Abstract
Background: The surgical outcomes in patients with scoliosis at two years following anterior thoracoscopic spinal in-strumentation and fusion have been reported. The purpose of this study was to evaluate the results at five years.
Methods: A consecutive series of forty-one patients with major thoracic scoliosis treated with anterior thoracoscopic spinal instrumentation was evaluated at regular intervals. Prospectively collected data included patient demographics, radiographic measurements, clinical deformity measures, pulmonary function, an assessment of intervertebral fusion, and the scores on the Scoliosis Research Society (SRS-24) outcomes instrument. Perioperative and postoperative complications were recorded. Patient data for the preoperative, two-year, and five-year postoperative time points were compared. In addition, a univariate analysis compared selected two-year radiographic, pulmonary function, and SRS-24 data of the study cohort and those of the patients lost to follow-up.
Results: Twenty-five (61%) of the original forty-one patients had five-year follow-up data and were included in the analysis. Between the two-year and five-year follow-up visits, no significant changes were observed with regard to the average percent correction of the major Cobb angle (56% ± 11% and 52% ± 14%, respectively), average total lung capacity as a percent of the predicted value (95% ± 14% and 91% ± 10%), and the average total SRS-24 score (4.2 ± 0.4 and 4.1 ± 0.7). Radiographic evaluation of intervertebral fusion at five years revealed convincing evidence of a fusion with remodeling and trabeculae present at 151 (97%) of the 155 instrumented motion segments. No postoperative infections or clinically relevant neurovascular complications were observed. Rod failure occurred in three patients, and three patients required a surgical revision with posterior spinal instrumentation and fusion.
Conclusions: Thoracoscopic anterior instrumentation for main thoracic idiopathic scoliosis results in five-year outcomes comparable with those reported previously for open anterior and posterior techniques. The radiographic findings, pulmonary function, and clinical measures remain stable between the two and five-year follow-up time points. Thoracoscopic instrumentation provides a viable alternative to treat spinal deformity; however, the risks of pseudarthrosis, hardware failure, and surgical revision should be considered along with the advantages of limited muscular dissection and improved scar appearance.
Level of Evidence: Therapeutic Level IV. See Instructions to Authors for a complete description of levels of evidence.
Although anterior scoliosis surgery was introduced more than three decades ago1, segmental posterior spinal instrumentation and fusion is the current gold standard for the treatment of patients with adolescent idiopathic scoliosis2. Mid-term and long-term follow-up with posterior techniques have shown consistent results in terms of deformity correction and fusion, with low complication rates3-8. However, the advantages of the anterior approach, primarily the ability to create thoracic kyphosis and better correct vertebral rotation with the fusion of fewer motion segments, continue to make this a viable surgical option9-12.
In the last decade, endoscopic techniques have been developed to instrument the anterior spine through a minimally invasive approach that is less detrimental to pulmonary function and more favorable in terms of postoperative pain and appearance13-15. Early studies have described high rates of pseudarthrosis, implant failure, and loss of fixation16-18; however, after the initial steep learning curve was overcome19,20, comparable results between thoracoscopic anterior procedures and alternate (posterior and open anterior) spinal techniques have been reported9,21-27. Those studies described results from twenty-four to forty-five months after surgery; however, to our knowledge, five-year follow-up data have not been reported previously.
The purposes of this study were to evaluate a single-surgeon, consecutive case series and to present the five-year follow-up data on the patients treated with thoracoscopic anterior instrumentation and arthrodesis at our institution. Specifically, the maintenance of deformity correction and spinal balance, the prevalence of complications, the ability to achieve fusion, pulmonary function data, and patient-based outcome measures were analyzed.
Patients with a main thoracic spinal deformity who were scheduled to undergo an endoscopic anterior spinal instrumentation were prospectively enrolled in this institutional review board-approved study. Radiographic and clinical evaluations were performed at regular intervals (preoperatively and at six weeks, three months, six months, one year, two years, and five years postoperatively). Data from the preoperative two-year and five-year follow-up visits were analyzed in this study. The prospectively collected data included demographic data (age, sex, and type of scoliosis), radiographs, clinical deformity measures, pulmonary function, an assessment of intervertebral fusion, and the scores on the Scoliosis Research Society outcomes instrument (SRS-24 questionnaire). Perioperative and postoperative complications, including deep wound infections, pulmonary dysfunction, implant failure, need for surgical revision, and clinically relevant neurologic compromise or vascular injuries, were recorded.
Twelve radiographic variables were evaluated. Coronal measures included the Cobb angles of the upper thoracic, thoracic, and thoracolumbar/lumbar curves, coronal and sagittal deviation of C7 from the center sacral vertical line, thoracic apical translation, and thoracolumbar/lumbar apical translation. Sagittal measures included proximal and distal junctional kyphosis, thoracic kyphosis (T5 to T12), sagittal alignment of the thoracolumbar junction (T10 to L2), and lumbar lordosis (T12 to sacrum). The percent correction of the major thoracic, upper thoracic, and thoracolumbar/lumbar Cobb angles was also determined at two years and five years postoperatively.
Clinical measures of the spinal deformity were collected preoperatively and at the two-year and five-year time points. Thoracic rib hump and lumbar prominence measurements were obtained with use of a standard scoliometer with the patient bending forward. Coronal decompensation was measured as the horizontal distance in centimeters between the prominence of C7 and the center of the posterior superior iliac spines, with the patient standing upright. Trunk shift was measured as the horizontal distance in centimeters between the center of the trunk and the center of the hips with the patient standing upright. Lastly, the vertical distance (in centimeters) between the right and left shoulder heights was also recorded.
