- Email: [email protected]
lumbar curve and coronal balance after posterior thoracic fusion for Lenke 1C and 2C adolescent idiopathic scoliosis
J Orthop Sci (2015) 20:31–37 DOI 10.1007/s00776-014-0655-7
ORIGINAL ARTICLE
Postoperative behavior of thoracolumbar/lumbar curve and coronal balance after posterior thoracic fusion for Lenke 1C and 2C adolescent idiopathic scoliosis Masayuki Ishikawa · Kai Cao · Long Pang · Kota Watanabe · Mitsuru Yagi · Naobumi Hosogane · Masafumi Machida · Yuta Shiono · Makoto Nishiyama · Yasuyuki Fukui · Morio Matsumoto
Received: 11 April 2014 / Accepted: 21 September 2014 / Published online: 13 October 2014 © The Japanese Orthopaedic Association 2014
Abstract Background Controversy still exists around surgical strategies for Lenke type 1C and 2C curves with primary thoracic and compensatory lumbar curves in adolescent idiopathic scoliosis (AIS). The benefit of selective thoracic fusion (STF) for these curve types is spontaneous lumbar curve correction while saving more mobile lumbar segments. However, a risk of postoperative coronal decompensation after STF has also been reported. This multicenter retrospective study was conducted to evaluate postoperative behavior of thoracolumbar/lumbar (TLL) curve and coronal balance after posterior thoracic fusion for Lenke 1C and 2C AIS. Methods Twenty-four Lenke 1C and 2C AIS patients who underwent posterior thoracic fusion were included. The mean age of patients was 15.7 years old at time of surgery. Constructs used for surgery in all cases were pedicle screw constructs ending at L3 or above. Radiographic measurements were performed on Cobb angles of the main thoracic
and TLL curves and coronal balance. Factors related to final Cobb angle of TLL curve and postoperative change of coronal balance were investigated. Results Mean Cobb angles for main thoracic and TLL curves were 59.0° and 43.9° preoperatively, and were corrected to 21.5° and 22.0° at final follow-up, respectively. Mean coronal balance was −5.6 mm preoperatively and was corrected to −14.6 mm at final follow-up. Final Cobb angle of TLL curve was significantly correlated with immediate postoperative Cobb angle of main thoracic curve and tilt of lowest instrumented vertebra (LIV). Postoperative change of coronal balance was significantly correlated with selection of LIV relative to stable vertebra. Conclusion Spontaneous correction of TLL curve occurred consistently by correcting the main thoracic curve and making the LIV more horizontal after posterior thoracic fusion for Lenke 1C and 2C AIS. The more distal fixation to stable vertebra resulted in coronal balance shifting more to the left postoperatively.
M. Ishikawa · Y. Shiono · M. Nishiyama · Y. Fukui Spine and Spinal Cord Center, Mita Hospital, International University of Health and Welfare, Tokyo, Japan
Introduction
M. Ishikawa · K. Cao · L. Pang · K. Watanabe · M. Yagi · N. Hosogane · Y. Shiono · Y. Fukui · M. Matsumoto Keio Spine Research Group, Keio University, Tokyo, Japan K. Cao · L. Pang · K. Watanabe · N. Hosogane · M. Matsumoto (*) Department of Orthopedic Surgery, School of Medicine, Keio University, 35 Shinanomachi Shinjuku-ku, Tokyo 160-8582, Japan e-mail: [email protected] M. Yagi · M. Machida Department of Orthopedic Surgery, Murayama Medical Center, Tokyo, Japan
The primary goal of surgical treatment for adolescent idiopathic scoliosis (AIS) is to achieve three-dimensional correction of structural curve(s) while maintaining coronal and sagittal balance, and to halt curve progression by solid arthrodesis. At the same time, it is thought that preserving motion segments decreases the risk of accelerated postoperative degenerative changes at unfused spinal segments. Thus, there is some controversy as to whether a minor unstructured curve with an apex that completely deviates from the midline should be treated by selective fusion of the structural curve, or by the fusion of both structural and minor unstructured curves.
13
32
After selective thoracic fusion (STF) for a primary thoracic curve with compensatory lumbar curve was advocated by Moe [1], numerous reports regarding surgical outcomes of STF for this curve type were published. King et al. [2] stated that spontaneous lumbar curve correction occurred after STF with Harrington rod instrumentation when the lowest instrumented vertebra (LIV) was at stable vertebra (SV) in their type II curves. Although other studies have also shown spontaneous unfused lumbar curve correction after STF for a primary thoracic curve with compensatory lumbar curve [3, 4], postoperative coronal decompensation was noted in some patients, resulting in re-operation with extension to the lower lumbar spine [5]. Several causative factors for postoperative coronal decompensation have been reported, including excessive correction of thoracic curve, improper selection of LIV, pre-existing coronal decompensation and inappropriate curve identification [5–9]. However, to our knowledge, many of previous studies on postoperative behavior of unfused lumbar curve and coronal balance after STF are based on various surgical approaches including an anterior approach and/or posterior approach with pedicle screw (PS), hook or hybrid constructs, and those with PS constructs are scant. Apparently, surgical procedures and corrective maneuvers would influence the postoperative course of an unfused lumbar curve after STF. The purposes of this study were (1) to evaluate postoperative behavior of thoracolumbar/lumbar (TLL) curve and coronal balance after posterior thoracic fusion with PS constructs for Lenke 1C and 2C AIS (Lenke 1C/2C-AIS), (2) to identify factors related to final Cobb angle of TLL curve and postoperative change of coronal balance, and finally (3) to determine the optimal surgical strategy for these curve patterns. We hypothesized that preoperative radiographic measurements, LIV selection and amount of main thoracic (MT) curve correction would have an impact on final Cobb angle of TLL curve and postoperative change of coronal balance.
Materials and methods This study was designed as a multicenter retrospective study reviewing radiographs and clinical charts. After IRB approval by the board of ethics committee at the university, radiographic measurements and clinical chart reviews were conducted. Twenty-four patients (2 males, 22 females), who underwent posterior thoracic fusion for Lenke 1C or 2C AIS at three institutes, were enrolled in this study. Inclusion criteria were as follows: (1) primary surgery by posterior thoracic fusion with PS constructs of the LIV ending at L3 or above for Lenke 1C/2C-AIS, (2) patients younger than 25 years old at the time of surgery.
