If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Thirty-two free vascularised fibula grafts performed at our unit have been assessed retrospectively with respect to success, bony union and percentage graft hypertrophy.
Between 1981 and 1998, there were 21 males and 11 females (aged 8–61 years) with follow-up of 5 months to 14.6 years. The mean bony defect bridged was 12.0 cm (standard deviation 4.8; range 5.0–21.0 cm). Bony union and hypertrophy were assessed radiographically.
Time to bony union was compared using the log-rank, Wilcoxon or likelihood ratio tests. Kaplan–Meier survival curves were constructed. Hypertrophy was compared with Wilcoxon's rank sum test.
Three flaps failed. Seventy-four percent of patients healed primarily at a median time of 4.75 months; five patients required further surgery to heal by 18 months (interquartile range 14–20 months). Complication rate and donor site morbidity were low. The stress fracture rate was 21%. Ninety percent of patients regained a functional limb by 12 months.
Hypertrophy was measured in 22 patients and ranged from 0 to 316% (median 71%; interquartile range 10–145%). Median hypertrophy in the lower limb was 76.5% (interquartile range 26.5–165%) compared to 33.5% in the upper limb (0–88%); p=0.16. Median hypertrophy in trauma cases was 80% (interquartile range 10–167%) compared to 70% in tumour cases (33–105%); p=0.62.
Our findings confirm that the fibula responds physiologically to biomechanical loading. Our results compare well with other series and alternative reconstructive modalities. We conclude that the free fibula flap can provide excellent results in the salvage of limbs with large bony defects.
Large bony defects resulting from trauma, tumour resection, congenital malformations or other bone disease present a difficult reconstructive challenge (Fig. 1) . There is no accepted standard procedure. Patients frequently undergo multiple operations, suffer prolonged immobilisation and may be left with considerable residual functional impairment.
Fig. 1Excision of osteosarcoma of the tibia, creating a 14 cm defect.
When the decision to reconstruct a limb has been taken, various methods are available to the reconstructive surgeon. Short bony defects up to a limited distance (4–6 cm), can be bridged by bone grafts, which then become incorporated by a process of ‘creeping substitution’.
Larger defects require the use of allografts or vascularised bone grafts. Vascularised grafts incorporate living bone into the defect. Healing of the bridged defect is similar to that of a double fracture, and this process is quicker than graft incorporation.
Distraction osteogenesis is becoming increasingly useful for reconstruction of bony defects, but is still the subject of outcome analysis.
After bone healing has been established, the vascularised graft can be seen to hypertrophy to a varying degree, increasing structural strength of the bone. Experimental and clinical studies have demonstrated that bone hypertrophies in response to mechanical forces,
although, for vascularised grafts, there does not appear to be a constant relationship to these forces.
After trauma, the role of any residual periosteum in new bone formation and hypertrophy, and repair around the ends of the transferred vascularised segment, has often been overlooked. However, many authors have agreed that the best results for bony union are obtained in the post-traumatic patient. We hypothesise that there is a potential for vascularised residual periosteum to be present around the bone defect in such patients, which could contribute to new bone formation, and rapid bony union, whereas there is no periosteum remaining after resection of a bony tumour.
The free vascularised fibula flap has become recognised as a useful technique for larger bony defects, since Taylor reported the first successful case in 1975.
We therefore wished to carry out an analysis of the long-term results of free fibula transfer, with particular emphasis on success in limb salvage and achievement of union. We wished to look at the amount of hypertrophy to have occurred in the fibula, and relate this to weight-bearing forces and the possible role of any residual periosteum. In the light of recent reconstructive options, such as bone transport, we wished to consider whether or not there is still a place for the free vascularised fibula flap.
2. Methods
This retrospective analysis was performed on a series of 32 consecutive patients who underwent free fibula flap transfer at Frenchay Hospital, Bristol, between 1981 and 1998. With one exception, all surgery was carried out by the senior author (PLT).
The information required for the study was obtained from hospital notes and radiographs. If the patient had been transferred, the referring hospital was contacted and the notes and radiographs were retrieved. Attempts were made to trace all patients, and where possible, they were reviewed in outpatient clinic.
Demographic details were recorded, including age at time of surgery, reason for the free flap, recipient site, size of bony defect, operative details, post-operative course and complications, subsequent follow-up and overall outcome.
