Deformity of the limb and limb-length discrepancy are well recognized sequelae of osteogenesis imperfecta. Operative treatment has usually consisted of multilevel osteotomy and placement of an intramedullary rod10,12. Recently, the Ilizarov fixator and method of lengthening7 have been used more frequently to correct limb-length discrepancy and multiplanar deformities of the lower extremity simultaneously. Several questions have been raised about the feasibility of using this method in the treatment of deformities of the lower limb in patients who have osteogenesis imperfecta. There have been concerns that the dysplastic bone may not withstand the presence of pins in the long term and that the regenerated bone may contain the same abnormal collagen as the original bone contained.
We report the results of the use of the Ilizarov method of lengthening to correct limb-length discrepancy and angular deformity in six patients who had osteogenesis imperfecta.
*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 Orthopaedic Surgery, Massachusetts General Hospital, ACC 529A, 15 Parkman Street, Boston, Massachusetts 02114.
‡Department of Orthopaedic Surgery, George Washington University Medical Center, 2150 Pennsylvania Avenue, N.W., Washington, D.C. 20037.
§Fondren Orthopedic Group, Joe W. King Orthopedic Institute, 6560 Fannin, Houston, Texas 77030.
¶Children's Hospital of Michigan, Wayne State University School of Medicine, 3901 Beaubien Boulevard, Detroit, Michigan 48201.
#Deceased.
We evaluated the results of the Ilizarov method for the correction of multiplanar deformity of the lower extremity in six (one male and five female) patients, who had type-I osteogenesis imperfecta, as characterized by blue sclerae and no deformity or fractures of the limbs at birth according to the criteria of Sillence et al. The average age was thirty-one years (range, fourteen to fifty-one years). All of the patients reported a decrease in the frequency of fractures with the end of puberty. However, one (Case 5) was still reporting the occurrence of fractures at the time of the index operation.
As there are no guidelines for the use of the Ilizarov method of limb-lengthening in patients who have osteogenesis imperfecta, we exercised great caution in the management of these patients. A slower rate of distraction, a longer-than-normal period of external fixation, and immobilization after the removal of the fixator were some of the measures that were employed to avoid complications.
Seven bones, four tibiae and three femora, were treated with the Ilizarov method. Five had a complex deformity: one femur had 30 degrees of posterior angulation, and the four tibiae had an average deformity of 28 degrees (range, 20 to 40 degrees). Three tibiae had a procurvatum deformity (anterior convexity), and the fourth tibia had a distal valgus deformity. Five of the six patients had an average of 5.6 centimeters (range, two to ten centimeters) of limb-length discrepancy. The sixth patient had generalized shortening of all of the long bones of both lower extremities as well as a non-union of the femur following multiple osteotomies and placement of an intramedullary rod for correction of an angular deformity. The Ilizarov circular external-fixator construct8 (Smith and Nephew Richards, Memphis, Tennessee) was used in five patients. The Orthofix monolateral femoral external fixator (EBI, Parsippany, New Jersey), designed to be used according to the technique of De Bastiani et al.1,3, was used in one patient.
The roentgenographic quality of the regenerated bone was evaluated with the criteria of Catagni. Radiodense new bone was seen to extend from the osteotomized bone ends into the radiolucent central area of distraction between twenty and thirty days after the initial operation. When distraction was completed, the bone columns bridged the central radiolucent area and the regenerated bone appeared more homogeneous. Maturation of the new bone then proceeded with the gradual formation of an external cortex and an intramedullary canal. The new bone was called hypotrophic when there was a delay in the appearance of radiodense new bone beyond thirty days after the initial operation, when there were breaks or deficient (radiolucent) areas within the bone columns, or when there were external concavities (an hourglass shape) or large focal external radiolucent areas in the region of the newly regenerated bone.
The patients were followed for an average of three years and four months (range, one year and seven months to six years). All six patients expressed over-all satisfaction with the outcome. The average correction of the deformity was 23 degrees (range, 20 to 30 degrees). Complete correction was achieved in three patients (Cases 2, 4, and 5), while two (Cases 3 and 6) had residual angular deformities of 15 and 10 degrees, respectively. One patient (Case 1) had no angular deformity.
The seven limb segments gained an average of 6.6 centimeters (range, two to eleven centimeters) in length. The average gain was 5.8 centimeters (range, two to nine centimeters) after the four tibial lengthenings and 7.7 centimeters (range, four to eleven centimeters) after the three femoral lengthenings. The preoperative limb-length discrepancy was eliminated in all but one patient (Case 4), who was left with a two-centimeter residual discrepancy after an eight-centimeter femoral lengthening.
