Restoration of Thoracic Kyphosis in Adolescent Idiopathic Scoliosis Over a Twenty-Year Period

Are We Getting Better?

Blake M. Bodendorfer, MD; Suken A. Shah, MD; Tracey P. Bastrom, MA; Baron S. Lonner, MD; Burt Yaszay, MD; Amer F. Samdani, MD; Firoz Miyanji, MD; Patrick J. Cahill, MD; Paul D. Sponseller, MD; Randal R. Betz, MD; David H. Clements 3rd, MD; Lawrence G. Lenke, MD; Harry L. Shufflebarger, MD; Michelle C. Marks, PT, MA; Peter O. Newton, MD

Disclosures

Spine. 2020;45(23):1625-1633. 

In This Article

Results

A total of 1063 patients with a diagnosis of main thoracic scoliosis (Lenke types 1–4) and preoperative thoracic hypokyphosis (<10°) and who had at least 2 years of follow-up were identified from the time frame between 1995 and 2015. After dividing patients into Periods 1, 2, and 3 cohorts, significant changes were seen among the three time periods for approach used. Additionally, change in 3D TK was significantly different among groups. When comparing individual cohorts, change in 3D TK was significantly different between Periods 1 and 2 and Periods 2 and 3, but not significantly different between Periods 1 and 3. Lastly, comparing the percentage of patients with 20° or more of TK at 2-year follow-up demonstrated significant differences among groups. When comparing individual cohorts, the percentage of patients with 20° or more of TK at 2-year follow-up was significantly different between Periods 1 and 2 and Periods 2 and 3, but not significantly different between Periods 1 and 3.

Significant differences were seen among the three time periods. Operatively treated AIS patients with thoracic hypokyphosis in the Period 1 cohort had excellent restoration of TK (25.5°±10.1° change in 3D TK), which slightly lessened in Period 2 (22.2°±9.3° change in TK) and improved to near the Period 1 cohort in Period 3 (24.4°±9.0° change in 3D TK). A similar trend was demonstrated when comparing the percentage of patients with 20° or greater TK at 2-year follow-up (71.2%, 54.5%, and 66.7%, sequentially). Improvement in TK was achieved in the latest time cohort compared with the middle cohort for both change in 3D TK and for the percentage of patients demonstrating 20° or greater TK at 2 years of follow-up. See Table 1 for all data and statistical analysis for all approaches combined. Similar trends were observed when combined PSF and anterior release approaches were excluded (Table 2).

Analyzing solely PSF approaches, change in 3D TK was significantly different among groups. When comparing individual cohorts, change in 3D TK was only significantly different between Periods 2 and 3. Lastly, comparing the percentage of patients with 20° or more of TK at 2-year follow-up demonstrated significant differences among groups. When comparing individual cohorts, the percentage of patients with 20° or more of TK at 2-year follow-up was only significant between Periods 2 and 3 (Table 3).

Because no isolated ASF with instrumentation approaches were performed in Period 3, comparison was only possible between Periods 1 and 2 when analyzing solely ASF approaches. Change in 3D TK approached significant difference between Periods 1 and 2, but comparing the percentage of patients with 20° or more of TK at 2 years of follow-up did not yield a significant difference (Table 4).

Comparison between isolated ASF with instrumentation (Periods 1–2), early PSF (Period 2), and contemporary PSF (Period 3) was also performed. Change in 3D TK was significantly different among groups. When comparing individual cohorts, change in 3D TK was significantly different between ASF and early PSF as well as between early PSF and contemporary PSF, while there was no significant difference between ASF and contemporary PSF. Next, comparing the percentage of patients with 20° or more of TK at 2-year follow-up demonstrated significant differences among groups. When comparing individual cohorts, the percentage of patients with 20° or more of TK at 2-year follow-up was significantly different between ASF and early PSF as well as between early PSF and contemporary PSF, while there was no significant difference between ASF and contemporary PSF (Table 5).

Because fusion levels may influence restoration of TK, comparison among groups was made whenever possible. The most frequently utilized upper instrumented vertebrae trended superiorly with time, with T5 (55.3%), T4 (21.1%), and T6 (21.1%) being most common for Period 1; T4 (30.9%), T2 (27.7%), and T3 (20.5%) for Period 2; and T3 (38.5%), T2 (32.1%), and T4 (24.2%) for Period 3 (P < 0.001). The most frequently utilized lower instrumented vertebrae trended inferiorly with time, with T12 (44.7%), L1 (26.3%), and T11 (21.1%) being most common for Period 1; L1 (25.6%), T12 (23.7%), and L2 (18.6%) for Period 2; L1 (29.4%), L2 (21.7%), and L3 (20.3%) for Period 3 (P < 0.001).

Because corrective techniques and instrumentation materials may influence restoration of TK, comparison among groups was made whenever possible. For those patients who underwent an anterior approach, there was a shift toward more approaches being performed thoracoscopically, with 68.4% open and 31.6% thoracoscopic during Period 1 and 35.2% open and 64.8% thoracoscopic during Period 2 (P < 0.001). Thoracoplasties were performed less frequently with time (P < 0.001), with 69.2% of patients during Period 1 (25% external, 75% internal) and 33.1% of patients during Period 2 (86.3% external, 13.7% internal) and 17.6% of patients during Period 3 (92.9% external, 7.1% internal). For anteriorly instrumented fusions, the most frequently utilized rod size in millimeters (mm) trended larger in diameter with time (P < 0.001), with 4 mm being most common in Period 1 (86.5%) and 4.75 mm in Period 2 (51.1%). Regarding anterior rod material utilization, Period 1 had predominately stainless steel rods reported (71.4%) and Period 2 had predominately titanium (66.7%) (P < 0.001). Period 1 had solely single rod constructs reported and Period 2 predominately had single rod constructs (71.4%) (P < 0.001). Only Period 2 had reported anterior rod strengths (in kilopounds per square inch, ksi), which were on average 169.0 ± 43.8 ksi.

For posteriorly instrumented fusions, the most frequently utilized rod size trended larger in diameter with time (P < 0.001), with 5 mm being most common in Period 1 (66.7%) and 5.5 mm in Periods 2 and 3 (95.2% and 96.9%, respectively). Regarding posterior rod material utilization, Period 1 had exclusively titanium rods reported, whereas Period 2 had 70.5% stainless steel, 9.7% titanium, and 8.1% cobalt chrome and Period 3 had 54.2% stainless steel, 40.8% cobalt chrome, and 4.6% titanium (P < 0.001). Posterior rod strength was only reported in Periods 2 and 3 at an average 180.9 ± 27.1 ksi and 183.8 ± 18.1 ksi, respectively, for the left rods (P = 0.064) and 171.6 ± 32.2 ksi and 179.6 ± 22.5 ksi, respectively, for the right rods (P < 0.001). Posterior constructs were predominately hybrid (92.9%) during Period 1, a minimum of 80% screw density or all screws (26.5% and 65.5%, respectively, for a total of 92.0%) during Period 2 and a minimum of 80% screw density or all screws (51.3% and 47.8%, total 99.1%) during Period 3 (P < 0.001).

Direct vertebral rotation maneuvers were performed in the vast majority of patients and increased with time with 90.9% of patients in Period 1, 94.4% of Period 2, and 94.2% of Period 3 (P < 0.001). Posterior column osteotomies were not performed in Period 1, but increased with time, with 60.7% of Period 2 patients and 87.3% of Period 3 patients (P < 0.001).

processing....