Outcomes of Interventions

There have been several reports that have suggest-ed that fracture healing among individuals with dia-betes is prolonged compared with the same types offractures that occur in individuals without diabe-

These clinical reports should be interpretedcautiously, since there is rarely adequate control ofall the experimental factors to make these compari-sons valid and because animal models of fracturehealing sometimes fail to support this often-held clin-ical suspicion.15 Similarly, not all foot fractures thatoccur in individuals with diabetes mellitus result indestructive neuropathic arthropathies and perma-nent di~ability.~'everal factors appear to influencethe short-term outcome (ie, healing rate) and thelong-term outcomes (ie, deformity, return to ambula-tion, disability).

The extent of the injury and the pattern are 2 fac-tors that influence both short-term and long-termoutcomes. If fractures are displaced (with disloca-tion), particularly of the hindfoot or ankle joints (in-volving talus, calcaneus, or tibia), there is often theneed for prolonged healing times, which frequentlyrequire immediate open reduction internal fixationor surgical fusion using arthrode~is.~~ractures withdislocations of the ankle treated by open reductioninternal fixation take 87% longer to heal and re-quire more vigilant compliance with strict non-weight-bearing in neuropathic individuals with diabe-

tes compared with similar types of fractures in s u bjects without diabetes.29

Another factor is the delay in seeking treatment orthe time interval before beginning protective immo-bilization. These factors appear to influence both theshort-term (healing times)3y nd long-term (disabilityand fixed foot deformity) outcomes." Sinacore o bserved a significantly shorter time to healing in aTCC in patients who sought protective immobiliza-tion in less than 14 days after the onset of theirsymptoms compared with patients who sought treat-ment after 14 days>9whereas Holmes and Hill foundmore severe Charcot joint changes in a small groupof subjects with diabetes ( n = 4) who had symptomsfor 59 days before seeking treatment.*' Since a largenumber of individuals frequently report only minorinjuries and are often unable to recall the specificonset of symptoms, there is limited information avail-able that documents whether delays in immobiliza-tion result in more permanent foot deformity, great-er disability, more frequent recurrences, or a greaterrisk for lower extremity amputation. The impact ofdelaying treatment on long-term outcomes needsmore thorough study.

METHODSWe summarize the outcomes of 15 published re-

ports from 1985-1999 located using the MEDLINEdatabase from 1966-present. Studies were selectedand included if the authors reported on (1) 2 ormore patients with diabetes mellitus and acute Char-cot arthropathies; (2) the short-term or long-termoutcomes, including the length of follow-up; and (3)

74 2 J Orthop Sports Phys Ther-Volume 29 . Number 12. December 1999


8/11

TABLE 2. Short-term and long-term outcomes in selected studies on management of acute Charcot arthropathies.Outcome*

Mean No. ofHealed months to healing Failed Nonunion Amputation Deaths Follow-upSource, year (% I (range) (% ) (%) (No.) (No.) (years) Orthosis or assistive deviceBrindley and Cofield,: 1985 84 3 (1-6) 16 5 0 0 3 3 required crutchesClohisy and Thompson," 100 N R NR NR 3 0 5 4 not walking 4 unilatera l AFO;1988 14 bilateral AFOLesko and Ma~rer,'~989 100 2-8 0 NR 0 0 2.6 1 required braceStuart and M ~ r r e y , ~990 85 4 (2-8) 15 15 3 1 3.5 9 walked with limpPappa et aL3' 1993 NR 7 (4-14) 31 34 1 1 3.5 2 used cane to walk;26 needed AFOThompson and Cl ~ h is y ,~ ~ 64 3 (9 for fusion) NR 40 2 5 2.9 11 required PTB orthosis1993Biehl et al,l 1993 100 3-1 8 50 0 0 0 1 1 needed walker, 2 used AFOMorgan et 1993 100 6 for cast, 12 for CROW 6 NR 1 0 1 NRMyerson et al," 1994 92 5 (4-9) 3 3 7 4 2 4 NRHolmes and Hill ," 1994 100 N R NR NR 2 0 2.25 NRBoninger and Le~nard,~ 50 15 (7-28) for CROW 17 NR 0 0 2-4 3 used AFO1996Armstrong et al,] 1997 100 4.3 2 2 0 0 1 NRSina~ore ,~~998 100 3 0 0 0 0 0.08 2 needed PTB; 2 needed AFOConnol ly and Csencsitz,I4 80 N R 20 0 0 0 1-1 0 2 needed AFO1998Young41999 50 3 50 0 1 0 0.25 1 BK prosthesisNR indicates not reported; CROW, Charcot restraint orthotic walker; AFO, ankle-foot orthosis; PTB, pate llar-tendon-bearing and BK, below-knee.* Percentages calcuiated based on the number of feet reported.

