Fracture after distraction osteogenesis - Bone & Joint

Fracture after distraction osteogenesis

A. H. R. W. Simpson, J. Kenwright

From the Nuffield Orthopaedic Centre, Oxford, England

We reviewed 173 patients undergoing distraction The introduction of methods of limb lengthening, which
osteogenesis to determine the incidence, location encourage effective formation of bone within the distracted
and timing of fractures occurring as a complication of segment, has reduced the incidence of fracture after length-
the procedure. ening.1-5 Unfortunately, however, this complication still
occurs2,5-10 and often leads to the need for further surgical
There were 17 fractures in 180 lengthened segments intervention which lengthens the time of treatment.
giving an overall rate of fracture of 9.4%.
Unexpectedly, the pattern and location of the fractures Our aim was to determine the incidence and pattern of
were very variable; six were within the regenerate fracture after distraction osteogenesis. The results of the
itself, six at the junction between the regenerate and treatment of the fractures are presented and predisposing
the original bone and five at distant sites in the limb. factors identified.
Of those occurring in the regenerate, five were noted
to be associated with compression and partial collapse Patients and Methods
of the regenerate. In three patients collapse and
deformity developed gradually in the distracted We reviewed the results of 180 procedures on 173 patients
segment over the six months after removal of the in whom distraction osteogenesis had been carried out. We
frame. used unilateral systems (Dynabraces; Smith & Nephew,
Cambridge, UK), limb lengtheners and limb reconstruction
The method of treatment of these fractures should rails (Intravent; Orthofix, Maidenhead, UK), Red Mono-
be chosen to take into account multiple factors, which tube with multi and single pin clamps (Howmedica,
are additional and often different from those to be Staines, UK) and circular external skeletal fixation frames
considered during management of acute traumatic (Ilizarov; Smith & Nephew). There was a minimum period
injuries. Internal fixation appears to be most of six months between the time of removal of the frame and
appropriate for displaced fractures, although in small review. We analysed all records, radiographs and ultra-
children, or in those in whom there has been, or is, sound examinations, made at two-weekly intervals. All the
infection of the screw tracks, a new period of fractures were examined, according to time of occurrence,
treatment using external fixation may be needed. site and pattern of bone failure. Table I gives details of the
Fixation by intramedullary nailing was associated with aetiology of shortening of the limb in those patients in
a risk of infection, even if screw tracks were assessed whom a fracture occurred.
as healthy at the time of insertion of the nail. Internal
fixation with the use of plates is safe for displaced, Results
unstable fractures in children.
There were 17 fractures in 15 patients (two patients each
J Bone Joint Surg [Br] 2000;82-B:659-65. had two fractures) after 180 procedures. The site of fracture
Received 22 February 1999; Accepted after revision 27 September 1999 varied widely; they also varied in time, in relation to the
removal of the frame and in their overall pattern.
A. H. R. W. Simpson, FRCS, Consultant Orthopaedic Surgeon, Clinical Time of fracture. Acute fractures occurred within six
Reader weeks of removal of the frame in nine instances, but in
J. Kenwright, FRCS, Nuffield Professor of Orthopaedic Surgery three patients an insidious late fracture occurred with
Nuffield Department of Orthopaedic Surgery, Nuffield Orthopaedic Cen- deformity noticed several months after the frame had been
tre, Headington, Oxford OX3 7LD, UK. removed (Fig. 1). One patient had a second fracture 12 days
after removal of the frame through the screw holes of a
Correspondence should be sent to Professor J. Kenwright. plate which had been applied for treatment of a previous
fracture which occurred two days after removal of the
©2000 British Editorial Society of Bone and Joint Surgery frame (Fig. 2a). Four patients had a fracture with the frame
0301-620X/00/59945 $2.00

VOL. 82-B, NO. 5, JULY 2000 659

660 A. H. R. W. SIMPSON, J. KENWRIGHT

Fig. 1 Fig. 2a Fig. 2b

Radiograph showing late valgus de- Radiographs showing a) a fracture which occurred with minimal trauma, through the last screw
formity, which was noticed several hole of a plate used to treat a fracture through a distracted segment in which the last screw was
months after apparent healing with- close to an old screw track and b) a subtrochanteric fracture through proximal screw tracks.
out angulation.

