CHAPTER 25 Thoracoscopic Approach for Spinal Conditions
Thoracoscopic Anterior Release and Fusion
Mack and colleagues1 first introduced endoscopic spine techniques in 1993. Since that time, thoracoscopy, also known as video-assisted thoracic surgery (VATS), has evolved to become a valuable tool in the treatment of spinal deformity and other spinal conditions. The goals of thoracoscopic anterior spinal surgery are essentially the same as the goals of open surgery, but they are accomplished with less invasive techniques. Specifically, the goal of VATS in the surgical management of idiopathic scoliosis is to perform a safe, reproducible, and effective procedure that results in improvement in spinal alignment and balance in all planes and axial derotation comparable to, or better than, that obtained with an open procedure.2 In addition to idiopathic scoliosis, thoracoscopy has been used for anterior releases in kyphosis, hemiepiphysiodeses and hemivertebrectomies, excision of spinal tumors, and treatment of spinal trauma.
Indications
Scoliosis
Thoracic scoliosis has various etiologies (idiopathic, neuromuscular, syndrome related) that are frequently not diagnosed and treated until the curve is relatively large and stiff. Anterior surgery has been most frequently used as a means to achieve a complete discectomy and release of the anterior spine; this results in greater curve flexibility and prevents the “crankshaft” phenomenon in young patients. Although no strict guidelines on the magnitude and flexibility of the spinal curvature that requires release have been established, generally curves with a Cobb angle greater than 70 to 75 degrees and a bend correction less than 50% are considered appropriate for release. When sufficient segmental mobility of the released vertebra has been achieved, posterior instrumentation is placed to correct the deformity (Fig. 25–1).
Lenke3 reported on a combined anterior VATS release and fusion followed by posterior instrumentation in the treatment of adolescent idiopathic scoliosis that had an average preoperative curve of 82 degrees (range 41 to 125 degrees) with postoperative correction of 70% to 28 degrees (range 5 to 60 degrees). Similarly, Newton and colleagues4 reported a series of 112 pediatric spinal deformity cases with an average preoperative curve of 80 degrees that received an anterior release combined with posterior instrumentation and found a 67% correction in idiopathic scoliosis and a 52% correction in neuromuscular scoliosis. More recent studies have called into question the utility of the anterior release, however, in the age of modern segmental pedicle screw instrumentation.5,6 Suk and colleagues6 found an average correction of 66% when posterior pedicle screws alone were used in preoperative thoracic curves of 80 degrees with a flexibility of 45%. Although modern pedicle screw constructs offer a similar correction, an anterior release may still be indicated to optimize coronal and axial plane correction, improve sagittal alignment by increasing thoracic kyphosis, and prevent crankshaft growth.
Children and young adolescents (Risser 0, open triradiate cartilage) with progressive scoliosis are known to be at risk for crankshaft deformity when treated with a posterior fusion alone.7 These results were reported for children using hook and wire fixation, however, and not modern pedicle screw instrumentation. Although there has been concern regarding adequate pedicle size in children to accommodate pedicle screws, Catan and colleagues8 performed a magnetic resonance imaging (MRI) analysis of thoracic pedicle morphology in preadolescent patients with idiopathic scoliosis and found that the anatomic measurements were compatible with pedicle screw instrumentation. Sarlak and colleagues9 reported more recently a series of seven children (average age 7.4 years) with scoliosis with an average preoperative thoracic curve of 56 degrees who were treated with posterior segmental pedicle screw instrumentation and 5 years of follow-up. These authors found a 57% correction rate with no evidence of crankshaft phenomenon in four patients but found a slight increase in Cobb angle and a significant increase in angle of trunk rotation (ATR) suggesting crankshaft phenomenon in two patients.9 Given the lack of clear evidence on the appropriate treatment of these cases, an anterior fusion may be a viable option to limit anterior growth and prevent this late increasing deformity.10 Thoracoscopic disc excision and fusion provides a minimally invasive option and minimizes the risk of pulmonary complications in these young patients with progressive deformity.11–13
Patients with spinal deformity associated with Marfan syndrome, neurofibromatosis 1, and prior spinal irradiation may have an increased risk of pseudarthrosis after an isolated posterior scoliosis correction. In cases such as these, an anterior fusion procedure may improve the odds of successful arthrodesis, especially when autogenous bone graft is used.4
Kyphosis
Controversy exists regarding the need for anterior procedures in cases of thoracic kyphosis.14–16 In previous studies, Papagelopoulos and colleagues17 and Sturm and colleagues18 found that posterior correction with hook or hybrid fixation alone did not provide adequate strength to maintain correction in patients with progressive kyphosis. Because of the lack of satisfactory results with posterior hook or hybrid instrumentation, combined anterior and posterior approaches to treatment of kyphotic deformity have been investigated. In a retrospective analysis of 32 patients with Scheuermann kyphosis treated with a combined anterior release followed by posterior segmental hybrid instrumentation, Lowe and Kasten19 found that a combined approach resulted in a 51% correction of the deformity with no major postoperative complications. Herrera-Soto and colleagues20 specifically investigated the use of thoracoscopy in these patients and found similar benefits to scoliosis patients, including decreased blood loss and less morbidity compared with an open thoracotomy.
