Limb-lengthening should be considered for patients who have a limb-length discrepancy of at least four centimeters, the potential to walk without the aid of an orthosis after the lengthening procedure, and the emotional stability to deal with the stress of prolonged treatment. Problems with the joints, such as contracture and stiffness, should be treated before lengthening is performed4,23,25. The early efforts of Putti, Abbott, and Anderson et al. have evolved into the Wagner, Ilizarov, and epiphysiolysis techniques6,14-16.
Limb-lengthening has been reported to be associated with a high prevalence of complications. Wagner documented various complications in twenty-six (45 per cent) of fifty-eight patients who had a femoral lengthening procedure. Hood and Riseborough reported a complication rate of 92 per cent in a series of thirty-eight patients.
We compared the results of the Wagner method of limb-lengthening with those of the Ilizarov method at one institution5. In addition, we compared the results of the Ilizarov method when the multiaxial Ilizarov fixator was used with those when the uniaxial Wagner fixator was used and the effects when allograft was used with those when autogenous bone graft was used.
*No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. No funds were received in support of this study.
†Department of Orthopaedics, Georgetown University Medical Center, 3800 Reservoir Road, N.W., Washington, D.C. 20007-2197.
‡Department of Orthopaedic Surgery, The Children's Hospital, 1056 East 19th Avenue, Denver, Colorado 80218-1088.
Wagner Method of Lengthening
A diaphyseal osteotomy is performed after attaching the Wagner apparatus to the bone. The bone ends are immediately distracted four to five millimeters, and the extremity is lengthened at the rate of 1.5 millimeters per day. The lengthening is suspended if the knee cannot be flexed more than 60 degrees or if there is an extension lag of 10 degrees or more. The patient is then managed with an intensive program of physical therapy to regain the range of motion before the lengthening is continued. When the desired length is reached, bone graft is placed in the gap and a plate is used to bridge the gap. Immediately after the operation, the extremity is immobilized in a cast extending from proximal to the knee to the toes. The cast is worn until the bone graft has been incorporated with the host bone. A knee-ankle-foot orthosis or ankle-foot orthosis is worn until normal muscle strength is regained. Progressive weight-bearing is encouraged while the lower extremity is immobilized in the cast and orthosis.
Ilizarov Method of Lengthening
A uniaxial (Wagner) or multiaxial (Ilizarov) external fixation device was used with the Ilizarov method of lengthening. Ilizarov used the term corticotomy to define his method for dividing the bone when the lengthening process is initiated. A narrow osteotome is introduced through a small incision down to the bone and is passed obliquely through the cortex, avoiding entry into the central portion of the medullary canal. The posterior cortex, which is not cut by the osteotome, is fractured by twisting the limb in order to complete the division of the bone. In theory, the endosteal blood supply is preserved by this technique. Ilizarov used angiograms to show that it is possible to cut the bone in laboratory animals while keeping the endosteal vessels intact.
Our method is, in fact, an osteotomy. Except for the small area of penetration, the periosteal sleeve is maintained. However, no special attempt is made to preserve the medullary vessels while the posterior cortex is cut with the narrow osteotome. The delay of three to five days before distraction is begun allows the endosteal and periosteal vessels to reconstitute. We have seen formation of bone on roentgenograms of all of our patients when this technique of osteotomy was used.
The bone is lengthened at the rate of one millimeter a day. The patient is allowed to bear weight on the involved extremity, and daily physical therapy is instituted to maintain a normal range of motion of all of the joints of the lower extremity. Once the desired length is achieved, the fixator is locked in place to permit maturation of the newly formed bone.
