Progressive scoliosis may be addressed with either orthotic management or surgical treatment. The exact indications for each remain debated as does the surgical approach when such treatment is undertaken. Both anterior and posterior instrumentation methods have been used successfully in the treatment of primary thoracic scoliosis.1–6 When the curve pattern is such that only thoracic instrumentation is required, the choice between anterior and posterior surgical approach exists. There are theoretical advantages and disadvantages to each of these approaches. The proponents of anterior thoracic instrumentation report fewer fusion levels, an anterior shortening procedure which results in a restoration of kyphosis as well as spontaneous transverse plane derotation.1–3 The procedure, however, traditionally has required a thoracotomy, which has approach-related morbidity, particularly with regards to pulmonary function. Pulmonary function is reduced after open anterior approaches to the thoracic spine for instrumentation and remains so up to 2 years after the index procedure.7,8 Additionally, the anterior procedure provides somewhat less rigid bony fixation compared to posterior instrumentation and as such has a greater incidence of loss of fixation and implant related failure.3
A thoracoscopic approach for insertion of anterior instrumentation has been developed in the past 10 years, which obviates many of the disadvantages of the open anterior thoracic approach.9–13 The morbidity associated with a thoracotomy is limited because of the minimal skin and chest wall dissection required with this method.11 Using several small incisions and an endoscope, visualization and anterior spinal surgery have been made possible through this minimally invasive technique. It has been used for over a decade in performing anterior thoracic spinal release and fusion operations with increasing success.14–17 However, early reports with this technique have been mixed, reporting difficulties with visualization and access into the disc space.18 The completeness of disc excision has been questioned; however, in several authors’ hands the results have been comparable with open surgery as experience has been gained.16,17,19–23 Reports regarding thoracoscopic anterior instrumentation have also been mixed with few reports of the 2-year outcomes in such patients.9,24,25
The purpose of this evaluation is to report a single surgeon’s experience with an initial series of 50 consecutive patients. All patients underwent thoracoscopic anterior thoracic instrumentation and fusion for scoliosis. The goal is to report the (greater than 2-year) outcomes with regards to the radiographic findings, pulmonary function, and the SRS Outcomes Instrument, as well as a review of the perioperative data and complications.
The primary author’s initial 50 thoracoscopic anterior spinal instrumentation patients were consecutively collected. All of the patients underwent surgery between July 1999 and September 2002. This data collection period insured greater than 2-year follow-up for all patients. There was an attempt to prospectively collect the data in all 50 cases.
Data collection included demographics, such as age, gender, and type of scoliosis (juvenile idiopathic, adolescent idiopathic, syringomyelia related). The pattern of scoliosis was characterized by the Lenke et al classification.26 Data regarding the surgical procedure included the operative time, estimated intraoperative blood loss, as well as the number of levels instrumented anteriorly. Normalized values for estimated blood loss and operative time were also created by dividing each value by the number of levels instrumented. Additionally, the lengths of the anterior thoracoscopic portal incisions were collected.
In the perioperative hospital period, data were collected with regard to the length of the hospital stay, the number of days in the ICU, the number of days of ventilator support, and the number of days after surgery when conversion from IV to PO pain medication occurred. The chest tube drainage was collected on the operative day, as well as postoperative day 1 and 2 and measured in milliliters. The total number of days the chest tube remained in place was also recorded.
Radiographic data were obtained systematically on each patient and measured by authors other than the surgeon. The data points collected included the upper thoracic, main thoracic, and lumbar Cobb measures, side bending flexibility of each curve, sagittal plane measures, and Risser sign. These radiographic measures were obtained before surgery as well as after surgery at 6 to 12 weeks, 1 year, 2 year, and at the most recent follow-up time beyond 2 years when such a visit existed. The SRS 22 and/or 24 Outcomes Questionnaire27 was administered to patients at similar intervals. Pulmonary function testing (spirometry) was performed before surgery and 2 years after surgery.
The indications for thoracoscopic anterior instrumentation varied slightly throughout the series; however, in general, idiopathic scoliosis patients with primary thoracic scoliosis in which only the thoracic levels were thought to require fusion were considered for this procedure.28–30 Patients were offered either an anterior thoracoscopic or posterior procedure with the decision left to the patient and family. Patients were advised that thoracoscopic surgery would require iliac crest bone graft harvest as well as 3 months of postoperative bracing. The patients who chose posterior instrumentation were not required to undergo postoperative bracing or an iliac crest bone graft harvest as local bone and cancellous allograft was the routine for these cases.