Table I. The pattern of the 17 secondary fractures which occurred in 15 patients after distraction osteogenesis

Fracture Maturity Aetiology Bone Length of Method of
(number of (number) (number) regenerate stabilisation*
fractures) (cm)

Within regenerate with collapse Adult (2) Trauma (1) Tibia (1) 5 N
Acute Osteomyelitis (1) Femur (1) 6 C
Adult (2) Congenital (2) Tibia (2) 5,4 N (2)
Gradual Child (1) Congenital (1) Tibia (1) 4 C (1)
Femur (1) 6
Within regenerate without collapse Child (1) Congenital (1) P
Femur (6) 3,5,5,6,6,6
Junctional Child (6) Trauma (1) C (3)
Osteomyelitis (1) Femur (2) 6,5 Ex (3)
Congenital (4)
Tibia (1) 6,3,1 P (2)
Through screw track Child (2) Congenital (2) Femur (2)
P
Distant site Child (3) Trauma (1) Ex
Metabolic (1) C
Congenital (1)

* N, intramedullary nail; Ex, external fixation; P, plate; C, cast

still in situ. In one of these, it occurred through the screw (Fig. 2b). A fracture was also seen through the upper tibial
tracks of the external fixator (Fig. 2b), in two the ipsilateral growth plate in one femoral lengthening (Fig. 3). In the 17
tibia fractured while the femur was being lengthened, and instances of fracture only two of the regenerates showed a
in one the fracture was in the contralateral femur when the narrowed portion producing a waist in the new bone of
patient was walking with a frame applied to the tibia. smaller diameter than the original bone (Fig. 4a). The
Site of fracture. The regenerate itself fractured in six fracture through one of these segments occurred within the
patients. A further six fractures occurred at the junction waist and in one, surprisingly, not at the waist, but at the
between the original bone and the regenerate. In the junction of the regenerate and the original bone (Fig. 4b).
remaining five patients the fracture occurred as a result of Pattern of fracture. Two radiological patterns of bone
minimal violence at levels separate from the distracted failure were observed. In one there was partial collapse of
segment. In two of these, they were through screw tracks the regenerate, which resulted in angulation. This happened

THE JOURNAL OF BONE AND JOINT SURGERY

FRACTURE AFTER DISTRACTION OSTEOGENESIS 661

Fig. 3

Radiograph showing a Salter- Fig. 4a Fig. 4b
Harris II fracture of the prox-
imal tibia in a boy under- Radiographs showing a) a wasted region of regenerate and b) a fracture occurring at
going lengthening of the fe- the junction of the regenerate and original bone (the arrows indicate the limits of the
mur. regenerate).

both acutely (Fig. 5) and gradually (Fig. 1). In the remain- In our series, fractures occurred in patients of all ages
ing fractures, no loss of length of the regenerate itself was and with both small and large lengths of regenerate. There
observed; complete non-communited fractures were sus- appeared to be no direct relationship between the risk of
tained in the region of the regenerate (Fig. 6). fracture and age or length of the segment (Table I).
Factors predisposing to fracture. The mean range of knee Classification of fracture. On the basis of the site, pattern
movement in the six patients whose fractures occurred and timing of the fracture we have developed a classifica-
through the junction of the regenerate and original bone tion as shown in Table II and Figure 7.
was 28°. In the remaining patients with a fracture it was Treatment. A variety of methods of treatment was used to
85° at the time when this occurred. allow for a number of individual factors, including skeletal
maturity, stability of the fracture and its suitability for
In the overall series of 180 lengthenings, ‘cysts’ devel- surgical fixation or treatment in a cast (Table I). Account
oped in 12, in five of which bone grafting of the distracted was also taken of the risk of developing infection from
segment had been performed before the frame was sepsis at the screw hole, the severity of stiffness of the knee
removed. They healed well. In two patients the ‘cysts’ were and the tolerance of the patient for a further protracted
small (less than 5 mm in diameter) and peripheral. They period of treatment. The quality of the regenerate was also
were left untreated and healed spontaneously without com- assessed to see whether bone grafting was needed.
plication. In four patients the ‘cysts’ were aspirated, length-
ening was stopped for a week and then recommenced at a All fractures eventually healed soundly, without loss of
slower rate. In one of these patients the ‘cyst’ did not the length gained, and without the development of perma-
reappear. In the other three patients, however, the ‘cysts’ nent deformity associated with the fracture, with the single
recurred although smaller in size; two of these segments exception of the patient who developed acute collapse and
healed, and one fractured. In one patient, treated early in was treated in a cast. This patient required a subsequent
the series, large ‘cysts’ were left untreated and were seen to corrective osteotomy.
persist even after the consolidation period. The bone subse- Complications. In adult patients, with displaced type-I
quently fractured. The persistence of large ‘cysts’ appears fractures and apparently satisfactory screw tracks, the limb
to be a predisposing factor to fracture. was realigned and an intramedullary nail inserted. Deep

VOL. 82-B, NO. 5, JULY 2000

662 A. H. R. W. SIMPSON, J. KENWRIGHT

Fig. 5a Fig. 5b

Radiographs showing a) a lengthened segment of the femur which b) had acute collapse
shortly after removal of the frame.

