The study design was a retrospective review and was approved by our hospital's institutional review board. Between May 1995 and January 2003, fifty-one individuals with idiopathic short stature and no history of medical illness, fracture, soft-tissue compromise, or bone infection underwent simultaneous bilateral leg lengthening at our institution. We lengthened the bones when the individual's height was below the third percentile for age and sex according to the standards for normal growth and development reported by the Korean Pediatric Association. The third percentile for men and women who are eighteen years or older is 163.5 and 152.5 cm, respectively.
It was the practice at our institution to lengthen the tibia only when the individual indicated that short stature had generated difficulties in carrying out activities of daily living, such as conducting business at counters, using public restrooms and public transportation, participating in sport activities, and engaging in hobbies that involved physical activities. Before being accepted as a candidate for leg lengthening, the individual had to undergo a psychological evaluation to determine the impact of short stature on his or her physical and social activities and the personal motivation for undergoing the lengthening procedure. The nature of the procedure, such as the duration of the treatment, the projected amount or limitation of the lengthening, and the potential complications or risks related to the surgery, was explained in detail. Individual factors such as the ability to understand the device and to tolerate the lengthening procedure and whether the individuals had the emotional maturity and good judgment to comply with follow-up and physical therapy requirements were assessed. Individuals with dysmorphophobia, neurosis, or another major depressive illness and those who had unrealistic goals regarding the gain in height were excluded after examination by a psychologist. All individuals on whom we operated were highly motivated and fully informed and understood the procedure and possible complications.
Individuals who had undergone conventional Ilizarov lengthening in one leg and lengthening over an intramedullary nail in the contralateral leg were excluded. A total of eighty-eight tibiae in forty-four individuals were included in this study. There were twenty-four men and twenty women with an average age of 22.7 years (range, eighteen to thirty-four years) at the time of surgery. The mean preoperative height (and standard deviation) was 153.7 ± 7.2 cm, with men averaging 158.8 cm (range, 152.0 to 163.1 cm) and women averaging 147.2 cm (range, 140.0 to 152.3 cm). Thirty-two leg-lengthening procedures in sixteen individuals were performed with use of the conventional Ilizarov method throughout both the distraction and the consolidation phases (Group A), and fifty-six lengthening procedures in twenty-eight individuals were performed over an intramedullary nail (Group B). The choice of procedure was based on the preference and informed consent of the individual and his or her family. However, as the National Public Health System and the private health insurance companies did not cover the cost of lengthening over an intramedullary nail, most of the individuals who underwent lengthening with the conventional Ilizarov method chose the procedure after consideration of their economic status. None of the individuals had an axial deformity of the tibia requiring acute or gradual correction.
The average age at the time of surgery was 23.4 years (range, eighteen to thirty-four years) in Group A and 22.3 years (range, eighteen to thirty-two years) in Group B. The mean preoperative height in Group A was 149.1 ± 7.8 cm, with men averaging 160.7 cm (range, 152.0 to 163.1 cm) and women averaging 146.2 cm (range, 140.0 to 152.3 cm). The mean preoperative height in Group B was 156.1 ± 5.6 cm, with men averaging 158.5 cm (range, 152.0 to 163.1 cm) and women averaging 148.9 cm (range, 145.0 to 152.2 cm). The average duration of follow-up after the completion of the lengthening procedure (complete consolidation of the regenerate) was forty-eight months (range, thirty-five to sixty-two months) in Group A and forty months (range, twenty-nine to fifty-nine months) in Group B.
Surgical Techniques and Postoperative Management
The conventional method included application of a tibial construct with three or four rings. With use of a multiple-drill-hole technique, fifteen monofocal and seventeen bifocal tibial corticotomies were performed distal to the tibial tuberosity and/or at the supramalleolar level. An osteotomy was done at the junction of the middle and distal thirds of the fibula.