The extent of intervertebral fusion was assessed on two-year and five-year postoperative posteroanterior and lateral radiographs at each motion segment where an anterior arthrodesis had been attempted. A four-level grading scale defined by Bridwell et al.28 was adapted to grade the extent of osseous fusion. Grade 1 was defined as complete obliteration of the disc space with bone. Grade 2 was defined as evidence of bridging bone formation over =50% of the disc space. Both Grade 1 and Grade 2 were considered to be fused radiographically. Grade-3 fusions were considered "questionable" and were defined as those that had <50% of the disc space filled with bone. Grade-4 fusions had no evidence of bone formation and/or had osteophytes, suggesting a pseudarthrosis, and were considered a radiographic failure of fusion.
Pulmonary function testing was performed at the preoperative evaluation and at the two-year and five-year follow-up visits. Each test was repeated three times, and the best result of the three was recorded for the forced vital capacity and the forced expiratory volume in one second. The total lung capacity was then calculated for each patient. In addition, the corresponding percent of the predicted value was calculated for each variable (forced vital capacity, forced expiratory volume in one second, and total lung capacity).
The SRS-24 questionnaire29 was administered to patients during the preoperative evaluation and at each follow-up visit. The SRS-24 scores were collected for the four preoperative domains (pain, general self-image, function from back condition, and level of activity) and three postoperative domains (self-image, function, and satisfaction). All questions were given a score from 1 to 5, with 5 being the optimal response, and each domain was analyzed with use of a mean score. The domain scores were normalized to account for a variation in the number of questions answered before and after surgical treatment. The normalized score was determined by dividing the domain scores by the number of questions answered.
Surgical Technique
The indications for surgery and the surgical technique have been described in detail in a previous study22. Briefly, patients with a primary thoracic scoliosis who were thought to require a spinal instrumentation and fusion of the thoracic spine were allowed to choose either an anterior thoracoscopic or a posterior procedure. Patients treated thoracoscopically were advised that they would require iliac crest bone-grafting, as well as three months of postoperative bracing. On the other hand, patients treated with a posterior spinal fusion did not receive iliac crest bone-grafting or postoperative bracing.
Endoscopically treated patients underwent single-lung ventilation and were placed in the lateral position. Somatosensory-evoked potential monitoring was performed in the upper and lower extremities in all patients. (Motor-evoked potential monitoring was not routinely performed prior to 2003 at our institution.) An image intensifier was first used to identify and confirm the levels for instrumentation. Three portals along the posterior axillary line and two anterior portals were used in the majority of patients to obtain complete exposure of the anterior vertebral body (rib head to rib head). The end vertebrae used to measure the thoracic Cobb angle were instrumented in each patient. A thorough discectomy was performed with careful excision of the superior and inferior vertebral end plates.
Bicortical vertebral body screws were placed at each level, and, after confirmation of the screw position, a rod contoured to the sagittal plane was placed in the screws in the cephalad vertebrae. After the rod was captured in the cephalad three vertebrae, the intervertebral disc spaces were filled with previously harvested and ground autogenous iliac crest bone. Structural allograft was used in the thoracolumbar region as needed to prevent iatrogenic hyperkyphosis. The rod was then sequentially cantilevered into the screws in the caudad vertebrae to obtain appropriate scoliosis deformity correction. A rod derotation maneuver was not performed in any of these cases. The pleura was then closed and a chest tube was placed.
Statistical Analysis
Repeated-measures analysis of variance was performed to compare differences in patient data among the preoperative, two-year, and five-year visits. Data were checked for normality and equal variances, and, in consideration of the number of comparisons performed, a Bonferroni post hoc analysis was used to set the level of significance for comparisons between the two and five-year follow-up data at a p value of 0.005. In addition, power and sample-size analyses were performed on the pulmonary function data to determine the minimum number of patients required to identify a significant difference between the two-year and five-year measurements. Lastly, a secondary analysis was performed with use of a univariate analysis of variance (p < 0.05) to compare selected two-year radiographic, pulmonary function, and SRS-24 data between the study cohort and the patients lost to follow-up.
The cases of forty-one consecutive patients from the initial series of thoracoscopic anterior instrumentation procedures performed by the primary author (P.O.N.), with potential five-year follow-up data, were reviewed. Only twenty-five patients (61%), however, had returned for a five-year visit and were included in the statistical analysis. Fourteen of the sixteen patients lost to five-year follow-up were last evaluated at the two-year postoperative visit and are discussed further. All patients had surgery between July 1999 and March 2002. During this same time period, twelve patients who were candidates for a thoracoscopic procedure chose a posterior spinal instrumentation and fusion instead.
Demographic data for the twenty-five patients included in this study cohort are shown in a table in the Appendix. The average age (and standard deviation) at the time of surgery was 13 ± 2 years (range, ten to sixteen years), and patients had an average follow-up of 64 ± 7 months (range, fifty-five to eighty months). All patients had a diagnosis of idiopathic scoliosis, except one who had a syrinx-related thoracic spinal deformity. Twenty-one patients were managed with a 4.0-mm stainless-steel instrumentation system, and four patients were managed with a 4.75-mm titanium-alloy construct. On the average, 6 ± 1 levels were instrumented in these patients. The cephalad instrumented vertebrae ranged from T4 to T7, and the caudad instrumented vertebra ranged from T10 to L1.
