Clinical application of individualized 3D-printed navigation template to children with cubitus varus deformity

Thirty-five cases of cubitus varus deformity were included in the study. The patients were 21 boys and 14 girls, aged 4.6–13.2 years (average, 7.5 years). The deformity was present on the left in 17 cases and on the right in 18 cases. The injuries were caused by supracondylar fracture of the humerus. After fracture, 24 cases underwent closed reduction and percutaneous Kirschner wire fixation, and 11 underwent manual reduction of plaster fixation. Patients that fulfilled the following criteria were eligible for the study: (i) all elbow varus deformities were secondary to the supracondylar fracture, (ii) all affected sides had carrying angles of more than − 15°, and (iii) the timing of cubitus varus orthopedic surgery was 12–43 months after the fracture occurred (mean, 16.4 months). Patients were excluded from the study when (i) the patient suffered from metabolic bone disease, (ii) other fractures were present, and (iii) the patient had previously undergone corrective surgery due to cubitus varus deformity. All patients were treated with wedge osteotomy of the lateral distal humerus. In 16 cases (8 on each side), 3D-printed navigation templates were used; in 19 cases (9 on the left and 10 on the right), traditional surgery was performed. Computed tomography (CT) scanning data of the bilateral upper extremities were collected from patient records. Preoperative CT images were acquired on a 64-MSCT scanner (Philips, Netherlands) with parameters of 120 kV, 160 mA, and 1-mm section thickness. Imaging data were stored in DICOM format. The hospital ethics committee approved this study, and the legal guardians of the children gave informed consent.

CT scanning data were imported into a computer, and the 3D reconstructed upper limb models were generated using Mimics 17.0 software (Materialise NV, Leuven, Belgium). The angle of the coronal osteotomy was calculated by computer software from the image overlap of the affected side and the contralateral side, and then, the angular difference between the two at the distal end of the humerus was directly measured. This angle difference was the angle at which the osteotomy was required (Fig. 1a). The two osteotomy planes were designed as isosceles triangles to ensure that the lateral edge of the osteotomy plane was continuous and did not form a step [25]. The lateral osteotomy plane was approximately 1.0 cm above the olecranon fossa (Fig. 1b). A 5-mm-thick process in the opposite direction was used to create a matching substrate, while the Kirschner wire guide pipe data were imported, combined, and reconstructed into an individualized navigation template. The rotation angle was calculated, and the positions of the upper and lower positioning pipes of the guide plate were determined (Fig. 1c). After completing the osteotomy, the two pipes were rotated and coincided to correct the distal internal rotation deformity. After passing the Boolean operation, the Kirschner wire guide pipe was penetrated to complete the guide plate design (Fig. 1d), the data were imported into the 3D printer, and the distal humerus model and navigation template (Fig. 1e) were printed using the medical polylactic acid (PLA) material.

For the distal humerus wedge osteotomy, after completely matching the individualized navigation template with the distal end of the humerus, the navigation template was fixed with a Kirschner wire to prevent slippage. The upper and lower positioning pipes of the osteotomy surface in the navigation template were marked with a Kirschner wire. A wedge-shaped osteotomy was then performed along the upper and lower osteotomy surfaces of the navigation template. The wedge-shaped osteotomy block was removed along with the fixed Kirschner wire and navigation template. The distal and proximal ends of the osteotomy surface were then joined, and the distal end was rotated outward to coincide with the two marked holes. Finally, two Kirschner wires were crossed to fix the distal and proximal ends of the osteotomy.

Patients (Fig. 2a, b) were anesthetized and placed in a supine position, and the affected elbow was exposed in vitro. The skin, fascia, and muscle layers were cut; then, the periosteum was dissected to expose the lateral cortical bone of the distal humerus. After matching the navigation template to the distal end of the humerus, the navigation template was fixed with a Kirschner wire to prevent slippage. The upper and lower positioning pipes of the osteotomy surface of the navigation template were marked with a Kirschner wire (Fig. 2c). A wedge-shaped osteotomy was performed along the upper and lower osteotomy surfaces of the navigation template. The wedge-shaped osteotomy block, the Kirschner wire for fixation, and the navigation template were then removed together (Fig. 2d). For the distal and proximal ends of the osteotomy plane, the distal end was rotated outward until the two marked holes coincided. The distal and proximal ends of the osteotomy were cross-fixed with two Kirschner wires, and the affected limb was observed. When the limb appearance was satisfactory, the surgical field was washed and sutured in layers. The elbow joint was placed at 20° flexion, and the long arm plaster was fixed to complete the operation. X-rays were reviewed regularly after surgery (Fig. 2e, f). At 6 weeks postoperation, continuous bone callus shadows were seen at the osteotomy site, and the osteotomy line was blurred. Therefore, the plaster and Kirschner wire were removed. The patients were instructed to perform functional exercises on the elbow joint. The carrying angle and elbow joint mobility were measured at follow-up (Fig. 2g). The carrying angle of cubitus varus deformity was recorded as a negative value. The treatment effect was evaluated according to Bellemore criteria [26], which was divided it into three levels, as follows: excellent, difference between the affected side and normal side of < 5° with elbow joint flexion and extension motion limited to < 10°; good, difference between the affected side and normal side of 6–10° with elbow joint flexion and extension motion limited to 11–20°; and poor, remaining varus deformity with difference between the affected side and normal side of > 20°, or complications requiring surgery.