To be treated with lengthening over an intramedullary nail, an individual had to have a tibial medullary diameter of at least 8 mm. Our technique was similar to that described by Herzenberg and Paley6, except that we inserted, without reaming, an AO tibial nail with a diameter 1 mm less than that of the tibial isthmus. In most cases, unreamed nailing could be performed without jamming. However, in order to make passage of the nail less traumatic by removing irregularities on the endosteal surface, "limited reaming" with a single pass of a reamer was done in seven tibiae with a tightly fitted 8-mm-diameter nail at the first trial of insertion. An 8-mm-diameter nail was used in forty tibiae, and a 9-mm-diameter nail was used in sixteen. Two proximal interlocking screws were inserted in a mediolateral direction. A preconstructed Ilizarov frame with two rings connected with telescoping rods was then applied, parallel to the nail. Two proximal tensioned wires were inserted posterior to the nail, and at least one wire at each ring passed the fibular head or the distal part of the fibula to prevent inadvertent migration of a fibular segment during lengthening. A tibial corticotomy was performed at the metaphyseal-diaphyseal junction with use of a multiple-drill-hole technique (Figs. 1-A and 1-B).
In both groups, lengthening was initiated on the seventh or tenth postoperative day, with a rate of 0.25 mm four times daily at each distraction site. Radiographs were made every week during the distraction phase and every four weeks during the consolidation phase. Callus formation was determined to have occurred when new bone formation was seen in the distraction gap on radiographs, especially on the lateral view. Individuals were allowed to bear partial weight with the use of two crutches. Patients treated with the conventional method had the fixator removed when more than three cortical regeneration bridges were confirmed on the anteroposterior and lateral radiographs. In patients treated with lengthening over a nail, when the desired length had been achieved, two distal interlocking screws and one distal tibiofibular transfixing screw were inserted, and the Ilizarov fixator was removed (Fig. 2). The distal tibiofibular transfixing screw was removed after consolidation of the fibula. The individuals were then examined every two months for the first year and every six months thereafter. The nail was removed on an elective basis at the request of the patient.
Data Analysis
The groups were compared with regard to several parameters: the increase in the length of the tibia, expressed in centimeters and as a percentage of the original length; the external fixation index (the duration of external fixation, in months, for each centimeter of lengthening); and the healing index (the time, in months, between the application of the external fixator, with or without a nail, and the radiographic appearance of consolidation for each centimeter of lengthening). As the imbalance between muscle forces on different sides of the bone or the instability of a construct can cause gradual deviation of the limb segment being lengthened2, we also assessed any axial deviation of the regenerate encountered during the lengthening procedures. Complications were defined, according to the system described by Paley2, as problems (difficulties that arise during treatment that can be fully solved nonoperatively), obstacles (difficulties that arise during treatment that can be fully solved operatively), or sequelae (difficulties that remain unresolved at the end of the treatment period).
Using our institution's lengthening protocol and patient-oriented evaluation form at each follow-up visit, we assessed three specific concerns related to the functional status of these patients: mobility during the lengthening procedure (the ability to carry out, without another's help, outdoor activities such as using public transportation or getting around the neighborhood as compared with the ability to perform only indoor activities or staying indoors most of the day)12, the time until the individual returned to previous activities after completion of the lengthening procedure (within six months compared with more than six months), and physical activity at the time of final follow-up (no limitations in any activity compared with some limitations in vigorous activities or strenuous sports such as running, lifting heavy objects, and mountain climbing)12. In addition, all individuals separately rated their satisfaction with the overall outcome according to five categories: extremely satisfied, moderately satisfied, no opinion, moderately unsatisfied, and extremely unsatisfied12.
The differences between the groups with regard to the heights of the men and women were determined with the Mann-Whitney U test. The independent t test or the Mann-Whitney U test was used to analyze the differences between the groups with regard to the radiographic parameters and the numbers of complications, and the Fisher exact test was used to analyze the differences in the proportions of patients responding to each category of the questionnaire. A p value of <0.05 was regarded as significant.
The mean final height (and standard deviation) of the patients was 160.1 ± 7.0 cm, with men averaging 165.3 cm (range, 157.4 to 172.1 cm) and women averaging 153.3 cm (range, 145.0 to 160.2 cm). The mean postoperative height in Group A (conventional lengthening) was 155.1 ± 7.9 cm, with men averaging 166.7 cm (range, 162.0 to 172.1 cm) and women averaging 152.3 cm (range, 145.0 to 160.0 cm). The mean postoperative height in Group B (lengthening over an intramedullary nail) was 162.6 ± 5.8 cm, with men averaging 165.1 cm (range, 157.4 to 171.8 cm) and women averaging 155.0 cm (range, 151.5 to 160.2 cm). There were no differences in the heights of the men or women between the groups (p > 0.05).