All available radiographs were examined to determine evidence of bony union, stress fracture and degree of hypertrophy of the transferred fibula. Bony union was said to have occurred when bone healing had taken place at both ends of the graft, and there was evidence of radiological union between the fibula and the recipient bone in two planes. If there were insufficient X-ray films to determine when this occurred, it was assumed to be the time at which it was documented in the case notes that the patient was able to use the limb without pain, or initiate weight-bearing.
A stress fracture was recorded if symptoms had been documented in the case notes and confirmed radiographically, or if a hairline break was seen in the cortex of the graft. An asymptomatic fracture was taken to have occurred if an area of localised callus was noted on later radiographs (Fig. 2) .
Fig. 2Localised callus identifying the site of previous stress fracture of the transferred fibula.
The amount of hypertrophy to have developed in the fibula at the time of final follow-up was calculated as follows. The maximum and minimum widths of the transferred fibula on the immediate post-operative X-ray films (AP and lateral views) were measured with callipers by two independent observers (Fig. 3) . The average value of these measurements was recorded as ‘original fibular width’. This procedure was repeated on the final films to obtain the measurement ‘final fibula width’. The amount of hypertrophy to have occurred during the follow-up period was therefore the difference between ‘original fibula width’ and ‘final fibula width’ and this was expressed as a percentage.
Fig. 3Hypertrophy is calculated by measurement of maximum and minimum transferred fibula widths on (A) the immediate post-operative X-ray film and (B) at last follow-up.
However, the fibula is triangular in cross-section, and De Boer has shown that rotation of the fibula by 10° to either side of neutral position, (a reasonable variation between radiographs), can produce an increase in the fibula width of as much as 20%,
Hence, we considered the bone to have hypertrophied only when the ‘final fibula width’ was at least 20% greater than the original.
If no original post-operative film was available, measurements were made from films of the patient's contralateral nontransferred fibula. A consistent anatomical feature was also measured on the first and all subsequent radiographs so that adjustments could be made for variations in magnification between different films. If such a consistent point was not present on all films to allow this adjustment to be made, the films were considered unsuitable for analysis. Patients with unsuitable X-ray films were excluded.
The amount of hypertrophy has been found to increase with time since surgery,
although, some authors believe that after 12 months, no further increase in hypertrophy occurs. We have therefore excluded patients who were not followed up for a minimum of 12 months.
2.1 Statistical analysis
In view of the small sample size only univariate analyses were undertaken. Time to bony union was analysed using survival methods. Patients for whom bony union did not occur were censored at last follow-up/death. For the categorical variables, time to bony union was compared using either the log-rank or Wilcoxon test. For continuous variables, Cox regression and the likelihood ratio test was used. Survival curves are constructed using the Kaplan–Meier method. Spearman's rank correlation was used to assess the association between continuous variables and hypertrophy and Wilcoxon's rank sum test was used to compare the degree of hypertrophy by each categorical variable.
3. Demographic data
Thirty-two free fibula flaps were performed in 32 patients over the 17-year period. There were 21 males and 11 females with an age range at time of surgery of 8–61 years (median 24 years; interquartile range 18–38 years). The mean length of bony gap bridged was 12.0 cm (standard deviation 4.8) with a range of 5.0–21.0 cm and the mean length of fibula harvested was 18.0 cm (standard deviation 4.6; range 6–27 cm).
Twenty-nine grafts were of the single strut type and three were double-barrelled, where the fibula was osteotomised and folded on itself, preserving the periosteum and vascular pedicle
(Fig. 4) . There were five osteocutaneous flaps, the remainder being bone only or bone with a cuff of muscle.
Fig. 4(A) Post-operative X-ray film showing a ‘double-barrelled’ free fibula graft used to reconstruct the distal femur. (B) X-ray film of the same patient 8 years later, showing marked hypertrophy of the fibula.
Several patients had undergone multiple surgical procedures prior to their referral for a free fibula flap, up to 14 in the case of one trauma patient. One patient returned abroad immediately after surgery and although we have a verbal report that the surgery was successful and the patient has a functional limb, we did not have sufficient records to include him in the outcome analyses. The length of time of follow-up in the remaining patients ranged from 5 months to 14.6 years (median 37 months (3.1 years); interquartile range 15–110 months (1.25–9.2 years)). Indications for the surgery are shown in Table 1.