13
M. Ishikawa et al.
Mean age and Risser grade at the time of surgery were 15.7 ± 3.4 (11–24) years old and 3.5 ± 1.5 (1–5), respectively. Mean follow-up period was 29.6 (6–87) months. Investigated clinical data were Lenke classification, number of fused vertebrae, and level of LIV. Radiographic coronal measurements included Cobb angles of proximal thoracic (PT), MT and TLL curves, apical vertebral translations (AVT) of MT (AVT-MT) and TLL (AVT-TLL) curves, LIV and L4 tilt, and coronal balance on preoperative, postoperative (between 1 and 4 weeks after surgery), and final follow-up standing PA radiographs. Preoperative curve flexibilities were evaluated on right and left side-bending films. Radiographic sagittal measurements included proximal thoracic kyphosis between T2 and T5, thoracic kyphosis between T5 and T12, thoracolumbar kyphosis between T10 and L2, lumbar lordosis between T12 and S1, and sagittal balance on preoperative, postoperative, and final follow-up standing lateral radiographs. Measurements for coronal curvature were performed by Cobb method. AVT-MT was defined as the horizontal distance between the center of apex (vertebra or disc) for MT curve and C7 plumb line, and AVT-TLL was defined as the horizontal distance between the center of apex (vertebra or disc) for TLL curve and center sacral vertical line (CSVL). Coronal balance was measured as the horizontal distance between C7 plumb line and CSVL, and was defined as a positive value when C7 plumb line locates at the right side to CSVL, and vice versa. Sagittal balance was measured as the horizontal distance between C7 plumb line and superior–posterior corner of S1 vertebra, and was defined as a positive value when C7 plumb line locates anteriorly to superior-posterior corner of S1 vertebra, and vice versa. Coronal decompensation was defined as the absolute value of coronal balance greater than 2 cm. Preoperative flexibilities (%) of PT, MT and TLL curves were calculated: flexibility (%) = (preoperative Cobb angle − preoperative Cobb angle on side-bending)/(preoperative Cobb angle) × 100. Correction rates of Cobb angles of PT, MT and TLL curves were calculated: correction rate (%) = (preoperative Cobb angle − postoperative Cobb angle)/(preoperative Cobb angle) × 100. The ratios of values of MT curve to those of TLL curve in preoperative Cobb angle and AVT were also calculated. Reference vertebrae including the lower end vertebra (EV) and the lower neutral vertebra (NV) of MT curve, SV at thoracolumbar lesion, and apex of TLL curve (ApexTLL) were recorded, and the gap difference between the LIV and reference vertebrae were counted (LIV-EV, LIV-NV, LIVSV, ApexTLL-LIV). Preoperative radiographic measurements including Cobb angles, and flexibilities of PT, MT and TLL curves, AVT of MT and TLL curves, ratios of MT curve to TLL curve in Cobb angle and AVT, coronal balance, and sagittal values,
33
Postoperative behavior of thoracolumbar
LIV selection and postoperative correction of MT curve and LIV tilt were evaluated whether correlating or not with final Cobb angle of TLL curve and postoperative change of coronal balance at final follow-up.
Surgical procedure All patients were operated on posteriorly. After exposure of the posterior elements of the spine, removal of inferior facet joints to be fused (in all cases) and Ponte osteotomy at the apical and juxta-apical segments with a rigid structural curve (in two cases) were performed. Pedicle screws were inserted bilaterally by free hand technique, with additional attachment of some hooks in eight cases. The majority of the LIVs were determined as SV on PA radiographs, but some of the LIVs were selected at a more distal level to partially correct the TLL curve. The corrective maneuver includes rod rotation on the concave side, followed by in situ rod contouring, placement of a rod on the convex side and segmental compression and distraction via PS. Intraoperative spinal cord monitoring with MEP was routinely performed. Bone grafting was carried out using local bone materials from facetectomies and osteotomies in all cases.
Statistical analyses Statistical analyses were performed using SPSS statistics 22 (IBM Japan, Tokyo, Japan). Radiographic measurement values at different time points were evaluated by analysis of variance. Unpaired t-test or Mann–Whitney U test were used to compare continuous or categorical variables, respectively. Correlation analyses were performed using Pearson’s correlation coefficients. Stepwise linear regression analysis was performed to predict final coronal balance. A P value less than 0.05 was set to be statistically significant.
Results Lenke classifications were 1C− in three, 1CN in eleven, 1C+ in one, 2C− in three, and 2CN in six. Mean number of fused vertebrae was 9.4 ± 2.3(6–14), and LIVs were T11 in seven, T12 in eight, L1 in five, L2 in three, and L3 in one (Table 1). Mean Cobb angles for PT, MT and TLL curves were 29.6 ± 8.9° (flexibility: 24.0 ± 18.6 %), 59.0 ± 13.8° (flexibility: 29.6 ± 15.6 %) and 43.9 ± 7.5° (flexibility: 73.7 ± 17.7 %) preoperatively, and were corrected to 15.7 ± 7.2°, 18.9 ± 8.2° and 20.8 ± 8.1° postoperatively, and 16.5 ± 7.4° (44.4 ± 19.6 % correction), 21.5 ± 9.3° (64.2 ± 10.3 % correction) and 22.0 ± 9.1°
Table 1 Clinical data Gender Age (years) Risser grade Follow-up period (months) Lenke classification
Mean fused vertebrae LIV
Male:2, Female:22 15.7 ± 3.4 3.5 ± 1.5 29.6 (6–87) 1C−:3 1CN:11 1C+:1 2C−:3 2CN:6 9.4 ± 2.3 T11:7 T12:8 L1:5 L2:3 L3:1
LIV lowest instrumented vertebra
(49.4 ± 20.1 % correction) at final follow-up, respectively. Mean preoperative Cobb angle of TLL curve on side-bending was 12.6 ± 10.2°. Preoperative Cobb angle ratio of MT curve to TLL curve was 1.3 ± 0.2. Postoperative changes of Cobb angles in PT, MT, and TLL curves were statistically significant (Table 2). Mean AVT-MT and AVT-TLL were 46.2 ± 16.7 and 23.1 ± 6.4 mm preoperatively, and were corrected to 3.9 ± 15.8 and 24.2 ± 10.9 mm postoperatively, and 7.6 ± 15.5 and 19.3 ± 8.1 mm at final follow-up. Preoperative AVT ratio of MT curve to TLL curve was 2.3 ± 1.2. Mean coronal balance was −5.6 ± 9.8 mm preoperatively and was corrected to −20.9 ± 12.9 mm postoperatively, and −14.6 ± 10.2 mm at final followup. Seven of 24 patients showed coronal decompensation (−21 ~ −40 mm) at final follow-up. None of these seven patients underwent revision surgery. Comparative study showed that of preoperative radiographic measurements, LIV selection, and amount of MT curve correction, only preoperative coronal balance was significantly different between patients with compensation and those without at final follow-up (−2.8 mm vs −12.7 mm, P < 0.05). Mean LIV tilt and L4 tilt were 20.5 ± 19.5° and 12.3 ± 4.3° preoperatively, and were corrected to 8.5 ± 8.2° and 12.4 ± 4.0° postoperatively, and 9.4 ± 8.8° and 9.8 ± 5.1° at final follow-up, respectively. Postoperative changes of AVT-MT, LIV tilt and coronal balance compared with preoperative values were statistically significant (Table 2). Proximal thoracic kyphosis significantly increased postoperatively (preoperative mean of 8.9 ± 4.4° vs final follow-up mean of 11.9 ± 5.0°). Preoperative thoracic kyphosis, thoracolumbar kyphosis and lumbar lordosis did not change significantly at final follow-up (Table 3). Mean sagittal balance was −9.5 ± 22.1 mm preoperatively
13
34
M. Ishikawa et al.
Table 2 Radiographic coronal measurements Preop. PT curve (°) Flexibility (%) Correction rate (%) MT curve (°) Flexibility (%) Correction rate (%) TLL curve (°) Flexibility (%) Correction rate (%) AVT-MT (mm) AVT-TLL (mm) LIV tilt (°) L4 tilt (°)
Table 4 Correlation analyses
Postop.