Fig. 6a Fig. 6b

Radiographs of a child with a congenitally short femur after bifocal lengthening showing a)
good consolidation and b) a fracture through the regenerate.

infection, which was confirmed bacteriologically, devel- fixation was between three and 25 weeks and infection
oped in all three patients in whom this method was used. still occurred. On the radiographs there was no sign of
In each instance, the screw tracks were well healed at the local infection although a widened screw track was visi-
time of insertion of the nail. The interval of time left ble. All infections settled completely after debridement.
between removal of the frame and intramedullary nail The nail was removed after bone healing and the infection

THE JOURNAL OF BONE AND JOINT SURGERY

FRACTURE AFTER DISTRACTION OSTEOGENESIS 663

Table II. Classification and number of fractures which occurred in 15 patients after distraction osteogenesis, and recommendations
for their prevention and treatment

Type Classification of the fracture Number of fractures Prevention Treatment of fracture

I Within the regenerate: 2 Bone graft large cysts Operative stabilisation
Collapse 3 before frame removal ± bone graft
Acute Accelerate maturation by Osteotomy and operative
1 dynamisation/compression stabilisation
Gradual of frame
Strengthen bone by Stabilisation: operative
No collapse dynamisation/compression for displaced and non-
± bone graft ± brace operative for undisplaced

II Junctional 6 Physiotherapy to maintain Stabilisation

range of movement of knee ± quadricepsplasty

III Through screw track 2 If possible use wires or Operative stabilisation

narrower external fixation screws

IV Distant site 3 Increase weight-bearing to Operative stabilisation

reduce osteoporosis for displaced and non-

operative for undisplaced

Total 17

had not recurred at the time of follow-up at least two years collapse of the regenerate when the frame was removed.
later. Aldegheri, Renzi-Brivio and Agostini22 stressed the

One refracture in the femur occurred after the use of a importance of loading the distracted segment for several
plate and screws. The plate bridged the distracted segment, weeks before removal of the frame. Similarly, Ilizarov14
but there was refracture at the level of the most distal screw recommended loading the distracted segment in order to
(Fig. 2a). accelerate bone consolidation. The results from a large
number of patients have shown low rates of fracture. In our
Discussion study weight-bearing was encouraged. Unilateral fixators
were also dynamised to allow axial slide. When circular
Fracture after distraction osteogenesis is a serious compli- frames were used distraction was reversed by making a few
cation which leads to a considerable increase in morbidity. turns of the screws when the regenerate had sufficient
Our study and that of Bernard et al10 have both shown that maturity to resist collapse.
fractures do occur, despite attempts to minimise the prob-
lem. Fractures in the region of the regenerate occurred early Despite the use of these measures to enhance osteogene-
after removal of the frame in 5% of our patients which is sis and the careful assessment of healing at the time of
comparable with the incidence described by DeBastiani et removal of the frame,5 fractures have occurred. Various
al.2 If all fractures are considered, the rate rises to 9.4%. patterns were seen and classified according to a system
This is similar to that reported by Stanitski et al11 who shown in Table II and Figure 7.
included all bony problems.
Approximately one-third of all fractures occured at the
The controllable factors which enhance effective bone
formation have already been described.12-16 They include a Fig. 7
low-energy osteotomy, a delay before distraction is com- The patterns of fracture as defined in Table II.
menced and a slow rate of distraction carried out in small
daily increments. Bone does not, however, always form
effectively and variable patterns of bone formation have
been described on radiographs by Hamanishi et al.17 If the
bone formed is not adequate as assessed by ultrasound,18,19
radiography17,20,21 or by measurement of stiffness, the
dynamics of distraction can be changed. Unfortunately,
such manoeuvres, which include the use of the accordian
technique, are not always effective, resulting in an insuffi-
cient quality or quantity of bone. In our group of patients
the formation of cysts was associated with a poor final
regenerate and a bone graft was often required. In the two
instances when this protocol was not employed there was

