VOL. 83-B, NO. 1, JANUARY 2001 55

B. Squires, FRCS, Specialist RegistrarP. E. Allen, FRCS, Specialist RegistrarJ. Livingstone, FRCS, Specialist RegistrarR. M. Atkins, MA, DM, FRCS, Consultant Orthopaedic SurgeonBristol Royal Infirmary, Marlborough Street, Bristol BS2 8HW, UK.

Correspondence should be sent to Mr B. Squires at 15 Fremantle Road,Cotham, Bristol, BS6 5SY, UK.

©2001 British Editorial Society of Bone and Joint Surgery0301-620X/01/111184 $2.00

Fractures of the tuberosity of the calcaneusB. Squires, P. E. Allen, J. Livingstone, R. M. AtkinsFrom Bristol Royal Infirmary, England

We describe 24 fractures of the tuberosity of thecalcaneus in 22 patients. Three were similar to

the type of avulsion fracture which has beenwell-defined but the remainder represent a groupwhich has been unrecognised previously. Using CTand operative findings we have defined the differentpatterns of fracture of the calcaneal tuberosity. Tenfractures extended into the subtalar joint, but did notfit the pattern of the common intra-articular fractureas described classically. We have defined a newpattern which consists of a fracture of the medialcalcaneal process with a further fracture whichseparates the upper part of the tuberosity in thesemicoronal plane.

Non-operative treatment of displaced fracturesresulted in a mis-shapen heel and a poor functionaloutcome. Open reduction and internal fixation witheither a plate or compression screw did not givesatisfactory fixation.

We prefer to use an oblique lateral tension-bandwire. This technique gave excellent fixation and werecommend it for the treatment of displaced fracturesof the tuberosity of the calcaneus.

J Bone Joint Surg [Br] 2001;83-B:55-61.Received 22 March 2000; Accepted after revision 13 July 2000

Several types of fracture of the calcaneus involving thetuberosity or body of the bone have been described. Themost common is the classical intra-articular fracture inwhich the tuberosity is broken into three pieces of varyingsize by a sagittal primary fracture and a transverse second-ary injury. CT has greatly enhanced the understanding ofthis fracture and allowed the development of a rationalapproach to operative treatment.1-6 There are two other rare

fractures which involve the tuberosity of the calcaneus andare said not to involve the posterior facet of the subtalarjoint. The beak or avulsion fracture occurs in osteoporosisand in the elderly. It separates the upper part of thetuberosity through a fracture line that is transverse orsemicoronal7-11 (Fig. 1a). Finally, an isolated extra-articularfracture of the medial process of the calcaneus has alsobeen reported (Fig. 1b).7,12

Our aim is to describe our experience with fractures ofthe tuberosity of the calcaneus. We will define the differentpatterns and indicate those which extend into the subtalarjoint, but which do not fit the pattern of the common intra-articular fracture of classic description.

Patients and Methods

Since 1988 all patients with fractures of the calcaneusreferred for treatment at the Bristol Royal Infirmary havebeen entered into a prospective database which includesdetails of the patient and of the mechanism of injury. Therewere 24 fractures which involved the tuberosity of the calca-neus, but did not follow the pattern of the common intra-articular fracture. The anatomy of the fracture was studied onplain radiographs, coronal and axial CT and, at the time ofsurgery in 11 patients. The calcaneal score, described by Kerr,Prothero and Atkins,13 was used to assess outcome.

The mean final follow-up was for 31 months (4 to 120).There were three simple extra-articular avulsion fractures

which were similar to those described previously.8,9 Theseoccurred in elderly patients with osteoporotic bone afterminimal trauma. In five fractures in four men aged 19 to 41years, there was an isolated fracture of the medial process.This followed a fall in three of the patients. The remaining16 fractures, occurring in 15 patients, were of a type notpreviously clearly identified (Figs 1c and 1d, 2 to 4). In all,a fracture of the medial process was present (Fig. 5a),which was similar to the isolated fracture of the medialprocess (Fig. 5b). In addition, all cases had a triangularfragment separated from the upper border of the tuberosityof the calcaneus, similar to an ‘avulsion’ or ‘beak’ fracture.In contrast to those who sustained a simple avulsion frac-ture, these patients were young, with a mean age of 30years (15 to 51). They were also significantly younger(p < 0.001) than patients with the common sagittal intra-

articular fracture whose mean age, according to our data-base of 250 patients, was 43 years (11 to 82). The genderdistribution was similar to that of the common sagittalfracture, there being 13 men and three women. The mecha-nism of injury was a significant fall from a mean height of4.2 m (0.6 to 7.5). The mechanism of injury and height offall were similar to those of the sagittal fracture.