The mean gain in tibial length in the total series was 6.2 cm (range, 2.5 to 8.4 cm), which represented a mean increase of 20.0% (range, 10.0% to 27.7%). An increase in length of 5.9 cm (range, 2.5 to 8.4 cm), or 21.3% (range, 10.0% to 32.7%), was achieved in Group A, and an increase of 6.4 cm (range, 3.5 to 8.0 cm), or 19.8% (range, 11.5% to 27.0%), was achieved in Group B (p > 0.05). The mean duration of external fixation was 12.4 months (range, 4.0 to 27.0 months) in Group A and 5.4 months (range, 3.0 to 10.0 months) in Group B, resulting in a mean external fixation index of 2.2 mo/cm (range, 0.8 to 6.0 mo/cm) in Group A and 0.9 mo/cm (range, 0.4 to 1.5 mo/cm) in Group B (p < 0.001). The mean healing index was 2.1 mo/cm (range, 1.0 to 5.7 mo/cm) in Group A and 1.7 mo/cm (range, 0.9 to 2.7 mo/cm) in Group B (p > 0.05) (Table I).
Group A had a total of eighty-two complications, including nineteen problems, sixty obstacles, and three sequelae, whereas Group B had a total of sixty-nine complications, including thirty-five problems, thirty-three obstacles, and one sequela (Table II). The average numbers of complications in Groups A and B were 2.56 and 1.24 per tibia, respectively (p < 0.001). Pin-track infection requiring conservative local measures and the administration of antibiotics developed in nine cases (28%) in Group A and in thirteen (23%) in Group B. However, no deep intramedullary infection occurred in either group. Transfixing wire(s) broke in thirty-eight cases in Group A (1.2 per tibia) and in twenty-two in Group B (0.4 per tibia). The wires usually broke at their junction with a bolt. Broken wire was exchanged with new wire and augmentation of a wire to the ring was performed in these cases so that effective distraction would not be jeopardized. A broken interlocking screw was found in seven (13%) of the fifty-six tibiae treated with a nail. However, all of these tibiae were found to be in a fairly advanced stage of consolidation, and the breakage of the screw did not seem to have affected the formation or consolidation of the regenerate (Fig. 3). There was also no shortening of the lengthened tibia due to screw breakage.
Axial deviation of the tibial regenerate was not observed clinically or radiographically in Group B. However, angulation at the lengthening site occurred in five cases (16%) in Group A. All five were treated with adjustment of the frame and tightening of wires and nuts, and an additional correctional osteotomy was performed in three of the five. Autologous bone-marrow injection or cancellous bone-grafting was performed because of delayed consolidation or atrophy of the new bone in five cases (16%) in Group A. Premature consolidation occurred in four cases (13%; two tibiae and two fibulae) in Group A and in five (9%; two tibiae and three fibulae) in Group B. All cases of premature consolidation, except for two in the fibula, were encountered within three weeks after the start of distraction and were treated with a repeat corticotomy. The premature tibial consolidations in Group B were associated with a nonparallel alignment of the distraction rods of the Ilizarov frame and the 8-mm-diameter nail, and with resultant eccentric distraction of the early regenerate (Fig. 4); these cases were rectified with reapplication of the Ilizarov apparatus.
Inferior subluxation and distal migration of the fibular head due to cutting out of a wire was seen in two cases in Group A and two in Group B. This was determined by comparing the positions of the fibular head relative to the tibial plateau on radiographs made at different time-points. This problem was treated with reinsertion of a wire. Ankle valgus with an inclination angle of the tibiotalar joint of >5° was observed in three cases in Group A and in two cases in Group B. This was due to a lengthening discrepancy between the tibia and fibula (two cases each in Group A and Group B) (Fig. 5) or to nonunion of a lengthened fibula and its subsequent proximal migration (one case in Group A). Fibular lengthening was performed with use of the Ilizarov method in two cases, and the others were not treated.