Final follow-up
15.7 ± 7.2*
16.5 ± 7.4*
59.0 ± 13.8 29.6 ± 15.6
18.9 ± 8.2*
44.4 ± 19.6 21.5 ± 9.3*
43.9 ± 7.5 73.7 ± 17.7
*
29.6 ± 8.9 24.0 ± 18.6
46.2 ± 16.7 23.1 ± 6.4 20.5 ± 19.5 12.3 ± 4.3
Coronal balance (mm) −5.6 ± 9.8
20.8 ± 8.1
3.9 ± 15.8* 24.2 ± 10.9 8.5 ± 8.2* 12.4 ± 4.0
64.2 ± 10.3 22.0 ± 9.1* 49.4 ± 20.1 7.6 ± 15.5* 19.3 ± 8.1 9.4 ± 8.8* 9.8 ± 5.1
−20.9 ± 12.9* −14.6 ± 10.2*
PT proximal thoracic, MT main thoracic, TLL thoracolumbar/lumbar, AVT apical vertebral translation, LIV lowest instrumented vertebra *
Statistical significance
Table 3 Radiographic sagittal measurements Preop. T2–T5 (°) T5–T12 (°) T10–L2 (°) T12–S1 (°) Sagittal balance (mm)
Postop.
Final follow-up
11.9 ± 5.0* 8.9 ± 4.4 11.5 ± 5.0* 17.2 ± 10.3 16.5 ± 8.3 19.3 ± 9.8 −5.0 ± 11.5 −3.2 ± 9.5 −2.8 ± 10.5 −52.4 ± 11.3 −44.6 ± 11.7* −49.2 ± 10.2 −9.5 ± 22.1
1.8 ± 21.1
−15.9 ± 14.9
*
Statistical significance
and was corrected to 1.8 ± 21.1 mm postoperatively, and −15.9 ± 14.9 mm at final follow-up (Table 3). The mean gap differences of LIV-EV, LIV-NV, LIV-SV, and ApexTLL-LIV were 0.9 ± 1.4, 0.9 ± 1.5, 0.5 ± 1.4, and 2.0 ± 1.2, respectively.
Correlation and stepwise linear regression analyses The final Cobb angle of the TLL curve was significantly correlated with the immediate postoperative Cobb angle of the MT curve (r = 0.51, P = 0.01) and LIV tilt (r = 0.67, P = 0.001), and MT curve correction rate (r = −0.56, P = 0.005), but not to LIV selection or preoperative radiographic measurements (Table 4). Although the preoperative Cobb angle of the TLL curve on side-bending tended to correlate with the final Cobb angle of the TLL curve, it was not statistically significant (r = 0.45, P = 0.06).
13
Final Cobb angle of TLL curve vs Postop. Cobb angle of MT curve Postop. LIV tilt angle MT correction rate Postop. change of coronal balance vs LIV-SV Preop. coronal balance
r
P value
0.51 0.67 −0.56
0.01 0.001 0.005
0.44
0.03
0.52
0.01
TLL thoracolumbar/lumbar, MT main thoracic, LIV lowest instrumented vertebra, SV stable vertebra
On the other hand, the postoperative change of coronal balance was significantly correlated with LIV-SV (r = 0.44, P = 0.03) and preoperative coronal balance (r = 0.52, P = 0.01), but not with the other preoperative radiographic measurements or amount of curve correction (Table 4). Stepwise linear regression analysis using parameters correlated with postoperative change of coronal balance developed the formula predicting final coronal balance. Final coronal balance (mm) = −10.6 + 0.4 (preoperative coronal balance) − 3.6 (LIV-SV). The model R2 = 0.38 (Fig. 1).
Discussion Which surgical strategy to use for a primary thoracic curve with compensatory lumbar curve is one of the most controversial issues in the treatment of AIS. Although the negative impact of posterior spinal instrumentation and fusion of the thoracic spine on unfused lumbar lesions, such as acceleration of disc degeneration during long-term follow-up has been pointed out, it is also recognized that, in cases with posterior thoracic fusion ending at an upper lumbar lesion, postoperative degenerative changes of lumbar discs are mild to moderate and incidence of low back pain is comparable to healthy people in the same generation [10]. Moreover, spontaneous correction of the lumbar spine after STF would provide better natural course of the unfused lumbar spine than when left uncorrected. Theoretically, when performing STF for a primary thoracic curve with compensatory lumbar curve, using a surgical strategy to induce maximum spontaneous lumbar curve correction, while maintaining coronal balance, is considered to be optimal. We attempted to define correlative factors to final Cobb angle of TLL curve and postoperative change of coronal balance in Lenke 1C/2C-AIS. Regarding postoperative behavior of the TLL curve, we found that correction of the TLL curve spontaneously occurred postoperatively in all cases with little correction
Postoperative behavior of thoracolumbar
35
Fig. 1 Representative case of a 12-year-old girl who underwent posterior thoracic fusion to a stable vertebra for Lenke 1C adolescent idiopathic scoliosis (a preoperative, b 1 week after surgery, c 3 years after surgery). Preoperative Cobb angle of lumbar curve (47°)
improved spontaneously at 3 years after surgery (12°). Preoperative coronal balance of −14.5 mm was corrected to −16.5 mm at 3 years after surgery. This measurement closely matched −16.4 mm of the predicted value by stepwise linear regression analysis
loss during follow-up periods, and that the final Cobb angle of the TLL curve (22.0°) closely matched the final Cobb angle of the instrumented and corrected MT curve (21.5°). These results suggest that TLL curves in Lenke 1C/2CAIS, which bend less than 25° on side-bending preoperatively, have compensatory nature to accommodate to the corrected thoracic spine. Furthermore, determination of which factors can influence spontaneous correction of TLL curve will be helpful for surgeons in predicting the postoperative behavior of TLL curve and may play a key role in planning surgical strategies. From the results of the correlation analyses, it was noted that the final Cobb angle of the TLL curve was positively correlated with the immediate postoperative Cobb angle of the MT curve and LIV tilt. With respect to preoperative radiographic measurements and LIV selection, only the preoperative Cobb angle of the TLL curve on side-bending tended to correlate with the final Cobb angle of the TLL curve; however, the other radiographic measurements and LIV selection had no impact on the final Cobb angle of the TLL curve. These results
suggest that achieving greater correction of the MT curve and making the LIV more horizontal can induce further spontaneous correction of TLL curve. Regarding coronal balance, this study showed that coronal balance in Lenke 1C/2C-AIS tends to shift to the left preoperatively, and it shifts further to the left immediately postoperatively. Although some improvement of coronal balance occurred during the follow-up period, seven of 24 patients had coronal decompensation (>20 mm) at final follow-up. Comparative study showed that only preoperative coronal balance shifting more to the left was detected as a causative factor for postoperative coronal decompensation. However, only two patients with coronal decompensation at final follow-up had coronal decompensation preoperatively, so further analysis on postoperative change of coronal balance was conducted. By correlation analyses in the current study, postoperative change of coronal balance was significantly correlated with preoperative coronal balance and LIV selection relative to SV, but not to the other preoperative radiographic measurements or
