Orthopedics

Materials and Methods

Between 2000 and 2008, fifty-two patients with posttraumatic cubitus varus were treated surgically by lateral closing-wedge osteotomy and followed up for at least 2 years (mean, 42 months; range, 24 to 72 months). The indication for corrective osteotomy in all cases was unacceptable cosmesis. After Institutional Review Board approval, medical records were reviewed and clinical results and radiological changes were analyzed. Thirty-eight boys and 14 girls had a mean age of 8 years and 9 months (range, 5 to 15 years) at surgery. Initial injuries were supracondylar fracture in 41 patients, distal epiphyseal separation in 9, lateral condylar fracture in 1, and intercondylar fracture in 1. Mean time from initial injury to surgery was 4 years and 6 months (range, 3 to 7 years).

Range of motion (ROM) of the elbow, humeroulnar angle, and shaft–condylar angle were assessed preoperatively and at final follow-up. Mean preoperative ROM was −4.5° in extension (range, −10° to 5°) to 132.4° in flexion (range, 115° to 145°). The humeroulnar angle was measured on plain anteroposterior radiographs of fully extended and supinated elbows to determine the severity of varus deformity. Mean preoperative humeroulnar angle was −17.4° (range, −43.4° to 3.2°) for affected elbows and 8.7° (range, 3.0° to 21.7°) for contralateral normal elbows. To quantify hyperextension deformities, the shaft–condylar angle was measured on plain lateral radiographs of 90° flexed elbows, and the mean value was 25.8° (range, 12.6° to 36.5°) for affected elbows and 37.6° (range, 29.4° to 45.0°) for contralateral normal elbows.

Surgical Technique

All surgeries were performed by 1 surgeon (S.J.S.). A lateral approach to the distal humerus was used in every patient. Under C-arm fluoroscopy, 2 K-wires were inserted at a preoperatively determined angle apart from each other in the distal humerus. While protecting the medial periosteum and cortex using Hohmann retractors, lateral closing-wedge osteotomy was performed, leaving the medial cortex intact. The preoperatively calculated angle was corrected with varus force, and the 2 K-wires were positioned parallel using the medial cortex as a lever (Figure 2). Depending on the sagittal shape of the deformity, osteotomy was performed with either flexion or extension simultaneously with valgus osteotomy. In no case was medial translation performed to reduce lateral condylar prominence, nor was nerve transposition performed in any case to prevent ulnar nerve neuropraxia.

Figure 2: Intraoperative fluoroscopic images of 2 preset Kirschner wires measuring the amount of correction (A), the medial cortex preserved as a stable hinge point after lateral closing-wedge osteotomy (B), and fixation with crossed Steinmann pins (C).

Fixation was performed using 2 to 4 (mean, 3.2) cross percutaneous Steinmann pins. Correction of the humeroulnar angle was confirmed by C-arm fluoroscopy, and surgery was completed after repairing the periosteum. A long arm cast was applied for a mean of 5.5 weeks (range, 4 to 6 weeks), and Steinmann pins were removed after a mean of 6.2 weeks (range, 5 to 10 weeks).

Evaluation of Lateral Condylar Prominence

The lateral condylar prominence index and lateral condylar prominence amount were used to quantify the lateral condylar prominence postoperatively. The lateral condylar prominence index was defined as the difference between the measured medial and lateral widths of the transepicondylar line from the longitudinal mid-humeral line and is expressed as a ratio of the total width of the distal humerus to minimize errors from magnification and variations in the size of individual humeri on anteroposterior radiographs (Figure 3).⁹ The lateral condylar prominence amount was defined as the difference between the measured medial and lateral widths of the transepicondylar line from the vertically extended line from lateral margin of the newly formed callus around the osteotomy site to the transepicondylar line and is expressed as a ratio on anteroposterior radiographs (Figure 4).

Figure 3: The lateral condylar prominence index is BC-AB/AC×100 (%), where AC is the transepicondylar line and B is the crossing point between the longitudinal midhumeral line and the transepicondylar line. The value decreases proportionately with remodeling.

Figure 4: The lateral condylar prominence amount is BC-AB/AC×100 (%), where AC is the transepicondylar line and BD is the vertically extended line from the lateral margin of the newly formed callus (arrows) around the osteotomy site to the transepicondylar line (A). The lateral condylar prominence amount closes to −100% with more remodeling because the vertical line (BD) moves to the more lateral side gradually (B).