Three knee flexion contractures (9%) and ten ankle plantar-flexion contractures (31%) developed in Group A. The range of knee motion was fully regained through physiotherapy in all three cases. Two ankle contractures were treated with Achilles tendon lengthening. Five cases resolved after application of an additional calcaneal ring and distraction of the regenerate bone; in two of them, progressive shortening of the regenerate was performed before application of a calcaneal ring. The range of ankle motion was fully regained with physiotherapy in one case, and two ankle contractures were not treated because the patients refused treatment. Four knee flexion contractures (7%) and ten ankle flexion contractures (18%) developed in Group B, and all but three fully resolved with physiotherapy. Achilles tendon lengthening was performed in three cases and was successful in two of them; in the third case, the limitation of motion was not fully corrected despite the tendon lengthening. Neurapraxia of the peroneal nerve during distraction of the regenerate developed in three cases (9%) in Group A and in four (7%) in Group B; however, all but one resolved after the rate of the distraction was slowed. In one case in Group A, a wire was reinserted after decompression of the peroneal nerve around the fibular neck and the surrounding muscles.
During the consolidation phase, the percentage of individuals who were able to perform outdoor activities without another's help was higher in Group B than in Group A (p = 0.004) (Table III). Although, compared with the individuals in Group A, the individuals in Group B tended to return more quickly to previous activities after completion of the lengthening procedure, this was not a significant difference. Many individuals had no limitation in physical activities at the time of final follow-up; however, 25% of each group was found to have some difficulties in pursuing vigorous activities or strenuous sports. Twelve individuals (75%) in Group A and twenty-two (79%) in Group B were satisfied with the overall result.
The Ilizarov method is a well-established technique for correction of limb-length discrepancy and for the lengthening of shortened limbs. However, soft-tissue transfixation by pins and wires can cause pin-site infections or joint stiffness, and a large amount of lengthening can induce secondary axial deformity of the regenerate. We have performed tibial lengthening by applying the Ilizarov apparatus over a nail with the aim of reducing the time required for external fixation, protecting against fracture of the callus, and enabling the individuals to return more quickly to previous daily activities. To the best of our knowledge, the current study is the largest series of lengthening procedures in which an Ilizarov external fixator was applied over an intramedullary nail exclusively in tibiae and the details regarding bone formation and complication rates were reported.
Our results showed that the mean external fixation index differed significantly between the groups, but the healing indices were similar. When we first started performing lengthening over an intramedullary nail, we were concerned that reaming might compromise the quality and quantity of the regenerate5,8. Two previous studies demonstrated poor regenerated bone formation in one of seven tibiae7 and three of nine tibiae8. Although reaming of the canal produces more space for a larger nail and increases mechanical stability4,6, it can also confer biological disadvantages such as disruption of the endosteal vascularity and an increased susceptibility to infection5,8,9,13-15. However, there was no delayed consolidation or atrophy of the new bone in our group treated with lengthening over a nail, which had been inserted without reaming, in contrast to the findings in the group with conventional lengthening. It is quite possible that the use of reamed or unreamed nailing is not an important issue and that early removal of the external fixator and maintenance of axial alignment without fracture are more important for patient comfort, especially when simultaneous bilateral leg lengthening is performed.
To prevent premature consolidation, overreaming to a diameter 1.5 mm larger than that of the nail has been suggested6. There were two cases of premature consolidation and failed distraction of the regenerate in the early distraction period due to nail impingement in the canal in our series. We think that insertion, without reaming, of a nail with a diameter 1 mm less than that of the tibial isthmus is sufficient to prevent premature consolidation and prevent the nail from binding in the canal during distraction, provided that the Ilizarov frame is placed exactly parallel to the nail.
The use of an intramedullary device necessitates lengthening along the anatomical axis, which theoretically is associated with the risk of axial deviation of the femur. However, lengthening over an intramedullary nail can easily be performed in the tibia, where the mechanical and anatomical axes are identical. We did not observe any axial deviation of the tibial regenerate related to lengthening over an intramedullary nail. However, a secondary axial deformity occurred during the lengthening in 16% of the cases treated with the conventional method, a finding that may be related more to an unstable construct than to the method of lengthening. Valgus deviation during lengthening has been a problem when a monolateral fixator has been used in conjunction with a nail6,7. We think that a unilateral fixator is not strong enough to overcome the resistance of the interosseous membrane. We did not observe axial deviation of the tibiae that were lengthened over a nail in our study, perhaps because of the additional stability supplied by the circular external fixator.