The Ilizarov external fixator provides a stable scaffold that permits distraction of the osteotomized ends. The device consists of two rings or more that are connected longitudinally by a threaded rod. The rod distracts the rings, which are placed on either side of an osteotomy site, thereby lengthening the extremity. The rings are attached to the bone through a series of thin wires that are passed transversely through the bone and surrounding soft tissues. Before they are affixed to the ring, the wires are tensioned with a tightening device in a manner that is similar to the tightening of the spokes on the rim of a bicycle tire. The transfixion wires can be placed anywhere along the bone as long as tendons, nerves, and vessels are avoided. The wires should be pushed through the soft tissues before drilling through bone. Irrigation of the operative site during drilling prevents thermal necrosis at the sites of the wires. Once a wire has traversed both cortices, a mallet may be used to pass it through the remaining soft tissues. Proper placement of the wires is important to maintain motion of the joint while the fixator is in place. To prevent impingement of the soft tissues, the muscles that are penetrated by the wires should be stretched at the time of transfixion. For example, passing a wire through the extensor surface necessitates flexion of the adjacent joint and, conversely, passing the wire into the flexor surface necessitates extension of the joint. After the first wire is placed, a ring of appropriate size is attached and the remaining wires are placed along the axis of the ring. The wires should not be placed obliquely in relation to the ring, as this may result in asymmetrical tension and displacement of the bone fragments during lengthening13.
Principles of Lengthening
The lengthening process is divided into two phases that are quite different from each other. In the initial phase, distraction of the bone and soft tissues is accomplished. In the second phase, consolidation and recanalization of the lengthened bone occurs. It appeared that the process of lengthening was more difficult for patients who had congenital shortening of the bone than for those who had shortening because of trauma, infection, or tumor. Various authors have expressed the amount of time needed to accomplish the desired lengthening as the lengthening rate2 (centimeters per month) or the healing index5 (days per centimeter). The index is determined by dividing the time that is needed for the bone to be fully healed and functional by the number of centimeters that the limb is lengthened. There was no measurable difference in the time needed for lengthening, but the time needed for consolidation was thirty-seven days per centimeter for the patients who had congenital shortening and thirty-three days per centimeter for the remaining patients (Table I).
In the first phase, the lengthening progresses at a rate of one millimeter per day with the Ilizarov method or 1.5 millimeters per day with the Wagner method. The use of a quarter-turn every six hours to achieve an increase of 0.25 to 0.4 centimeter in length reduces the degree of pain. Therefore, in theory, there should be a gain of one centimeter in length every ten days. In practice, the progression may be delayed by pain or by increasing restriction of joint motion. The first few centimeters of distraction progress more easily than the last few centimeters. The rate for the lengthening phase was twelve days (range, six to thirty days) per centimeter for the Ilizarov technique and nine days (range, four to twenty-one days) per centimeter for the Wagner technique.
The second phase, during which consolidation and recanalization occur, is tolerated more easily by the patient, and the irritation in the pin tracks has usually subsided. Motion of the joints and muscle strength are gradually increased as the patient has much less pain. Maturation and remodeling of bone proceed much more rapidly once weight-bearing is permitted. The consolidation phase requires approximately twice the time that is required for the distraction phase. In practice, consolidation and recanalization required twenty-six days (range, eight to eighty-three days) per centimeter when the Ilizarov technique was used and forty-four days (range, twelve to 159 days) per centimeter when the Wagner technique was used. Therefore, the external fixator was used for an average of thirty-eight days (twelve days for the first phase and twenty-six days for the second phase) per centimeter when the Ilizarov method was performed and fifty-three days (nine days for the first phase and forty-four days for the second phase) per centimeter when the Wagner method was performed.
Patients
Thirty-four patients at The Children's Hospital in Denver, Colorado, had forty-one limb-lengthening procedures performed between 1977 and 1989 (Table I). The Wagner technique of lengthening was used in nineteen patients (ten tibiae and ten femora), and the Ilizarov technique was used in eighteen patients (ten tibiae and eleven femora). The Wagner external fixation device was used in all twenty lengthening procedures that were performed with the Wagner technique and in fourteen procedures that were performed with the Ilizarov technique. The Ilizarov external fixation device was used in the remaining seven lengthening procedures that were performed with the Ilizarov technique. The method for lengthening was selected on the basis of the availability of the technique. Lengthenings in the earlier years of the study were performed with the Wagner method, and the Ilizarov method was used in the later years.