The average time from placement of the external fixator to full, unsupported weight-bearing was twelve months (range, five to twenty-three months). The healing index was calculated by dividing the number of months from the initial operation until full, unsupported weight-bearing by the total number of centimeters of lengthening9. The average healing index was 2.12 months per centimeter (range, 1.3 to 4.00 months per centimeter). The average healing index was 2.34 months per centimeter (4.00, 1.45, 1.89, and 2.00 months per centimeter in Cases 2, 3, 5, and 6) after the four tibial corrections with the Ilizarov technique and 1.68 months per centimeter (1.3 in Case 4 and 2.05 in Case 5) after the two femoral corrections with the Ilizarov technique.
There were eighteen complications, which included stiffness of the knee in two patients; a peroneal nerve palsy in two; a superficial pin-track infection in three; and a deep pin-track infection, greater-than-normal loss of blood intraoperatively, loosening of two pins, worsening of the instability of the knee, and an infection in the knee in one patient each. In one patient, a Rush rod that had been placed before correction of the deformity migrated proximally and was removed after the correction was completed. There were five fractures. One occurred through the newly regenerated bone in a patient from whom the fixator had been removed before a distinct intramedullary canal and cortices had developed; one fibula fractured at the site of insertion of a wire; one femur fractured during placement of an intramedullary rod after removal of the external frame; and two fractures (one through the site of a pin for the femoral external fixator and the other in the untreated tibia) occurred in the one patient who was still reporting the occurrence of fractures at the time of the index operation.
Follow-up roentgenograms revealed osseous union of all of the lengthened segments, the femur with the non-union (Case 5), and the four fractures that occurred after the operation on the extremity (Cases 1, 4, and 5). The roentgenographic appearance of the regenerated bone in each patient was similar to the host bone adjacent to the site of the corticotomy (Figs. 1-A, 1-B, 1-C, 1-D, 1-E, 1-F, 1-G through 1-H).
CASE 1. A fourteen-year-old girl had four centimeters of shortening of the right lower extremity, without an angular deformity, as a result of three fractures (one of the femur and two of the tibia). She was able to walk, ride a bicycle, and swim without difficulty, but she desired correction of the limb-length discrepancy. The discrepancy was corrected completely with use of the technique of De Bastiani et al.1,3. The external fixator was in place for five months and one week, and the healing index was 1.30 months per centimeter. Complications included loss of mobility of the ipsilateral knee and a fracture through the newly regenerated bone. The preoperative range of motion was regained after manipulation of the knee with the patient under anesthesia, at the time that the external frame was removed, and a program of physical therapy. The fixator was removed before a distinct medullary canal and cortices were seen on the roentgenograms, and the patient sustained a fracture of the femur five days after the removal. The frame was reattached and left in place for twelve weeks, until there was evidence of healing of the fracture and the regenerated bone was mature. At the follow-up examination six years after the operation, the patient had regained the preoperative functional status and the extremity appeared more normal.
CASE 2. A thirty-seven-year-old woman who had two centimeters of shortening and a 20-degree valgus angulation of the distal aspect of the tibia, secondary to an old injury of the distal tibial physis, desired correction of the deformity. She could walk only very short distances because of pain in the right ankle, and she used a cane. The limb-length discrepancy and angular deformity were corrected completely with use of the Ilizarov technique. The fixator was in place for six months, after which the extremity was placed in a below-the-knee cast for two months. The resulting healing index was 4.00 months per centimeter. Complications included a superficial pin-track infection, which resolved with the oral use of antibiotics; a fracture of the fibula at the location of a smooth Ilizarov wire, which did not interrupt the correction of the deformity in the adjacent tibia and which healed with the frame in place; and an infection of the knee joint, which was treated successfully with serial aspirations and lavages as well as parenteral administration of antibiotics. At a follow-up evaluation two years and ten months after the operation, the patient was walking without the use of assistive devices but she complained of persistent pain in the ankle with prolonged weight-bearing.
CASE 3. A fifty-one-year-old woman had sustained twenty fractures of the right tibia as a child and as an adolescent, resulting in a residual shortening of nine centimeters and a procurvatum deformity of 40 degrees (Figs. 1-A, 1-B, 1-C, 1-D, 1-E, 1-F, 1-G through 1-H). She had difficulty walking even with the use of an orthosis and shoe-lift, and she desired correction of the deformity. An arthrodesis of the spine had been done previously for scoliosis. The patient also had an unstable knee joint and roentgenographic evidence of osteoarthrosis.