13
36
amount of MT curve correction. Previous studies in the literature reported that causative factors for postoperative coronal decompensation after STF for primary thoracic curve with compensatory lumbar curve include excessive correction of thoracic curve, improper selection of LIV, pre-existing coronal decompensation, and improper curve identification [5–9]. Concerning excessive correction of the thoracic curve as the causative factor for postoperative coronal decompensation, it has been well-speculated since the introduction of Cotrel-Dubousset instrumentation [5, 6]; however, recent reports with PS constructs found that better correction of MT curve did not result in coronal decompensation [11, 12]. This discrepancy may be partially caused by the different mechanism of corrective forces in deformity correction. During the derotation maneuver in the posterior approach, derotation force applied on MT curve with PS constructs would transmit less derotation force to lumbar lesion, compared to with hook constructs, because pedicle screws at lower foundation could regulate aggravation of TLL curve. As for LIV selection, only the relative position of the LIV to SV (LIV-SV) was found to have significant impact on postoperative change of coronal balance in this study. A more distal fixation to SV resulted in coronal balance shifting more to the left postoperatively. Although relative positions of LIV to EV, NV and ApexTLL (LIV-EV, LIVNV, ApexTLL-LIV) tended to correlate with postoperative change of coronal balance, they were not statistically significant. Relative positions of EV, NV, SV and ApexTLL to each other vary in cases with primary thoracic curve with compensatory lumbar curve; however, only the SV is determined consistently with reference to CSVL, so postoperative change of coronal balance may correlate more strongly with relative position of LIV to SV. As demonstrated in previous studies and the current study, preoperative coronal balance tends to shift to the left and pre-existing coronal imbalance is a causative factor of postoperative coronal decompensation in Lenke 1C/2C-AIS, so selecting the LIV as the SV would be optimal to prevent aggravation of coronal imbalance postoperatively. In cases with preoperative severe coronal imbalance to the left, selecting the LIV beyond the apex of the TLL curve should also be considered to maintain coronal balance by correcting both MT and TLL curves, because the ability of STF to maintain coronal balance after surgery may be limited. Stepwise linear regression analysis performed in this study predicts that if the SV is selected as the LIV, preoperative coronal balance would need to be less than −24 mm for achieving a coronal balance of less than −20 mm at final follow-up. However, this study included only one patient with preoperative coronal balance more than −24 mm, so this predicted value should be assessed by the evaluation of
13
M. Ishikawa et al.
prospectively collected data containing a larger number of patients with severe coronal imbalance. This retrospective study contains several limitations, including a small sample size and relatively short followup periods. Prospective studies with a larger number of patients and longer follow-up period (at least until skeletal maturity) may further clarify radiographic and clinical outcomes after posterior thoracic fusion for a primary thoracic curve with compensatory lumbar curve patterns with more statistical power. In conclusion, the current study suggests that a surgical strategy which achieves greater correction of the MT curve and makes the LIV more horizontal, while selecting the LIV as the SV, would be optimal when performing posterior thoracic fusion with a PS construct for Lenke 1C/2C-AIS. Conflict of interest The authors declare that they have no conflict of interest.
References 1. Moe JH. A critical analysis of methods of fusion for scoliosis. An evaluation in two hundred and sixty-six patients. J Bone Joint Surg Am. 1958;40(3):529–54. 2. King HA, Moe JH, Bradford DS, Winter RB. The selection of fusion levels in thoracic idiopathic scoliosis. J Bone Joint Surg Am. 1983;65(9):1302–13. 3. Lenke LG, Betz RR, Bridwell KH, Harms J, Clements DH, Lowe TG. Spontaneous lumbar curve coronal correction after selective anterior or posterior thoracic fusion in adolescent idiopathic scoliosis. Spine. 1999;24(16):1663–72. 4. Dobbs MB, Lenke LG, Walton T, Peelle M, Rocca GD, StegerMay K, Bridwell KH. Can we predict the ultimate lumbar curve in adolescent idiopathic scoliosis patients undergoing a selective fusion with undercorrection of the thoracic curve? Spine. 2004;29(3):277–85. 5. Lenke LG, Bridwell KH, Baldus C, Blanke K. Preventing decompensation in King type II curves treated with Cotrel-Dubousset instrumentation. Strict guidelines for selective thoracic fusion. Spine. 1992;17(8S):S274–81. 6. Thompson JP, Transfeldt EE, Bradford DS, Ogilvie JW, BoachieAdjei O. Decompensation after Cotrel-Dubousset instrumentation of idiopathic scoliosis. Spine. 1990;15(9):927–31. 7. McCance SE, Denis F, Lonstein JE, Winter RB. Coronal and sagittal balance in surgically treated adolescent idiopathic scoliosis with the King II curve pattern. A review of 67 consecutive cases having selective thoracic arthrodesis. Spine. 1998;23(19):2063–73. 8. Demura S, Yaszay B, Bastrom TP, Carreau J, Newton PO, Harms Study Group. Is decompensation preoperatively a risk in Lenke 1C curves? Spine. 2013;38(11):E649–55. 9. Wang Y, Bünger CE, Wu C, Zhang Y, Hansen ES. Postoperative trunk shift in Lenke 1C scoliosis. What causes it? How can it be prevented? Spine. 2012;37(19):1676–82. 10. Green DW, Lawhorne TW III, Widmann RF, Kepler CK, Ahern C, Mintz DN, Rawlins BA, Burke SW, Boachie-Adjei O. Long-term magnetic resonance imaging follow-up demonstrates minimal transitional level lumbar disc degeneration after
Postoperative behavior of thoracolumbar posterior spine fusion for adolescent idiopathic scoliosis. Spine. 2011;36(23):1948–54. 11. Chang KW, Chang KI, Wu CM. Enhanced capacity for spontaneous correction of lumbar curve in the treatment of major thoraciccompensatory C modifier lumbar curve pattern in idiopathic scoliosis. Spine. 2007;32(26):3020–9.
37 12. Takahashi J, Newton PO, Ugrinow VL, Bastrom TP. Selective thoracic fusion in adolescent idiopathic scoliosis. Factors influencing the selection of the optimal lowest instrumented vertebra. Spine. 2011;36(14):1131–41.