Complete demographic data were available for each patient and included the underlying cause of the limb-length discrepancy, the indications for operative intervention, the duration that the fixator was used, the use of allograft or autogenous graft, the age of the patient at the onset and at the completion of the lengthening, and the development of complications. The preoperative and postoperative length of the extremity was measured with use of orthoroentgenograms10, and the data of Anderson et al. were used to predict the length of the extremity at maturity. Anteroposterior and lateral roentgenograms were made every six weeks to document the progression of callus formation and the incorporation and remodeling of the bone graft. Additional oblique roentgenograms were made if the plate obscured the site of the osteotomy. Roentgenograms that showed a continuous cortical outline extending proximal and distal to the lengthened segment of bone suggested complete incorporation and consolidation of the segment.
Wagner Method of Lengthening
The etiology of the limb-length discrepancy in the nineteen patients (thirteen boys and six girls) who had the twenty Wagner limb-lengthening procedures was poliomyelitis (five patients); congenital short femur (four patients); fracture (three patients); fibular hemimelia (four patients); and Ollier disease, radiation therapy, and osteomyelitis (one patient each). Ten femora and ten tibiae were lengthened. Joint contractures and angulatory deformities were corrected before the extremity was lengthened. The Wagner technique was not used to lengthen the femur and tibia simultaneously in any patient. All nineteen patients had had at least one operation before the lengthening procedure, and they had an average of four procedures to correct problems, such as a residual deformity, after the lengthening. For example, patients who had congenital short extremity usually had four to nine procedures after the lengthening.
In fifteen patients, the Wagner external fixator was kept in place for approximately six weeks. In the other four, the fixator was kept in place until consolidation of the bone was complete because signs that the bone ends and the graft had incorporated obviated the need for additional operative intervention. The plate that was used for internal fixation was removed eighteen to twenty-four months after all external support was discontinued.
The average age at the onset of the lengthening was 10.2 years (range, 8.2 to 14.3 years), and the average age at the completion of treatment was 11.0 years (range, 8.8 to 15.2 years). The average duration of hospitalization was fifty days (range, twenty to 120 days). The time to consolidation of the ends of the osteotomized bone averaged 8.9 months (femur, 10.3 months and tibia, 7.5 months) (Table II).
The average initial limb-length discrepancy in these nineteen patients was 5.5 centimeters (range, 2.9 to 9.8 centimeters). The average residual discrepancy was 0.14 centimeter (range, -2.4 to +1.9 centimeters). The length of the bone increased an average of 5.4 centimeters (range, 2.8 to 8.0 centimeters). The average length of the femur before the lengthening was 31.0 centimeters (range, 22.6 to 44.5 centimeters), and the average final length was 36.8 centimeters (range, 30.6 to 49.0 centimeters). Before the lengthening, the tibia was an average of 28.6 centimeters (range, 24.0 to 34.4 centimeters) long, and the average final length was 33.7 centimeters (range, 29.7 to 39.8 centimeters). The length of the tibia increased an average of 18 per cent (range, 10 to 26 per cent), and the length of the femur, an average of 19 per cent (range, 8 to 35 per cent) in comparison with the original length. Two patients who had congenital short femur and a large limb-length discrepancy (8.0 centimeters in one and 9.8 centimeters in the other) needed two femoral lengthening procedures in sequence.
The type of bone graft that was used did not alter the duration of treatment with the Wagner method. The treatment lasted 9.5 months when the allograft was used and 9.8 months when the autogenous graft was used. The group of patients who had received allograft bone had a total of twelve complications, including delayed fracture (six patients), infection (two patients), delayed union (two patients), and angulation and non-union (one patient each). The group who had an autogenous graft had a total of ten complications, including delayed fracture (four patients), angulation (four patients), and infection and delayed union (one patient each).
Ilizarov Method of Lengthening
The underlying cause of the limb-length discrepancy in the eighteen patients who were managed with the Ilizarov method of lengthening was congenital short femur (four patients); trauma (three patients); osteomyelitis (three patients); Ollier disease (Figs. 1, 2, and 3) (two patients); Russell-Silver dwarfism (two patients); and poliomyelitis, osteogenesis imperfecta, fibular hemimelia, and fibrous dysplasia (one patient each). Twelve boys and six girls had eleven femoral and ten tibial lengthenings with the Ilizarov method. Either the Wagner or the Ilizarov external fixation device was used, according to the preference of the orthopaedic surgeon. An average of four operative procedures were performed before the lengthening. The patient who had fibrous dysplasia (Case 12) needed ten operations, and a patient who had congenital short femur (Case 5) needed fifteen operations.