For the two patients, who had avascular necrosis of the head of the femur, the fibula was used to attempt revascularisation of the femoral head and prevent its further collapse. In one case, the fibula was placed as a strut to reduce pain in fibrous dysplasia of the femur with coxa vara (Fig. 5) . In another patient, who had previously undergone a fibula pro tibia procedure for congenital pseudoarthrosis, the free fibula was used as an interpositional graft to try and reduce residual pain.
Fig. 5Post-operative X-ray film showing use of the free fibula, in a patient with fibrous dysplasia of the femur with coxa vara.
Overall, 29 flaps were successful, giving a success rate for the series of 91% (95% confidence intervals 74–99%).
4.2 Bony union
Bony union was assessed in 27 cases. Those excluded were the three failed flaps, the patient from abroad and one patient who died from recurrence, whose X-ray records were not traceable.
In 20 patients, bony union was uncomplicated and achieved primarily at both ends of the graft at a median time of 4.75 months (interquartile range 3.75–7; range 2–12 months). Five patients required secondary surgery to achieve healing at a median of 18 months (interquartile range 14–20; range 11–22 months). Therefore, overall union was achieved in 25 out of 27 cases at a median of 5.5 months (interquartile range 4–11 months).
There were two cases of nonunion at one end of the graft, both failing to unite at the distal end. Time to healing at the proximal end was two and a half months in one patient, and not recorded for the other.
4.2.1 Factors affecting time to bony union
There was no evidence to suggest that the time to bony union differed significantly between trauma, tumour and other cases (p=0.59, Wilcoxon test) or according to the position of the transferred fibula (p=0.68, Wilcoxon test). No association with age (p=0.30) or bony gap (p=0.75) was found (Fig. 6) .
Fig. 6Graph showing the relationship between time to bony union and bony gap.
One patient was lost to follow-up after 3 weeks and so we have looked at limb salvage in the remaining 31 patients. Of the three failed flaps, two required above knee amputations but the third has a functional, if painful, arm. Of the two cases of nonunion, one patient although having an apparently successful flap later required above knee amputation for infected nonunion. With the other case of nonunion, the patient has a useful limb and has declined further surgery. At 12 months follow-up, 28 patients (90%; 95% confidence intervals 74–98%) had regained a functional limb.
In four patients, the free fibula was performed for reasons other than trauma, tumour or chronic osteomyelitis (i.e. congenital pseudoarthrosis of the tibia, fibrous dysplasia of the femur and avascular necrosis of the head of the femur). The two former patients did well, but of the patients with avascular necrosis of the femoral head, one went onto a total hip replacement 18 months after original surgery and the other required arthrodesis for pain after 15 months. Therefore, at last follow-up, 26 out of 31 patients (84%; 95% confidence intervals 66–95%) regained long term a functional limb following their flap surgery.
4.4 Hypertrophy
Measurement of hypertrophy was possible in 22 patients. Those excluded were the three failed flaps, five patients with unsuitable X-ray records and the two cases where the fibula was used for revascularisation of the femoral head.
The range of hypertrophy in the 22 patients was 0–316% (median 71%; interquartile range 10–145%). To investigate the role of periosteum, we compared the hypertrophy measured in trauma patients with that in tumour patients; and to look at the effect of mechanical loading, we compared the amount in upper limbs versus that in lower limbs.
4.4.1 Hypertrophy in trauma patients versus tumour patients (n=19)
The one patient with chronic osteomyelitis showed hypertrophy of 261%. This patient had originally suffered trauma, but the infected segment was excised complete with periosteum as in the tumour cases. As he did not fit clearly into either category, we therefore excluded this patient from comparison of trauma versus tumour. We also excluded the patients with congenital pseudoarthrosis and fibrous dysplasia of the femur. The comparative measurements of hypertrophy in cases performed for trauma or tumour are shown in Table 2. There was no evidence to suggest that the degree of hypertrophy differed significantly between trauma and tumour patients (p=0.62).