Complete correction of the limb-length discrepancy and 25 degrees of correction of the procurvatum deformity were achieved with use of the Ilizarov technique. The fixator was in place for ten months, after which the extremity was immobilized in an above-the-knee cast for three months. The healing index was 1.45 months per centimeter. The patient used an external knee brace for support during the period of lengthening and for six months after the correction of the deformity. She had symptoms of transient dysfunction of the peroneal nerve during the period of lengthening. The rate of correction was thus slowed to one-quarter millimeter per day for a two-week period, followed by a gradual return over a one-month period to a rate of one millimeter per day; the symptoms resolved. Three years and nine months after the operation, the patient was walking without the use of assistive devices.
CASE 4. A twenty-nine-year-old woman had sustained numerous fractures of the left femur and tibia, resulting in a ten-centimeter limb-length discrepancy and a 30-degree posterior angulation of the distal third of the femur at the distal limit of a previously placed intramedullary rod. She had to use a ten-centimeter shoe-lift and two crutches to walk, and she desired correction of the deformity. With use of the Ilizarov technique, complete correction of the angular deformity and correction of the limb-length discrepancy to within two centimeters of that of the contralateral limb (a gain of eight centimeters in length) were achieved. Distraction proceeded at a rate of 0.5 millimeter per day in two daily increments. The fixator was in place for ten months, and the healing index was 1.3 months per centimeter.
At the time of removal of the external fixator, an attempt was made to place a custom Russell-Taylor proximal locking intramedullary rod to provide more stability as well as to prevent another fracture. Insertion of the rod was difficult, and the femur fractured at the site of a cerclage wire that had been placed during a previous procedure for fixation of a fracture. The new fracture was successfully treated with placement of the extremity in a modified AO external fixator. The patient had another complication in that a Rush rod, inserted before the index procedure, migrated proximally and caused pain in the hip. It had to be removed after completion of the correction. At a follow-up examination at one year and seven months, the patient was walking without assistive devices, except for a single crutch only when she walked a long distance or in a crowded environment.
CASE 5. A twenty-three-year-old woman had sustained multiple fractures of both lower and both upper extremities. At the time of presentation, she had generalized bowing and shortening of the long bones of both lower extremities. She was only four feet (1.2 meters) tall and wanted bilateral limb-lengthening. There was a non-union of the mid-part of the femoral shaft despite placement of an intramedullary rod in the right femur after a previous procedure for correction of the deformity. She also had a 20-degree anteromedial angular deformity of the right tibia. She could walk short distances with a walker.
The intramedullary rod was removed and replaced with a circular external fixator to apply axial compression to the ends of the fractured bone. Once the fracture had healed, the frame was readjusted, a corticotomy was performed in the distal third of the femur, and distraction was begun at the rate of one millimeter per day (in four daily increments of 0.25 millimeter) at both sites. The limb was lengthened eleven centimeters. Eighteen months after the operation, the external fixator was removed and a Hex-fix monolateral external fixator (Smith and Nephew Richards) was placed on the femur for further immobilization. This remained in place for four and one-half months, and the resulting healing index was 2.05 months per centimeter. Complications related to the femoral lengthening included a superficial pin-track infection, which resolved with oral administration of antibiotics; loosening of two pins, one of which was removed; a non-displaced fracture through a pin site in the distal part of the femur one week after removal of the Orthofix device, which healed without deformity with immobilization in a cast; and loss of mobility of the ipsilateral knee. The range of motion of the knee was regained after manipulation with the patient under anesthesia, at the time that the frame was removed, and with physical therapy.
With use of the Ilizarov technique, the angular deformity of the left tibia was corrected completely and the extremity was lengthened nine centimeters. The fixator was in place for fifteen months, and a cast was worn for two months. The healing index was 1.89 months per centimeter. A non-displaced fracture was found in the right tibia and was treated with an above-the-knee cast until it healed. Three years and nine months after the femoral procedure and three years and four months after the tibial procedure, the patient still used a walker.
CASE 6. A thirty-four-year-old man had sustained numerous fractures and resultant mild deformity of all of the long bones of the lower extremities. He had a 30-degree anteromedial angular deformity of the left tibia and three centimeters of shortening of the extremity. He could walk short distances with the aid of two crutches, but he wanted the deformity to be corrected. With use of the Ilizarov technique, complete correction of the limb-length discrepancy and a 20-degree correction of the angular deformity (with 10 degrees of residual deformity) were achieved. The fixator was in place for four months, and a cast was used for two months. The resulting healing index was 2.00 months per centimeter. Complications included intraoperative loss of 500 milliliters of blood from the proximal fragment. The blood loss was controlled with pressure, collagen, and topical administration of thrombin.