Radiographic Measures
The Lenke classification system was used to describe the spinal deformity in each patient30. The values for all coronal and sagittal radiographic parameters are shown in Table I. Preoperatively, the major Cobb angle averaged 53.2° ± 9.8° (range, 41° to 80°), with a measurement of the main thoracic deformity on a bending radiograph that averaged 25.6° ± 8.5° (range, 9° to 41°) and an average flexibility of 52% ± 14% (range, 23% to 78%). At the two-year visit, the average major Cobb angle was 23.4° ± 6.0° (range, 11° to 33°), resulting in an average deformity correction of 55.8% ± 10.6% (range, 29.8% to 73.8%). This was maintained at the five-year visit, with an average major Cobb angle of 25.6° ± 8.1° (range, 12° to 40°) and an average percent correction of 51.5% ± 14.1% (range, 28.8% to 77.8%) from preoperative values (Figs. 1-A through 1-E). There was no clinically important difference in the major Cobb angle between the two follow-up visits (p = 0.11), with an average increase in curve magnitude of only 2.2° ± 6.7° (range, -18.0° to 14°). A negative value indicates that the major Cobb angle decreased over time. The average values and percent correction of the compensatory upper thoracic and thoracolumbar/lumbar curves are shown in Table I. Again, there was no significant difference in the percent correction of the compensatory curves between the two follow-up visits (p > 0.12). Of note, the preoperative best measurements of the upper thoracic and thoracolumbar/lumbar bending radiographs averaged 15.7° ± 6.8° (range, 0° to 29°) and 7.9° ± 7.8° (range, 0° to 28°), respectively, with an average percent flexibility of 44% ± 22% (range, 4% to 100%) and 76% ± 22% (range, 30% to 100%), respectively.
In terms of the sagittal alignment, the average thoracic kyphosis increased significantly from 18.7° ± 9.3° (range, 2° to 30°) preoperatively to 28.8° ± 10.5° (range, 10° to 48°) at two years (p = 0.04). At five years, it was maintained at 25.8° ± 10.5° (range, 8° to 43°) with no clinically significant difference between the two follow-up visits (p = 0.12). The average lumbar lordosis also increased from 58.2° ± 8.6° (range, 43° to 78°) preoperatively to 65.0° ± 10.0° (range, 44° to 84°) at two years; however, this change was not found to be significant (p = 0.07). At five years, the average lumbar lordosis was maintained at 65.4° ± 12.7° (range, 43° to 86°) (p = 0.79). The preoperative, two-year, and five-year sagittal measures of the thoracolumbar junction (T10 to L2) are shown in Table I, and none changed significantly between the two follow-up visits (p = 0.52).
Clinical Measures
Clinical measures of the spinal deformity are shown in Table II. Between the preoperative and the two-year postoperative visits, there were significant improvements in the average coronal decompensation (1.1 ± 0.4 cm to 0.4 ± 0.6 cm) and average rib hump measurements (15.2° ± 4.1° to 8.2° ± 4.1°) (p < 0.04). Lumbar prominence, trunk shift, and shoulder height measurements also improved at two years; however, the changes were not significant (p > 0.12). Between the two and five-year follow-up visits, there were no clinically significant differences in any of the deformity measures analyzed (p > 0.35).
Assessment of Intervertebral Fusion
Autogenous iliac crest bone graft was used in all twenty-five patients, and twenty patients had supplemental structural allograft. Allograft placement was used from the T9-T10 disc space to the T12-L1 disc space. Segmental assessment of the five-year postoperative radiographs revealed a solid osseous fusion (Grade 1 or 2) with clear bridging and trabecular bone at 151 (97%) of 155 total motion segments. Twenty-one patients (84%) were given Grade-1 or 2 ratings over all motion segments instrumented (see Appendix). One patient was noted to have a so-called questionable fusion (Grade 3) at one motion segment, and three patients (12%) were noted radiographically to have a pseudarthrosis (Grade 4) at one motion segment each. A clinically symptomatic pseudarthrosis was present in only one patient, and it required a posterior revision procedure secondary to persistent pain at that level.
For comparison, an assessment of the radiographs of the twenty-five patients made at two years postoperatively showed a solid osseous fusion (Grade 1 or 2) in 146 (94%) of 155 motion segments. Sixteen (64%) of the twenty-five patients were given ratings of Grade 1 or 2 over all motion segments instrumented. Six patients (24%) were noted to have a questionable fusion (Grade 3) at one motion segment, and three patients (12%) were noted radiographically to have a pseudarthrosis (Grade 4) at one motion segment each.
Complications
No postoperative deep wound infections or clinically relevant neurovascular or pulmonary complications were noted in any of these patients. Implant failure (rod breakage) occurred in three (7%) of forty-one patients; all three had instrumentation with use of a 4.0-mm stainless-steel rod. The first patient (Case 23) was noted to have a broken rod (at T8-T9) at the three-month postoperative visit, and she had an asymptomatic pseudarthrosis at that level. She eventually required a surgical revision because of deformity progression (described below). The second patient (Case 25) was noted to have a broken rod (at T9-T10) more than three years after surgery, and the third patient (Case 27) was noted to have a broken rod (at T8-T9) at the one-year postoperative visit. Both of these patients were asymptomatic at the time of the latest follow-up and had not required a surgical revision.
Surgical Revisions
Posterior spinal instrumentation and arthrodesis was required in three (7%) of forty-one patients. The first patient (Case 7) had intermittent back and shoulder pain in the postoperative period and, despite activity modification, anti-inflammatory medication, and physical therapy, the symptoms persisted. Computed tomography confirmed the presence of a pseudarthrosis at T8-T9 (Fig. 2), and the patient elected to undergo a posterior instrumentation and arthrodesis (T5 to L2). The second patient (Case 16) was noted to have postoperative curve progression to a main thoracic coronal deformity of 49° ("adding on" three levels below the prior fusion) after growing >20 cm during the postoperative period. At four years postoperatively, when she was fourteen years old, she underwent a posterior spinal instrumentation and arthrodesis from T3 to L3, with correction of the thoracic curve to 15°. The third patient (Case 23) was noted to have a broken rod at T8-T9 with progression of the thoracic deformity to 46° and of the thoracolumbar deformity to 14°. A revision posterior spinal instrumentation and arthrodesis from T4 to L2 was performed two years and six months after the primary procedure.