13
ORIGINAL ARTICLE
Postoperative behavior of thoracolumbar/lumbar curve and coronal balance after posterior thoracic fusion for Lenke 1C and 2C adolescent idiopathic scoliosis Masayuki Ishikawa · Kai Cao · Long Pang · Kota Watanabe · Mitsuru Yagi · Naobumi Hosogane · Masafumi Machida · Yuta Shiono · Makoto Nishiyama · Yasuyuki Fukui · Morio Matsumoto
Received: 11 April 2014 / Accepted: 21 September 2014 / Published online: 13 October 2014 © The Japanese Orthopaedic Association 2014
Abstract Background Controversy still exists around surgical strategies for Lenke type 1C and 2C curves with primary thoracic and compensatory lumbar curves in adolescent idiopathic scoliosis (AIS). The benefit of selective thoracic fusion (STF) for these curve types is spontaneous lumbar curve correction while saving more mobile lumbar segments. However, a risk of postoperative coronal decompensation after STF has also been reported. This multicenter retrospective study was conducted to evaluate postoperative behavior of thoracolumbar/lumbar (TLL) curve and coronal balance after posterior thoracic fusion for Lenke 1C and 2C AIS. Methods Twenty-four Lenke 1C and 2C AIS patients who underwent posterior thoracic fusion were included. The mean age of patients was 15.7 years old at time of surgery. Constructs used for surgery in all cases were pedicle screw constructs ending at L3 or above. Radiographic measurements were performed on Cobb angles of the main thoracic
and TLL curves and coronal balance. Factors related to final Cobb angle of TLL curve and postoperative change of coronal balance were investigated. Results Mean Cobb angles for main thoracic and TLL curves were 59.0° and 43.9° preoperatively, and were corrected to 21.5° and 22.0° at final follow-up, respectively. Mean coronal balance was −5.6 mm preoperatively and was corrected to −14.6 mm at final follow-up. Final Cobb angle of TLL curve was significantly correlated with immediate postoperative Cobb angle of main thoracic curve and tilt of lowest instrumented vertebra (LIV). Postoperative change of coronal balance was significantly correlated with selection of LIV relative to stable vertebra. Conclusion Spontaneous correction of TLL curve occurred consistently by correcting the main thoracic curve and making the LIV more horizontal after posterior thoracic fusion for Lenke 1C and 2C AIS. The more distal fixation to stable vertebra resulted in coronal balance shifting more to the left postoperatively.
M. Ishikawa · Y. Shiono · M. Nishiyama · Y. Fukui Spine and Spinal Cord Center, Mita Hospital, International University of Health and Welfare, Tokyo, Japan
Introduction
M. Ishikawa · K. Cao · L. Pang · K. Watanabe · M. Yagi · N. Hosogane · Y. Shiono · Y. Fukui · M. Matsumoto Keio Spine Research Group, Keio University, Tokyo, Japan K. Cao · L. Pang · K. Watanabe · N. Hosogane · M. Matsumoto (*) Department of Orthopedic Surgery, School of Medicine, Keio University, 35 Shinanomachi Shinjuku-ku, Tokyo 160-8582, Japan e-mail: [email protected] M. Yagi · M. Machida Department of Orthopedic Surgery, Murayama Medical Center, Tokyo, Japan
The primary goal of surgical treatment for adolescent idiopathic scoliosis (AIS) is to achieve three-dimensional correction of structural curve(s) while maintaining coronal and sagittal balance, and to halt curve progression by solid arthrodesis. At the same time, it is thought that preserving motion segments decreases the risk of accelerated postoperative degenerative changes at unfused spinal segments. Thus, there is some controversy as to whether a minor unstructured curve with an apex that completely deviates from the midline should be treated by selective fusion of the structural curve, or by the fusion of both structural and minor unstructured curves.
13
32
After selective thoracic fusion (STF) for a primary thoracic curve with compensatory lumbar curve was advocated by Moe [1], numerous reports regarding surgical outcomes of STF for this curve type were published. King et al. [2] stated that spontaneous lumbar curve correction occurred after STF with Harrington rod instrumentation when the lowest instrumented vertebra (LIV) was at stable vertebra (SV) in their type II curves. Although other studies have also shown spontaneous unfused lumbar curve correction after STF for a primary thoracic curve with compensatory lumbar curve [3, 4], postoperative coronal decompensation was noted in some patients, resulting in re-operation with extension to the lower lumbar spine [5]. Several causative factors for postoperative coronal decompensation have been reported, including excessive correction of thoracic curve, improper selection of LIV, pre-existing coronal decompensation and inappropriate curve identification [5–9]. However, to our knowledge, many of previous studies on postoperative behavior of unfused lumbar curve and coronal balance after STF are based on various surgical approaches including an anterior approach and/or posterior approach with pedicle screw (PS), hook or hybrid constructs, and those with PS constructs are scant. Apparently, surgical procedures and corrective maneuvers would influence the postoperative course of an unfused lumbar curve after STF. The purposes of this study were (1) to evaluate postoperative behavior of thoracolumbar/lumbar (TLL) curve and coronal balance after posterior thoracic fusion with PS constructs for Lenke 1C and 2C AIS (Lenke 1C/2C-AIS), (2) to identify factors related to final Cobb angle of TLL curve and postoperative change of coronal balance, and finally (3) to determine the optimal surgical strategy for these curve patterns. We hypothesized that preoperative radiographic measurements, LIV selection and amount of main thoracic (MT) curve correction would have an impact on final Cobb angle of TLL curve and postoperative change of coronal balance.
Materials and methods This study was designed as a multicenter retrospective study reviewing radiographs and clinical charts. After IRB approval by the board of ethics committee at the university, radiographic measurements and clinical chart reviews were conducted. Twenty-four patients (2 males, 22 females), who underwent posterior thoracic fusion for Lenke 1C or 2C AIS at three institutes, were enrolled in this study. Inclusion criteria were as follows: (1) primary surgery by posterior thoracic fusion with PS constructs of the LIV ending at L3 or above for Lenke 1C/2C-AIS, (2) patients younger than 25 years old at the time of surgery.