The surgical procedure was initiated with single-lung ventilation. This was performed throughout the series with a variety of techniques, although in most cases a double-lumen endotracheal tube was used and placed with bronchoscopic visualization. The lateral position was used in all cases with the convex side of the curve up. Somatosensory-evoked potential monitoring was performed in the upper and lower extremities in all cases.
The image intensifier was used before prepping and draping to plan portal placement. A direct lateral view of the spine was obtained marking the midlateral aspect of each vertebra along the lateral chest. A direct anteroposterior view was performed as well, marking the orientation of each of the vertebra to be instrumented in the frontal plane. Using the intersection of these lines, a trajectory for each vertebral screw was determined. In general, three portals were placed along the posterior axillary line directly lateral to the vertebral bodies with two additional anterior portals used for visualization, retraction, and discectomy.
Initially, two anterior portals and the most inferior posterior portal were placed using 11.5-mm thoracoports. Standard three-port thoracoscopic exposure of the spine was performed: one port used for retraction, one for visualization, and one as a working port. A longitudinal incision over the pleura approximately 6 to 8 mm anterior to the rib heads was performed over the levels to be instrumented. The segmental vessels overlying each of the selected vertebrae were divided using an ultrasonic device. The levels for instrumentation were confirmed with the image intensifier and circumferential exposure of the spine performed. Raytec sponges were placed between the cut edge of the pleura and the anterior spine, protecting the great vessels and esophagus. Discectomy was initiated with a circumferential annulotomy using the ultrasonic device. The discectomy was completed with rongeurs, carefully removing the endplate cartilage from the superior and inferior aspects of the adjacent vertebrae. At the most inferior 1 or 2 levels, a structural graft, which typically consisted of a fibular allograft, was used (in addition to iliac autograft) to obtain/maintain appropriate sagittal alignment in the lower thoracic segments.
Following the discectomy, the two remaining posterior portals were established. Screws were placed in the proximal vertebrae first, sequentially moving inferiorly. A 15-mm posterior thoracoport was used for screw placement. The rib head was maintained in all cases without other disruption to the ribs. The screw hole was started with an awl placing the screw in the midlateral vertebral body, parallel to the endplates, just anterior to the rib head. A calibrated tap, followed by a depth gauge were used to determine exact screw length with bicortical purchase attempted. Two to three screws were placed through each skin incision, moving the portal into the appropriate rib inter-space in order to align each screw parallel to its corresponding vertebral body (Figure 1A). Screw position was confirmed using the image intensifier in anteroposterior and lateral views and rod length measured with a calibrated template. Early in the series, a 4-mm stainless steel system was used; however, in the latter part of the series, a 4.75-mm diameter titanium alloy rod system was used. In each case, the rods were contoured to the appropriate sagittal plane and some degree of scoliosis placed in the rod based on the ultimate scoliosis correction that was thought would be achievable. The rods were placed in the proximal screws first, capturing these screws and bone grafting the intervertebral space with previously harvested autogenous iliac crest bone graft, which had been placed through a bone mill (Figure 1B). With the proximal three levels captured, bone grafted and compressed, the rod was then sequentially cantilevered into the lower screws. Each of the subsequent levels was similarly grafted and compressed. Following the instrumentation, pleural closure was performed using the Endostitch device (Figure 1C). The chest was irrigated and a chest tube placed in all cases.
The series consisted of 44 females and 6 males with a mean age of 14 years (range, 9–48 years). The diagnosis was adolescent idiopathic scoliosis (AIS) in 44 of the patients, juvenile idiopathic scoliosis in 5, and in a sixth juvenile patient the scoliosis was associated with syringomyelia. Additionally, 1 of the 44 AIS patients had previously undergone an L4–S1 spondylolisthesis fusion. The Lenke et al26 curve types included 21 Type 1A, 8 Type 1B, 13 Type 1C, 5 Type 2A, 1 Type 2C, and 2 Type 3C curves. Forty-five of the 50 patients were available for clinical and radiographic follow-up at greater than or equal to 2 years after surgery. The average length of follow-up for these 45 patients was 33 months (range, 2–5 years). Of the 5 patients with incomplete follow-up data, 2 have moved out of the area and 3 were seen at 1 year after surgery, but not since. Three of the five have been contacted by phone or e-mails and report “no problems.”