Pulmonary Function
The preoperative pulmonary function testing revealed an average forced vital capacity of 2.9 ± 0.6 L, an average forced expiratory volume in one second of 2.5 ± 0.5 L, and an average total lung capacity of 3.6 ± 0.5 L. The corresponding preoperative percent of the predicted values compared with norms were 89% ± 20% for forced vital capacity, 83% ± 17% for forced expiratory volume in one second, and 90% ± 12% for total lung capacity. The preoperative, two-year, and five-year data on pulmonary function testing are shown in Figure 3. A significant increase in the absolute and the percent of predicted total lung capacity was observed between the preoperative and two-year visits (p < 0.03). Between the two-year and five-year visits, however, there were no significant differences in pulmonary function (p > 0.05). The average total lung capacity as a percent of the predicted value was 95% ± 14% at two years and 91% ± 10% at five years. A power analysis of forced vital capacity and total lung capacity revealed ß values of 0.14 and 0.13, respectively, with a small effect size of 0.05. According to this analysis, a sample size of at least seventy patients would be required to demonstrate >80% power and definitively conclude that pulmonary function did not change between the two follow-up visits.
SRS-24 Data
SRS outcomes instrument scores are graphically shown in Figure 4. The average pain scores improved significantly from 3.9 ± 0.6 preoperatively to 4.4 ± 0.5 at the time of the two-year follow-up (p = 0.007). No significant differences were observed in any of the preoperative or postoperative domains from the two-year to the five-year visit (p > 0.10).
Lost to Follow-up Analysis
Fourteen of the sixteen patients lost to five-year follow-up were seen at the two-year visit. Two (5%) of forty-one patients were lost to follow-up less than six months postoperatively. A secondary analysis compared selected two-year radiographic, pulmonary function and SRS-24 data of the patients lost to follow-up and those of the study cohort (see Appendix). No significant differences between these groups were found for any of the data compared. Of the patients lost to follow-up, one (Case 23) required a surgical revision with posterior instrumentation for a broken rod and asymptomatic pseudarthrosis at T8-T9 and another two patients (Cases 25 and 27) were noted to have broken rods but they did not require a revision.
An entirely endoscopic spinal deformity correction, instrumentation, and fusion for primary thoracic scoliosis was first reported by Picetti et al.16,17. An average coronal deformity correction of 50.2% was achieved with a minimum two-year follow-up17. Five-year follow-up data with use of current thoracoscopic anterior instrumentation techniques have not been published previously as far as we know. In the present study, thoracoscopic instrumentation resulted in a similar average thoracic correction of 55.8% (range, 29.8% to 73.8%) at two years. Additionally, the thoracic deformity correction was found to be maintained at 51.5% (range, 28.8% to 77.8%) at an average follow-up of 5.3 years.
These current observations on the maintenance of deformity correction are similar to mid-term data reported previously for traditional posterior and open anterior instrumentation techniques. The correction of a main thoracic deformity with posterior instrumentation systems averaged between 46% and 55% at two years and was maintained between 41% and 44% at eight to thirteen years of follow-up5. Following open anterior thoracic instrumentation, Potter et al. observed that the average correction of the main thoracic curve was 52% at twenty-four to eighty months postoperatively4. Similarly, at an average follow-up of five years after open anterior treatment of thoracolumbar-lumbar scoliosis, Burton et al. reported that the average curve correction was 63% compared with the preoperative value31.
In the surgical treatment of scoliosis, it is important to consider, in addition to deformity correction in the coronal plane, the postoperative sagittal alignment. Traditionally, thoracic scoliosis has been described as a hypokyphotic deformity, resulting in lateral deviation and rotation of the spinal column32. Investigators have suggested that thoracic hypokyphosis occurs secondary to unbalanced anterior spinal overgrowth33. Anterior scoliosis procedures have been shown to be effective at creating thoracic kyphosis by shortening the anterior portion of the spine through discectomy and correcting the anterior overgrowth by achieving a fusion9,11,21. In the present cohort of patients, the average thoracic kyphosis increased significantly from 18.7° ± 9.3° preoperatively to 28.8° ± 10.5° at two years. In conjunction, the average lumbar lordosis also increased (from 58.2° ± 8.6° to 65.0° ± 10.0°); however, this change was not significant. This correlation between thoracic kyphosis and lumbar lordosis has been described previously and is thought to occur as an attempt to maintain sagittal balance (the head centered over the pelvis)34-36. Sagittal alignment (thoracic kyphosis, thoracolumbar junction, and lumbar lordosis) was not found to change significantly between the two-year and five-year follow-up visits (p > 0.05).
Most radiographic measures in this cohort remained unchanged between the two and five-year visits. Only lateral translation of C7 from the central sacral vertical line was found to be significantly different between the two latest follow-up time points. This change is likely of no clinical significance and is secondary to a change in the positioning used for lateral radiographs between the two follow-up periods. Recent studies have attempted to determine the optimal positioning for making long-cassette (36-in [91-cm]), standing lateral radiographs in patients with adolescent scoliosis. The fist-on-clavicle position has been shown to cause a less negative shift in the sagittal vertical axis and less compensatory posterior rotation of the pelvis and to produce substantially better overall visualization of critical vertebral landmarks37,38. During the five-year follow-up visit, most radiographs of the patients were made with use of the fist-on-clavicle position, whereas in the previous radiographs, the patients had been positioned with the arms flexed forward.
Improvements in the clinical measures of spinal deformity are another important component of surgical success after scoliosis treatment. It has been shown that the appearance associated with the deformity (rib hump, trunk shift, and shoulder imbalance) is one of the most important reasons why patients seek surgical treatment39,40. In the present cohort, thoracoscopic instrumentation significantly improved average coronal decompensation (from 1.1 ± 0.4 cm preoperatively to 0.4 ± 0.6 cm at two years) and the average thoracic rib hump (from 15.2° ± 4.1° preoperatively to 8.2° ± 4.1° at two years). Lumbar prominence, trunk shift, and shoulder height imbalance also improved postoperatively; however, these changes were not significant most likely because of a limited sample size and relatively small initial deformity. The improvements in all five of these measures were maintained between the two-year and five-year visits, with no clinically significant differences noted between the two follow-up time points.