Individual fracture fragmentsThe fragment of the medial process. The medial processwas fractured in each patient apart from the three withosteoporotic avulsion fractures. In the five in whom thisfragment occurred in isolation, the fracture was essentiallyundisplaced. In the remaining 16 fractures, the fragmentwas displaced and associated with a further fracture of thesuperior body which is described below.

The fragment of the medial process was of a consistentshape, whether it was an isolated injury or associated witha more complicated fracture pattern. This fragment is clear-ly demonstrated on three-dimensional CT reconstructions(Fig. 4). In the coronal plane, the fracture began relativelysuperiorly on the medial wall of the tuberosity and thenpassed inferiorly in the lateral direction. It ran obliquelydownwards, either to the inferior surface of the calcaneus(Fig. 4) or to the lateral wall, albeit at a level lower than themedial side; thus the fragment of the medial process in-cluded the lateral process in 60% of cases.

The fragment, when displaced, was impacted into themain body of the calcaneus and displaced medially (Fig.6).The fracture of the superior body. The fracture line, whichgave rise to this fragment, lay in the semicoronal plane andpassed posteriorly, involving the whole width of the calca-

56 B. SQUIRES, P. E. ALLEN, J. LIVINGSTONE, R. M. ATKINS

THE JOURNAL OF BONE AND JOINT SURGERY

Fig. 1

Diagram of the fractures of the calcaneal tuberosity showing a) an extra-articular avulsion fracture in an elderly patient with osteoporotic bone, b) afracture of the medial process occurring in isolation, c) a fracture of the medial process occurring in conjunction with a secondary ‘avulsion’ fracturewhich is extra-articular and d) a fracture of the medial process occurring in conjunction with a secondary ‘avulsion’ fracture which is intra-articular.

Fig. 2

A lateral radiograph of an extra-articular fracture of the tuberosity with afracture of the medial process.

Fig. 3a Fig. 3b

Figure 3a – A lateral radiograph of an intra-articular fracture of the tuberosity with a fracture of the medial process. Figure 3b – Sagittal CTreconstruction of a similar fracture.

neus (Fig. 4c). The fragment was extra-articular in sixfractures, but in a further ten the anterior part of the fractureline passed across the whole width of the posterior facet ofthe subtalar joint. In these cases, when the fracture wasintra-articular, the posterior facet was divided into anteriorand posterior parts in the coronal plane (Figs 1d, 2 and 3).The intra-articular fracture passed through the middle of theposterior facet of the subtalar joint.

The superior fragment displaced upwards, presumablydue to the pull of tendo Achillis, and rotated so that theposterosuperior border moved upwards and the postero-inferior edge moved posteriorly, compressing the skin atthe back of the heel. The fragment hinged on its anteriorapex as it displaced so that a fracture which had minimal

displacement anteriorly was often significantly displacedposteriorly, potentially giving rise to a heel boss. Thisphenomenon explains the disability resulting from so-called‘undisplaced’ fractures.Operative technique. It is our policy to treat all displacedcalcaneal fractures by open reduction and internal fixationusing the extended lateral approach.14 In this series, theindication for open reduction and internal fixation wasdisplacement of the fragment which comprises the upperborder of the calcaneal tuberosity. Surgery was undertakenafter the initial swelling had resolved, with the patient inthe lateral position, under antibiotic cover and with a thightourniquet. It is not usually necessary to expose the poster-ior facet of the subtalar joint. This allows maximum preser-

57FRACTURES OF THE TUBEROSITY OF THE CALCANEUS

VOL. 83-B, NO. 1, JANUARY 2001

Fig. 4a Fig. 4b Fig. 4c

The three-dimensional spiral CT reconstructions of a fracture of a medial process with an associated avulsion fracture of the superior body. Figure 4a– A medial view showing the displaced fragment of the medial process and the secondary fracture of the superior body extending into the posterior facetof the subtalar joint. Figure 4b – A lateral view showing that the lateral process has not been fractured. The fracture in the superior body fragment isin the semicoronal plane. Figure 4c – A posteroinferior view showing the shape of the fragment of the medial process. The fragment of the superior bodyis significantly displaced posteriorly due to the pull of tendo Achillis.