VOL. 82-B, NO. 5, JULY 2000

664 A. H. R. W. SIMPSON, J. KENWRIGHT

junction between the regenerate and the original bone. It external fixation was an essential component of leg length-
was noted that these regenerated segments appeared to be ening. Despite precautions to ensure that the tracks were
of good quality at the time of removal of the frame. well healed, infection was seen in three patients. Radio-
Experimental studies23 have shown that regenerates with graphs showed wide screw tracks at the time of nailing in
poor bone formation will fail within their substance at low each instance, and it is probable that there was low-grade
loads, whereas those with strong bone fail at higher loads infection and some necrotic bone lying within these. Tracks
and at the junction of the regenerate and the original bone. should be formally curetted at the time of removal of the
We also noted that there was an association between loss of frame in adults considered to be at risk of sustaining a
movement of the knee and the junctional fracture in the subsequent fracture. Nailing is probably still the treatment of
femur. If the knee is stiff, the lever arm, which lies on one choice for such fractures, although special precautions are
side of the regenerate, is greatly lengthened and relatively required. In a recent study, Antich-Adrover et al27 recom-
low forces acting upon the limb may result in large bending mended that there should be an interval of at least ten days
moments; these forces may exceed breaking point. Efforts between removal of the frame and intramedullary nailing,
should be made to regain joint movement before removal and that antibiotics need to be given at the time of the
of the frame, but if it has to be removed when the knee is nailing; they recorded a low incidence of infection after
stiff (less than 45° of flexion), precautions must be taken to secondary nailing.
prevent fracture. For example, splints should be used and
physiotherapy designed to mobilise the knee while restrict- In children, open reduction and plate fixation gave reli-
ing the load on the regenerate. able results. The plate should extend well beyond the
distracted segment and the screws must avoid the previous
Fractures were also encountered at sites remote from the tracks of external fixation; insertion through or adjacent to
distracted segment, often after minimal violence. Patients such a site risks subsequent fracture.
who do not progress to weight-bearing develop disuse
osteoporosis, which puts them at risk for this type of In a child in whom rapid healing of the fracture is
fracture. This emphasises the importance of ensuring that expected, or in those with infected screw tracks, repeated
the limb is loaded, particularly in patients in whom the use of an external fixation frame should be considered.
frames have to remain in place for many months. Most patients, however, are reluctant to undergo a further
period of treatment with the frame and onlay plate fixation
Some fractures occurred through screw tracks (type-III is the first choice for fracture after leg lengthening.
fractures). In each instance these were in children being
treated by unilateral fixators. The use of bone screws of Patients with fracture who have had a protracted period
smaller diameter or transfixion wires should reduce this of treatment, and in whom there is grossly restricted knee
risk. flexion at the time of open reduction, require quadriceps-
plasty. In each case in which this has been performed with
Our study has shown considerable variation in the pat- the use of postoperative continuous passive motion, the
tern of fractures which occur in association with distraction range of knee movement obtained at operation was
osteogenesis. When selecting the method of treatment, maintained.
many factors should be considered. Most patients will have
undergone a programme of treatment in the frame lasting We would like to thank Vicki Ponting for preparation of the manuscript.
between three and 12 months. A method of management of No benefits in any form have been received or will be received from a
the fracture which would allow rapid rehabilitation must
offer considerable advantage. commercial party related directly or indirectly to the subject of this
article.
Treatment must vary according to the type of fracture. A
cast can rarely be used, except for undisplaced fractures with References
no risk of loss of length. Operative fixation is indicated to
control the position of the fracture and to allow rapid 1. Hood RW, Riseborough EJ. Lengthening of the lower extremity by
rehabilitation. Diaphyseal fractures in adults were deemed the Wagner method: a review of the Boston Children’s Hospital
suitable for intramedullary nailing, but infection occurred in experience. J Bone Joint Surg [Am] 1981;63-A:1122-31.
all three instances in which nails were used, despite well-
healed screw tracks. Many studies have shown that intra- 2. De Bastiani G, Aldegheri R, Renzi-Brivio L, Trivella G. Limb
medullary infection can develop after the use of external lengthening by callus distraction (callotasis). J Pediatr Orthop
fixation for displaced tibial fractures.24-28 The recorded 1987;7:129-34.
incidence of this complication varies considerably, between
4.8%28 and 44%.24 There is a correlation between infection 3. Paley D. Problems, obstacles, and complications of limb lengthening
after nailing and the presence of past pin-track infection, by the Ilizarov technique. Clin Orthop 1990;250:81-104.
seen during the treatment by external fixation. Such earlier
infection, although healed at the time of intramedullary 4. Velazquez RJ, Bell DF, Armstrong PF, Babyn P, Tibshirani R.
nailing, increased the risk of medullary infection to 71%.25 Complications of use of the Ilizarov technique in the correction of
In our small group of patients, prolonged treatment by limb deformities in children. J Bone Joint Surg [Am] 1993;75-A:1148-
56.