13
M. Ishikawa et al.
Mean age and Risser grade at the time of surgery were 15.7 ± 3.4 (11–24) years old and 3.5 ± 1.5 (1–5), respectively. Mean follow-up period was 29.6 (6–87) months. Investigated clinical data were Lenke classification, number of fused vertebrae, and level of LIV. Radiographic coronal measurements included Cobb angles of proximal thoracic (PT), MT and TLL curves, apical vertebral translations (AVT) of MT (AVT-MT) and TLL (AVT-TLL) curves, LIV and L4 tilt, and coronal balance on preoperative, postoperative (between 1 and 4 weeks after surgery), and final follow-up standing PA radiographs. Preoperative curve flexibilities were evaluated on right and left side-bending films. Radiographic sagittal measurements included proximal thoracic kyphosis between T2 and T5, thoracic kyphosis between T5 and T12, thoracolumbar kyphosis between T10 and L2, lumbar lordosis between T12 and S1, and sagittal balance on preoperative, postoperative, and final follow-up standing lateral radiographs. Measurements for coronal curvature were performed by Cobb method. AVT-MT was defined as the horizontal distance between the center of apex (vertebra or disc) for MT curve and C7 plumb line, and AVT-TLL was defined as the horizontal distance between the center of apex (vertebra or disc) for TLL curve and center sacral vertical line (CSVL). Coronal balance was measured as the horizontal distance between C7 plumb line and CSVL, and was defined as a positive value when C7 plumb line locates at the right side to CSVL, and vice versa. Sagittal balance was measured as the horizontal distance between C7 plumb line and superior–posterior corner of S1 vertebra, and was defined as a positive value when C7 plumb line locates anteriorly to superior-posterior corner of S1 vertebra, and vice versa. Coronal decompensation was defined as the absolute value of coronal balance greater than 2 cm. Preoperative flexibilities (%) of PT, MT and TLL curves were calculated: flexibility (%) = (preoperative Cobb angle − preoperative Cobb angle on side-bending)/(preoperative Cobb angle) × 100. Correction rates of Cobb angles of PT, MT and TLL curves were calculated: correction rate (%) = (preoperative Cobb angle − postoperative Cobb angle)/(preoperative Cobb angle) × 100. The ratios of values of MT curve to those of TLL curve in preoperative Cobb angle and AVT were also calculated. Reference vertebrae including the lower end vertebra (EV) and the lower neutral vertebra (NV) of MT curve, SV at thoracolumbar lesion, and apex of TLL curve (ApexTLL) were recorded, and the gap difference between the LIV and reference vertebrae were counted (LIV-EV, LIV-NV, LIVSV, ApexTLL-LIV). Preoperative radiographic measurements including Cobb angles, and flexibilities of PT, MT and TLL curves, AVT of MT and TLL curves, ratios of MT curve to TLL curve in Cobb angle and AVT, coronal balance, and sagittal values,
33
Postoperative behavior of thoracolumbar
LIV selection and postoperative correction of MT curve and LIV tilt were evaluated whether correlating or not with final Cobb angle of TLL curve and postoperative change of coronal balance at final follow-up.
Surgical procedure All patients were operated on posteriorly. After exposure of the posterior elements of the spine, removal of inferior facet joints to be fused (in all cases) and Ponte osteotomy at the apical and juxta-apical segments with a rigid structural curve (in two cases) were performed. Pedicle screws were inserted bilaterally by free hand technique, with additional attachment of some hooks in eight cases. The majority of the LIVs were determined as SV on PA radiographs, but some of the LIVs were selected at a more distal level to partially correct the TLL curve. The corrective maneuver includes rod rotation on the concave side, followed by in situ rod contouring, placement of a rod on the convex side and segmental compression and distraction via PS. Intraoperative spinal cord monitoring with MEP was routinely performed. Bone grafting was carried out using local bone materials from facetectomies and osteotomies in all cases.
Statistical analyses Statistical analyses were performed using SPSS statistics 22 (IBM Japan, Tokyo, Japan). Radiographic measurement values at different time points were evaluated by analysis of variance. Unpaired t-test or Mann–Whitney U test were used to compare continuous or categorical variables, respectively. Correlation analyses were performed using Pearson’s correlation coefficients. Stepwise linear regression analysis was performed to predict final coronal balance. A P value less than 0.05 was set to be statistically significant.
Results Lenke classifications were 1C− in three, 1CN in eleven, 1C+ in one, 2C− in three, and 2CN in six. Mean number of fused vertebrae was 9.4 ± 2.3(6–14), and LIVs were T11 in seven, T12 in eight, L1 in five, L2 in three, and L3 in one (Table 1). Mean Cobb angles for PT, MT and TLL curves were 29.6 ± 8.9° (flexibility: 24.0 ± 18.6 %), 59.0 ± 13.8° (flexibility: 29.6 ± 15.6 %) and 43.9 ± 7.5° (flexibility: 73.7 ± 17.7 %) preoperatively, and were corrected to 15.7 ± 7.2°, 18.9 ± 8.2° and 20.8 ± 8.1° postoperatively, and 16.5 ± 7.4° (44.4 ± 19.6 % correction), 21.5 ± 9.3° (64.2 ± 10.3 % correction) and 22.0 ± 9.1°
Table 1 Clinical data Gender Age (years) Risser grade Follow-up period (months) Lenke classification
Mean fused vertebrae LIV
Male:2, Female:22 15.7 ± 3.4 3.5 ± 1.5 29.6 (6–87) 1C−:3 1CN:11 1C+:1 2C−:3 2CN:6 9.4 ± 2.3 T11:7 T12:8 L1:5 L2:3 L3:1
LIV lowest instrumented vertebra
(49.4 ± 20.1 % correction) at final follow-up, respectively. Mean preoperative Cobb angle of TLL curve on side-bending was 12.6 ± 10.2°. Preoperative Cobb angle ratio of MT curve to TLL curve was 1.3 ± 0.2. Postoperative changes of Cobb angles in PT, MT, and TLL curves were statistically significant (Table 2). Mean AVT-MT and AVT-TLL were 46.2 ± 16.7 and 23.1 ± 6.4 mm preoperatively, and were corrected to 3.9 ± 15.8 and 24.2 ± 10.9 mm postoperatively, and 7.6 ± 15.5 and 19.3 ± 8.1 mm at final follow-up. Preoperative AVT ratio of MT curve to TLL curve was 2.3 ± 1.2. Mean coronal balance was −5.6 ± 9.8 mm preoperatively and was corrected to −20.9 ± 12.9 mm postoperatively, and −14.6 ± 10.2 mm at final followup. Seven of 24 patients showed coronal decompensation (−21 ~ −40 mm) at final follow-up. None of these seven patients underwent revision surgery. Comparative study showed that of preoperative radiographic measurements, LIV selection, and amount of MT curve correction, only preoperative coronal balance was significantly different between patients with compensation and those without at final follow-up (−2.8 mm vs −12.7 mm, P < 0.05). Mean LIV tilt and L4 tilt were 20.5 ± 19.5° and 12.3 ± 4.3° preoperatively, and were corrected to 8.5 ± 8.2° and 12.4 ± 4.0° postoperatively, and 9.4 ± 8.8° and 9.8 ± 5.1° at final follow-up, respectively. Postoperative changes of AVT-MT, LIV tilt and coronal balance compared with preoperative values were statistically significant (Table 2). Proximal thoracic kyphosis significantly increased postoperatively (preoperative mean of 8.9 ± 4.4° vs final follow-up mean of 11.9 ± 5.0°). Preoperative thoracic kyphosis, thoracolumbar kyphosis and lumbar lordosis did not change significantly at final follow-up (Table 3). Mean sagittal balance was −9.5 ± 22.1 mm preoperatively
13
34
M. Ishikawa et al.
Table 2 Radiographic coronal measurements Preop. PT curve (°) Flexibility (%) Correction rate (%) MT curve (°) Flexibility (%) Correction rate (%) TLL curve (°) Flexibility (%) Correction rate (%) AVT-MT (mm) AVT-TLL (mm) LIV tilt (°) L4 tilt (°)
Table 4 Correlation analyses
Postop.