Fig. 5a Fig. 5b

CT showing a) how the fracture of themedial process in conjunction with acoronal ‘avulsion’ fracture, is similar tob) the fracture of the medial processwhich occurs in isolation.

vation of soft-tissue attachments to the avulsed fragment.The fracture is reduced and held with large reductionforceps.

Our method of fixation has changed over the course ofthe series. Initially, we used a single 6.5 mm cancellousscrew as advocated by the AO group, but subsequently thiswas changed to a plate and screws. Dissatisfaction withboth of these techniques led us to develop a method ofoblique tension-band wiring (Fig. 7).

Two 2 mm Kirschner wires are passed across the fracturefrom superior and posterior to inferior and anterior. Theentry point for the wires is just anterior to tendo Achillis,which is protected by a soft-tissue guide during insertion.Inferiorly, the wires pass just lateral to abductor digitiminimi, which must be protected by a bone lever. A figure-of-eight tension band wire is passed around the ends of theKirschner wires over the lateral wall of the calcaneus andtightened to obtain further compression. Before the applica-

tion of the tension band, the ends of the Kirschner wires areturned over outwards; the wires can then be rotated through180° so that the ends lie snugly against the calcaneusbefore the tension band is tightened.

Postoperatively, a below-knee plaster cast is applied withthe foot at a right angle. The use of an equinus plaster hasproved unnecessary and such is the strength of the fixationthat immobilisation in plaster is probably not required. It isour policy to keep all patients with intra-articular fracturesnon-weight-bearing for six weeks.

The four patients with displaced fractures which weretreated non-operatively had contraindications to surgery.

Results

The fractures which were treated non-operatively healedrelatively slowly. As may be expected, the four patientswith displaced fractures who were managed non-oper-

58 B. SQUIRES, P. E. ALLEN, J. LIVINGSTONE, R. M. ATKINS

THE JOURNAL OF BONE AND JOINT SURGERY

Fig. 6

Coronal CT of bilateral fractures of thetuberosity showing how the fragment ofthe medial process displaces mediallyand impacts into the body of the calca-neus. The fracture on the right extendedinto the posterior subtalar joint, but thaton the left was extra-articular.

Fig. 7a Fig. 7b

Lateral (a) and axial (b) radiographs ofan intra-articular fracture of the tuber-osity treated by the tension-band wiringtechnique.

atively fared poorly as judged by the calcaneal fracturescore. They complained of misshapen heels which inter-fered with footwear (Fig. 8). The eight patients who hadnon-operative treatment because the fracture was not sig-nificantly displaced, did not have uniformly good results.One developed severe pain at the insertion of tendo Achillisand, in a further two, there was a painful heel bump. Thiswas due to backward and upward movement of the infer-oposterior edge of the upper fragment of the fracture, someof which was, in retrospect, visible on the initial radio-graph. Further minor displacement may have occurredduring the healing period. Small displacements at this siteproduce a heel bump which may cause marked localdiscomfort.

Operative fixation using a one-third tubular plate or acompression screw proved to be unsatisfactory. These twotechniques did not provide a stable reduction, and a pro-longed period in plaster was necessary to achieve soundunion.

In contrast to the other techniques, the use of an obliquetension band was technically simple and provided strongfixation, so that union occurred rapidly in a satisfactoryposition. Loss of position or problems with wound-healingwere not seen in this group. No patient treated with atension band has complained of symptoms from the fixationdevice, none of which has required removal.

All of the patients treated operatively had a good finaloutcome (Table I).

Discussion

Our experience of fractures of the tuberosity of the calca-neus differs from that previously described. Although thereis no feature common to all the fractures they do appear torepresent variations of a similar pattern. In elderly, osteo-

porotic patients, the fracture represents an avulsion of tendoAchillis with its bony insertion. It seems, however, that inyounger patients, the fracture of the medial process repre-sents the first stage in the production of a more complexinjury. This fracture has been thought previously to resultfrom vertical shear after an injury when the heel is in avalgus position.12 CT reveals that the isolated fracture isidentical to that which occurs in conjunction with the morecomplex pattern, the only difference being the degree of thedisplacement.