the pattern or location of the arthropathy. Studies se-lected are summarized in Tables 1 and 2.RESULTS

Patients in the studies reviewed (N = 301) aver-aged 50 ? 7 years of age and had diabetes mellitusfor an average of 17 ? 4 years (Table 1). The loca-tion of arthropathies, described in the table, wereclassified according to the Sanders' patterns? al-though some authors may have used other systems.Descriptions of patterns in the studies selected repre-sented all locations in the foot and ankle. Several re-ports stated outcomes based on the number of feetinvolved or the number of fracture sites when multi-ple locations were o b s e r ~ e d . ~ ~ ~ ' . ~ ~ . ~ ~e classified in-tervention types into nonsurgical immobilization andsurgical interventions (all types). We summarize theshort-term outcomes, including the percentage of ar-thropathies healed, average time to healing, the per-centage of arthropathies that failed to heal, and thepercentage of nonunions when reported. The long-term outcomes were the number of amputations,deaths, and individuals who required an orthosis,brace, or assistive device to ambulate at the follow-upperiod. Individuals who reportedly required aids forambulation were assumed to have a mobility limita-tion. Those patients with displaced, unstable frac-tures of patterns 111, IV, or V and treated with surgi-cal interventions were immobilized for a variable pe-riod before surgical stabilization. The healing timeslisted in Table 2 do not always reflect these addition-

al immobilization periods, since they were not uni-formly reported.

In the studies reviewed, healing was achieved in86% (range, 50-100%) of the feet (304 of 353 feetreported) with acute Charcot arthropathies. The av-erage healing time for nonoperative immobilizationwas 5.7 ? 3.5 months, with a range of 3-18 months.Based on their clinical experiences, several authori-t i e s ~ ~ . ~ ~dvocate immobilization periods nearly dou-ble the healing times observed for similar types offractures in individuals without diabetes mellitus. Thehealing times summarized herein confirm these clini-cal experiences and give the rehabilitation specialista time frame when most individuals will typicallyachieve satisfactory healing. Not surprising, neuro-pathic fractures involving midfoot, hindfoot, or theankle (patterns 111, IV, or V) take longer to heal thanforefoot fractures because of the greater weight-bear-ing loads required in these parts of the foot.:iY

Studies that report immobilization after surgicalstabilization take 6-18 months to heal, with an aver-age of 8.8 2 12.1 month^.^*.^'-^^^.^^ andother^^'.^' suggest a thorough period of strict non-weight-bearing immobilization, followed by a gradualreintroduction to partial then more full weight-bear-ing over several months to ensure complete fracturehealing with bony ~tab ili ty.~"~ '.~'

Failed nonsurgical management occurs in approxi-mately 53 (21%) of 244 feet reviewed. Stable non-unions and pseudoarthrosis were a complication in23 (10.5%) of 220 feet that were reported failures,although improved limb salvage procedures and de-

J Orthop Sports Phys Ther.Volume 29. Number 12. December 1999


9/11

laying surgery until the chronic Eichenholtz stagehave dramatically reduced these complicating out-c o m e ~ . ~ ~ . ~ ~

Although this question has not been addressed ad-equately in controlled clinical trials, studies that re-port immobilization using serial TCC3.3'~J9ppear tohave a faster healing time than methods using cus-tom-made A F O S .~ .~ell-fitted TCCs are particularlyeffective for nondisplaced neuropathic fractures withplantar ulceration of the forefoot or m i d f o ~ t . " ~ ~ . ~ ~Walking TCCs may not be as effective for fractures ofthe ankle or fractures with ulceration in the hindfootbecause of the difficulty in spreading plantar heelpressures using casting. For ankle or hindfoot frac-tures, patellar-tendon-bearing AFOs have been advo-cated to relieve weight-bearing forces and promotehealing.2J.45

It has been difficult to assess the contribution ofstrict non-weight-bearing versus partial weight-bear-ing or full weight-bearing on the short-term andlong-term outcomes from published studies. Al-though most authorities recommend a period ofstrict non-weight-bearir~g~*~~~~he compliance tothese recommendations is reported to be poor.J7s3gSinacore observed healing times of acute Charcot ar-thropathies were shorter when subjects were judgedto be adherent to partial weight-bearing instructionsusing an assistive device.39The effect of weight-bear-ing status during immobilization on longer-term out-comes (ie, greater than 6 months) needs to be inves-tigated to help clinicians provide a more thoroughand realistic recommendation to their patients.