Surgical data (all 50 patients) demonstrated that the number of instrumented levels averaged 7.3 ± 0.8 (range, 6–9). The most frequent proximal level was T5 (range, T4–T7), while the most common distal level was T12 (range, T10–L1). The type of instrumentation in the initial 35 patients was with a 4-mm stainless rod system (MOSS-Miami, DePuy Spine, Raynham, MA), while 14 of the 15 remaining patients underwent instrumentation with a 4.75-mm diameter titanium alloy rod system (Frontier, DePuy Spine). The mean operative time for the procedure was 350 ± 50 minutes and ranged from 265 to 528 minutes. The time for anesthesia before the initiation of surgery averaged 57 ± 23 minutes with a range of 29 to 106 minutes. In 43 of the 50 cases, the surgery was performed using five skin incisions on the chest. In 2 cases, four portals were used; and in 5 others, six portals were required. The total length of the portal incisions averaged 10.8 ± 2.0 cm with a range of 8.3 to 17.5 cm. Autogenous bone graft was harvested from the iliac crest in all patients and supplemented with structural allograft in 42 of the 50 cases. The structural allograft was used with autogenous bone and placed only in the inferior levels to prevent iatrogenic thoracolumbar kyphosis. The levels of allograft placement ranged from the T9–T10 disc space to the T12–L1 disc space.
The estimated intraoperative blood loss averaged 431 ± 273 mL (range, 75–1,400 mL). A blood transfusion was given in 34 of the 50 cases, with cell saver blood returned in 25 patients, predonated autologous blood returned in 12 patients, and allogenic blood transfused in 10 patients. Normalizing the operative time and estimated blood loss based on the number of levels treated resulted in an average operative time per level of 48 ± 6 minutes per level and an estimated intraoperative blood loss per level of 60 ± 37 mL per level.
Review of the perioperative hospital course demonstrated that all 50 patients were extubated on the date of surgery, although 3 patients spent from 1 to 6 days in the ICU. The remaining 47 patients were sent to the Intermediate Care Unit (n = 1) or the Orthopedic Ward (n = 46). The chest tube remained in place for all patients for at least 2 days with an average output on the operative day, postoperative day 1 and postoperative day 2 of 355 ± 158 mL, 348 ± 115 mL, and 226 ± 120 mL, respectively. The chest tube was maintained on average for 3 ± 1 day (range, 2–7 days). Pain management was initiated in all patients using intravenous narcotics. On average, they were converted to oral pain medication on postoperative day 4 (range, 2–8 days). The length of hospital stay averaged 6 ± 1 day with a range of 3 to 11 days. The patient requiring an 11-day hospital stay was diagnosed after surgery with sleep apnea and underwent tonsillectomy during his hospitalization, which roughly doubled his length of hospitalization.
All patients were initially treated during their hospitalization with incentive spirometry and oxygen supplementation. Two patients developed atelectasis requiring more aggressive treatment. In 1 patient, mucus plugs developed in the left lung (dependent side during surgery), which required removal by bronchoscopy. An additional patient developed atelectasis on the operative side and was successfully managed with intermittent positive pressure breathing. Additional early complications included 1 patient with reduced bilateral upper extremity sensation in the ulnar nerve distribution of both arms. The etiology of this was unclear and possibly related to intraoperative positioning, although no upper extremity somatosensory-evoked potential changes were noted at the time. The symptoms resolved within 2 days after surgery without treatment. There were no other postoperative neurologic deficits recognized. Postoperative wound complications developed in 2 patients. In 1 patient, an iliac bone graft site infection developed, which required return to the operating room for irrigation and debridement. In a second patient, the iliac bone graft site had partial dehiscence thought related to a stitch abscess. This was successfully managed with local wound care and a course of oral antibiotics.