The configuration of the fracture of the medial processand its medial displacement indicate that it is an injurywhich occurs when the heel is in the varus rather than in thevalgus position as has been previously suggested.12

If the injury is sufficiently severe, the fragment of themedial process displaces medially and impacts into theremaining intact tuberosity. The combination of the weak-ening of the tuberosity due to fracture of the medial processand the pull of tendo Achillis is sufficient to produce afurther fracture line which splits off the superior fragmentof the tuberosity, even in bone which is not osteoporotic.Depending on factors as yet undetermined, the anterioredge of this second fracture line may pass behind or withinthe posterior facet of the subtalar joint.

In summary, we have observed four types of fracture ofthe calcaneal tuberosity (Fig. 1): a) the osteoporotic extra-articular avulsion fracture occurring in the elderly; b) theisolated undisplaced fracture of the medial process; c)fracture of the medial process which occurs in conjunctionwith a secondary ‘avulsion’ fracture, the secondary fracturebeing either extra-articular; and d) fracture of the medialprocess which occurs in conjunction with a secondary‘avulsion’ fracture, the secondary fracture extending intothe posterior facet of the subtalar joint.

The fractures from the last three groups, which include

59FRACTURES OF THE TUBEROSITY OF THE CALCANEUS

VOL. 83-B, NO. 1, JANUARY 2001

Fig. 8a Fig. 8b

Unsatisfactory outcome from non-operative treatment of a patient who presented many years after the fracturefor consideration of reconstruction because of difficulty with footwear. Figure 8a – A lateral radiographshowing extensive remodelling. Figure 8b – A photograph showing the large heel boss.

disruption of the medial calcaneal process, are all part ofthe same ‘family’ and result from a similar mechanism ofinjury.

The plain radiographs of an intra-articular fracture of thetuberosity show changes similar to the more common intra-articular tongue-type fracture, although with experience,differentiation is usually possible. CT examination is neces-sary for accurate analysis and to allow operative planning.A precise study is important since different methods ofsurgical treatment may be necessary.

In our experience, non-operative treatment of displacedfractures of the tuberosity has resulted in a poor outcomeand optimal fixation of the fracture has been achieved using

the oblique lateral tension-band wire. The alternativemethods of fixation were unreliable, but nevertheless resul-ted in a good final outcome, although one patient requiredsecondary reconstructive surgery.

We recommend that displaced fractures of the calcanealtuberosity are treated by open reduction and internal fixa-tion using an oblique tension-band wire.

We would like to express our gratitude to Dr Roland Watura, Dr MarkCobby and staff at the Department of Radiology, Frenchay Hospital for thethree-dimensional spiral CT reconstruction used for Figure 5.

No benefits in any form have been received or will be received from acommercial party related directly or indirectly to the subject of thisarticle.

60 B. SQUIRES, P. E. ALLEN, J. LIVINGSTONE, R. M. ATKINS

THE JOURNAL OF BONE AND JOINT SURGERY

Table I. Epidemiological data and results of treatment of 22 patients with fractures of the tuberosity of the calcaneum

Age ReviewCase Gender (yr) Mechanism Displaced Treatment Complications Calcaneal score (yr)

Extra-articular avulsion fractures1 F 76 Direct blow Yes Tension-band wire 75 12 mth

2 M 78 Tripped Yes Non-operative 26 9 mth(peripheral vascular disease) (comorbidity)

3 F 87 Tripped Yes Non-operative 12 6 mth(Debilitated) (comorbidity)

Fractures of medial process

4 M 30 Fall 3.6m No Non-operative 100 7 yr

5 M 26 Fall 1.8m No Non-operative 85 7 mth

6 M 41 Tripped No Non-operative Painful heel boss 79 6 mth(ankle arthrodesed)

7 (Right) M 19 Fall 1.8m No Non-operative 100† 4 mth

(Left) M 19 Fall 1.8m No Non-operative 100† 4 mth

Fracture of medial process with a secondary extra-articular ‘avulsion’ fracture8 M 25 Fall 4.5m Yes Lateral plate Partial separation 100 6 yr

