Injury, Int. J. Care Injured 32 (2001) S-D-99–S-D-106

The use of trans-articular and extra-articular external fixation formanagement of distal tibial intra-articular fractures

M. El-Shazly∗, J. Dalby-Ball, M. Burton, M. SalehSection of Orthopaedics and Traumatology, Division of Clinical Sciences, Clinical Sciences Centre,

University of Sheffield, Northern General Hospital, Herries Road, Sheffield S5 7AU, UK

Accepted 24 August 2001

Abstract

Twenty-nine consecutive cases of distal tibial intra-articular fractures treated by trans-articular or extra-articular external fixation tech-niques have been reviewed. Eleven cases were treated initially with a trans-articular dynamic axial fixator. Of these seven were convertedto an extra-articular SHF, for a combination of poor ankle motion and delayed healing of the metaphyseo-diaphyseal dissociation (MDD).Three of these cases (two patients) required bone grafting for delayed healing of the diaphyseal component of the fracture. Apart fromone refracture through the MDD, no major complications were seen. No deep infections and no angular malalignments were noted. Therewere 11 pin track infections. Subjective assessment using short form-36 (SF-36) questionnaires, however, revealed significant differencescompared to a normal population particularly in physical function and pain at a mean follow-up of 21 months. Using Bone’s criteria for as-sessment of range of motion there were 62% excellent and good results, which dropped to 53% when fractures with a metaphyseo-diaphysealextension were included.

The use of minimally invasive techniques of internal fixation and stabilisation with a Sheffield hybrid frame in the management ofdistal tibial intra-articular fractures has minimal complications. Trans-articular external fixation is a good primary treatment for badlycomminuted articular fractures with poor soft tissue condition. Conversion to extra-articular external fixation is recommended for slowerhealing fractures allowing ankle movement and early weight-bearing. The presence of a MDD dissociation lengthens the treatment timesignificantly, adds to the morbidity and affects final outcome. © 2001 Elsevier Science Ltd. All rights reserved.

1. Introduction

The ‘traditional’ approach to the treatment of severe distaltibial intra-articular fractures including pilon fractures usingopen reduction and internal fixation techniques has graduallyfallen out of favour in the last decade [1]. The good resultsproduced by Ruedi and co-worker [2–4] and others havenot been consistently reproduced. A number of reports ofserious complications have followed [1,5–7].

The use of minimally invasive external fixation tech-niques has gained considerable support more recently, withfewer complications [8–11]. Previous reports describe thetrans-articular technique [12]. The Sheffield hybrid fixator(SHF) has been used since January 1995 in the treatment ofover 100 cases of complex trauma, non-unions and malu-nions [13]. This study aims to evaluate the early results oftreatment of distal tibial intra-articular fractures with theextra-articular approach using the SHF.

∗ Corresponding author. Present address: Droitwich Knee Clinic Limited,St. Andrews Road, Droitwich Spa, Worcestershire WR9 8YX, UK.Tel.: +44-1905-79485.E-mail address: mohi.el.shazly@kneeclinics.com (M. El-Shazly).

2. Material and methods

We reviewed retrospectively 31 consecutive cases of distaltibial intra-articular fractures treated by the Sheffield AdultLimb Reconstruction Service in the period from January1995 to September 1998. Of these, one had delayed primarytrans-tibial amputation for an associated Gustillo and An-derson type III-C open fracture of the tibial shaft. Anotherhad a primary arthrodesis using a trans-articular SHF, for aRuedi and co-workers [2–4] type III (AO type C3.2) fracturein a young manual labourer. Of the 29 remaining cases in27 patients, there were 20 males and 7 females with a meanage of 47 years (range: 26–68 years). Cases treated by bothtrans-articular and extra-articular techniques were included.

Nineteen patients (65.5%) had high-energy injuries, ofwhich nine had axial loading injuries as a result of a fall fromheight. The rest were either indirect low energy twistinginjuries or simple falls. Five were multiply injured and eight(28%) had open fractures. Twenty-six patients had a fullset of initial pre-operative X-rays and CT scans. Using theRuedi and co-workers classification [2–4], there were 6 typeI, 8 type II and 12 type III fractures.