In a series of 39 patients with Scheuermann kyphosis, Lee and colleagues15 compared posterior-only thoracic segmental pedicle screw constructs with combined anterior and posterior constructs. These authors found a similar correction rate between both techniques, with an increase in complications in the combined anterior and posterior group. Although posterior segmental pedicle instrumentation seems to offer a similar correction rate to a combined anterior and posterior approach, no recommendations can be made based on the current evidence.14 The optimal treatment approach to progressive kyphosis must be based on the surgeon’s judgment and experience for each individual patient.
Congenital Deformity
Operative management of congenital scoliosis depends on the type of vertebral anomaly, its location, the age of the patient, and the potential continued growth of the child. Although various techniques are available, thoracoscopy also has been applied to the treatment of congenital scoliosis.4,12,21 Several methods of surgical treatment have been employed in these patients with limited results. In a study with 12 years of follow-up in patients with congenital scoliosis, Kesling and colleagues22 found that 15% of 54 patients who received a posterior arthrodesis developed crankshaft phenomenon.
Given the lack of satisfactory results with previous posterior-only techniques, anterior surgery may be considered in these patients The anterior portion of circumferential arthrodesis and growth-modifying hemiepiphysiodesis are theoretically possible via the endoscopic approach. First reported by Roaf,23 hemiepiphysiodesis has been described more recently by Samdani and Storm,24 who reported that a convex hemiepiphysiodesis is best performed on children 5 years of age with a short curve less than 40 degrees and scoliosis involving five or fewer segments using a combined anterior and posterior approach. In this combined anterior and posterior hemiepiphysiodesis procedure, the convex halves of the discs are removed anteriorly, followed by a posterior arthrodesis and casting.25 Although a lower thoracic level hemivertebra occasionally may be indicated for excision, doing so thoracoscopically is challenging. Many patients undergoing treatment for congenital scoliosis are younger than 5 years of age and require anterior fusion over very few levels of the spine and may not benefit from a thoracoscopic approach. The use of VATS in these patients must be decided on a case-by-case basis.
Contraindications
As suggested earlier, the small size of the patient is a relative contraindication to thoracoscopy. Lung deflation is more difficult in these cases because standard size double-lumen and bronchial blocking endotracheal tubes are too large. Another contraindication is the presence of a markedly reduced working distance between the chest wall and the spine. In severe cases of scoliosis, the reduced distance limits the field of vision (the endoscope is too close to the spine to obtain any perspective) and the maneuverability of the working instruments. A minimum distance of 2 to 3 cm of working space between the rib cage and the spinal column is required to provide adequate visualization. If the distance is less than 2 cm, thoracoscopy is not advised. Body weight of less than 30 kg is a relative contraindication because the relative benefits of thoracoscopy seem to be reduced in small patients.26
Visualization of the surgical field is mandatory in all surgical approaches, and it may be compromised in thoracoscopic surgery by incomplete deflation of the lung or pleural adhesions that prevent collapse away from the chest wall and spine. Pleural adhesions (Fig. 25–2) can be anticipated in patients with a history of prior ipsilateral thoracic surgery or significant pulmonary infection, both of which should discourage the surgeon from using the thoracoscopic approach.
Surgical Technique
Patient positioning has traditionally been in the lateral decubitus position (Fig. 25–3). Some studies have suggested that in select cases prone positioning may be possible, avoiding the need to reposition the patient for the posterior procedure or even allowing simultaneous anterior release and posterior instrumentation.27–29 Although the ability to convert to an open approach may be restricted, it has been shown that the prone position does not adversely affect postoperative pulmonary function.30 This approach necessitates a more posterior portal placement, however, and may limit the anterior extent of spinal exposure and disc excision.