Final follow-up
15.7 ± 7.2*
16.5 ± 7.4*
59.0 ± 13.8 29.6 ± 15.6
18.9 ± 8.2*
44.4 ± 19.6 21.5 ± 9.3*
43.9 ± 7.5 73.7 ± 17.7
*
29.6 ± 8.9 24.0 ± 18.6
46.2 ± 16.7 23.1 ± 6.4 20.5 ± 19.5 12.3 ± 4.3
Coronal balance (mm) −5.6 ± 9.8
20.8 ± 8.1
3.9 ± 15.8* 24.2 ± 10.9 8.5 ± 8.2* 12.4 ± 4.0
64.2 ± 10.3 22.0 ± 9.1* 49.4 ± 20.1 7.6 ± 15.5* 19.3 ± 8.1 9.4 ± 8.8* 9.8 ± 5.1
−20.9 ± 12.9* −14.6 ± 10.2*
PT proximal thoracic, MT main thoracic, TLL thoracolumbar/lumbar, AVT apical vertebral translation, LIV lowest instrumented vertebra *
Statistical significance
Table 3 Radiographic sagittal measurements Preop. T2–T5 (°) T5–T12 (°) T10–L2 (°) T12–S1 (°) Sagittal balance (mm)
Postop.
Final follow-up
11.9 ± 5.0* 8.9 ± 4.4 11.5 ± 5.0* 17.2 ± 10.3 16.5 ± 8.3 19.3 ± 9.8 −5.0 ± 11.5 −3.2 ± 9.5 −2.8 ± 10.5 −52.4 ± 11.3 −44.6 ± 11.7* −49.2 ± 10.2 −9.5 ± 22.1
1.8 ± 21.1
−15.9 ± 14.9
*
Statistical significance
and was corrected to 1.8 ± 21.1 mm postoperatively, and −15.9 ± 14.9 mm at final follow-up (Table 3). The mean gap differences of LIV-EV, LIV-NV, LIV-SV, and ApexTLL-LIV were 0.9 ± 1.4, 0.9 ± 1.5, 0.5 ± 1.4, and 2.0 ± 1.2, respectively.
Correlation and stepwise linear regression analyses The final Cobb angle of the TLL curve was significantly correlated with the immediate postoperative Cobb angle of the MT curve (r = 0.51, P = 0.01) and LIV tilt (r = 0.67, P = 0.001), and MT curve correction rate (r = −0.56, P = 0.005), but not to LIV selection or preoperative radiographic measurements (Table 4). Although the preoperative Cobb angle of the TLL curve on side-bending tended to correlate with the final Cobb angle of the TLL curve, it was not statistically significant (r = 0.45, P = 0.06).
13
Final Cobb angle of TLL curve vs Postop. Cobb angle of MT curve Postop. LIV tilt angle MT correction rate Postop. change of coronal balance vs LIV-SV Preop. coronal balance
r
P value
0.51 0.67 −0.56
0.01 0.001 0.005
0.44
0.03
0.52
0.01
TLL thoracolumbar/lumbar, MT main thoracic, LIV lowest instrumented vertebra, SV stable vertebra
On the other hand, the postoperative change of coronal balance was significantly correlated with LIV-SV (r = 0.44, P = 0.03) and preoperative coronal balance (r = 0.52, P = 0.01), but not with the other preoperative radiographic measurements or amount of curve correction (Table 4). Stepwise linear regression analysis using parameters correlated with postoperative change of coronal balance developed the formula predicting final coronal balance. Final coronal balance (mm) = −10.6 + 0.4 (preoperative coronal balance) − 3.6 (LIV-SV). The model R2 = 0.38 (Fig. 1).
Discussion Which surgical strategy to use for a primary thoracic curve with compensatory lumbar curve is one of the most controversial issues in the treatment of AIS. Although the negative impact of posterior spinal instrumentation and fusion of the thoracic spine on unfused lumbar lesions, such as acceleration of disc degeneration during long-term follow-up has been pointed out, it is also recognized that, in cases with posterior thoracic fusion ending at an upper lumbar lesion, postoperative degenerative changes of lumbar discs are mild to moderate and incidence of low back pain is comparable to healthy people in the same generation [10]. Moreover, spontaneous correction of the lumbar spine after STF would provide better natural course of the unfused lumbar spine than when left uncorrected. Theoretically, when performing STF for a primary thoracic curve with compensatory lumbar curve, using a surgical strategy to induce maximum spontaneous lumbar curve correction, while maintaining coronal balance, is considered to be optimal. We attempted to define correlative factors to final Cobb angle of TLL curve and postoperative change of coronal balance in Lenke 1C/2C-AIS. Regarding postoperative behavior of the TLL curve, we found that correction of the TLL curve spontaneously occurred postoperatively in all cases with little correction
Postoperative behavior of thoracolumbar
35
Fig. 1 Representative case of a 12-year-old girl who underwent posterior thoracic fusion to a stable vertebra for Lenke 1C adolescent idiopathic scoliosis (a preoperative, b 1 week after surgery, c 3 years after surgery). Preoperative Cobb angle of lumbar curve (47°)
improved spontaneously at 3 years after surgery (12°). Preoperative coronal balance of −14.5 mm was corrected to −16.5 mm at 3 years after surgery. This measurement closely matched −16.4 mm of the predicted value by stepwise linear regression analysis
loss during follow-up periods, and that the final Cobb angle of the TLL curve (22.0°) closely matched the final Cobb angle of the instrumented and corrected MT curve (21.5°). These results suggest that TLL curves in Lenke 1C/2CAIS, which bend less than 25° on side-bending preoperatively, have compensatory nature to accommodate to the corrected thoracic spine. Furthermore, determination of which factors can influence spontaneous correction of TLL curve will be helpful for surgeons in predicting the postoperative behavior of TLL curve and may play a key role in planning surgical strategies. From the results of the correlation analyses, it was noted that the final Cobb angle of the TLL curve was positively correlated with the immediate postoperative Cobb angle of the MT curve and LIV tilt. With respect to preoperative radiographic measurements and LIV selection, only the preoperative Cobb angle of the TLL curve on side-bending tended to correlate with the final Cobb angle of the TLL curve; however, the other radiographic measurements and LIV selection had no impact on the final Cobb angle of the TLL curve. These results
suggest that achieving greater correction of the MT curve and making the LIV more horizontal can induce further spontaneous correction of TLL curve. Regarding coronal balance, this study showed that coronal balance in Lenke 1C/2C-AIS tends to shift to the left preoperatively, and it shifts further to the left immediately postoperatively. Although some improvement of coronal balance occurred during the follow-up period, seven of 24 patients had coronal decompensation (>20 mm) at final follow-up. Comparative study showed that only preoperative coronal balance shifting more to the left was detected as a causative factor for postoperative coronal decompensation. However, only two patients with coronal decompensation at final follow-up had coronal decompensation preoperatively, so further analysis on postoperative change of coronal balance was conducted. By correlation analyses in the current study, postoperative change of coronal balance was significantly correlated with preoperative coronal balance and LIV selection relative to SV, but not to the other preoperative radiographic measurements or