0020-1383/01/$ – see front matter © 2001 Elsevier Science Ltd. All rights reserved.PII: S0020-1383(01)00159-0

S-D-100 M. El-Shazly et al. / Injury, Int. J. Care Injured 32 (2001) S-D-99–S-D-106

Eleven cases were treated initially with a trans-articulardynamic axial fixator (Orthofix Srl, Verona), with an articu-lated ankle distraction unit, allowing early motion in nine ofthese. Of these, seven were converted to an extra-articularSHF. In five of these, this was due to delayed healing ofthe associated metaphyseo-diaphyseal dissociation (MDD).In one case, it was due to poor ankle motion with the articu-lated unit. One 68-year-old lady was unable to comply withpartial weight-bearing and had repeated falls which eventu-ally lead to re-displacement of the MDD.

There were four late referrals for treatment. In these casessurgery was performed 2–10 weeks from injury. The rest ofthe patients had their surgery at a mean of 3.6 days frominjury (range: 0–9 days). The delay in surgery was mainlydue to poor skin condition and in one case due to lack ofallocated theatre time.

Two patients did not attend for follow-up. One was trans-ferred to a special unit with severe brain damage as a resultof associated head injury. The second patient who had sus-tained her injury from a suicide attempt by jumping from aheight, was in a special psychiatric unit. This left 27 cases in25 patients who were followed-up for a mean of 21 months(14–36 months).

At final follow-up, the patients were assessed subjectivelyusing the short form-36 (SF-36) questionnaire. Results werecompared to a normal Sheffield population in an age matchedfashion. The paired samplest-test was used for statisticalanalysis and a 95% confidence interval was designated asthe threshold for statistical significance. All patients were as-sessed objectively by a single observer (MB) using standardcriteria. Pain was recorded on a visual analogue scale (VAS)and range of motion was assessed using Bone’s criteria.

3. Surgical technique

Preoperative AP, lateral (and mortise views where avail-able) were reviewed and classified according to Ruedi andco-workers [2–4]. The level of the fibular fracture whenpresent was recorded according to Weber’s classification.Preoperative planning was done using standard view ra-diographs but mainly relying on CT scans. The axial cutsaround and immediately above the dome of the talus wereanalysed and reconstruction of the various fragments of thetibial pilon was planned accordingly.

Of the 18 fibular fractures, only 4 of the earlier cases in thisseries required internal fixation. Three patients had platingof their fibular fracture. The fourth patient had percutaneousinsertion of an intramedullary Bundle nail. In one case, openreduction was needed to assist with reduction of the fibularfracture and no internal fixation was applied.

Intra-articular elements of the fracture were attended tofirst. Joint distraction was applied either manually or usinga trans-articular fixator. The individual joint fragments werethen reduced percutaneously under image intensifier controland internal fixation performed using cannulated screws or

fragment fixation screws (Orthofix Srl, Verona). However,in nine cases, open reduction was performed through limitedincisions planned over fracture windows, through which thearticular fragments were reduced under direct vision. Thesecases were either too comminuted and deemed unsuitablefor percutaneous reduction or an attempt at percutaneousreduction had failed and, therefore, the surgeon immedi-ately proceeded to ORIF. In one delayed case in whichprevious internal fixation, performed elsewhere, had to berevised at 7 weeks form injury, a formal open reductionwas performed through a long conventional antero-medialexposure. Primary bone grafting of a metaphyseal defectwas performed in two cases.

If a trans-articular fixator was used then this was utilisedto reduce the reconstructed articular block to the metaph-ysis and locked in position. If a SHF was applied, 3–42 mm wires were attached to the distal ring of a pre-builttwo ring frame. These were then tensioned. The articularblock was then reduced to the metaphysis and two Orthofixcortical screws were used, attached to the proximal ringthrough the Sheffield clamp. A further cortical screw wasthen inserted orthogonal to the plane of the Sheffield clampand attached directly to the proximal ring. The patient wasgiven a removable foot support attached to the distal ringto prevent equinus deformity and active movement of theankle was encouraged. Partial weight-bearing was allowedafter 3 weeks progressing to full weight-bearing at 6 weeks.Fixator removal was determined according to the radiolo-gical appearances and performed in the out-patients’ clinicwithout anaesthesia.