Table 2Comparative measurements of hypertrophy in the transferred fibula in trauma cases versus tumour cases
4.4.2 Hypertrophy in lower limb versus upper limb (n=22)
Fourteen out of 16 fibulas in the lower limb showed hypertrophy (87.5%), compared to three out of the six fibulas in the upper limb (50%). Table 3 shows the comparison of measured hypertrophy in the upper and lower limbs. Although, the degree of hypertrophy was generally higher in the lower limbs, the difference was not significant at the 5% level (p=0.16).
Table 3Comparative measurements of hypertrophy in the transferred fibula in the upper limb versus the lower limb
There was low correlation between the length of time, since bony union and degree of hypertrophy (p=0.28, n=22), but no correlation between age and degree of hypertrophy (p=−0.09, n=22) or between bony gap and hypertrophy (p=−0.21, n=22) was found.
4.5 Complications
Four patients developed a haematoma at the anastomosis site that required evacuation. There were no sequelae. Three patients developed superficial wound infection, defined as clinical evidence of infection with positive wound culture swab, which resolved with antibiotic treatment, and three patients required antibiotic therapy for pin-site infections. Four patients required surgical intervention to deal with deeper infection. These included drainage of infected haematoma (n=1), drainage of abscess (n=1) and excision of sinus tracts (n=2). One patient developed a deep venous thrombosis in the contralateral limb necessitating 3 months warfarin therapy. There was one case of urinary retention in a child who subsequently required a dorsal slit and catheterisation. One patient had revision of internal fixation at the hip, and one required manipulation under anaesthetic to improve range of motion at the wrist.
There were no deaths related to the reconstruction, although, one patient with an osteosarcoma of the ileum, died of localised recurrence 10 months after surgery. There were no other cases of recurrence.
4.6 Stress fractures
Excluding the three failures and the patient who was lost to follow-up after 3 weeks, there were six cases of fracture of the transferred fibula (21%; 95% confidence intervals 8–41%). One was asymptomatic, being detected on a late radiograph and had healed without intervention; four required plating and bone-grafting to achieve healing and one has gone onto nonunion and is currently being treated in an Ilizarov frame.
4.7 Donor site morbidity
Three patients exhibited extensor hallucis longus weakness. In two cases this was transient, and one patient has mild residual weakness.
5. Discussion
5.1 Bony union
Overall bony union in our series was achieved primarily in 20 out of 27 patients (74%; 95% confidence intervals 53–89%) at a median time of 4.75 months, and five out of the remaining seven achieved secondary union at a median time of 18 months, with two cases of nonunion. This compares favourably with results demonstrated by other large series.
In our group of patients, there was no difference in the time to bony union for those patients undergoing reconstruction following trauma or bone tumour resection, thus our hypothesis regarding residual periosteum is not confirmed. It may be that the bone healing potential of the healthy periosteum and bone of the transferred vascular fibula was not enhanced by the possible presence of any residual periosteum in trauma cases. It may however, be due to a lack of statistical power, as the series is very small and hence, has the power to pick up only large differences.
Conversely, a delayed time to bony union may be suspected to occur in tumour patients because, despite a healthy soft tissue envelope being usually preserved during surgery, subsequent adjuvant treatment in the form of radiotherapy or chemotherapy may have a detrimental effect on union. Numbers of patients receiving adjuvant therapy were small, but it did not appear to have a detrimental effect on bony union in our tumour group.
5.2 Bone fixation
A vascularised bone graft obtains more rapid union if there is rigid fixation of the bone ends,
but fixation should not interfere with blood flow. One of the difficulties we experienced initially was holding the transferred fibula in satisfactory alignment with the recipient bone, whilst maintaining fixation. Over the years, we have used a number of fixators and have recently found the use of the modified Ilizarov frame
helpful. This consists of a system of carbon fibre rings and olive pins. Pre-operative planning requires collaboration between orthopaedic and plastic surgeons. The fixator's half-ring construction means that the anterior half of the frame can be removed to facilitate the fibula transfer surgery, providing free access to the recipient vessels, for example, the posterior tibial artery in the lower limb (Fig. 7) . It is reapplied at the end of the operation and modified in order to provide a definitive stabilisation. Gradual tightening of the frame produces compression of the bone ends and stimulates healing. The frame allows transmission of axial symmetrical load and patients can initiate weight-bearing as soon as pain allows—usually after a few weeks. As the patient progressively puts more weight through the affected limb, so the support provided by the frame can be gradually reduced by destabilisation, until the fibula has hypertrophied to the extent that the patient can fully weight-bear.