The radiographic analysis demonstrated a primary structural thoracic curve in all cases. In 18 patients the Risser sign was 0, with the remaining 32 varying between Risser 1 and 5. The preoperative thoracic Cobb averaged 53° ± 9° (range, 40°–80°). This corrected on side bending to 28° ± 10° (range, 9°–50°). This correlated to a flexibility percentage in the thoracic curve on average of 52% ± 16% (range, 23%–96%). The initial correction achieved, as measured on an upright film within the first 3 months of surgery demonstrated a correction of the thoracic Cobb to 20° ± 7° (range, 10°–38°). At the 1- and 2-year postoperative time points, the thoracic deformity measured 22° ± 6° and 24° ± 8°, respectively. Twelve patients had radiographs after the 2-year time point and at most recent follow-up the thoracic Cobb (≥2 years) averaged 24° ± 7.5° (range, 10°–46°) (Figure 2; Table 1). This most recent follow-up data exclude 2 patients who underwent posterior revision surgery (both more than 2 years after the thoracoscopic procedure). The details of these cases are described later.
Sagittal alignment was assessed based on the thoracic kyphosis measured from T5–T12. Before surgery this averaged 19° ± 10° (range, −17°–50°). On the initial postoperative upright radiograph, the kyphosis averaged 24° ± 8° (range, 3°–41°). The T5–T12 thoracic kyphosis at 1 year, 2 year, and latest follow-up was 27° ± 9°, 29° ± 9°, and 28° ± 10°, respectively (Figure 3; Table 1).
The thoracic apical vertebra was translated in the frontal plane relative to the center sacral vertical line, 42 ± 25 mm before surgery and 14 ± 1.2 mm at most recent follow-up. Lowest instrumented vertebra tilt in the coronal plane was reduced from 25° ± 7.6° before surgery to 8.4° ± 6.9° at most recent follow-up. However, the angulation of the disc below the lowest instrumented vertebra was 2.7° ± 3.8° before surgery (range, −5.0°–13°) and at most recent follow-up averaged −1.6° ± 4.6° (range, −11°–8.0°).
Junctional kyphosis was assessed at the levels proximal and distal to the instrumentation. The proximal junctional kyphosis at most recent follow-up averaged 8.4° ± 4.2° with a range of 0° to 19°. There were 9 patients with 11° or more of proximal junctional kyphosis at most recent follow-up. Distal junctional kyphosis at most recent follow-up averaged −1.0° ± 5.1° (range, −17°–10°). In addition to radiograph measures of deformity, scoliometer measures of trunk rotation improved from 15° ± 4° before surgery to 9° ± 4° at most recent follow-up (Figure 4).
Implant failure occurred in three cases. The first (Case 23) was a juvenile patient with syringomyelia and a hyperkyphotic thoracic spine before surgery. She was Risser 0 and presented at her 3-month follow-up visit with a broken 4-mm stainless steel rod at the apex of the deformity. The patient had refused postoperative brace wear. She was subsequently revised with posterior instrumented fusion (Figure 5). The second patient (Case 28) was a relatively small adolescent patient (weight, 42 kg) in which intraoperatively three apical vertebral bodies fractured during screw insertion. The procedure was completed with modest correction and the patient casted for 3 months after surgery. The implants were intact at the 6-month visit, but she presented with an apical rod fracture at the 1-year postoperative check. Her fusion appears stable, and there has been no progression of her scoliosis up to 3 years after surgery. No additional surgical procedures are planned. A third patient (Case 26) with AIS presented with an asymptomatic apical level implant failure at 3.5 years after surgery. The implants were intact at the 2-year postoperative check and radiographically the fusion appears solid, and there is no plan for additional surgical intervention.
Revision posterior spinal fusion procedures were performed in 2 of the 50 cases. One patient (Case 23), as noted above, and a second juvenile patient (Case 16) with an initial thoracoscopic fusion from T5–T10 had modest early loss of proximal fixation. Four years after surgery and after more than 20 cm of growth, she had substantial progression of her deformity. This occurred primarily in the segments distal to her fusion “adding on” three levels. She was revised with a posterior instrumentation and fusion from T3 to L3.
In reviewing the radiographic outcomes, the goal of providing lasting scoliosis correction without additional surgery appears to have been achieved in 47 of the 50 cases. Several cases in addition to the above implant failures and revisions, however, had an outcome less than ideal. These technical complications did not appear to affect the clinical result. Five patients were considered to have been fused at least one level short of ideal (Figure 2, Case 21). Proximal screw plowing or loss of fixation occurred in 2 patients (Figure 2, Case 11). Loss of correction greater than 10° occurred in 5 patients, while hyperkyphosis ≥40° developed after surgery in 5 patients.