5. Dahl MT, Gulli B, Berg T. Complications of limb lengthening: a
learning curve. Clin Orthop 1994;301:10-8.

6. Mezhenina EP, Roulla EA, Pechersky AG, et al. Methods of limb
elongation with congenital inequality in children. J Pediatr Orthop
1984;4:201-7.

7. Dwyer JSM, Owen PJ, Evans GA, Kuiper JH, Richardson JB.
Stiffness measurements to assess healing during leg lengthening: a
preliminary report. J Bone Joint Surg [Br] 1996;78-B:286-9.

8. Trivella GP, Brigadoi F, Aldegheri R. Leg-lengthening in Turner
dwarfism. J Bone Joint Surg [Br] 1996;78-B:290-3.

THE JOURNAL OF BONE AND JOINT SURGERY

FRACTURE AFTER DISTRACTION OSTEOGENESIS 665

9. Noonan KJ, Leyes M, Forriol F, Canˆ adell J. Distraction osteogene- 19. Eyres KS, Bell MJ, Kanis JA. Methods of assessing new bone
sis of the lower extremity with the use of monolateral external formation during limb lengthening: ultrasonography, dual-energy X-
fixation: a study of two hundred and sixty-one femora and tibiae. ray absorptiometry and radiography compared. J Bone Joint Surg [Br]
J Bone Joint Surg [Am] 1998;80-A:793-806. 1993;75-B:358-64.

10. Bernard J, Eyres KS, Bell MJ, Saleh M. Fractures in limb lengthen- 20. Green SA. Postoperative management during limb lengthening.
ing procedures: a preliminary analysis. J Bone Joint Surg [Br] Orthop Clin North Am 1991;22:723-34.
1998;80-B:Suppl 1, 58.
21. Aquerreta JD, Forriol F, Canadell J. Complications of bone length-
11. Stanitski DF, Shahcheraghi H, Nicker DA, Armstrong PF. Results ening. Int’l Orthop (SICOT) 1994;18:299-303.
of tibial lengthening with the Ilizarov technique. J Pediatr Orthop
1996;6:168-72. 22. Aldegheri R, Renzi-Brivio L, Agostini S. The callotasis method of
limb lengthening. Clin Orthop 1989;241:137-45.
12. Ilizarov GA. The tension-stress effect on the genesis and growth of
tissue: Part I: the influence of stability of fixation and soft-tissue 23. Wood MK, Li G, Brownlow H, Simpson AHRW. Which test best
preservation. Clin Orthop 1989;238:249-81. predicts the strength of the regenerate?: x-rays, DEXA or stiffness?
J Bone Joint Surg [Br] 1998;80-B:Suppl 3, 264.
13. Ilizarov GA. The tension-stress effect on the genesis and growth of
tissue: Part II. The influence of the rate and frequency of distraction. 24. McGraw JM, Lim EVA. Treatment of open tibial-shaft fractures:
Clin Orthop 1989;239:263-85. external fixation and secondary intramedullary nailing. J Bone Joint
Surg [Am] 1988;70-A:900-11.
14. Ilizarov GA. Clinical application of the tension-stress effect for limb
lengthening. Clin Orthop 1990;250:8-26. 25. Maurer DJ, Merkow RL, Gustillo RB. Infection after intramedullary
nailing of severe open tibial fractures initially treated with external
15. White SH, Kenwright J. The importance of delay in distraction of fixation. J Bone Joint Surg [Am] 1989;71-A:835-8.
osteotomies. Orthop Clin North Am 1991;22:569-79.
26. Blachut PA, Meek RN, O’Brien PT. External fixation and delayed
16. Schwartsman V, Schwartsman R. Corticotomy. Clin Orthop intramedullary nailing of open fractures of the tibial shaft: a sequential
1992;280:37-47. protocol. J Bone Joint Surg [Am] 1990;72-A:729-35.

17. Hamanishi C, Yasuwaki Y, Kikuchi H, Tanaka S, Tamura K. 27. Antich-Adrover P, Mart´ı-Garin D, Murias-Alvarez J, Puente-
Classification of the callus in limb lengthening: radiographic study of Alonso C. External fixation and secondary intramedullary nailing of
35 limbs. Acta Orthop Scand 1992;63:430-3. open tibial fractures: a randomised, prospective trial. J Bone Joint
Surg [Br] 1997;79-B:433-7.
18. Derbyshire NJ, Simpson AHRW. The role of ultrasound in limb
lengthening. Brit J Radiology 1992;65:576-80. 28. Wheelwright EF, Court-Brown CM. Primary external fixation and
secondary intramedullary nailing in the treatment of tibial fractures.
Injury 1992;23:373-6.

VOL. 82-B, NO. 5, JULY 2000