13
36
amount of MT curve correction. Previous studies in the literature reported that causative factors for postoperative coronal decompensation after STF for primary thoracic curve with compensatory lumbar curve include excessive correction of thoracic curve, improper selection of LIV, pre-existing coronal decompensation, and improper curve identification [5–9]. Concerning excessive correction of the thoracic curve as the causative factor for postoperative coronal decompensation, it has been well-speculated since the introduction of Cotrel-Dubousset instrumentation [5, 6]; however, recent reports with PS constructs found that better correction of MT curve did not result in coronal decompensation [11, 12]. This discrepancy may be partially caused by the different mechanism of corrective forces in deformity correction. During the derotation maneuver in the posterior approach, derotation force applied on MT curve with PS constructs would transmit less derotation force to lumbar lesion, compared to with hook constructs, because pedicle screws at lower foundation could regulate aggravation of TLL curve. As for LIV selection, only the relative position of the LIV to SV (LIV-SV) was found to have significant impact on postoperative change of coronal balance in this study. A more distal fixation to SV resulted in coronal balance shifting more to the left postoperatively. Although relative positions of LIV to EV, NV and ApexTLL (LIV-EV, LIVNV, ApexTLL-LIV) tended to correlate with postoperative change of coronal balance, they were not statistically significant. Relative positions of EV, NV, SV and ApexTLL to each other vary in cases with primary thoracic curve with compensatory lumbar curve; however, only the SV is determined consistently with reference to CSVL, so postoperative change of coronal balance may correlate more strongly with relative position of LIV to SV. As demonstrated in previous studies and the current study, preoperative coronal balance tends to shift to the left and pre-existing coronal imbalance is a causative factor of postoperative coronal decompensation in Lenke 1C/2C-AIS, so selecting the LIV as the SV would be optimal to prevent aggravation of coronal imbalance postoperatively. In cases with preoperative severe coronal imbalance to the left, selecting the LIV beyond the apex of the TLL curve should also be considered to maintain coronal balance by correcting both MT and TLL curves, because the ability of STF to maintain coronal balance after surgery may be limited. Stepwise linear regression analysis performed in this study predicts that if the SV is selected as the LIV, preoperative coronal balance would need to be less than −24 mm for achieving a coronal balance of less than −20 mm at final follow-up. However, this study included only one patient with preoperative coronal balance more than −24 mm, so this predicted value should be assessed by the evaluation of
13
M. Ishikawa et al.
prospectively collected data containing a larger number of patients with severe coronal imbalance. This retrospective study contains several limitations, including a small sample size and relatively short followup periods. Prospective studies with a larger number of patients and longer follow-up period (at least until skeletal maturity) may further clarify radiographic and clinical outcomes after posterior thoracic fusion for a primary thoracic curve with compensatory lumbar curve patterns with more statistical power. In conclusion, the current study suggests that a surgical strategy which achieves greater correction of the MT curve and makes the LIV more horizontal, while selecting the LIV as the SV, would be optimal when performing posterior thoracic fusion with a PS construct for Lenke 1C/2C-AIS. Conflict of interest The authors declare that they have no conflict of interest.
References 1. Moe JH. A critical analysis of methods of fusion for scoliosis. An evaluation in two hundred and sixty-six patients. J Bone Joint Surg Am. 1958;40(3):529–54. 2. King HA, Moe JH, Bradford DS, Winter RB. The selection of fusion levels in thoracic idiopathic scoliosis. J Bone Joint Surg Am. 1983;65(9):1302–13. 3. Lenke LG, Betz RR, Bridwell KH, Harms J, Clements DH, Lowe TG. Spontaneous lumbar curve coronal correction after selective anterior or posterior thoracic fusion in adolescent idiopathic scoliosis. Spine. 1999;24(16):1663–72. 4. Dobbs MB, Lenke LG, Walton T, Peelle M, Rocca GD, StegerMay K, Bridwell KH. Can we predict the ultimate lumbar curve in adolescent idiopathic scoliosis patients undergoing a selective fusion with undercorrection of the thoracic curve? Spine. 2004;29(3):277–85. 5. Lenke LG, Bridwell KH, Baldus C, Blanke K. Preventing decompensation in King type II curves treated with Cotrel-Dubousset instrumentation. Strict guidelines for selective thoracic fusion. Spine. 1992;17(8S):S274–81. 6. Thompson JP, Transfeldt EE, Bradford DS, Ogilvie JW, BoachieAdjei O. Decompensation after Cotrel-Dubousset instrumentation of idiopathic scoliosis. Spine. 1990;15(9):927–31. 7. McCance SE, Denis F, Lonstein JE, Winter RB. Coronal and sagittal balance in surgically treated adolescent idiopathic scoliosis with the King II curve pattern. A review of 67 consecutive cases having selective thoracic arthrodesis. Spine. 1998;23(19):2063–73. 8. Demura S, Yaszay B, Bastrom TP, Carreau J, Newton PO, Harms Study Group. Is decompensation preoperatively a risk in Lenke 1C curves? Spine. 2013;38(11):E649–55. 9. Wang Y, Bünger CE, Wu C, Zhang Y, Hansen ES. Postoperative trunk shift in Lenke 1C scoliosis. What causes it? How can it be prevented? Spine. 2012;37(19):1676–82. 10. Green DW, Lawhorne TW III, Widmann RF, Kepler CK, Ahern C, Mintz DN, Rawlins BA, Burke SW, Boachie-Adjei O. Long-term magnetic resonance imaging follow-up demonstrates minimal transitional level lumbar disc degeneration after
Postoperative behavior of thoracolumbar posterior spine fusion for adolescent idiopathic scoliosis. Spine. 2011;36(23):1948–54. 11. Chang KW, Chang KI, Wu CM. Enhanced capacity for spontaneous correction of lumbar curve in the treatment of major thoraciccompensatory C modifier lumbar curve pattern in idiopathic scoliosis. Spine. 2007;32(26):3020–9.
37 12. Takahashi J, Newton PO, Ugrinow VL, Bastrom TP. Selective thoracic fusion in adolescent idiopathic scoliosis. Factors influencing the selection of the optimal lowest instrumented vertebra. Spine. 2011;36(14):1131–41.
13