Our technique of corticotomy after insertion of a nail may be challenging; however, we did not encounter any technical problem or complication related to the corticotomies in this series. Another difference between our techniques and those of others6,8,10 is the use of a two-ring construct. We do not think that a three-ring Ilizarov frame necessarily adds stability because we observed no complications that may have been related to instability of the two-ring frame. Interestingly, in each of our groups, there were two cases of ankle valgus associated with a less-elongated fibula and its relative proximal migration. These findings could partly be explained by incompetence of the distal tibiofibular transfixing screw; however, additional study is necessary to identify any other factors that could be responsible for a lengthening discrepancy between the two bones.
Our results suggest that complications related to pins or wires, such as pin-site infections and joint stiffness, are the rule rather than the exception, even with lengthening over an intramedullary nail. The most serious potential disadvantage of lengthening over an intramedullary nail is a deep infection8,9 and, for this reason, some authors have discontinued using this technique and returned to employing the conventional Ilizarov method8. In order to prevent deep infection, Herzenberg and Paley recommended that at least 2 mm of actual space be present between any pin and nail6. We tried to position proximal wires well posterior to the nail and to obtain purchase in the fibular head as well. We treated any superficial pin-site infections as early and aggressively as possible, and there were no deep intramedullary infections in our series.
There were two cases of a wire cutting out during the distraction period, which might have been due in part to a relatively narrow divergence angle between the inserted wires. The angle of 9° between the insertion point of an AO tibial nail, inserted without reaming, and the long axis of the tibia allows more space for insertion of wires than is possible with other nails6, the insertion angle of which is typically 15°. However, use of a nail in which an interlocking screw can be placed in an oblique direction in order to capture the proximal end of the fibula may be better when there is not enough space between the nail and the wires to be inserted6. We observed seven cases with a broken proximal interlocking screw. This finding might have been related to the use of a screw with a smaller diameter (3.9 mm for nails inserted without reaming compared with 4.9 mm for nails inserted with reaming) and increased weight-bearing by the individual during the consolidation period. Nonetheless, the broken screws were not related to either shortening or axial deviation of the regenerated bone.
Our study had several limitations. Its retrospective design did not allow direct comparison of the two techniques with regard to function. Although there were no differences between the groups with respect to the heights of the men and women, issues such as the perceived inconveniences of each method might be affected by the characteristics of the individuals in the groups, since the selection of the lengthening procedure was based in part on the patient's economic circumstances. Furthermore, we did not evaluate the patients' final general health status with use of a validated instrument. To the best of our knowledge, there is no validated instrument with which to assess the patient's mobility or convenience during these lengthening procedures. We assessed the parameters (mobility and physical activity) included in the physical function subscale of the Children Health Information Service Rand Scale12,16; however, those measures are traditionally used to assess outcomes after treatment of complex lower-limb and foot deformities in children and adolescents. Nevertheless, our results suggest that individuals undergoing lengthening over an intramedullary nail can cope with outdoor activities without the help of others more easily than can individuals undergoing conventional lengthening. Progressive joint contractures are expected to occur because of the lengthening procedure per se. However, our findings of a lower prevalence of ankle flexion contracture necessitating tendon lengthening and of permanent ankle contracture in Group B suggest that early removal of the external fixator made rehabilitation easier for these individuals and allowed a faster resumption of ankle motion.
Recently, the indications for leg lengthening have been expanded to include short stature without underlying disease17,18. Catagni et al. reported the results of cosmetic leg lengthening with use of the Ilizarov method17. Although they did not use a validated instrument to assess the quality of life, all of their patients were satisfied with the improvement in self-esteem, decrease in distress or shyness, and improvement in the quality of life. In contrast, however, 25% of the individuals in Group A and 21% of the individuals in Group B in our series were not satisfied with the overall result, and there were many complications and reoperations. We think that use of cosmetic limb lengthening in a short but otherwise normal patient without an underlying disease cannot escape criticism because of the burden of complications. Furthermore, any complication can be considered as "major" in a normal patient. Although the period required for external fixation was shorter and the complication rate was lower in Group B, the individuals in both groups returned to previous activity at a similar time and there was no difference in the limitation of physical activity at the time of final follow-up. Nevertheless, our results demonstrated that tibial lengthening over an intramedullary nail, with use of the Ilizarov external fixator, produces new bone formation of a quality equal to that obtained with conventional Ilizarov lengthening but reduces the duration of external fixation and the complication rate. 