Postoperatively, the patient had sensory changes in the first web space of the ipsilateral foot, which resolved spontaneously in the succeeding seventy-two hours. The patient also had a superficial pin-track infection, which was treated with oral administration of antibiotics. In addition, a chronic draining sinus developed in association with a ring sequestrum at the site of a smooth Ilizarov wire. Staphylococcus aureus grew on culture of specimens of the drainage, and despite oral treatment with repeated courses of antibiotics the drainage was still present at the follow-up examination two years and four months after the operation. At that time, the patient could walk without the use of assistive devices, although he did use a single crutch when walking a long distance or in a crowded environment. He continued to have pain at the site of the drainage.
The techniques of distraction histogenesis, a term coined by Ilizarov, are based on the concept that tissues regenerate when subjected to tension11-13. The method involves minimum exposure of tissues, and weight-bearing is encouraged while the extremity is immobilized in the external frame. These factors are believed to hasten the recovery of muscle tone and strength4. The technique has made it possible to correct the most complex multiplanar deformity and simultaneously lengthen the shortened extremity7. Although there is a high prevalence of complications with this technique, there are few alternatives, and patients who have functional limitations as a result of deformities may consider it worthwhile to accept the risks associated with these complex procedures.
The use of the Ilizarov technique in patients who have underlying skeletal dysplasia has raised a concern that both the quantity and the quality of the regenerated bone may be lacking. In patients who have osteogenesis imperfecta, the newly formed bone, like the host bone, is fragile as a result of the production of abnormal type-I collagen, which leads to abnormal structure and mineralization of the organic matrix. However, cellular proliferation and migration are apparently normal19,20. Despite the lack of information about the effects of stress and tension on the bone of patients who have osteogenesis imperfecta, our results suggest that the newly regenerated bone has a roentgenographic appearance that is similar to that of the adjacent host bone and is able to withstand normal functional loading. We believe that the fracture that occurred through the newly regenerated bone soon after removal of the external fixation frame in one patient (Case 1) was probably a result of premature removal of the frame rather than of abnormal bone. On the basis of the roentgenographic criteria of Catagni, the regenerated bone had a density similar to that of the adjacent host bone.
The healing index in the present study was longer than that in previous reports of limb-lengthening9, but the increased time was secondary to the simultaneous correction of the deformity and the limb-length discrepancy. The lack of information about the effect of this technique in patients who have osteogenesis imperfecta may have also caused the surgeons to leave the fixator frame on for a longer period.
The complications that occurred in this group of patients were substantial, but many reflect the difficulties associated with the Ilizarov technique9. Problems related to the pins, stiffness or instability of joints, and transient nerve dysfunction have been reported in other studies of this technique9. However, several complications in our study may have been related to the underlying osteogenesis imperfecta. One patient (Case 6) had intraoperative hemorrhage, a problem previously identified during spinal arthrodesis in patients who had osteogenesis imperfecta13. Fragility of the bones proved to be a problem in one patient (Case 5), who was still having frequent fractures at the time of the index procedure. This patient had a fracture in the untreated tibia during the rehabilitation phase as well as a fracture through a previous pin site in the treated extremity.
Our patients who had a severe deformity and only mild fragility of the bones (Cases 2, 3, 4, and 6) ultimately had a good result, and the complications that occurred did not detract from the improvement in function. However, for the patients in whom the indications for operative intervention were less compelling (Cases 1 and 5), the difficulties, the expense, and the minimum improvements in function and appearance made the risks unjustifiable. Because of the technical difficulties that were encountered and the increased risk of complications, we consider these procedures to be contraindicated in patients who are still having frequent fractures (Case 5).
While the Ilizarov technique provides a viable option for the correction of deformity in patients who have osteogenesis imperfecta, for whom there is no other alternative, surgeons should be mindful of the difficulties that we encountered in this series. It appears that the fixator may need to be left in place for a longer period, until the regenerated bone clearly appears mature on roentgenograms. In addition, the extremity should be protected after the frame has been removed, with use of external support such as a cast or brace, with care taken to avoid overloading of the contralateral lower limb during rehabilitation. This technique is probably contraindicated in patients who have a greater degree of fragility of the bones (as judged by the age at the time of the first fracture and the total number, frequency, distribution, and circumstances of occurrence of the fractures) and in those who continue to sustain fractures.
NOTE: The authors thank Dr. Cato T. Laurencin and Dr. Anam K. Kour for their contributions to this work.