The average age of the patients when the lengthening was begun was 12.5 years (range, 10.5 to 14.8 years). The average age when the process of lengthening was completed was 13.0 years (range, 11.2 to 15.8 years) (Table II). The patients were in the hospital for an average of thirty-eight days (range, three to 120 days). The external fixator was left in place for an average of 6.3 months (range, 3.0 to 14.5 months), until consolidation was seen on roentgenograms. Protective bracing was not used because all of the patients regained normal motion of the joints and were bearing full weight on the extremity with the fixator in place. An average of three (range, zero to eight) operations were performed after the Ilizarov lengthening procedure.
The initial limb-length discrepancy was an average of 4.9 centimeters (range, 1.0 to 16.4 centimeters). The average residual discrepancy of the entire extremity was 1.0 centimeter (range, -13.7 to 10.2 centimeters), that of the femur was 1.0 centimeter (range, -13.7 to 3.0 centimeters), and that of the tibia was 1.3 centimeters (range, 0.5 to 10.2 centimeters). The average increase in length was 5.9 centimeters (range, 1.0 to 16.8 centimeters) and 20 per cent (range, 4 to 72 per cent). The average length of the femur before the lengthening was 34.7 centimeters (range, 28.4 to 44.0 centimeters), and the average final length was 40.2 centimeters (range, 31.6 to 50.0 centimeters). The average length of the tibia before the lengthening was 29.8 centimeters (range, 23.0 to 35.8 centimeters), and the average final length was 36.3 centimeters (range, 25.9 to 40.1 centimeters). A simultaneous lengthening of both the femur and the tibia in one patient who had congenital short femur increased the length of the extremity by 10.7 centimeters, and one patient who had Ollier disease had an increase of 19.5 centimeters.
Autogenous Graft Compared with Allograft
Bone graft was used only in the lengthenings with the Wagner method. The type of bone graft used (allograft in nine lengthening procedures and autogenous graft in eleven) did not affect the duration of treatment, the rate of complications, or the quality of bone generated. Unincorporated allograft was seen in three extremities, when a biopsy was performed at the time that the plate was removed twelve to eighteen months after the initial operation9.
Complications
We classified every instance in which drainage from the pin sites was treated with antibiotics as a pin-track infection. All of the nerve palsies that were noted involved the peroneal nerve, and the paralysis was classified as minor or major on the basis of the final outcome. Problems of alignment of the extremity in the fixator and contractures of the knee or ankle joint were classified as minor complications because they resolved after closed manipulation with the patient under anesthesia. Major complications included delayed fracture, angulation at the osteotomy site, extended nerve palsy, subluxation of the hip or knee, osteomyelitis, delayed union, non-union, malunion, wound dehiscence, deep infection, and loss of length of the extremity.
The group of patients who had limb-lengthening with the Ilizarov method had thirteen major complications. The most common complication was an angulatory deformity (four instances). There were two instances each of delayed fracture, fracture through a pin site, and subluxation of the hip, and one instance each of skin slough of a digit, shortening, and permanent nerve palsy. There were forty-three minor complications, including thirteen pin-track infections, thirteen soft-tissue contractures, eight problems related to alignment of the fixator, six instances of premature consolidation of the ends of the osteotomized bone, and three transient nerve palsies. Early union of the bone ends before the desired length had been attained was observed only when the Ilizarov technique had been used (Table III).
The group of patients who had limb-lengthening with the Wagner method had thirty major and twenty-eight minor complications. The major complications included ten delayed fractures of the involved bone; five angulatory deformities; three instances each of extended nerve palsy, osteomyelitis, and delayed union; one instance each of non-union, deep infection, subluxation of the knee, and fracture through a pin site; and two instances of wound dehiscence. The minor complications included twelve pin-track infections, nine problems related to alignment of the fixator, four soft-tissue contractures, and three transient peroneal-nerve palsies (Table III).