SRS Outcomes Questionnaire
The SRS Outcomes Questionnaire results suggested patient were highly satisfied with the outcomes 2 years after surgery with average scores for this domain of 4.8 ± 0.6. The remainder of the domain scores are presented in Table 2.
Pulmonary Function Results
Pulmonary function testing before surgery (n = 27 patients) revealed forced vital capacity (FVC) and forced expiratory volumes in one second (FEV1) values of 2.9 ± 0.6 L and 2.5 ± 0.5 L, respectively. The corresponding percent predicted compared with norms was 92% ± 18% for FVC and 84% ± 12% for FEV1. Two years after surgery, the FVC was 3.1 ± 0.5 L (92% ± 15% predicted). These were unchanged from preoperative values (P = 0.06 and P = 0.7, respectively). FEV1 values 2 years after surgery were also similar to those before surgery at 2.7 ± 0.4 L (P = 0.02) and 84% ± 12% predicted (P = 0.4), although the small increase in the FEV1 value was statistically significant.
To evaluate the potential “learning curve” effect, the series was divided into five groups of 10 sequential cases each. Group 1 consisted of Cases 1 to 10, Group 2 of Cases 11 to 20, and so on. The first case from each group in addition to Case 50 is displayed in Figures 2 to 4. The perioperative data analyzed by group are presented in Table 3. There was a trend toward decreasing normalized operative time with less variability as the series progressed. The estimated blood loss when normalized to the number of surgical levels treated demonstrated a significant decline as well (Figure 6). The percentage correction of the thoracic curve increased as the series progressed (Figure 7).
Complications occurred throughout the series; however, major complications requiring revision posterior spinal fusion primarily occurred in the first half of the series (Cases 16 and 23). Two additional patients with implant failures who did not require revision occurred early in the second half of the series (Cases 26 and 28). There have been no implant failures after changing from the 4.0-mm stainless steel rod system to the 4.75-mm titanium alloy rod. This may relate not only to the increased resistance to fatigue failure of the larger titanium alloy rod but also to better fusion technique, which has been gained with additional experience. In addition, the length of follow-up is shorter in those patients instrumented with the 4.75-mm titanium alloy rod, although there is greater than 2-year follow-up in each case. In addition, 4 of the 5 patients fused too short and both of the patients with proximal screw migration occurred in the first half of the series.
The results of this initial series of patients suggests the thoracoscopic approach is a viable option in the treatment of AIS. The results of this series of patients compares favorably with the results reported by Betz et al3 for a series of patients treated with an open anterior thoracic approach. The types of curve treated (both in curve type and magnitude) were similar, as were the radiographic outcomes. The correction achieved 2 years after surgery was 55% in the present series and 58% in historical open series of Betz et al.3 Although the implant failure rate was higher in the open series (31%), many of the patients were treated with smaller threaded rods. In a series by Sweet et al,2 using solid rods similar to the present series, the incidence of implant failure was comparable 5% versus 6%. Two other authors have reported on 2-year outcomes of thoracoscopic instrumentation for AIS, Wong et al9 and Picetti and Pang24 in 2004. Wong et al reported on 12 cases comparing their results to a group of similar historical cases treated with posterior instrumentation.9 The radiographic outcomes were similar between the groups and were comparable to the present series as well. Picetti and Pang24 noted greater difficulties with the earlier cases in their series with a high rate of nonunion (60%) in their first 15 cases. However, in the latter half of their series, the results approximated those of Wong et al and the present series. Thus, the results of this consecutive series of patient treated with thoracoscopic instrumentation and fusion match those of prior open thoracotomy series and the thoracoscopic series reported to date.
Although this and other reports10–13,25,31 agree that anterior thoracic scoliosis correction can be accomplished thoracoscopically, the question of “Should it be done by this approach?” remains. The procedure is technically demanding and offers little room for error. Single anterior rod instrumentation is much less resistant to fatigue failure than posterior constructs, thus requiring early fusion for the operation to be successful. Early fusion demands thorough discectomy technique, which when performed thoracoscopically requires substantial experience.16,17 The discectomy also plays heavily into the correction achieved by the instrumentation as well. Three-dimensional thoracic scoliosis correction with the anterior approach requires anterior column shortening that is possible only when the full circumference of the anulus fibrosis and nucleus is removed. Complete disc excision allows transverse plane derotation, restoration of thoracic kyphosis, and coronal plane correction. The thoracic vertebrae are generally large enough to easily accommodate a single transverse screw; however, the placement within the vertebra is critical both for screw purchase and ultimate rod placement. There is no good way to recover from a misplaced screw, making selection of the starting point and direction critical to the success of the procedure.