Traditionally, the concerns associated with thoracoscopic spinal instrumentation procedures have been related to the potential for vascular or neurologic injury, the initial steep learning curve, implant failure, and pseudarthrosis16,18,20. In analyses of the placement of bicortical vertebral body screws following endoscopic instrumentation, recent studies have concluded that anterior screws are safe with respect to the spinal canal; however, the screws can be in close proximity to the major vessels and, occasionally, have even created a contour deformity in the aorta41,42. Although major injuries to the thoracic aorta following the placement of vertebral body screws in adolescent patients have not been reported previously, as far as we know, several case reports of such injuries in adults following anterior spinal instrumentation with older implant systems have been published43-45. In the present series of patients, clinically relevant vascular injuries or neurologic compromise were not observed. Implant failure, however, was observed in three patients and required a surgical revision in one patient.
In terms of intervertebral fusion ratings, a Grade-3 or 4 fusion (an incomplete union) was observed in four (16%) of twenty-five patients at five years. This incidence is within the range for pseudarthroses reported in previous studies of anterior spinal approaches (0% to 20%)16,18,26. Every patient in this case series received autogenous iliac crest bone graft, which is thought to be the ideal graft material and was shown in a goat model to lead to a more consistent fusion compared with allograft demineralized bone46. In addition, most fusions were supplemented with structural allograft around the thoracolumbar junction to maintain sagittal alignment. Of note, fusion ratings were found to improve somewhat between the two and five-year visits. At two years, nine (36%) of twenty-five patients had a Grade-3 or 4 fusion at one motion segment. All three Grade-4 fusions remained unchanged at five years; however, five of the six Grade-3 fusions had improved by at least one grade. In addition, three Grade-2 fusions had improved to Grade 1 between the two follow-up visits. These improvements in the fusion ratings are expected as endochondral ossification continues to mature over time, and the cartilaginous matrix is replaced by bone. It is interesting, however, that improvements in fusion were not noted at the three motion segments determined to be Grade 4. These established pseudarthroses are likely secondary to an inherent motion or instability that exists at these levels. To date, only one (Case 7) of these three patients required a surgical revision.
Both complications of rod breakage and pseudarthrosis are likely secondary to inadequate construct stiffness. All three patients who had an implant failure and three of the four patients with incomplete fusions at five years were treated with the smaller-diameter (4.0-mm) stainless-steel rods. In the fourteen patients lost to follow-up after two years, all six patients with a Grade-3 or 4 fusion were also treated with the smaller-diameter instrumentation. Several strategies have been developed to increase the stiffness of the anterior spinal instrumentation construct. Kaneda et al.34 introduced a dual-rod, dual-screw system; however, it can be quite large for use in the thoracic spine. Sweet et al.12 advocated the use of a structural interbody cage to act as a load-sharing device and to increase construct stability in flexion. An alternate approach would be to modify the material properties of the implants used. During this case series, the instrumentation was changed from 4.0-mm stainless steel to 4.75-mm titanium alloy. A recent biomechanical study evaluated the construct stiffness and fatigue properties of these two instrumentation systems47. Both implants had similar construct stiffness; however, the titanium alloy was found to have greater fatigue resistance because of its higher yield stress. A comparison of clinical outcomes with use of these two different instrumentation systems (4.0-mm stainless steel compared with 4.75-mm titanium alloy) is currently underway.
Surgical revisions were required in three patients in this case series: one patient with a symptomatic pseudarthrosis, one patient with a progressive deformity ("adding-on"), and one patient with a broken rod. It is important to recognize that this series represents the initial forty-one patients to undergo anterior thoracoscopic instrumentation at our institution. Therefore, it includes all patients involved in the learning curve for this technique. However, our rate of surgical revision is similar to previously reported rates following traditional posterior and open anterior instrumentation48-52. In 2004, Asher et al. reviewed the cases of 179 patients with mid-term follow-up after traditional posterior instrumentation48. In that series, fourteen patients (7.8%) required a surgical revision for fifteen postoperative complications, including late pain at the operative site in seven patients, pseudarthroses in four patients, delayed deep wound infections in two patients, and implant prominence and adjacent segment degeneration in one patient each. In 2007, Kuklo et al. reported surgical revision rates in patients with a minimum two-year follow-up after the use of four different posterior spinal instrumentation techniques (hooks only in 389 patients, a hybrid system in 423, pedicle screws in 295, and combined anterior-posterior constructs in 321), ranging from 2.4% to 7.5%52. The revision rate with more recent pedicle screw instrumentation was lower (2.4%), although complications associated with malpositioning of pedicle screws were not discussed.
Pulmonary function data after thoracoplasty in patients with adolescent idiopathic scoliosis have demonstrated an initial substantial decline with either a return to preoperative values53 or a persistent decline54 two years after surgery. Recently, five-year follow-up pulmonary function data were reported by Kim et al.55. In an evaluation of 118 patients, they observed no change in the absolute pulmonary function values and a modest decline in the percent of the predicted values at five years following surgical treatment for adolescent idiopathic scoliosis that involved any type of chest cage disruption (thoracotomy). Faro et al. reported that the thoracoscopic approach caused a smaller decline in pulmonary function up to one year after surgery compared with open anterior approaches56. In our cohort, there was no significant difference between the absolute or percent of the predicted preoperative and two-year values for forced vital capacity and forced expiratory volume in one second, and, in fact, the total lung capacity values increased significantly. This finding is similar to that reported by Izatt et al.57, who concluded that endoscopic anterior scoliosis surgery has no lasting negative effect on pulmonary function at two years of follow-up. At the five-year follow-up, our current data did not show any significant absolute differences in forced vital capacity, forced expiratory value in one second, and total lung capacity; however, considering the small ß values, further investigation with a larger sample size is warranted.
