Achondroplasia, the most common type of skeletal dysplasia, is caused by an autosomal-dominant mutation in the fibroblast growth factor receptor 3, which impairs endochondral bone formation . Clinical features include rhizomelic shortening of the limbs, short stature, midface hypoplasia, and, commonly, genu varum . Kopits  reported that approximately half of the patients with achondroplasia had substantial tibial bowing. The cause of genu varum in achondroplasia is currently unknown, but it results in pain, joint instability, and limitation of function . Genu varum may be more likely to occur in boys than in girls .
A subset of patients with achondroplasia and genu varum requires osteotomies to eliminate or reduce clinically significant pain and fibular thrust. Although osteotomy can provide relief for such symptoms, it is not without complications, including compartment syndrome, peroneal palsy, recurrence of the angular deformity, nonunion, malunion, infection, growth-plate damage, and vascular injury .
Traditionally, radiographs are used to determine the center of rotation of angulation and, thereby, the appropriate site for the osteotomy procedure . However, because of the large cartilage envelope in young people, a radiograph alone may not be sufficient for locating the deformity site. Inan et al.  found poor interobserver agreement using the mechanical axis deviation method to measure the medial proximal tibial angle and lateral distal tibial angle in children aged below 6 years. They determined that significant measurement errors in the proximal and distal tibial joints can occur, especially in young children with achondroplasia, and suggested the use of additional clinical modalities to improve accuracy . In a study comparing supine fluoroscopy with full-length standing radiographs for the assessment of lower limb alignment, Sabharwal and Zhao  found that fluoroscopy was useful in certain populations, but noted a trend toward greater variability between the modalities in terms of estimation of mechanical axis deviation between the two modalities in younger patients, although the difference was not statistically significant. The authors suggested that this trend was a result of incomplete ossification with greater soft-tissue laxity in younger patients.
We postulated that using arthrography to visualize the epiphyseal cartilage, especially in children, might change the preoperative interpretation of the center of rotation of angulation, thus affecting the type of osteotomy performed. Our goals were to determine: (i) whether the choice of osteotomy based on radiographic findings alone would have differed from that based on combined radiographic and arthrographic findings, and (ii) whether the difference was more prevalent in one age group than another.
Of approximately 700 patients with achondroplasia treated by the senior author (M.C.A.) between January 1996 and January 2006, we identified 41 pediatric patients (74 limbs) who had undergone a corrective osteotomy for genu varum; who had preoperative hip, knee, and ankle joint radiographs and arthrograms; and who had a minimum of 2 years of follow-up. Indications for surgery included clinically significant pain secondary to malalignment and/or fibular thrust. The average age of the 27 men and 14 women was 8.85 years (range: 3–17 years) (Fig. 1).
We retrospectively reviewed the standing anteroposterior and lateral radiographs of each patient to determine the center of rotation of angulation, and the senior author (M.C.A.) identified the appropriate proximal osteotomy location based on those radiographic findings alone. For each patient, the lateral distal femoral angle, proximal medial tibial angle, and distal tibial angle were drawn to determine the most appropriate site for the osteotomy. On a separate occasion, the senior author (M.C.A.) blindly reviewed each patient's arthrograms using the same method as described above. The arthrograms had been obtained a few days or immediately before surgery (stance had been reproduced by applying pressure to the bottom of the feet of the anesthetized patient in an attempt to obtain realistic images). Each anteroposterior radiograph and arthrogram had been obtained with the patient's patella forward, thereby providing the same view of rotation in both studies. We noted the type of osteotomy that had been used to correct the deformity and compared it with the choice based on radiographic findings alone. All osteotomies were in the metaphyseal region of the distal femur, the proximal tibia, or the distal tibia. A change in procedure was defined as a change in the site of osteotomy between these three locations.
Procedures performed were closing wedge, opening wedge, dome, and modified Rab (oblique) proximal tibial osteotomies. Our analysis included 55 proximal tibial osteotomies (26 of which used a modified Rab procedure), 32 distal tibial osteotomies, 10 distal femoral osteotomies, and 2 proximal femoral osteotomies. For the modified Rab procedure, a dual screw technique was used in place of the traditional single screw with a wedging cast . With the large degree of laxity often present in the knee ligaments of these patients, the correction afforded with wedging cast may occur at the ligaments of the knee rather than at the site of the osteotomy. The dual screw technique provides a more precise correction that can be locked into place, potentially improving healing by reducing the mobility allowed by the wedging cast.
We used a χ2 test of association to determine whether the patient's age was associated with a change of osteotomy procedure based on the addition of arthrographic findings. We compared the younger half of the patient population (21 patients aged below 8 years) to the older half (20 patients aged above 8 years) with respect to the proportion of patients for whom the inclusion of arthrographic findings in the decision-making process changed the procedure that would have been performed based on only radiographic findings. The level of significance was set at a P value of less than 0.05.
We found that, in 26 out of 41 patients (63.4%; 45 of 74 limbs, 60.8%), inclusion of arthrographic findings in the decision-making process changed the procedure that would have been performed if only radiographic findings were available.
