Long-term clinical outcomes of wedge osteotomy of hemivertebra for adolescent patients with congenital scoliosis

To the Editor: The hemivertebra is an unusual but not rare spinal deformity, which could result in progressive congenital scoliosis (CS). Owing to its poor response to brace treatment, early surgical intervention is usually required for correcting deformity and preventing curve progression.[1] Hemivertebra resection has been proved to be an effective surgical treatment for CS secondary to hemivertebra. The best indication is isolated hemivertebra in young children.[2] Unfortunately, some CS patients, especially in the rural areas, remain under-diagnosed and untreated until reaching their adolescence. Aggravated spinal deformity and associated back pain urge them to seek for surgical treatment. Clinically, the adolescent case usually presents a large hemivertebra, which means conventional hemivertebra resection would leave a large residual gap between the vertebrae. Thus, it is a high-demanding technique to perform subsequent mechanical compression and shortening procedure, which may lead to failed bone fusion or other neurological complications.[3] As is known, progression of the curve is most rapid during the adolescent growth spurt and it stops at skeletal maturity. For a nearly matured spine, its remaining growth potential to cause further deformity is limited. Does a late adolescent patient really need a complete hemivertebra excision for CS? The other suitable surgical option is still worth investigating. Compared with conventional hemivertebra resection, wedge osteotomy (WO) may be another feasible option for some adolescent patients. This study is designed to compare the long-term clinical outcomes of different surgical strategies for adolescent patients with congenital hemivertebra. This retrospective study was approved by the Ethics Committee of Renji Hospital, Shanghai Jiaotong University School of Medicine (No. KY2022-002-B) and with a waiver of consent. This study included all adolescent patients with CS who underwent correction surgeries in Renji Hospital between January 2009 and June 2015. The inclusion criteria were as follows: (1) Diagnosed with CS due to hemivertebra; (2) Age between 14 years and 20 years; (3) Risser sign ≥grade 4; (4) Coronal scoliosis curve <60° and sagittal kyphosis <30°; and (5) With a minimum of 8-year follow-up. Patients with previous spinal surgery or incomplete medical data were excluded. Medical records were reviewed to identify demographic data, operating time, average blood loss, and any complications in the intraoperative and follow-up periods. This initial search yielded 67 CS patients due to hemivertebra. Of them, 10 patients lacked incomplete medical data and 14 patients (20.1%) were lost to follow-up. Finally, a total of 43 patients who satisfied the aforementioned criteria were enrolled in this study. Correction surgeries were indicated for their late-diagnosed spinal prominence and trunk imbalance. All patients’ spinal growth was beyond the phase of adolescent growth spurt. Because there is no agreed-on standard for the surgical intervention for hemivertebra, the patients were treated by different surgical strategies according to the spine surgeons’ own experience. They were divided into two categories: WO group and complete resection (CR) group. All surgeries were performed by a senior surgeon in our department. In the CR group, the patients were treated with CR of hemivertebra through posterior approach. The upper and lower disks of hemivertebra, including the cartilage endplates, were removed completely. In the WO group, the patients were treated by partial resection of hemivertebra. A wedge-shaped bone resection of the hemivertebra was performed on the guidance of preoperative designed angle of osteotomy [Figure 1]. During the procedure of osteotomy, the base was the posterior border of the vertebral body, and the apex of the osteotomy was the midpoint of the anterior margin of the hemivertebra body. The intervertebral disk and little cancellous bone of the hemivertebra were preserved.[4]Figure 1: The degree of the WO (W) is approximately equal to the coronal Cobb angle of the segmental curve (S). After wedge osteotomy of hemivertebra, the correction of the scoliosis and trunk imbalance was satisfied and maintained well after 10-year follow-up. The CT reconstruction show solid fusion. Line a/b: A vertical line is drawn along the inferior endplate of the upper end vertebral body and the superior endplate of the lower end vertebral body, respectively, named lines a and b, and the angle between the two lines is defined as coronal Cobb angle (S). CT: Computed tomography; WO: wedge osteotomy.Radiographic measurements were taken from standing posteroanterior and lateral radiographs before surgery, 1 week after surgery, and during the last follow-up. Proximal junctional kyphosis (PJK) was defined as the proximal junction angle at least 10° greater than the preoperative value. Radiographs taken at the latest follow-up were reviewed to determine the implant position, any implant failure, or pseudarthrosis. All radiographic parameters were independently measured by two spine surgeons. All patients were required to fulfill the Scoliosis Research Society-22 (SRS-22) Questionnaire both at preoperation and at the last follow-up in order to assess the improvement in health-related quality of life (HRQoL). The SRS-22 covers five domains including function, pain, self-image, satisfaction with treatment, and mental health. The WO group consisted of 23 patients with an average age of 16.4 years (range: 14.0–20.0 years) at surgery. The CR group comprised 20 patients with an average age of 15.9 years (range: 14.0–19.0 years). The hemivertebra was located in the thoracic spine in 9 cases, in the thoracolumbar region in 22 cases, and in the lumbar spine in 12 cases. All hemivertebra were non-incarcerated [Supplementary Table 1, https://links.lww.com/CM9/C313]. The average follow-up was 106.8 months (98.0–130.0 months) and all cases reached skeletal maturity at the final follow-up. The mean fusion level was 3.8 segments in the WO group and 3.5 segments in the CR group. The mean operation time was 193.8 ± 72.4 min (range: 145.0–275.0 min) in the WO group and 315.2 ± 100.6 min (range: 180.0–390.0 min) in the CR group. The average estimated blood loss was 435.2 ± 360.7 mL (range: 330.0–850.0 