We assessed the development of complications in relation to the type of fixator that was used with the Ilizarov method of lengthening. There were thirty-one minor and nine major complications with use of the Wagner fixation device. Ten were pin-track infections and eight were problems with alignment. There were twelve minor and four major complications with use of the Ilizarov fixator. Three were pin-track infections. There were no problems with alignment.
The Ilizarov technique for the lengthening of bones was developed in the former Soviet Union, and it was promoted in the West by De Bastiani7,17,20,21. Ilizarov performed a diaphyseal osteotomy and used bicycle spokes fixed to a circular frame to axially distract the early callus. Readily accessible bones, such as the tibia, can be lengthened with the use of circular fixation. However, because circular fixation cannot be used in the proximal third of the femur, a combination of half-pins inserted proximally and thin wires inserted distally is used. De Bastiani first employed this method to lengthen the limbs of patients who had achondroplasia and subsequently used the procedure in children who had a limb-length discrepancy that was due to other causes7.
During the last two decades, the Wagner technique has been the most widely accepted method for the lengthening of long bones25. In 1978, Wagner reported an average increase in length of 6.8 centimeters, with twenty-six complications in fifty-eight patients. He did not identify the rate of lengthening of the tibia, but he found no difference between the rates for the femur and tibia in thirteen simultaneous lengthenings. Hood and Riseborough found a rate of complications of 92 per cent in thirty-eight patients who had an average increase in length of 6.6 centimeters in the femur and of 4.8 centimeters in the tibia. Stephens documented an average increase of 5.7 centimeters in twenty-four patients who had associated complications that varied in severity. Paterson et al. suggested the use of a modification to the Wagner technique in an effort to lower the high rate of complications. They found that an osteotomy (referred to as a corticotomy by some authors13,17) with distraction of the bone ends after eight to twenty days could promote early osteogenesis and obviate the need for subsequent bone-grafting and use of a plate. In a study of fifty patients, Paterson et al. reported an average increase of 5.5 centimeters in the length of the limb, with a rate of complications of 24 per cent.
The histological appearance of the regenerated bone that is seen after Ilizarov lengthening is similar to that of bone seen during the process of fracture-healing or intramembranous bone formation. An inflammatory reaction develops after the osteotomy. When distraction is performed, fibroblasts appear and produce collagen along the direction of the distraction. As the distraction continues, the collagen condenses into bundles interposed with capillaries, and this vascularization promotes osteoid production. Ilizarov used biopsy specimens and an experimental model to show that, with further distraction, the osteoid matures into lamellar bone, which, in turn, remodels into cortical bone with an intramedullary canal13.
Authors who have reported that the desired increase in the length of the limb was reached within five to seven months did not document the actual length that was obtained18,24. Several authors have recommended that the bone should not be distracted more than 15 per cent of the initial length in one stage23,25. We attained an average increase in one stage of 20 and 18 per cent in the tibia and femur, respectively. Two patients with congenital short femur who both needed two serial procedures for femoral lengthening had an increase in the length of the bone of 11.1 and 12.8 centimeters. The combined rate of lengthening and healing does not double when the tibia and femur are lengthened simultaneously. Simultaneous lengthening may, in fact, slow the osteogenic potential of each bone. However, there are no laboratory data to support this belief.
We assessed the effect of the type of external fixation used with the Ilizarov method on the development of complications. The uniaxial Wagner frame is easy to apply, is fixed in place with large Schanz screws, and has a narrow range within which angular adjustments can be made. The increased prevalence of problems with alignment, and the difficulty with preventing or minimizing angular deformity, may be related to the uniaxial design of the apparatus. It is possible that the Wagner device was associated with more complications because the larger Schanz screws were more resistant to soft-tissue excursion during the lengthening process, resulting in localized ischemia and infection. This necessitated the release of the pins from time to time, which increased the pain and anxiety of the patient12. The multiaxial Ilizarov device is more difficult to apply, is fixed in place with small (1.5-millimeter) transosseous tension wires, and can be used for the correction of a multiplanar deformity as well as an angulatory deformity. The smaller pins of the Ilizarov fixator can accommodate a greater degree of soft-tissue stretching and are associated with less tissue ischemia. Release of the pins is not necessary because they cut their own track. These pins also resist bending because they are set at a tension of 90 to 100 pounds (forty-one to forty-five kilograms). However, they can interfere with the function of the quadriceps and hamstrings as well as cause pain and loss of motion in the knee joint during the period of lengthening because they cross the entire diameter of the limb.
