Herniation of thoracic intervertebral discs occurs commonly but infrequently leads to symptomatic presentation. Historically, posterior treatment of herniated thoracic discs was limited to laminectomy and accompanied by a high incidence of paraplegia or other neurologic dysfunction. More modern posterolateral techniques can be performed safely and adequately, and often have advantages over transthoracic approaches. This chapter discusses the history of posterior treatment of thoracic herniated discs, the clinical presentation and evaluation, methods of posterior treatment, and treatment outcomes.
The prevalence rate of thoracic herniated discs is 11% to 37%.
Only 0.15% to 4% of all disc surgery involves thoracic discs.
Thoracic herniated discs can be classified according to the anatomic location of the herniated fragment or according to the clinical presentation of symptoms.
Posterior decompression is indicated for symptomatic paracentral discs or central, uncalcified discs and can be performed through a costotransversectomy, transpedicular, or transfacet, pedicle-sparing approach. The need for concomitant posterior decompression is a relative indication for a posterior approach for disc resection.
Posterior surgery is contraindicated for central discs that are either large or calcified and often adherent to the dura. A transthoracic surgical approach should be performed.
Laminectomy is not recommended for decompression of herniated thoracic discs because of the high risk for iatrogenic neurologic injury.
Thoracic disc herniation occurs most commonly between T8 and L1, with the highest incidence at T11-12.
The decision to operate should always be based on clinical symptoms. Many thoracic discs are asymptomatic.
The pedicle is the key anatomic landmark to assure proper orientation.
The choice of posterolateral surgical approach to herniated thoracic discs should be based on both the surgeon’s experience and the characteristics of the disc fragment. Disc location and degree of calcification are important considerations.
The counting technique used to determine the level of the herniated disc in the operating room must be the same as was used in the diagnostic study.
Transthoracic discectomy likely provides the least risk for iatrogenic neurologic injury.
Retraction of the spinal cord for exposure of herniated disc fragments should be avoided at all costs to limit the risk for iatrogenic neurologic injury.
Thoracic intervertebral disc herniation occurs commonly but infrequently leads to clinical symptoms. Imaging studies and autopsy reviews have indicated an 11% to 37% prevalence rate of thoracic intervertebral disc herniation in healthy, asymptomatic volunteers. Thoracic disc herniation, however, comprises only 0.15% to 4% of all disc surgeries. This disparity suggests that the sophisticated imaging modalities available today reveal considerably more pathology than previously studies have revealed and must be viewed with caution. The ultimate decision to offer surgery rests on a careful clinical evaluation. The goals of surgical treatment are to adequately decompress the involved neural elements and to avoid iatrogenic damage to either the spinal cord or the spinal column stability. These goals can be achieved with both anterior transthoracic and posterolateral surgical approaches. This chapter discusses the history of posterior treatment of thoracic herniated discs, the clinical presentation and evaluation, methods of posterior treatment, and treatment outcomes.
HISTORY AND EPIDEMIOLOGY
The earliest reports of thoracic intervertebral disc herniation have been attributed to Key in 1838 with the description of a herniated thoracic disc causing myelopathy in a patient. Nearly a century elapsed before surgeons began to perform operative treatment for herniated thoracic discs. In an early study of the outcomes of disc herniation, Mixter and Barr report a series of three cases of thoracic disc herniation treated by laminectomy. Two patients were left completely paraplegic caudal to the level of decompression, revealing the risk for posterior surgical treatment. These results were confirmed by others, and led to the development of other posterior methods of treatment for safe and effective treatment of herniated thoracic discs.
Initial evaluation of patients undergoing surgery for disc herniation revealed thoracic discs to be involved in only 0.15% to 4% of patients. This led early investigators to believe that thoracic disc herniation occurred with extremely low frequency. Autopsy studies, however, and later investigation of asymptomatic volunteers has revealed the incidence of herniated thoracic discs to be much greater. Haley and Perry have found protrusion of thoracic discs to be present in 11% of patients at autopsy. Awwad et al. have performed computed tomography after myelography on asymptomatic patients and also found an 11% prevalence rate of disc protrusion. Notably, 88% of the protrusions caused some deformity of the spinal cord. The additional work of Wood and coworkers using magnetic resonance imaging (MRI) of asymptomatic volunteers revealed an even greater prevalence of asymptomatic disc pathology. In their evaluation of 90 healthy volunteers, 37% were noted to have a frank disc herniation, 53% a disc bulge, and 58% an annular tear. Deformation of the spinal cord was noted in 29% of the asymptomatic volunteers. These results confirm previous work with other imaging modalities, as well as the work of Boden and others with MRI of the lumbar spine, that disc pathology is extremely common and often asymptomatic. Imaging studies, therefore, should be ordered with caution; a careful clinical examination should guide the interpretation of pathologic findings.