When patients were grouped by age at the time of surgery, we found that the inclusion of arthrographic findings changed the surgical procedure significantly more often in the younger patients (i.e. aged below 8 years; 20 out of 21 patients, 95.3%; 34 of 36 limbs, 94.44%) than in the older patients (i.e. aged above 8 years; six of 20 patients, 30.0%; 11 of 38 limbs; 28.95%).
In answer to our first investigational question (did the choice of osteotomy based on radiographic findings alone differ from that based on combined radiographic and arthrographic findings?), we found that a different procedure would have been performed in 60.8% of limbs if the arthrogram had not been performed. In most patients, the inclusion of the arthrogram led to a change in the site at which the osteotomy was performed. We recognize that internal rotation of tibiae can also be an important component of achondroplastic deformity of the lower extremity. Malrotation was addressed in each patient individually, depending on where the osteotomy was performed. For example, in cases where malalignment occurred in the femur and malrotation was identified in the tibia, two osteotomies were performed to address both deformities.
We believe that our claim about the inclusion of arthrography in the decision-making process contributes to the selection of the most appropriate choice for each patient is supported by our high rates of excellent (Fig. 2) and good outcomes (as determined by each patient's postoperative alignment, lessening of pain, and improvement in function; by the lack of any cases of recurrence to date at a minimum per-patient follow-up of 2 years; and by only one malunion). For example, on the basis of standard radiography alone, the patient shown in Fig. 2 would likely have undergone a classical proximal tibial osteotomy using standard radiographs; however, with the addition of the arthrogram, the choice of procedure was changed, resulting in the excellent outcome shown. Historically, there is a generally acknowledged recurrence rate of 30 to 50%. In a study evaluating osteotomy for the correction of genu varum in patients with achondroplasia, Beals and Stanley  reported 16 cases of recurrence in 39 procedures, two of which required reoperation. In our study, we had no recurrences and no reoperations, which we believe reflects improved decision-making and subsequent patient outcomes as a result of the inclusion of the arthrogram. To our knowledge, there are no other publication studies assessing the surgical correction of genu varum in patients with achondroplasia; therefore, our study represents the largest series to date.
In answer to our second investigational question (was the choice related to patient age?), we found a significant relationship between these parameters: most patients for whom the choice would have changed were aged below 8 years, but the choice also would have changed for a notable proportion of older patients (six out of 20, 30.0%). In a study by Inan et al. , a greater degree of inconsistency in measurements of lower extremity alignment in younger (≤6 years old) than in older patients was observed. Our study has provided a potential method for reducing that variability.
Our investigation had a few limitations. First, not all patients were followed to skeletal maturity (to date, some are still not skeletally mature). However, all patients did have a minimum follow-up of 2 years. Second, a more ideal study design might have been to have had two separate preoperative evaluation groups, one in which radiographs alone were used and one in which both radiographs and arthrograms were used, which would have allowed for a direct comparison of outcomes between the two methods. However, this comparison was not an option because of the retrospective nature of our study.
Despite these limitations, it could be argued that, based on our results of no recurrence to date, one factor of such an outcome is the accurate assessment of a patient's center of rotation of angulation. Therefore, our outcomes suggest that the use of arthrography in conjunction with radiography, especially in the pediatric population, appears to afford more complete visualization of the cartilage of the hip, knee, and ankle joints than radiographs alone; the most accurate interpretation of each patient's center of rotation of angulation; and the most appropriate choice of corrective osteotomy, leading to a good postoperative outcome.
This study was supported in part by the Johns Hopkins School of Medicine Dean's Fund.
1. Richette P, Bardin T, Stheneur C. Achondroplasia
: from genotype to phenotype. Joint Bone Spine 2008;75:125–130
2. Kopits SE. Correction of bowleg deformity in achondroplasia
. Johns Hopkins Med J 1980;146:206–209
3. Bailey JA II. Orthopaedic aspects of achondroplasia
. J Bone Joint Surg Am 1970;52:1285–1301
4. Ain MC, Shirley ED, Pirouzmanesh A, Skolasky RL, Leet AI. Genu varum
. J Pediatr Orthop 2006;26:375–379
5. Pinkowski JL, Weiner DS. Complications in proximal tibial osteotomies in children with presentation of technique. J Pediatr Orthop 1995;15:307–312
6. Paley D, Herzenberg JE, Tetsworth K, McKie J, Bhave A. Deformity planning for frontal and sagittal plane corrective osteotomies. Orthop Clin North Am 1994;25:425–465
7. Inan M, Jeong C, Chan G, Mackenzie WG, Glutting J. Analysis of lower extremity alignment in achondroplasia
. Interobserver reliability and intraobserver reproducibility. J Pediatr Orthop 2006;26:75–78
8. Sabharwal S, Zhao C. Assessment of lower limb alignment: supine fluoroscopy compared with a standing full-length radiograph. J Bone Joint Surg Am 2008;90:43–51
9. Rab GT. Oblique tibial osteotomy
for Blount's disease (tibia vara). J Pediatr Orthop 1988;8:715–720
10. Beals RK, Stanley G. Surgical correction of bowlegs in achondroplasia
. J Pediatr Orthop B 2005;14:245–249