4. Results

4.1. Methods

At final follow-up the patients were asked to fill out anSF-36 questionnaire [14].

The SF-36 results were compared to scores obtainedfrom a normal population who had filled out similar ques-tionnaires. The patients were divided into six categoriesaccording to age. Significant differences (P < 0.05) werefound between the two age matched populations in physicalfunction, social function, pain and to a lesser extent healthperception (Fig. 1). The difference was non-significant formental and energy levels.

Twenty patients marked their pain levels on a VAS. Themean of the VAS score recorded was 2.55, with a (range:0–7.2). Four patients had no pain at final follow-up.

Two patients had an equinus deformity. One required apercutaneous lengthening of his Tendo Achilles. Overall pa-tients had a mean dorsiflexion of 5◦ (range:−5–15◦), a meanplantar flexion of 33◦ (range 15–60◦). Using Bone et al. cri-teria [9], there were 53% good and excellent results (Fig. 2a).

All fractures went on to eventual union. However, threecases in two patients required bone grafting of the MDD

M. El-Shazly et al. / Injury, Int. J. Care Injured 32 (2001) S-D-99–S-D-106 S-D-101

Fig. 1. SF-36 analysis compared to a normal population. Differencessignificant at 95% confidence interval except for mental scores and energylevels (P > 0.05): iphys, physical functioning; isocial, social functioning;irolep, role limitations due to physical problems; irolem, role limitationsdue to emotional problems; imental, mental health; ienergy, energy/vitality;ipain, bodily pain; ighp, general health perceptions.

before progressing to full union in 14, 18 and 21 months,respectively. Overall treatment time, calculated as time toremoval of fixator and unprotected weight-bearing, rangedfrom 3.5 to 21.5 months (mean 9.4 months). However, frac-tures without diaphyseal elements were found to have a meantreatment time of 4.5 months (Fig. 3). Low energy fractureswith a diaphyseal extension had longer healing times thanhigh-energy metaphyseal fractures, using Muller’s defini-tion of the metaphyseo-diaphyseal junction [15]. Excludingfractures with diaphyseal extension, the percentage of goodand excellent results rose from 53 to 61% (Fig. 2b). Patientswith good or excellent range of motion using Bone’s crite-ria, had a mean treatment time of 10.4 months (range 5–18

Fig. 4. Box and Whiskers chart of cases scoring good and excellent range of motion index [9] (satisfactory) and fair and poor (unsatisfactory) plottedagainst treatment time in months. Difference non-significant (P = 0.06).

Fig. 2. Range of motion index Bone et al. [9]: (a) whole population; (b)cases with MDD excluded.

Fig. 3. Effect of MDD and mechanism of injury on length of treat-ment in months: HE, high-energy injury; LE, low energy injury; MDD,metaphyseo-diaphyseal dissociation.

months), compared with 19.1 months (range 7–40 months)in the fair to poor range of motion group. This effect wastested using the independent samplest-test and found to bestatistically insignificant (P = 0.06) (Fig. 4).

S-D-102 M. El-Shazly et al. / Injury, Int. J. Care Injured 32 (2001) S-D-99–S-D-106

One case with a grade II open fracture, sustained a com-partment syndrome which was treated by fasciotomy. Thispatient recovered but had residual grade IV weakness of herextensor hallucis longus. Apart from this and the two equinus

Fig. 5. a) RA III fracture, b) treated by minimal internal fixation and a trans-articular fixator, c) delayed healing of MDD, d) immediate full weightbearing, e) clinical and radiological healing at 8 months (14 months from injury); f) refracture 6 months later; monolateral dynamic axial fixator andearly partial weight-bearing; g) healing in 4 months.

deformities referred to above, one patient had a refracturethrough his MDD, while playing cricket 6 months after fullhealing (Fig. 5). He was treated using a monolateral DAF andwent on to full union in anatomical position without incident.

M. El-Shazly et al. / Injury, Int. J. Care Injured 32 (2001) S-D-99–S-D-106 S-D-103

Fig. 5 (Continued ).