This study demonstrates the complications that can be expected during the adoption of such a technique, this despite the primary authors experience with over 100 thoracoscopic anterior discectomy and fusion procedures and 2 years of instrumentation technique development in an animal model before initiating this series of clinical cases. Revision surgery was required for 4% (2 patients) of the 50 patients. Technical complications occurred in another 16% (8 patients), although the clinical outcomes were not substantially altered. More than half of these cases involved improper selection of the distal level of instrumentation. Refinements in patient selection and surgical technique/ability will likely reduce the frequency of these technical complications.
Given these challenges, back to the question: “Should it be done?” The results of this series will not allow that question to be completely answered. There are potential advantages of this operation over traditional open anterior and posterior approaches, the benefit of which remain to be critically measured. The cosmetic nature of scoliosis surgery cannot be overlooked, and many patients (given the procedure can be done safely without revision) are interested in small, less obvious scars that the thoracoscopic approach clearly offers. The potential functional benefits, however, remain the primary driving force in the development of this technique. Capitalizing on the advantages of open anterior thoracic instrumentation, principally shorter fusions, without the chest wall morbidity or posterior muscle dissection have been the goals of thoracoscopic instrumentation. A comprehensive comparison of the outcomes of these three surgical approaches will be required to ultimately define the role of thoracoscopic AIS treatment.
As for all procedures, success is and will remain dependent on patient selection as well as technical ability. Improving technical ability was suggested in the present series; however, patient selection also evolved. The author’s current practice is to offer this procedure to adolescent (rather than juvenile) girls with Lenke 1 curves less than 70°. If their T5–T12 kyphosis measures greater than 30°, a posterior procedure is suggested. The length of fusion is rarely less than 7 levels, and if the distal end vertebra could be either of two levels (i.e., parallel discs), the instrumentation is extended to the distal of the two vertebrae.
Thoracoscopic anterior instrumentation for adolescent idiopathic scoliosis is a viable surgical option. The outcomes of this consecutive series of patients are comparable to prior open and endoscopic series presented in the literature. The technical challenges of this operation are evident in the learning curve effect, which has been demonstrated. The role of thoracoscopic scoliosis correction is beginning to be defined, yet room for continued analysis clearly remains.
- Anterior thoracoscopic instrumentation is feasible for the correction of thoracic adolescent idiopathic scoliosis.
- The results improved as the series progressed, suggesting a learning curve effect.
- Implant failure occurred in 3 (6%) of the 50 patients, with revision posterior instrumentation required for 2 patients (4%).
1.Lowe T, Betz R, Lenke L, et al. Anterior single-rod instrumentation of the thoracic and lumbar spine: saving levels. Spine
2.Sweet FA, Lenke LG, Bridwell KH, et al. Prospective radiographic and clinical outcomes and complications of single solid rod instrumented anterior spinal fusion in adolescent idiopathic scoliosis. Spine
3.Betz RR, Harms J, Clements DH, et al. Comparison of anterior and posterior instrumentation for correction of adolescent thoracic idiopathic scoliosis. Spine
4.Remes V, Helenius I, Schlenzka D, et al. Cotrel-Dubousset (CD) or Universal Spine System (USS) instrumentation in adolescent idiopathic scoliosis (AIS): comparison of midterm clinical, functional, and radiologic outcomes. Spine
5.Asher M, Lai SM, Burton D, et al. Safety and efficacy of Isola instrumentation and arthrodesis for adolescent idiopathic scoliosis: two- to 12-year follow-up. Spine