Fig. 7To maintain the fibula in correct alignment, we use a modified Ilizarov frame, designed by an orthopaedic colleague.26 During surgery to transfer the fibula, part of the frame can be removed.
This technique seems to provide firm fixation with three-dimensional adjustment, unlike unilateral fixators. It also allows fixation of smaller segments of bone. It is our experience that for example, the proximal pins on a unilateral fixator on the tibia tend to become loose over a long time and can contribute to misalignment or nonunion (Fig. 8) . Minami found that in grafts which were properly aligned, there were no fractures.
We believe that if this method had been employed for our earlier cases, then some of the complications we observed might have been reduced and the incidence of stress fracture lower.
Fig. 8Bowing of the transferred fibula can be seen in this lateral X-ray film, taken 6 years after surgery. This was due to use of a unilateral fixator.
One of the benefits of vascularised bone, including the fibula, as a reconstructive option is its ability to hypertrophy. Several factors are thought to influence this, including mechanical loading,
Although, hypertrophy has been noted by previous authors, it has not often been quantified. In those studies that have addressed measurement of hypertrophy, the variation in methods used precludes direct comparisons. For example, some authors merely measure an amount in mm
and none acknowledges the practical difficulty in determining the exact point on subsequent films to take measurements for valid comparison.
We found that the fibula hypertrophied in all cases of lower limb reconstruction, where there was mechanical stress through the recipient bone. Where there was no transmission of force, as in the one case of pseudoarthrosis of the tibia, no hypertrophy was recorded, even after 16 months. In this case, the fibula was acting as a strut to try to reduce pain. This patient had previously undergone transfer of the ipsilateral fibula into the tibia, and this had hypertrophied. The free fibula was therefore not subjected to full weight bearing forces, as the mechanical load was shared also by the previous graft. Likewise, where the fibula is acting as a strut to try and reduce the deformity caused by fibrous dysplasia of the femur, no hypertrophy was seen. Our findings corroborate those of other authors, who find that stress-shielding does not result in hypertrophy.
In three out of six cases, we could also demonstrate measurable hypertrophy in the upper limb. Previous investigators have failed to find hypertrophy in the upper limb and suggested that this was because the fibula and recipient forearm bone are of equivalent width (Fig. 9) . Fujimaki et al. conducted a series of experiments on puppies, and found that in no case did hypertrophy exceed the width of the recipient bone.
On the basis of this, they suggested that the processes of normal bone healing by callus formation, and the development of fibular hypertrophy are different, as in reactive callus formation, overgrowth beyond the contours of the original bone may occur. They believed that the size of the recipient bone was an influential factor in the amount of hypertrophy formed.
Fig. 9(A) The post-operative views of fibula transfer into the radius. The fibula appears to be the same width as the recipient bone. (B) Pre-operative views of the same patient.
We also found that in the three cases where the recipient site was the radius, and we could assess hypertrophy, none was recorded. However, in the humerus, all three transferred fibulae developed increased girth. It has been unequivocally demonstrated that hypertrophy does occur in upper limbs in response to mechanical loading.
We suggest that the hypertrophy seen in the humeri is due to the loading that occurs with daily use and exercise, the humerus transmitting the whole of the weight-bearing load in comparison to the radius, where the load is shared between the two forearm bones. Moreover, we have measured hypertrophy that extends beyond the diameter of the recipient bone for the whole length of the fibula, not just at the site of a fracture (Fig. 10) as have other authors,
and therefore we do not believe that size of recipient bone is a factor, but that load-bearing is more important.
Fig. 10Hypertrophy of the transferred fibula 9 years after surgery, is seen to extend beyond the original diameter of the recipient femur. This occurs along the whole length of the fibula, and is not the result of localised callus.
De Boer proposed three different entities of hypertrophy: ‘periosteal’ (bone formation around the graft), ‘ endosteal’ (the cortex and medullary canal have increased in diameter); and a mixture of the two.