Finally, the SRS-24 scores observed in our cohort were comparable with those reported previously58,59. Crawford et al. reported an improvement from a mean preoperative pain score of 3.76 ± 0.69 to a mean postoperative score of 4.39 ± 0.48 at twenty-four months58. A similar increase in the mean pain value was noted in our cohort from 3.9 ± 0.6 preoperatively to 4.4 ± 0.5 at two years postoperatively. Scores in the other three preoperative domains also improved between the preoperative and two-year visits; however, these changes were not significant. Between the two and five-year visits, no significant difference was noted in the average total score (4.2 ± 0.4 and 4.1 ± 0.7, respectively) or in any of the seven domains.
The high proportion of patients lost to follow-up (39%) is a weakness of this study; however, we presented data from the minimum two-year follow-up evaluation, including all complications (pseudarthrosis and rod breakage) that we are aware of, for fourteen of the sixteen patients lost to follow-up. In addition, the secondary analysis comparing two-year data for the cohort lost to follow-up after two years (fourteen patients) and the cohort with five-year follow-up (twenty-five patients) demonstrated no significant differences between the groups at that time point. While this analysis does not give us any information on what may have transpired between the two and five-year visits, it is reassuring to know that these groups were similar at the two-year visit.
In conclusion, thoracoscopic anterior instrumentation for main thoracic idiopathic scoliosis results in five-year outcomes comparable with those reported previously for open anterior and posterior techniques. Radiographic, pulmonary function, and clinical measures remain stable between the two and five-year follow-up time points. The relatively small 4.0-mm stainless-steel rod is at risk for implant failure, especially if a pseudarthrosis develops; however, the surgical revision rate is similar to alternate approaches. Although it is a technically challenging method for the correction of thoracic scoliosis, requiring careful patient selection, meticulous discectomy, and thorough bone-grafting, thoracoscopic anterior instrumentation remains a useful surgical approach.
Tables showing descriptive data on all patients, intervertebral fusion assessment at all levels at two and five years, and a comparison of patients in the study cohort and those lost to follow-up at five years are available with the electronic versions of this article, on our web site at (go to the article citation and click on "Supplementary Material") and on our quarterly CD/DVD (call our subscription department, at 781-449-9780, to order the CD or DVD). 
Dwyer AF, Newton NC, Sherwood AA. An anterior approach to scoliosis. A preliminary report. Clin Orthop Relat Res.1969;62:192-202.62192
1969
[PubMed]
Arlet V, Reddi V. Adolescent idiopathic scoliosis. Neurosurg Clin N Am.2007;18:255-9.18255
2007
[CrossRef]
Helenius I, Remes V, Yrjönen T, Ylikoski M, Schlenzka D, Helenius M, Poussa M. Harrington and Cotrel-Dubousset instrumentation in adolescent idiopathic scoliosis. Long-term functional and radiographic outcomes. J Bone Joint Surg Am.2003;85:2303-9.852303
2003
Potter BK, Kuklo TR, Lenke LG. Radiographic outcomes of anterior spinal fusion versus posterior spinal fusion with thoracic pedicle screws for treatment of Lenke Type I adolescent idiopathic scoliosis curves. Spine.2005;30:1859-66.301859
2005
[CrossRef]
Remes V, Helenius I, Schlenzka D, Yrjönen T, Ylikoski M, Poussa M. Cotrel-Dubousset (CD) or Universal Spine System (USS) instrumentation in adolescent idiopathic scoliosis (AIS): comparison of midterm clinical, functional, and radiologic outcomes. Spine.2004;29:2024-30.292024
2004
[CrossRef]
Bridwell KH, Hanson DS, Rhee JM, Lenke LG, Baldus C, Blanke K. Correction of thoracic adolescent idiopathic scoliosis with segmental hooks, rods, and Wisconsin wires posteriorly: it's bad and obsolete, correct? Spine.2002;27:2059-66.272059
2002
[CrossRef]
Kim YJ, Lenke LG, Cho SK, Bridwell KH, Sides B, Blanke K. Comparative analysis of pedicle screw versus hook instrumentation in posterior spinal fusion of adolescent idiopathic scoliosis. Spine.2004;29:2040-8.292040
2004
[CrossRef]
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;23:2063-73.232063
1998
[CrossRef]
Betz RR, Harms J, Clements DH 3rd, Lenke LG, Lowe TG, Shufflebarger HL, Jeszenszky D, Beele B. Comparison of anterior and posterior instrumentation for correction of adolescent thoracic idiopathic scoliosis. Spine.1999;24:225-39.24225
1999
[CrossRef]
Lowe TG, Alongi PR, Smith DA, O'Brien MF, Mitchell SL, Pinteric RJ. Anterior single rod instrumentation for thoracolumbar adolescent idiopathic scoliosis with and without the use of structural interbody support. Spine.2003;28:2232-42.282232
2003
[CrossRef]
Lowe TG, Betz R, Lenke L, Clements D, Harms J, Newton P, Haher T, Merola A, Wenger D. Anterior single-rod instrumentation of the thoracic and lumbar spine: saving levels. Spine.2003;28:S208-16.28S208
2003
[CrossRef]