6.Edwards C, Lenke L, Peelle M, et al. Selective thoracic fusion for adolescent idiopathic scoliosis with C modifier lumbar curves: 2 to 16 year radiographic and clinical results. Spine
7.Graham E, Lenke L, Lowe T, et al. Prospective pulmonary function evaluation following open thoracotomy for anterior spinal fusion in adolescent idiopathic scoliosis. Spine
8.Lenke LG, Newton PO, Marks MC, et al. Prospective pulmonary function comparison of open versus endoscopic anterior fusion combined with posterior fusion in adolescent idiopathic scoliosis. Spine
9.Wong H, Hee H, Yu Z, et al. Results of thoracoscopic instrumented fusion versus conventional posterior instrumented fusion in adolescent idiopathic scoliosis undergoing selective thoracic fusion. Spine
10.Picetti GD 3rd, Bueff HU. Endoscopic instrumentation, correction, and fusion of idiopathic scoliosis. Spine J
11.Newton PO, Marks M, Faro F, et al. Use of video-assisted thoracoscopic surgery to reduce perioperative morbidity in scoliosis surgery. Spine
12.Sucato DJ. Thoracoscopic anterior instrumentation and fusion for idiopathic scoliosis. J Am Acad Orthop Surg
13.Picetti G, Pang D, Beuff H. Thoracoscopic techniques for the treatment of scoliosis: early results in procedure development. Neurosurgery
14.Mack MJ, Regan JJ, McAfee PC, et al. Video-assisted thoracic surgery for the anterior approach to the thoracic spine. Ann Thorac Surg
15.Regan JJ, Mack MJ, Picetti GD 3rd. A technical report on video-assisted thoracoscopy in thoracic spinal surgery: preliminary description. Spine
16.Newton PO, White KK, Faro F, et al. The success of thoracoscopic anterior fusion in a consecutive series of 112 pediatric spinal deformity cases. Spine
17.Al-Sayyad MJ, Crawford AH, Wolf RK. Early experiences with video-assisted thoracoscopic surgery: our first 70 cases. Spine
18.Arlet V. Anterior thoracoscopic spine release in deformity surgery: a meta-analysis and review. Eur Spine J
19.Newton PO, Shea KG, Granlund KF. Defining the pediatric spinal thoracoscopy learning curve: sixty-five consecutive cases. Spine
20.Early S, Newton P, White K, et al. The feasibility of anterior thoracoscopic spine surgery in children under 30 kilograms. Spine
21.Sucato D, Elerson E. A comparison between the prone and lateral position for performing a thoracoscopic anterior release and fusion for pediatric spinal deformity. Spine
22.Niemeyer T, Freeman BJ, Grevitt MP, et al. Anterior thoracoscopic surgery followed by posterior instrumentation and fusion in spinal deformity. Eur Spine J
23.King AG, Mills TE, Loe WA Jr, et al. Video-assisted thoracoscopic surgery in the prone position. Spine
24.Picetti G, Pang D. Thoracoscopic techniques for the treatment of scoliosis. Childs Nerv Syst
25.Sucato D, Flohr R. Accurate preoperative rod length measurement for thoracoscopic anterior instrumentation and fusion for idiopathic scoliosis. J Spinal Disord Tech
26.Lenke LG, Betz RR, Harms J, et al. Adolescent idiopathic scoliosis: a new classification to determine extent of spinal arthrodesis. J Bone Joint Surg Am
27.Haher TR, Gorup JM, Shin TM, et al. Results of the Scoliosis Research Society instrument for evaluation of surgical outcome in adolescent idio-pathic scoliosis: a multicenter study of 244 patients. Spine
28.Newton PO. The use of video-assisted thoracoscopic surgery in the treatment of adolescent idiopathic scoliosis. In: Pellegrini J, ed. Instructional Course Lectures
. Rosemont, IL: American Academy of Orthopaedic Surgeons, 2005:551–8.
29.Newton PO, Betz R, Clements D, et al. Thoracoscopic approach for a deformity with frontier instrumentation. In: Kim DH, Fessler RG, Regan JJ, eds. Endoscopic Spine Surgery and Instrumentation, Percutaneous Procedures
. New York: Thieme, 2005:162–70.
30.Newton PO. Alternative approaches to thoracoscopic anterior spinal release and fusion for spine deformity. In: Regan JJ, Lieberman IH, eds. Atlas of Minimal Access Spine Surgery
. St Louis: Quality Medical, 2004:385–97.
31.Sucato DJ, Kassab F, Dempsey M. Analysis of screw placement relative to the aorta and spinal canal following anterior instrumentation for thoracic idiopathic scoliosis. Spine