Results

Range of Motion

Mean elbow ROM in extension improved from −4.5° (range, −10° to 5°) preoperatively to 1.5° (range, −5° to 5°) at final follow-up (P<.05). Mean elbow ROM in flexion improved from 132.4° (range, 115° to 145°) preoperatively to 140.5° (range, 125° to 150°) at final follow-up (P<.05) (Table 1).

Table 1: Results of Supracondylar Lateral Closing-wedge Osteotomy

Lateral Condylar Prominence Remodeling

More severe preoperative cubitus varus deformities were associated with a larger lateral condylar prominence index and a lateral condylar prominence amount immediately after lateral closing-wedge osteotomy (P<.05 and P<.05, respectively). Mean lateral condylar prominence index improved from 38.8% (range, 24.5% to 58.9%) immediately postoperatively to 3.4% (range, −5.8% to 20.9%) at final follow-up (P<.05). No significant difference in mean lateral condylar prominence index existed between elbows at final follow-up. Mean lateral condylar prominence amount also improved from −31.6% (range, −59.2% to 1.4%) to −65.0% (range, −89.2% to −37.8%) (P<.05) (Table 2).

Table 2: Remodeling of the Lateral Condylar Prominence

Lateral condylar prominence index changes were proportional to increases in preoperative deformity severity (humeroulnar and preoperative shaft–condylar angle) and follow-up period (P<.05 for all). Lateral condylar prominence amount changes increased with preoperative deformity severity (humeroulnar and preoperative shaft–condylar angle), follow-up period, and younger age (P<.05 for all). Mean lateral condylar prominence index decreased rapidly in the early post-operative period, mostly within 2 years (Figure 5). Changes in mean lateral condylar prominence amount in patients younger than 11 years (36.6%) were significantly greater than changes in mean lateral condylar prominence amount in patients aged 12 years or older (20.4%) (P=.001) (Figure 6). No early or late complications, such as loss of correction or infection, occurred.

Figure 5: Graph showing that the mean lateral prominence amount (LPA) decreased rapidly within 2 years postoperatively.

Figure 6: Graph showing the change in the lateral prominence amount (LPA). The amount of lateral condylar prominence remodeling was significant in young patients, particularly those 11 years old or younger.

Discussion

The lateral closing-wedge osteotomy is one of the easiest and most widely used techniques for treating cubitus varus deformity, and hyperextension deformity can be corrected simultaneously. However, Griffin⁵ reported unsatisfactory cosmetic problems after lateral closing-wedge osteotomy because of lateral condylar prominence and secondary serpentine deformity, and some authors have reported that the majority of patients were dissatisfied with postoperative scarring when the lateral approach was used.^9,10 To offset the problems of lateral condylar prominence and serpentine deformity, some have attempted dome-shaped osteotomy.^4,11,12 Some authors have tried to correct the deformity with a 3-dimensional method^13–15 or by using an external fixator.^9,16 These methods increase the success rate of correction in the immediate postoperative period and reduce the lateral condylar prominence and secondary serpentine deformity, but they leave bigger scars, are technically more difficult, and have a longer surgery time than lateral closing-wedge osteotomy. The current authors speculate that these surgeries are necessary for the older patient who is finished growing.

Many authors have mentioned the possibility of gradual remodeling of the lateral condylar prominence and improvement of serpentine deformity in children. Wong et al⁸ reported that the lateral condylar prominence was observed in 14 of 22 patients treated by lateral closing-wedge osteotomy and that the severity of the lateral prominence was reduced more in younger patients by remodeling. Barrett et al⁶ reported that 1 in 17 patients had an unsatisfactory lateral condylar prominence and postoperative scarring after lateral closing-wedge osteotomy. Furthermore, Voss et al⁷ observed no lateral condylar prominence in any of their 36 patients.

The results of the current study show that the lateral closing-wedge osteotomy is able to solve the 3- and 2-dimensional deformities effectively with no recurrence of varus deformity requiring corrective repeat osteotomy. The humeroulnar angle and shaft–condylar angle were corrected significantly at final follow-up, and they were similar to the angles of the contralateral normal side (Figure 7). Range of motion was also significantly increased compared with preoperative values.