Sweet FA, Lenke LG, Bridwell KH, Blanke KM, Whorton J. Prospective radiographic and clinical outcomes and complications of single solid rod instrumented anterior spinal fusion in adolescent idiopathic scoliosis. Spine.2001;26:1956-65.261956
2001
[CrossRef]
Landreneau RJ, Hazelrigg SR, Mack MJ, Dowling RD, Burke D, Gavlick J, Perrino MK, Ritter PS, Bowers CM, DeFino J. Postoperative pain-related morbidity: video-assisted thoracic surgery versus thoracotomy. Ann Thorac Surg.1993;56:1285-9.561285
1993
[CrossRef]
Landreneau RJ, Wiechmann RJ, Hazelrigg SR, Mack MJ, Keenan RJ, Ferson PF. Effect of minimally invasive thoracic surgical approaches on acute and chronic postoperative pain. Chest Surg Clin N Am.1998;8:891-906.8891
1998
Newton PO, Marks M, Faro F, Betz R, Clements D, Haher T, Lenke L, Lowe T, Merola A, Wenger D. Use of video-assisted thoracoscopic surgery to reduce perioperative morbidity in scoliosis surgery. Spine.2003;28:S249-54.28S249
2003
[CrossRef]
Picetti GD 3rd, Ertl JP, Bueff HU. Endoscopic instrumentation, correction, and fusion of idiopathic scoliosis. Spine J.2001;1:190-7.1190
2001
[CrossRef]
Picetti GD 3rd, Pang D, Bueff HU. Thoracoscopic techniques for the treatment of scoliosis: early results in procedure development. Neurosurgery.2002;51:978-84.51978
2002
Sucato DJ. Thoracoscopic anterior instrumentation and fusion for idiopathic scoliosis. J Am Acad Orthop Surg.2003;11:221-7.11221
2003
Lonner BS, Scharf C, Antonacci D, Goldstein Y, Panagopoulos G. The learning curve associated with thoracoscopic spinal instrumentation. Spine.2005;30:2835-40.302835
2005
[CrossRef]
Newton PO, Shea KG, Granlund KF. Defining the pediatric spinal thoracoscopy learning curve: sixty-five consecutive cases. Spine.2000;25:1028-35.251028
2000
[CrossRef]
Lonner BS, Kondrachov D, Siddiqi F, Hayes V, Scharf C. Thoracoscopic spinal fusion compared with posterior spinal fusion for the treatment of thoracic adolescent idiopathic scoliosis. J Bone Joint Surg Am.2006;88:1022-34.881022
2006
[CrossRef]
Newton PO, Parent S, Marks M, Pawelek J. Prospective evaluation of 50 consecutive scoliosis patients surgically treated with thoracoscopic anterior instrumentation. Spine.2005;30(17 Suppl):S100-9.30S100
2005
[CrossRef]
Graham EJ, Lenke LG, Lowe TG, Betz RR, Bridwell KH, Kong Y, Blanke K. Prospective pulmonary function evaluation following open thoracotomy for anterior spinal fusion in adolescent idiopathic scoliosis. Spine.2000;25:2319-25.252319
2000
[CrossRef]
Grewal H, Betz RR, D'Andrea LP, Clements DH, Porter ST. A prospective comparison of thoracoscopic vs open anterior instrumentation and spinal fusion for idiopathic thoracic scoliosis in children. J Pediatr Surg.2005;40:153-7.40153
2005
[CrossRef]
Muschik MT, Kimmich H, Demmel T. Comparison of anterior and posterior double-rod instrumentation for thoracic idiopathic scoliosis: results of 141 patients. Eur Spine J.2006;15:1128-38.151128
2006
[CrossRef]
Wong HK, Hee HT, Yu Z, Wong D. Results of thoracoscopic instrumented fusion versus conventional posterior instrumented fusion in adolescent idiopathic scoliosis undergoing selective thoracic fusion. Spine.2004;29:2031-9.292031
2004
[CrossRef]
Norton RP, Patel D, Kurd MF, Picetti GD, Vaccaro AR. The use of thoracoscopy in the management of adolescent idiopathic scoliosis. Spine.2007;32:2777-85.322777
2007
[CrossRef]
Bridwell KH, Lenke LG, McEnery KW, Baldus C, Blanke K. Anterior fresh frozen structural allografts in the thoracic and lumbar spine. Do they work if combined with posterior fusion and instrumentation in adult patients with kyphosis or anterior column defects? Spine.1995;20:1410-8.201410
1995
Haher TR, Gorup JM, Shin TM, Homel P, Merola AA, Grogan DP, Pugh L, Lowe TG, Murray M. Results of the Scoliosis Research Society instrument for evaluation of surgical outcome in adolescent idiopathic scoliosis. A multicenter study of 244 patients. Spine.1999;24:1435-40.241435
1999
[CrossRef]
Lenke LG, Betz RR, Harms J, Bridwell KH, Clements DH, Lowe TG, Blanke K. Adolescent idiopathic scoliosis: a new classification to determine extent of spinal arthrodesis. J Bone Joint Surg Am.2001;83:1169-81.831169
2001
Burton DC, Asher MA, Lai SM. Patient-based outcomes analysis of patients with single torsion thoracolumbar-lumbar scoliosis treated with anterior or posterior instrumentation: an average 5- to 9-year follow-up study. Spine.2002;27:2363-7.272363
2002
[CrossRef]
Dickson RA. The aetiology of spinal deformities. Lancet.1988;1:1151-5.11151
1988
Guo X, Chau WW, Chan YL, Cheng JC, Burwell RG, Dangerfield PH. Relative anterior spinal overgrowth in adolescent idiopathic scoliosis—result of disproportionate endochondral-membranous bone growth? Summary of an electronic focus group debate of the IBSE. Eur Spine J.2005;14:862-73.14862
2005
[CrossRef]
Kaneda K, Shono Y, Satoh S, Abumi K. New anterior instrumentation for the management of thoracolumbar and lumbar scoliosis. Application of the Kaneda two-rod system. Spine.1996;21:1250-62.211250
1996
[CrossRef]
Mac-Thiong JM, Labelle H, Charlebois M, Huot MP, de Guise JA. Sagittal plane analysis of the spine and pelvis in adolescent idiopathic scoliosis according to the coronal curve type. Spine.2003;28:1404-9.281404
2003