However, his studies to calculate hypertrophy did not look specifically at measurement of these separate entities and reported series show little convincing evidence for endosteal new bone formation. Of those who comment on hypertrophy at all, the majority take it to be thickening of the cortex from periosteal formation of new bone, with no change in the medullary cavity. Indeed, Jones et al. investigating hypertrophy in the dominant humeri of tennis players,
While it would be reasonable to expect the new bone formed during hypertrophy to originate both from the periosteum and endosteum, we found it impossible to measure the endosteal diameter of the fibulae consistently with any accuracy, particularly in cases which had been followed up for longer than 5 years, due to the dense appearance of the hypertrophied bone.
Panuel et al. used imaging studies of the vascularised fibula in children to demonstrate involvement of vascular periosteum in the hypertrophy process.
They showed peripheral new bone formation, thickening of cortex and incorporation into host bone (even without weight-bearing). They proposed that asymmetrical periosteal reaction is secondary to subtle injuries of the periosteum during harvesting of the graft. We compared the amount of hypertrophy in those cases where the defect had resulted from trauma, and hence there may be residual periosteum around the bone defect, with those cases where the bone and periosteum had been completely excised for tumour. We did not find that there was any difference between these two groups and therefore if recipient bone periosteum was present around traumatic bone defects, it did not contribute significantly to graft hypertrophy.
We found that hypertrophy occurred in adults as well as children and that in our series, age was not a significant influencing factor.
5.4 Stress fracture
Our fracture rate of 21% is comparable with other studies.
Of our six cases of fracture, one was asymptomatic and picked up on later X-ray films, one occurred following a fall, one was due to asymmetrical pressure on the graft as a result of bowing and two fractured beneath internal fixation. Age and site were not significant factors. Adequate support of the fibula is required until sufficient hypertrophy has occurred to allow weight bearing. As mentioned earlier, we believe that the stress fracture rate will be reduced by use of the modified Ilizarov frame.
Double-barrelling of the fibula appears to reduce the incidence of stress fractures.
Limb salvage in this heterogeneous group of patients compares favourably with other series. It was achieved in 84% of patients (95% confidence intervals 66–95%), reconstructing bony gaps ranging from 5 to 21 cm, using free vascularised fibulas with lengths ranging from 10 to 27 cm. When donor bone was shortened, this was done sub-periosteally, to allow a vascularised periosteal cuff to drape over the ends of the bony junction. There was no significant correlation between time to bony union and the length of bony gap (correlation between time to bony union and bony gap was −0.05), demonstrating that large bony gaps can be effectively and expeditiously bridged using this technique.
A direct comparison of this technique against others is not possible without a double-blind randomised trial. However, the indications for a free fibula transfer as a reconstructive method for long bony defects should be determined by comparative review of the outcome of this procedure against that of other reconstructive techniques.
Alternative reconstruction modalities include conventional cancellous bone grafts, allografts, other vascularised bone transfers and bone transport.
Cancellous grafts are useful for smaller defects, up to 2.5 cm, obtaining union in 90% of patients.
In medium-sized defects, 2.5–6 cm, union rates have not been reported. However, it is generally accepted that union rates are lower and there are more complications in this group.
For these reasons, this technique is not suitable for larger defects. Enneking showed that for defects of 12–25 cm, a union rate of 68% was achieved, but 30% of patients required a second operation and there was a stress fracture rate of 58%.
They have the advantages of unlimited size, potential to preserve joint motion and lack of donor site morbidity. However, drawbacks include expense and the possibilities of disease transmission and host rejection. There is limited data regarding the mechanism of healing of such massive allografts,
and, although, there may be an element of ‘creeping substitution’, massive allografts may be thought of as acting as inert endoprostheses. Their success rate has been unpredictable and there is a high rate of complications such as infection, fracture, delayed union and potential for late failure.
Endoprostheses are reasonably inexpensive and provide immediate stability, rapid restoration of function and strength. There is certainly a place for them in limb salvage surgery in experienced hands,
but again, there is a high complication rate, including infection, mechanical loosening, implant failure, amputation and revision surgery and approximately 5 cm of bone either side of the defect is required for fixation.
which still has a role in reconstruction of bony defects, especially around the ankle, but its curvature limits its useful length to a maximum of 10 cm.
Comparison of vascularised iliac crest and vascularised fibula transfer for reconstruction of segmental and partial bone defects in long bones of the lower extremity.