Figure 7: Clinical photograph (A) and radiograph (B) of an 8-year-old boy with a cubitus varus deformity (humeroulnar angle, −16°) of the right elbow after supracondylar fracture malunion. Lateral closing-wedge osteotomy was performed, and the initial lateral condylar prominence index and lateral condylar prominence amount were measured as 24.3% and −46.8%, respectively (C). Five weeks postoperatively, a callus appeared around the concave side of the osteotomy (arrows). Although the lateral condylar prominence index worsened from 24.3% to 47.9%, the lateral condylar prominence amount improved from −45.8% to −75.0% (D). Two years postoperatively, remodeling of the lateral condylar prominence was almost completed, with a lateral condylar prominence index of 30.3% and a lateral condylar prominence amount of −81.5% (arrows=callus) (E). Clinical photograph taken at final follow-up (F). Radiograph taken at final follow-up showing the corrected cubitus varus (humeroulnar angle, 12°) and the remodeled lateral condylar prominence (lateral condylar prominence index, 12.6%; lateral condylar prominence amount, −87.9%) (G). Abbreviations: AC, transepicondylar line; B1, the crossing between AC and the vertically extended line from the lateral margin of the newly formed callus around the osteotomy site to the transepicondylar line; B2, the crossing between AC and the longitudinal midhumeral line.

The lateral condylar prominence index, introduced by Wong et al,⁸ is measured using the longitudinal midhumeral axis and closes to 0% with remodeling progression. The immediate postoperative value of the lateral condylar prominence index may be affected by the amount of deformity correction because the longitudinal midhumeral axis is shifted. For more exact remodeling analyses, the lateral condylar prominence amount was also used. The lateral condylar prominence amount is measured with the line from the lateral margin of the newly formed callus around the osteotomy site instead of the longitudinal midhumeral axis. It closes to −100% with remodeling progression. The authors speculate that lateral condylar prominence amount may be closely correlated with the change of bone formation around the concave osteotomy site of the lateral condyle and reflect the amount of remodeling. Furthermore, the immediate postoperative lateral condylar prominence amount is not affected by the amount of deformity correction.

The lateral condylar prominence and secondary serpentine deformity were remodeled in most cases. The reason for remodeling is explained by Wolf’s law: More remodeling was expected in younger patients, especially those younger than 11 years, who had greater growth capacity and a thick periosteum (Figure 6).¹⁷ However, because a wide range of growth exists in children of preadolescent age, the remodeling amount varied greatly. Therefore, it must be taken into account when choosing the operative method that children of the same age with different growth are expected to have different remodeling results (Figure 6). Furthermore, preoperative deformity severity and longer follow-up significantly affected the remodeling amount.

Complication rates in lateral closing-wedge osteotomy are reported to be high. Oppenheim et al¹⁸ reported that 24% of patients who underwent lateral closing-wedge osteotomy experienced complications, such as neuropraxia, sepsis, and unacceptable scarring. Furthermore, loss of fixation or fixation failure has been reported.^2,19,20 Griffin⁵ reported that the possibility of loss of correction was increased by medial soft tissue tension and insecure fixation of the osteotomy site, and that it is more likely than metaphyseal deformation. Ippolito et al²¹ reported loss of correction of supracondylar osteotomy in posttraumatic cubitus varus for 17 of 19 patients after a mean follow-up of 23 years. For more secure supracondylar osteotomy fixation, some have tried other fixation methods involving a plate,⁴ staples,²² or an external fixator^9,16 rather than Steinmann pins.

However, many authors have reported that fixation with Steinmann pins or K-wire with tension band wire loop was enough to secure the osteotomy site in children with cubitus varus.^7,17,18,23 The current authors consider Steinmann pins to be firm enough to secure the osteotomy site because they observed no loss of correction prior to radiographic union. Furthermore, in children, ulnar nerve transposition may be unnecessary.

Lateral closing-wedge osteotomy is a simple and cosmetically effective method of treating cubitus varus deformity in children due to remodeling of the lateral condylar prominence in patients aged 11 years or younger. Accordingly, extensive surgical methods, such as dome osteotomy, 3-dimensional osteotomy, or osteotomy using an external fixator, are not necessary in young patients.