Upasani VV, Tis J, Bastrom T, Pawelek J, Marks M, Lonner B, Crawford A, Newton PO. Analysis of sagittal alignment in thoracic and thoracolumbar curves in adolescent idiopathic scoliosis: how do these two curve types differ? Spine.2007;32:1355-9.321355
2007
[CrossRef]
Faro FD, Marks MC, Pawelek J, Newton PO. Evaluation of a functional position for lateral radiograph acquisition in adolescent idiopathic scoliosis. Spine.2004;29:2284-9.292284
2004
[CrossRef]
Horton WC, Brown CW, Bridwell KH, Glassman SD, Suk SI, Cha CW. Is there an optimal patient stance for obtaining a lateral 36? radiograph? A critical comparison of three techniques. Spine.2005;30:427-33.30427
2005
[CrossRef]
Sanders JO, Polly DW Jr, Cats-Baril W, Jones J, Lenke LG, O'Brien MF, Stephens Richards B, Sucato DJ; AIS Section of the Spinal Deformity Study Group. Analysis of patient and parent assessment of deformity in idiopathic scoliosis using the Walter Reed Visual Assessment Scale. Spine.2003;28:2158-63.282158
2003
[CrossRef]
Smith PL, Donaldson S, Hedden D, Alman B, Howard A, Stephens D, Wright JG. Parents' and patients' perceptions of postoperative appearance in adolescent idiopathic scoliosis. Spine.2006;31:2367-74.312367
2006
[CrossRef]
Bullmann V, Fallenberg EM, Meier N, Fischbach R, Schulte TL, Heindel WL, Liljenqvist UR. Anterior dual rod instrumentation in idiopathic thoracic scoliosis: a computed tomography analysis of screw placement relative to the aorta and the spinal canal. Spine.2005;30:2078-83.302078
2005
[CrossRef]
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.2004;29:554-9.29554
2004
[CrossRef]
Ohnishi T, Neo M, Matsushita M, Komeda M, Koyama T, Nakamura T. Delayed aortic rupture caused by an implanted anterior spinal device. Case report. J Neurosurg.2001;95(2 Suppl):253-6.95253
2001
Jendrisak MD. Spontaneous abdominal aortic rupture from erosion by a lumbar spine fixation device: a case report. Surgery.1986;99:631-3.99631
1986
Woolsey RM. Aortic laceration after anterior spinal fusion. Surg Neurol.1986;25:267-8.25267
1986
[CrossRef]
Newton PO, Lee SS, Mahar AT, Farnsworth CL, Weinstein CH. Thoracoscopic multilevel anterior instrumented fusion in a goat model. Spine.2003;28:1614-20.281614
2003
Wedemeyer M, Parent S, Mahar A, Odell T, Swimmer T, Newton P. Titanium versus stainless steel for anterior spinal fusions: an analysis of rod stress as a predictor of rod breakage during physiologic loading in a bovine model. Spine.2007;32:42-8.3242
2007
[CrossRef]
Asher M, Lai SM, Burton D, Manna B, Cooper A. Safety and efficacy of Isola instrumentation and arthrodesis for adolescent idiopathic scoliosis: two- to 12-year follow-up. Spine.2004;29:2013-23.292013
2004
[CrossRef]
Carreon LY, Puno RM, Lenke LG, Richards BS, Sucato DJ, Emans JB, Erickson MA. Non-neurologic complications following surgery for adolescent idiopathic scoliosis. J Bone Joint Surg Am.2007;89:2427-32.892427
2007
[CrossRef]
Coe JD, Arlet V, Donaldson W, Berven S, Hanson DS, Mudiyam R, Perra JH, Shaffrey CI. Complications in spinal fusion for adolescent idiopathic scoliosis in the new millennium. A report of the Scoliosis Research Society Morbidity and Mortality Committee. Spine.2006;31:345-9.31345
2006
[CrossRef]
Ho C, Skaggs DL, Weiss JM, Tolo VT. Management of infection after instrumented posterior spine fusion in pediatric scoliosis. Spine.2007;32:2739-44.322739
2007
[CrossRef]
Kuklo TR, Potter BK, Lenke LG, Polly DW Jr, Sides B, Bridwell KH. Surgical revision rates of hooks versus hybrid versus screws versus combined anteroposterior spinal fusion for adolescent idiopathic scoliosis. Spine.2007;32:2258-64.322258
2007
[CrossRef]
Vedantam R, Lenke LG, Bridwell KH, Haas J, Linville DA. A prospective evaluation of pulmonary function in patients with adolescent idiopathic scoliosis relative to the surgical approach used for spinal arthrodesis. Spine.2000;25:82-90.2582
2000
[CrossRef]
Lenke LG, Bridwell KH, Blanke K, Baldus C. Analysis of pulmonary function and chest cage dimension changes after thoracoplasty in idiopathic scoliosis. Spine.1995;20:1343-50.201343
1995
Kim YJ, Lenke LG, Bridwell KH, Kim KL, Steger-May K. Pulmonary function in adolescent idiopathic scoliosis relative to the surgical procedure. J Bone Joint Surg Am.2005;87:1534-41.871534
2005
[CrossRef]
Faro FD, Marks MC, Newton PO, Blanke K, Lenke LG. Perioperative changes in pulmonary function after anterior scoliosis instrumentation: thoracoscopic versus open approaches. Spine.2005;30:1058-63.301058
2005
[CrossRef]
Izatt MT, Harvey JR, Adam CJ, Fender D, Labrom RD, Askin GN. Recovery of pulmonary function following endoscopic anterior scoliosis correction: evaluation at 3, 6, 12, and 24 months after surgery. Spine.2006;31:2469-77.312469
2006
[CrossRef]
Crawford JR, Izatt MT, Adam CJ, Labrom RD, Askin GN. A prospective assessment of SRS-24 scores after endoscopic anterior instrumentation for scoliosis. Spine.2006;31:E817-22.31E817
2006
[CrossRef]
Merola AA, Haher TR, Brkaric M, Panagopoulos G, Mathur S, Kohani O, Lowe TG, Lenke LG, Wenger DR, Newton PO, Clements DH 3rd, Betz RR. A multicenter study of the outcomes of the surgical treatment of adolescent idiopathic scoliosis using the Scoliosis Research Society (SRS) outcome instrument. Spine.2002;27:2046-51.272046
2002
[CrossRef]