S-D-104 M. El-Shazly et al. / Injury, Int. J. Care Injured 32 (2001) S-D-99–S-D-106

In one case, a poor anatomical reduction of the articu-lar surface was noted. There was one case with early OAchanges on follow-up X-ray with narrowing of the jointspace. No cases of angular malalignment of more than 5◦ ineither plane were noted in this series.

There were 11 cases of pin-site infection, which weretreated with simple pin-site care or oral antibiotics. One caserequired exploration and washout of a pin-site followingremoval of the fixator. She recovered fully with no clinicalor radiological evidence of residual deep infection.

5. Discussion

Early results of internal fixation for tibial pilon fractureswere disappointing. It was not until Ruedi and co-workerpublished their first studies in 1969 and 1973 [3,4] thatinternal fixation gradually returned to favour with a smallnumber of other studies reporting similar results. Theyreported 74–85% good and excellent results. These figuressubsequently became the gold standard that other studieshave attempted to reproduce. However, it must be said thatthe majority of their cases resulted from relatively low en-ergy skiing injuries and that only 25% of their populationhad high-energy injuries. They later published further workin a new series in 1979 [2] in which 50% of their patientshad high-energy injuries and had 80% good and excellentoverall results. However, the results of their Ruedi andAllgower type III cases were significantly worse with only52% good and excellent cases. Subsequent studies usinginternal fixation reported a disastrous number of majorcomplications [1,6,7] mainly related to soft tissue break-down and deep infection. In our study 65.5% of cases weredue to high-energy injury. Almost half of these were due tofalls from a height. There were no cases of deep infectionat final follow-up in this series.

A number of studies have reported good results usingtrans-articular techniques of external fixation [8,9,12]. In thisstudy, due to the high incidence of metaphyseo-diaphysealinvolvement, only four of the eleven cases treated initiallywith a trans-articular fixator finished their treatment withoutexchange. The other seven had to have the fixator replacedby a SHF due to slow healing of the MDD, difficultyweight-bearing and in one case poor ankle motion usingthe articulated ankle unit.

Trans-articular techniques have several advantagesparticularly in the more comminuted RA III cases. Theycan be applied in a relatively short period and are, therefore,useful for the multiply injured patient. They also facilitatearticular reconstruction by ligamentotaxis. Most impor-tantly they provide a good primary treatment, with goodstability, especially for patients with poor skin condition,allowing time for proper assessment of the patient and frac-ture, including planning for articular reconstruction. Theycan always be converted to an extra-articular circular fix-ator later if necessary. Trans-articular fixators are perfectly

suited for severely comminuted articular fractures with nometaphyseal or diaphyseal extension.

However, when there is an associated MDD element, wehave demonstrated in this study that the fracture is muchslower in healing. As loosening of the distal talar screwsin trans-articular fixation is common after 4 months [16],circular external fixation using the SHF which allows imme-diate ankle motion and early weight-bearing is more suitedfor this purpose. The SHF is also ideally suited for these frac-tures since it demonstrates beam loading and self-stiffeningunder load, similar to the Ilizarov all wire fixator [17], butwithout diaphyseal transfixion, thus, encouraging high lev-els of activity [18]. Exchange from a trans-articular fixatorto the SHF is relatively simple and does not necessitateremoval of the cortical screws from the diaphyses. They aresimply attached to the Sheffield clamp on the proximal ringof a pre-built frame and wires are then placed in the distaltibia 3–5 mm from the articular surface, attached to the distalring.

The other major contribution made by Ruedi andco-workers [2–4] is that they laid out clear principles andsteps for reconstruction of these fractures, starting withreconstruction of the fibular fracture. The advantage of thisis that it tends to pull the tibia out to near its normal length,thus, facilitating the next step of articular reconstruction.While many authors using external fixation techniquesfor tibial pilon fractures continue to advocate this prac-tice [9,11,19,20], others have deemed this step not onlyunnecessary but also potentially harmful [21–23]. It in-creases the potential morbidity with added skin incisionsand contradicts the principles of non-rigid fixation allowingmicromotion between the main fragments. In our opinion,if we have rigid fixation as with a plate on one side anda fixator allowing dynamic axial motion on the oppositeside the fracture can only collapse into varus, reducing theisotropic properties of the system. In this series, only threeof the earlier patients required open reduction of which twohad plating. Another patient had an intramedullary Bundlenail inserted percutaneously.