of fragments

9 M 26 Fall 3.0m Yes 6.5mm compression screw Partial separation 88† 3 yrof fragments

10 M 20 Fall 4.5m Yes Tension-band wire 91 3 yr

11 (Right) M 38 Fall 7.5m Yes Closed reduction 94† 2 yr

12 M 18 Crush injury Yes Tension-band wire 100 5 mth

13 M 15 Fall 3.6m No Non-operative 100† 4 mth

Fracture of medial process with a secondary intra-articular ‘avulsion’ fracture14* M 33 Fall 3.0m Yes Non-operative Mis-shapen heel with 58 9 yr

resultant footwear problems

15 F 20 Fall 6.0m Yes Lateral plate 76 7 yr

16 F 40 Fall 0.45m Yes 6.5mm compression screw Wound breakdown 77 6 yr

11 (Left) M 38 Fall 7.5m Yes Closed reduction 94† 2 yr

17 M 38 Fall 6.0m Yes Tension-band wire 79 18 mth

18 F 21 Fall 3.6m Yes Tension-band wire 94† 18 mth

19 M 39 Fall 1.8m Yes Non-operative as scars Painful heel boss 40 18 mthfrom previous injury

20 M 36 Fall 6.0m No Non-operative Pain at insertion of 20† 15 mthtendo Achillis

21 M 51 Fall 3.0m No Non-operative Painful heel boss 30 6 mth

22 M 30 Fall 6.0m No Non-operative Lost to follow-up(no fixed abode)

* the original radiographs for this patient had been destroyed and were not available for study. The description of the fracture in this Table is as recorded inhospital records† these records were obtained from patients with bilateral calcaneal fractures

References1. Eastwood DM, Maxwell-Armstrong CA, Atkins RM. Fracture of

the lateral malleolus with talar tilt: primarily a calcaneal fracture notan ankle injury. Injury 1993;24:109-12.

2. Langdon IJ, Kerr PS, Atkins RM. Fractures of the calcaneum: theanterolateral fragment. J Bone Joint Surg [Br] 1994;76-B:303-5.

3. Eastwood DM, Gregg PJ, Atkins RM. Intra-articular fractures of thecalcaneum. Part I: pathological anatomy and classification. J BoneJoint Surg [Br] 1993;75-B:183-8.

4. Eastwood DM, Langkamer VG, Atkins RM. Intra-articular fracturesof the calcaneum. Part II: open reduction and internal fixation by theextended lateral transcalcaneal approach. J Bone Joint Surg [Br]1993;75-B:189-95.

5. Sanders R, Fortin P, DiPasquale T, Walling A. Operative treatmentin 120 displaced intra-articular calcaneal fractures: results using aprognostic computed tomography scan classification. Clin Orthop1993;290:87-95.

6. Zwipp H, Tscherne H, Thermann H, Weber T. Osteosynthesis ofdisplaced intra-articular fractures of the calcaneus: results in 123cases. Clin Orthop 1993;290:76-86.

7. Bohler L. Diagnosis, pathology, and treatment of fractures of the oscalcis. J Bone Joint Surg 1931;13:75-89.

8. Lowy M. Avulsion fractures of the calcaneus. J Bone Joint Surg [Br]1969;51-B:494-7.

9. Protheroe K. Avulsion fractures of the calcaneus. J Bone Joint Surg[Br] 1969;51-B:118-22.

10. Hedlund LJ, Maki DD, Griffiths HJ. Calcaneal fractures in diabeticpatients. J Diabetes Complications 1998;12:81-7.

11. Kathol MH, El-Khoury GY, Moore TE, Marsh JL. Calcanealinsufficiency avulsion fractures in patients with diabetes mellitus.Radiology 1991;180:725-9.

12. Watson-Jones R. Fractures and other bone and joint injuries. 1st ed.Edinburgh: E & S Livingstone, 1940.

13. Kerr PS, Prothero DL, Atkins RM. Assessing outcome followingcalcaneal fracture: a rational scoring system. Injury 1996;27:35-8.

14. Freeman BJC, Duff S, Allen PE, Nicholson HD, Atkins RM. Theextended lateral approach to the hindfoot: anatomical basis and surgi-cal implications. J Bone Joint Surg [Br] 1998;80-B:139-42.

61FRACTURES OF THE TUBEROSITY OF THE CALCANEUS

VOL. 83-B, NO. 1, JANUARY 2001