All measurement data were presented as means ± SD, and Student’s t test was used to examine the differences between groups. All count data of the two groups were compared using the chi-square test or Fisher’s exact test analysis. All statistical analyses were conducted using Stata software version 9.0 (StataCorp LP, College Station, TX, USA) and SPSS software version 17.0 (SPSS Inc., USA). Statistical significance was set at p < 0.05.

The bilateral elbow joint parameters were accurately measured via 3D reconstruction of the CT scanning data (Fig. 1a). The angles that needed to be turned out and rotated were determined (Fig. 1b, c); then, a computer was used to simulate the osteotomy plane and the fixed needle of the Kirschner wire. Navigation templates were successfully designed for all patients. All navigation templates matched the distal humerus model exactly in simulated surgery. The osteotomy was successfully completed along the direction of the osteotomy surface of the navigation template, and the osteotomy block was removed. The two ends of the model were rotated after the osteotomy to coincide with the positioning points, then fixed with two crossed Kirschner wires. The appearance of the distal humerus returned to normal, which was consistent with the computer planning and design results.

For the 16 children with cubitus varus deformity that received the 3D-printed individualized navigation template, the intraoperative navigation template and the distal radius model were completely matched (Fig. 2c), and the osteotomy along the navigation template was successful. The time from initial placement of the navigation template to secure fixation of the two Kirschner wires was 11.69 ± 2.21 min. Patients were followed up with for 6–12 months after surgery (Table 1). The postoperative appearances of the patients were well corrected, and the prominence of the lateral condyle was significantly improved compared with preoperation (Fig. 2g). Anteroposterior and lateral radiographs showed no varus, extension, or flexion deformity 10 weeks after operation (Fig. 2f). The carrying angle of the affected side preoperation was − 23.50 ± 5.75°, while the corresponding healthy side carrying angle was 5.19 ± 2.79°. The carrying angle of the affected side postoperation was 4.94 ± 1.53°, and the difference in angle to that of the healthy side was 1.13 ± 1.20°. At the final follow-up, the maximum extension angle of elbow joint mobility on the affected side was 1.00 ± 6.24°, and the maximum flexion angle was 126.3 ± 5.33°. With respect to the Bellemore criteria, 14 cases were excellent and two cases were good. No complications, such as incision infection, nonunion, neurovascular injuries, pin tract infection, and loss of reduction, occurred.

In the 19 children with cubitus varus deformity that received traditional surgery, the time from clear exposure to completion of the Kirschner wire fixation was 22.89 ± 3.94 min. Although the varus deformity was well corrected by traditional surgery, the prominence of the lateral condyle became more apparent. The carrying angle of the affected side preoperation was − 25.26 ± 5.35°, while that of the healthy side was 5.21 ± 2.37°. The carrying angle of the affected side postoperation was 3.94 ± 3.25°, and the difference in angle to that of the healthy side was 4.21 ± 2.27°. At the final follow-up, the maximum extension angle of the elbow joint mobility on the affected side was 2.00 ± 6.51°, and the maximum flexion angle was 126.8 ± 5.08°. With respect to the Bellemore criteria, 12 cases were excellent, and seven cases were good. No complications, such as incision infection, nonunion, neurovascular injuries, pin tract infection, or loss of reduction, occurred in any of the 19 cases. The operating time for the patients who received the 3D-printed navigation template was significantly shorter than that of the patients who underwent traditional surgery (p < 0.001). In the 3D-printed navigation template group, the difference between the affected side and the healthy side was significantly smaller than that of the traditional surgery group (p < 0.001). Elbow joint function did not significantly differ between the two groups (p > 0.05).