On analysis of the SF-36 questionnaires, Sands et al. [24]found significant differences in general health perception,physical function, emotional role function, pain and energylevels when compared with age matched population dataregardless of method of treatment. Similarly, in this seriesmost of these categories listed above were significantly im-paired. This was particularly true for physical function andto a lesser extent pain. The difference was non-significantfor mental status and energy levels. Only four patients weretotally pain free. Marsh stated that excellent long term painfree movement is not predictably achieved in these caseswith any treatment strategy [22].

Limitation of motion is a predictable problem withintra-articular fractures of the distal tibia and is proportion-ate to the severity of damage to the articular surface. Inthis series, however, it seems that due to aggressive reha-bilitation, including immediate movement, whether using a

M. El-Shazly et al. / Injury, Int. J. Care Injured 32 (2001) S-D-99–S-D-106 S-D-105

trans-articular fixator with an articulated ankle unit or anextra-articular SHF, our results in terms of range of motioncompare favourably with the literature. Griffiths et al. [11]used limited internal fixation and hybrid external fixators totreat 16 intra-articular fractures. At 10 months follow-up,there were 50% good and excellent results using Bone’srange of motion index. In our series this figure was 53%using the same classification system, but rose to 62% whenfractures with MDD elements were excluded, for compari-son with the above mentioned study.

Thus, cases with fractures mainly affecting the meta-physis and having less intra-articular comminution fairedworse in terms of range of motion despite equally aggres-sive early post-operative motion. We attribute this to thelengthy treatment times associated with these cases.

Barbieri et al. [25] used hybrid fixators to treat 37 distaltibial fractures including 17 with articular involvement.Three patients (9%) had loss of reduction necessitatingrevision of the frame. At a mean follow-up of 15 months,they reported three non-unions (9%). In our series, therewere three cases requiring bone grafting of their MDDprior to progressing to union at 14–21 months. Strictlyspeaking these could be considered diaphyseal non-unionsalthough their treatment was completed with the originalSHF in place. These cases have significantly affected themean healing time in this series. Olive wires were not usedin any of these cases to stabilise the MDD. Recent exper-imental and early clinical evidence from Sheffield suggestthat the use of olive wires improves the mechanical stabilityin oblique diaphyseal fractures and shortens treatment time[26]. There were no cases of postoperative loss of reductionor late malunions in this series.

Despite limitation of motion and persistent pain in asignificant number of cases, there were only two majorcomplications. One compartment syndrome was treated sur-gically with minimal residual dysfunction. The only othermajor complication was a refracture through the MDD 6months following the end of treatment.

Barbieri et al. [25] reported two cases of deep infection(5.8%). Similarly, Griffiths and Thordarson [11] had twocases (12%) and Tornetta et al. [10] had one case of deepinfection (3.8%) and only one superficial pin track infec-tion. Our study had 11 pin track infections, of which onerequired exploration and washout in theatre. There were nodeep infections. This compares favourably with the abovestudies and poses a significant improvement if compared tocomplications following internal fixation.

Also in this study, apart from one poor anatomicalreduction of the articular element, there were no cases ofaxial malalignment. This again compares favourably withthe literature, where cases of malunion were reported inup to 25% of cases [27]. While in some cases the articularsurface cannot be anatomically restored, intra-operative andpost-operative vigilance is essential to restore the properaxial alignment, using modern external fixator techniquesand a versatile and stable fixator.

Not only is the quality of the articular reconstructionimportant, but also accurate reduction of the MDD andrestoration of overall alignment.

References

[1] Teeny SM, Wiss DA. Open reduction and internal fixation oftibial plafond fractures: variable contributing to poor results andcomplications. Clin Orthopaed 1993;292:108–17.

[2] Ruedi TP, Allgower M. The operative treatment of intra-articularfractures of the lower end of the tibia. Clin Orthopaed 1979;138:105–10.

[3] Ruedi T. Fractures of the lower end of the tibia into the ankle joint:results 9 years after open reduction and internal fixation. Injury1973;5:130.