mL) in the WO group and 789.3 ± 581.4 mL (range: 380.0–1200.0 mL) in the CR group. The durations of surgery and blood loss were significantly lower in the WO group than that in the CR group (all P <0.05). The correction results in the two groups are presented in Supplementary Table 2, https://links.lww.com/CM9/C313. In the WO group, the mean segmental coronal curve was corrected from 42.5 ± 7.5 preoperatively to 14.3 ± 4.6 postoperatively (66.4% correction rate), and 16.8 ± 5.1° (60.5% correction rate) at the final follow-up. For patients in the CR group, the correction rate of coronal curve was 73.2%, and 62.2% at the final follow-up. In the sagittal plane, the segmental kyphosis was improved from 14.3 ± 12.3° to 7.2° ± 6.7° at the last follow-up with a correction rate of 49.7% in the WO group, and from 15.0° ± 10.8° to 6.3° ± 7.2° with a correction rate of 58.0% in the CR group. In both groups, compensatory cranial and caudal curve were spontaneously corrected well after surgery. Trunk shift was also significantly diminished after operation. Three patients (one in the WO group and two in the CR group) were found to have PJK. The average proximal junctional angle increased from 8.7° to 16.5° at 12 months postoperation. At the last follow-up, the average proximal junctional angle was 21.9°, showing a tendency of stability during the long-term follow-up. Fortunately, no revision surgery was needed. During the follow-up, X-ray and computed tomography (CT) scan showed that the scoliosis correction maintained well without any sign of implant failure or pseudarthrosis. Solid fusion of the posterior elements and osteotomy space were achieved in all patients. Postoperatively, all five domains of the SRS-22 scale showed a general improvement overall. In the WO group, the total SRS-22 score, including the pain domain (P = 0.009), self-image domain (P <0.001), and the satisfaction domain (P = 0.024) demonstrated significant difference compared with preoperative data [Supplementary Table 3, https://links.lww.com/CM9/C313]. In the CR group, two patients experienced transient neurological complications of lower extremity numbness. Postoperative CT and magnetic resonance imaging(MRI) examinations did not reveal dislocation of the pedicle screw, and thus the neurological deficits were attributed to the procedure of hemivertebra resection. While in the WO group, none of the patients suffered neurological compromise. CS secondary to hemivertebra is a complicated spinal deformity in young children and adolescents. Hemivertebra with normal end plates cause the asymmetric growth of the spine, which progressively lead to a wedge-shaped deformity. In our series, each patient was beyond the phase of the adolescent growth spurt. All the 43 patients included were in late adolescence aged between 14 years and 20 years and their Risser signs were 4 or 5 on presentation. Their spines were almost matured and the hemivertebra had limited growth potentials to cause progressive deformity. In previous practice, we noticed that the residual gap after CR of a large hemivertebra cannot be closed completely with a bone-on-bone contact. This defect may result in overload of the convex pedicles and postoperative spinal instability, increasing the risk of implant failures. Considering the above condition, WO of hemivertebra was performed, in which the hemivertebra was partial excised. The retained disk and a little cancellous bone of hemivertebra could fill the residual gap in osteotomy space, and functions as a hinge preventing critical translation of the vertebra when the convexity is compressed down and the gap is closed. This bone-on-bone closure not only decreases the stress of the instrumentation but also enhances the postoperative spinal stability, which could consequently reduce the risk of pseudarthrosis and implant-related complications. Besides, to further decrease the remaining growth potential of hemivertebra, both the cephalic and caudal endplates of hemivertebra were destructed by curette during the procedure. After long-term follow-up observation, all patients achieved solid spinal fusion, without implant failure or pseudarthrosis needed for revision surgery. For patients undergoing WO of hemivertebra, the average correction rates were 66.4%, with only 3° loss at the last follow-up. The compensatory curves, coronal imbalance, and segmental kyphosis also obtained satisfactory correction and maintained well during the follow-up. Moreover, HRQoL assessment demonstrated significant improvement of the SRS-22 score compared with preoperative status, especially in the self-image domain score. Taken together, these results reveal that WO can achieve satisfactory correction in adolescent CS patients. Notably, the average operation time was 193.8 min in the WO group and 315.2 min in the CR group. The average estimated blood loss was 435.2 mL in the WO group and 789.3 mL in the CR group. Compared with hemivertebra resection, WO strategy has a shorter operation time and lower blood loss. This advantage attributes to a simpler surgical procedure, without excessive excision of hemivertebra endplates and intervertebral disk. Hemivertebra resection, involving the complete removal of a vertebral level and both adjacent disks, should be defined as grade V osteotomy.[5] It is still a technically high demanding procedure and is usually associated with more risks of neurological deficits.[3,5] In the present study, two patients in the CR group suffered transient neurological compromises of sensory deficits. Whereas in the WO group, no neurologic complications occurred. Theoretically, the WO of hemivertebra can be classified as a grade III osteotomy, involving only partial resection of the posterior vertebral body and the posterior elements with pedicles. Therefore, this less invasive surgical strategy may be another safer surgical option for adolescent CS patients. In summary, WO of hemivertebra can be indicated for adolescent CS patients. Compared with hemivertebra resection, WO of hemivertebra is an advisable surgical strategy with advantages of reduced trauma, less blood loss, and a shorter operation time. Based on the long-term clinical outcomes, WO of hemivertebra is a relatively simpler and safer procedure with reliable correction effect. Funding This work was supported by grants from National Natural Science Foundation (Nos. 81802216 and 82272573). Conflicts of interest None.