The age distribution of symptomatic thoracic disc herniation follows the pattern of other disc pathology. In a classic study of herniated thoracic discs, Arce and Dohrmann found the greatest incidence of herniated discs to occur in the fifth decade of life, with most herniations occurring between the fourth and sixth decades of life. Wood and coworkers, however, did not find a similar age association. They found thoracic disc pathology to be equally prevalent among patients younger or older than 40 years. Similarly, they did not find patient sex to significantly influence the prevalence of thoracic disc herniation.
CLASSIFICATION AND CLINICAL PRESENTATION
Although no universally accepted classification scheme exists, various authors have characterized thoracic disc herniation by anatomic location of the disc fragment or by the clinical syndrome that it produces. A thoracic intervertebral disc can herniate into a central, centrolateral, or lateral position, with 70% to 90% of discs herniating centrally or centrolaterally. The central or centrolateral position appears to remain consistent across the spectrum of clinical pathology, with similar prevalence rates in both symptomatic and asymptomatic patients.
The location of thoracic herniated discs can be further anatomically characterized by the level of the thoracic spine involved. Although Wood and colleagues found disc herniations to occur at all levels of the thoracic spine, others have found a predominance of herniation to occur at the lower thoracic intervertebral discs. In their evaluation of 280 thoracic disc herniations, Arce and Dohrmann found 26% of cases to occur at the T11-12 disc. Seventy-five percent of disc herniations were found to occur between T8 and L1. These findings have been confirmed by both autopsy and MRI studies. Disc herniations also occur in the upper region of the thoracic spine, however, with much lower frequency.
Thoracic disc herniation can also be characterized by the pattern of symptoms that are observed clinically. Many authors suggest that patients follow a typical pattern of symptomatic presentation that begins with axial or radiating thoracic pain. Muscular weakness typically occurs next. The clinical progression finally leads to bowel or bladder dysfunction. Similar to the presentation of disc herniation of other segments of the spine, the symptom pattern will depend on the location of the herniation and the neural elements involved. Consistent with the typical pattern of central or centrolateral disc herniation, however, most patients will present with long tract signs of myelopathy. Tokuhashi et al. have found that the location of a disc herniation in the lower thoracic spine and thoracolumbar junction predict the pattern of symptoms. Herniations found between T10 and T12 typically produce myelopathic symptoms, whereas herniations at T12 and below produce a lower motor neuron pattern. In their classic report on thoracic disc herniation, Arce and Dohrmann found 90% of symptomatic patients to have signs of spinal cord compromise, with 30% having bowel or bladder dysfunction. The acuteness of symptom presentation can vary. Patients may progress slowly over time in a manner similar to clinical pattern of cervical spondylotic myelopathy or with precipitous neurologic decline.
Although many authors have documented the clinical patterns of thoracic disc herniation, it should be noted that the diagnosis can be confounded by many other conditions. Thoracic radicular pain can mimic pain of cardiac or visceral origin. Lower extremity findings can similarly be produced by lower thoracic disc herniations. In addition, long tract signs can be present with disc herniation and spondylosis of the cervical spine. Finally, neoplastic conditions can also occupy space within the spinal canal and produce symptoms of compression of the neural elements. These conditions can often be confused with symptomatic disc herniation. A thorough neurologic examination is, therefore, mandatory to differentiate patterns of thoracic disc herniation from other conditions.