Several observations regarding long-term outcomescan be made from the studies reviewed. Elevendeaths (3.65%) in 301 patients were reported withinthe average follow-up period of 2.5 years after treat-ment for Charcot arthropathy. Partial or completefoot amputation occurred in 20 (6.6%) of 301 su bjects reviewed during follow-up. Eighty-three (28%)of 301 patients reviewed required AFOs, permanentbracing, or assistive devices for ambulation a t thetime of follow-up (Table 2). Most of these outcomesoccurred in patients who required surgical interven-tions, particularly after displaced ankle or hindfootfractures (Table 2).

We could not adequately determine the frequencyor severity of residual foot deformity after operativeor nonoperative interventions, since most reports (11of 15) did not include this outcome. Presumably allpatients were neuropathic, and nearly all patientswere routinely prescribed therapeutic footwear aftertheir initial healing. Unfortunately, the impact oflong-term use of protective footwear in patients withneuropathic fixed foot deformities remains un-known. It can be concluded that chronic neuropath-ic arthropathy is a devastating and disabling long-term complication of diabetes mellitus that may re-sult in a poor long-term outcome.

DISCUSSIONIdentifying and Preventing Acute CharcotArthropathies

It may be possible for physical therapists and otherrehabilitation specialists to identify those individualswho are at high risk for fracture, joint dislocation, orsubluxation. Individuals with chronic diabetes melli-tus of 10 years or longer, whose blood sugar levelsare not well controlled and who have peripheralneuropathy, appear to be at highest risk. Individualswith chronic diabetes, high plantar pressures (ex-ceeding 65 N/cm2), and functional equinus deformi-ties (ie, limited passive range of motion of ankle dor-siflexion) appear to be particularly susceptible toforefoot and midfoot arthropathies with plantar ul-c e r a t i o n ~ . ~ . ~ ~

The loss of bone mass, particularly in the smallbones of the foot, may predispose the individual withchronic diabetes to neuropathic fracture, dislocation,or subluxation. Recently, individuals with acute Char-cot arthropathy were observed to have less bone den-sity in the lower extremities (but not in the lumbarspine) than neuropathic patients with no evidence ofCharcot foot ar th r~ pa th y. ~~urthermore, the bonedensity of the lower limb with the Charcot arthropa-thy was significantly less than the limb with no Char-cot ar thr~pa thy.~' standard radiographl%r bonedensity assessment via dualenergy X-ray absorptiome-try or broad-band ultrasound attenuation may helpdetect the early loss of critical bone mass beforeovert f r a ~ tu r e .~ .~ ~

Individuals with diabetes who wear inadequatefootwear (too flexible and not supportive) may be athigh risk for Charcot changes. A poor choice of foot-wear in an insensitive individual may contribute tohigh foot pressures and daily trauma, in turn causingprogressive bony and ligamentous destruction. S u pportive and protective footwear with soft, accommo-dating insoles and rigid soles could reduce excessiveplantar pressures, thereby preventing unperceivedtrauma to the foot during normal weight-bearing ac-tivities. Physical therapists, pedorthists, and shoe fit-ters should recommend and prescribe protectivefootwear that minimizes plantar pressures on theneuropathic foot.22

Physical therapists should also screen the individu-al with diabetes mellitus for the loss of protectivesensation and signs of autonomic (sympathetic) neu-ropathy to help identify those individuals most at riskf o r a r t h r ~ p a t h y . ~ ~ . ~nce identified, aggressive strate-gies that can reduce these risk factors should be in-stituted for these individuals. These strategies includeeducating the patient in the need for daily foot in-spections and wearing protective footwear at alltimes.

744 J Orthop Sports Phys Therevolume 29 Number 12 December 1999


10/11


11/11

december1999 lr sinacore[1]

Documents