The greatest differences between the two types of fixators were the number of pin-track infections and the problems of angulation. The Wagner fixator was used in fourteen lengthening procedures that were done with the Ilizarov technique, and the Ilizarov fixator was used in seven. There were ten pin-track infections and eight instances of malalignment with use of the Wagner apparatus, whereas there were only three pin-track infections and no problems with alignment with use of the Ilizarov fixation device.
The use of allograft or autogenous bone graft in the patients who had the Wagner lengthening did not greatly alter the number and type of complications. There was a total of twelve complications in the group of patients who had an allograft and ten complications in the group of patients who had an autogenous bone graft.
The increase in length that can be obtained with the Ilizarov method is not influenced by the type of external fixator that is used. A shorter hospitalization is needed with the Ilizarov technique because the patient can monitor the blood pressure and carry out the distraction at home, with the help of ancillary services. The care of the pins and physical therapy can also be provided on an outpatient basis. The major advantages of the Ilizarov method are the elimination of the need for a second procedure such as bone-grafting and application and removal of a plate as well as a decrease in the number of procedures that are needed after the lengthening. These factors, in turn, may decrease the psychological stress associated with repeated hospitalizations12. There were also fewer complications associated with the use of the Ilizarov external fixator. We used the Ilizarov device when rotation and angulation had to be corrected simultaneously and for isolated tibial lengthening. We believe that a multiaxial deformity needs to be treated with a multiplanar device to achieve and maintain correction. We prefer to use the Wagner device for femoral lengthening with the Ilizarov method of distraction because we found that this combination made it easier to regain motion of the knee.
The primary problems during placement of the external fixator include soft-tissue impingement, instability of the fixator frame, and malalignment. Soft-tissue problems can be avoided by careful placement of the wires, and skin releases around pin sites at the time of the operation can help to prevent pin-track infections. Preoperative planning and the use of fluoroscopy or plain roentgenograms in the operating room can help to prevent problems of malalignment.
In order to avoid fractures through the segment of lengthened bone, the quality and quantity of bone that is generated must be sufficient to ensure stability. It is often difficult to determine accurately when the bone is sufficiently mature to permit unrestricted weight-bearing and when the external fixation device can be removed safely. The initial amorphous mass of ossification that is seen on the roentgenograms gradually remodels into a tubular structure resembling the diaphyseal portion of a long bone. We call this process recanalization as it is similar to the process by which a new canal is formed within thrombosed veins. Not only is it important that the surface cortex be formed but the medullary canal must be formed as well. To our knowledge, there have been no reports of late fractures or progressive angulation when the external fixator was removed after the formation of the medullary canal. On the basis of the empirical observation that weight-bearing stimulates the process of maturation of bone, patients should achieve full weight-bearing on the involved extremity without external support at least six weeks before the device is removed.
We did not divide the complications in our study into problems, obstacles, and complications as some authors have recommended17,21. If we had done so, the prevalence of complications would have been 25 per cent, and the minor complications would have been considered problems while the major complications would have been considered obstacles and complications. The prevalence of complications (1.5 per lengthening) in our series may appear to be high because every instance of the use of antibiotics was classified as an infection. In addition, the inclusion of patients who had congenital short femur increased the prevalence of angulatory deformities. The large number of complications that were seen in patients who had congenital short femur may be related to the presence of associated anomalies such as the absence of an anterior cruciate ligament, lateral angulation of the lower extremity, and soft tissues that were relatively resistant to lengthening4,8,25.
Studies have shown that lengthening of an extremity has the potential to provide the patient with equal limb lengths. Although limb-lengthening procedures are associated with impressive gains in length, our study shows that the rate of complications is equally impressive. Therefore, great caution and consideration must be exercised before embarking on the correction of a major discrepancy in limb length.