INDICATIONS FOR SURGERY
Although little evidence is available to document the long-term outcome of herniated thoracic intervertebral discs, the pattern of symptoms at presentation has been found to aid treatment planning. Patients with bilateral symptoms, particularly of the lower extremities, have been found to have a poor nonoperative course with progressive neurologic dysfunction that is often irreversible. This course contrasts that of patients with unilateral thoracic radicular signs at presentation. The disparity in outcome between these clinical syndromes has led many to advocate surgery for myelopathy that shows signs of progression including lower extremity weakness or paralysis. Additional indications for surgery include painful radicular symptoms that do not respond to conservative treatment.
A multitude of posterolateral approaches to the thoracic spine exist. The first involves a midline skin incision and an initial dissection that is familiar to most spine surgeons. The second is a paramedian approach that is gaining in popularity and can be accomplished in a minimally invasive fashion. An endoscopic transpedicular technique has been described but has not been widely embraced, although excellent results can be achieved. Other minimally invasive approaches are more directly related to their open counterparts and are becoming increasingly popular.
A midline approach can be used to perform thoracic discectomy through a translaminar, transpedicular, or transfacet, pedicle-sparing approach. The patient is positioned prone on a Jackson table. A midline skin incision is designed with reference to planning a radiograph. In the upper thoracic spine, this is done with a posteroanterior (PA) film, and the counting is done from T1, which can be identified by the presence of the first transverse process and rib. For the midthoracic spine, a PA film is obtained including the 12th rib. For lower lesions, it is usually easier to count from the sacrum. It is vitally important that the landmarks used in the design of the incision are the same ones used in establishing the level on the preoperative diagnostic study. It is important for the surgeon to verify the level independently and not to simply rely on the level as dictated by the reporting radiologist.
Intravenous antibiotics are given, and a midline incision is made to include two levels above and below the level of the lesion. The paraspinal muscles are reflected laterally as a subperiosteal exposure of the spine is performed. The medial aspect of the transverse process is exposed, as well as the complete facet joint. A second radiograph is obtained in the same projection as the planning film with a metallic marker cranial to the pedicle and caudal the disc. The subsequent resection is tailored to the specific anatomy. In general, a lateral laminotomy is performed with a high-speed drill and is extended laterally to perform a medial facetectomy. More room lateral to the spinal cord is obtained by removing a combination of the remainder of the facet and the superomedial aspect of the pedicle. A combination of high-speed drill, curettes, and thin-footplate Kerrison punches are used. This is done in an extensile way until enough room is made to access the disc fragment without any manipulation of the spinal cord. In the extreme case of complete facetectomy and pedicle resection, the incidence of instability after surgery has been low.
The disc fragment is mobilized with a nerve hook and retrieved with a small pituitary rongeur. Great care is taken to apply only a lateral vector to the fragment until it is clear of the spinal cord. The defect in the annulus is then identified and opened further with a #11 blade. Any loose fragments are then retrieved with curettes and a small pituitary rongeur. The closure of the incision is performed in the standard manner.
Paramedian Minimally Invasive Approach
A paramedian minimally invasive approach can be used to perform thoracic discectomy through a lateral, extracavitary approach including costotransversectomy. The location of the paramedian incision is designed as described earlier but is placed approximately 2 cm lateral to the midline. The rib typically can be palpated through the skin at this location. The thoracolumbar fascia is opened sharply, and the fibers of the trapezius, rhomboids, and erector spinae are separated bluntly until a portion of the rib is exposed. The dissection is then advanced medially in the subperiosteal plane of the rib until the lateral projection of the transverse process is exposed. A localizing radiograph is obtained with a metallic clamp estimating the location of the pedicle below the disc ( Fig. 11-1 ). The dorsal aspect of the rib is removed using a high-speed drill, leaving only a thin cortical shell on the ventral side. This portion of the rib can be removed in its entirety if needed. Next, the lateral aspect of the transverse process is removed with the drill. At this point, the lateral aspect of the pedicle is visible, and the neuroforamen can be defined with a forward-angled curette. The superior aspect of the pedicle is thinned with a high-speed drill and removed with a thin-footplate, 2-mm Kerrison. The interface between the dura and the herniated disc is now visible and can be dissected with a Penfield #4 instrument ( Fig. 11-2 ). This exposure can be achieved through a midline approach, albeit with a larger incision and more extensive subperiosteal dissection.