This procedure allows the production of new bone between two vascular bone surfaces, by creating a corticotomy or traditional osteotomy, and separating the two ends gradually after a latent period of 2–10 days, by distraction.
The rate of distraction varies between 0.5 and 2 mm per day and new bone forms by a process of intra-membranous ossification. Supplementary bone-grafting is frequently needed. This practice is often successful but is a lengthy and painful treatment and the duration of treatment for some of our cases, for example, defects of over 12 cm, if osteogenesis had been used, would have been unrealistically long. Tsuchiya et al. report of treating defects up to 18 cm by this method and the mean external fixation index was 39.5 days/cm (with a range of 27.9–65.5).
Post-operative care is extensive, involving regular clinic attendances, mechanical adjustments of the Ilizarov frame with incremental lengthening, care of pin-sites and exercise. The complications associated with distraction osteogenesis can also be considerable, for example, pin-site inflammation, infection and local osteomyelitis, premature or delayed consolidation, nonunion, axial deviation, late bending or fracture. During lengthening, joint mobility may be compromised and neurovascular damage may result from pins or osteotomes or as a result of the stretching itself.
Comparisons of bone transport with other conventional techniques have shown that there are some advantages
the free fibula flap has become the most frequently used source of vascularised bone for limb reconstruction, due to its favourable characteristics and low donor site morbidity.
It is a tubular bone, formed by endochondral ossification, and it consists of dense cortical bone along its length. This, coupled with its triangular cross-section allows it to resist angular and rotational stress. It possesses a strength that allows internal fixation by a variety of methods and its diameter matches that of the radius and ulna. It can often fit into the medullary cavities of the humerus, femur and tibia, and can be raised with a skin paddle, making it a good choice for limb reconstruction.
and these include the necessity for two surgical teams, experience with microvascular technique, lengthy surgery, necessity for a donor site and the occasional need for secondary procedures. Some authors feel that the weakness of the grafted fibula is one of the drawbacks of this technique.
They use supplementary conventional iliac bone grafts. We did not do this and the hypertrophy we measured shows that the bone can reach a strength to allow return to full function without the need for supplementary grafting.
Although a large clinical series, our sample sizes for analyses are small for statistical purposes, and power to detect significant associations with the potential risk factors is low. A larger series is needed to investigate associations further.
However, in our hands, the free vascularised fibula has proved a useful tool in the salvage of limbs with large bony defects. It produced rapid bone union in the majority of cases. Our findings contribute to the theory that the free fibula has the ability to respond physiologically to biomechanical loading by hypertrophy in both the upper and lower limbs, and that hypertrophy can occur irrespective of the age of the patient. Hypertrophy can extend beyond the diameter of the recipient bone. The possible presence of residual periosteum does not appear to influence hypertrophy or bony union.
The free fibula flap is therefore a useful technique for complex reconstructive cases, where distraction osteogenesis would take an unfeasibly long time and amputation may be the only alternative. It permits the creation of large defects—allowing adequate debridement of infected bone or en bloc resection of tumour, without worsening prognosis for these patients or reducing quality of life, and with acceptable donor site morbidity. We propose that it is the gold standard for reconstruction of bony limb defects greater than 10 cm and should be considered for less sizeable defects in selected patients.
Acknowledgements
Funding for RSDU was received from DHSC (South).
References
Phemister D.B.
The fate of transplanted bone and regenerative power of its various constituents.
Comparison of vascularised iliac crest and vascularised fibula transfer for reconstruction of segmental and partial bone defects in long bones of the lower extremity.
☆This work was presented at BAPS Winter Meeting in association with the Plastic Surgery Educational Foundation of the USA, Royal College Surgeons, London, 4 December 1998.
00186-3&id=gr1.jpg){kind=link}
00186-3&id=gr2.jpg){kind=link}
00186-3&id=gr3.jpg){kind=link}
00186-3&id=gr4.jpg){kind=link}
00186-3&id=gr5.jpg){kind=link}
00186-3&id=gr6.gif){kind=link}
00186-3&id=gr7.jpg){kind=link}
00186-3&id=gr8.jpg){kind=link}
00186-3&id=gr9.jpg){kind=link}
00186-3&id=gr10.jpg){kind=link}