The role of the anesthesiologist is crucial in the success and safety of thoracoscopic surgery.31 Spinal cord monitoring is advised using somatosensory and transcranial motor evoked potentials. Complete ipsilateral lung deflation is essential to prevent lung parenchymal injury from passing instruments and to allow visualization of the spine. Double-lumen endotracheal tubes are preferred in patients large enough (>45 kg) to accept these devices. In children (<45 kg), selective intubation of a single lung is often required as an alternative. A small balloon advanced into the main stem bronchus blocks ventilation to the lung on the operative side. In nearly all patients with normal preoperative pulmonary function, single-lung ventilation can be tolerated. The surgeon and anesthesiologist should be aware of the increased risk of developing postoperative mucous plugs as a result of single-lung ventilation.
After lung deflation, portals are established through the chest wall (Fig. 25–4). The orientation of these portals may vary depending on the pathology, although in most cases of deformity release and fusion they are best placed in a linear relationship along the anterior axillary line. Owing to the site of diaphragm insertion, the inferior portals require a slightly more posterior placement to maintain an intrathoracic position. Initial exposure of the spine often requires gentle retraction of the lung, at least until it becomes completely atelectatic (Fig. 25–5). The vasculature including the azygos vein and subclavian artery is identified before the introduction of surgical instruments to prevent inadvertent injury (Fig. 25–6). The vertebral levels are confirmed by identifying the first rib partially hidden beneath the subclavian artery and counting down distally (Fig. 25–7). Division of the pleura overlying the spine may be performed either longitudinally, over the length of the spine to be fused, or transversely, at each disc space.
FIGURE 25–4 Proper placement of portals is necessary for multilevel discectomies with the patient in lateral position.
FIGURE 25–5 Intraoperative endoscopic view of fan retractor (A) placed on lung (B) to aid in complete deflation.
Treatment of the segmental vessels (Fig. 25–8) may be similarly individualized with either division or preservation, depending on the needs of the case or preference of the surgeon. In most cases, the authors prefer a longitudinal pleural exposure with division of the segmental vessels using Harmonic laparoscopic coagulating sheers (Ethicon Endo-Surgery, Cincinnati, Ohio). Division of the segmental vessels allows greater anterior spinal exposure for more complete annular release. Blunt dissection of the pleura to the contralateral side of the spine is performed exposing approximately 270 degrees of the disc perimeter. After division of the pleura, any remaining areolar tissue is divided, and packing sponges are used to create a space between the anterior spine and the pleura.
FIGURE 25–8 After division of pleura (A), Harmonic laparoscopic coagulating sheers are applied to segmental vessel (B).
Possible levels that can be accessed thoracoscopically are T2-L1. Exposure of the T12-L1 disc and L1 vertebral body requires division of a small segment of the diaphragm insertion, which can be accomplished by extending the pleural incision distally into the diaphragm. The proximal thoracic spine in the right chest is often covered by the confluence of the segmental veins, which may appear daunting at the T3 and T4 levels. With slow cautious use of the ultrasonic devices, these vessels can be sealed and divided safely, however, exposing the upper thoracic spine. Disc excision techniques are similar to techniques used in open surgery. An annulotomy is performed with the electrocautery or Harmonic scalpel. A rongeur is an excellent tool for most of the disc excision. Specially designed endoscopic rongeurs are available in extended lengths with various angles (straight, up, right, left) to reach the depths of each disc space (Fig. 25–9).
When the discectomy is complete, either allogeneic cancellous or autogenous (rib or iliac crest) bone graft is placed into the disc space with an endoscopic tubular plunger (Fig. 25–10). The method and type of bone grafting also seems to be important to the success of arthrodesis. This may be crucial only in selected cases; however, all patients are at some risk for pseudarthrosis after posterior instrumentation and fusion procedures. In a study of 112 patients treated with an anterior release followed by posterior instrumentation, Newton and colleagues4 compared the grade of arthrodesis between patients who received autogenous versus allogeneic bone graft. These authors found the disc space was fused in 88% of the autograft group compared with 72% of the allograft group at 2-year follow-up.4 When autograft is not available, either allograft bone or demineralized bone matrix may be used because they have been shown to result in similar fusion rates.4,32 Although autogenous bone graft is optimal for patients at greatest risk for pseudarthrosis, the risk-to-benefit ratio must be analyzed on a case-by-case basis.
Outcomes
The thoracoscopic approach has several advantages compared with an open thoracotomy approach. These advantages, including pulmonary function, reduced recovery period, less pain, and improved cosmesis, are realized, however, only if the efficacy of the spinal procedure equals that of open surgery. Experimental animal and clinical studies suggest comparable efficacy in experienced hands. Several studies have been conducted to evaluate the learning curve necessary to be experienced in this technique.12,33,34 Newton and colleagues12 found that there was a slight decrease in operative time throughout the course of the first 65 patients treated at their institution and concluded that thoracoscopy had a steep but not prohibitive learning curve. In a more recent study by Son-Hing and colleagues,33