[4] Ruedi TP, Allgower M. Fractures of the lower end of the tibia intothe ankle joint. Injury 1969;1:92.

[5] Bourne RB, Rorabeck CH, McNab J. Intra-articular fractures of thedistal tibia: the pilon fracture. J Trauma 1983;23:591–6.

[6] McFerran MA, Smith SW, Boulas HJ, et al. Complicationsencountered in the treatment of pilon fractures. J Orthopaed Trauma1992;6:195–200.

[7] Ovadia DN, Beals RK. Fractures of the tibial plafond. J Bone JointSurg 1986;68-A:543.

[8] Bonar S, Marsh JL. Unilateral external fixation for severe pilonfractures. Foot Ankle Int 1993;14(2):57–64.

[9] Bone LB, Stegemann P, McNamara K, et al. External fixation ofseverely comminuted and open tibial pilon fractures. Clin Orthopaed1993;292:101–7.

[10] Tornetta P, Weiner L, Bergman M, et al. Pilon fractures: treatmentwith combined internal and external fixation. J Orthopaed Trauma1993;7(6):489–96.

[11] Griffiths GP, Thordarson DB. Tibial plafond fractures: limited internalfixation and a hybrid external fixator. Foot Ankle Int 1996;17(8):444–8.

[12] Saleh M, Shanahan MDG, Fern E. Intra-articular fractures of thedistal tibia: surgical management by limited internal fixation andarticulated distraction. Injury 1993;24(1):37–40.

[13] Saleh M. The Sheffield hybrid fixator design: considerations andclinical experience. Orthopead Prod News 1998;5/6:33–5.

[14] Jenkinson C, Layte R, Wright L, et al. The UK SF-36: an analysisand interpretation manual. Oxford: University of Oxford, 1996.

[15] Muller ME, Nazarian S, Koch P, et al. The comprehensiveclassification of fractures of long bones. Berlin: Springer, 1990.

[16] Marsh JL. Distal tibial and pilon fractures (use of the articulated bodyfor the ankle). In: Orthofix operative technique manuals. Verona:Orthofix Srl, p. 13.

[17] Yang L, Saleh M. Fracture site motion with hybrid external fixators.J Bone Joint Surg 1999;81-B(Suppl):330.

[18] Saleh M, Yang L. Limb reconstruction after high-energy trauma. BritMed Bull Trauma 1999;55(4):870–84.

[19] Hutson JJ, Wiss DA. Small wire circular fixators in the managementof pylon fractures. In: Proceedings of the 64th Annual Meeting of theAmerican Academy of Orthopaedic Surgeons. Instructional courselectures. San Francisco: AAOS, 1997.

[20] Brumback RJ, McGarvey WC. Fractures of the tibial plafond:evolving treatment concepts for the pilon fracture. Orthoped ClinNorth Am 1995;26(2):273–85.

[21] Renzi-Brivio L. In: Proceedings of the 3rd Riva Congress. 2000.[22] Marsh JL. Large pin external fixators in the management of pylon

fractures. In: Proceedings of the American Academy of OrthopaedicSurgeons. Instructional course lectures, San Francisco: AAOS, 1997.

[23] Williams TM, Marsh JL, Nepola JV, et al. External fixation oftibial plafond fractures: is routine plating of the fibula necessary? JOrthopaed Trauma 1998;12(1):16–20.

S-D-106 M. El-Shazly et al. / Injury, Int. J. Care Injured 32 (2001) S-D-99–S-D-106

[24] Sands A, Grujic L, Byck DC, et al. Functional outcomes in tibialpilon fractures. Clin Orthopaed 1998;347:131–7.

[25] Barbieri R, Schenk R, Koval K, et al. Hybrid external fixation in thetreatment of tibial plafond fractures. Clin Orthopaed 1996;332:16–22.

[26] Shellbrooke K, Yang L, Ali F, et al. Additional stabilisation ofoblique tibial fractures treated by external fixation. In: Proceedings

of the 11th Annual Scientific Meeting of the Limb Lengthening andReconstruction Society. San Francisco, 2001.

[27] Court-Brown C, Walker C, Garg A, et al. Half ring external fixationin the management of tibial plafond fractures. J Orthopaed Trauma1999;13(3):200–6.