Avoiding Catastrophic Intraoperative Neurologic Events in Spine Deformity Surgery

Eduardo C. Beauchamp, MD

Lawrence G. Lenke, MD

INTRODUCTION

The most feared complication in spine deformity surgery is permanent neurologic deficit and paralysis. Potentially devastating for affected patients, the fear of neurologic complications leads many to avoid necessary surgery. Moreover, these complications place a great burden on affected families and society in general. They are a prime cause of medical malpractice and can have far-reaching effects to surgeons involved. All of these reasons speak to the critical importance of doing everything possible to minimize the risk of these complications.

Intraoperative neuromonitoring (IONM) has evolved as the mainstay in making spine surgery safer from a neurologic perspective. The use of multimodal spinal cord monitoring has been shown to decrease the incidence of these complications allowing the surgeon to respond on a timely manner to intraoperative alerts. During a neuromonitoring alert, the use of an intraoperative checklist may help the surgeon respond optimally under these stressful and chaotic situations.

Despite the advances in spinal cord monitoring, surgical techniques, and imaging guidance, permanent neurologic injuries during spine surgery may still occur, though true incidence varies by type of surgery and is difficult to accurately ascertain. Given the relatively low rate of new neurologic deficit after spine surgery, a large cohort of patients is needed to provide an accurate estimate of neurologic complications. Using the Scoliosis Research Society morbidity and mortality database, Hamilton et al.¹ assessed the rate of neurologic deficit in the adult and pediatric population following 108 419 spine surgery procedures. The rate of new neurologic deficit across all diagnosis was 0.83% and 1.32% for the adult and pediatric population, respectively. When stratified by anatomical location of surgery, the rate of new neurologic deficit in the adult population, ranging from nerve root to spinal cord deficit, was 0.7%, 2.67%, and 0.52% for cervical, thoracic, and lumbar spine surgeries, respectively.¹

In contrast, the incidence of new neurologic deficit following complex spine deformity surgery has been reported to be from <1% to more than 20%.²,³,⁴,⁵,⁶,⁷ A recent prospective multicenter study evaluated the neurologic complications associated with
surgical correction of complex adult spinal deformity.⁵ A total of 273 patients from 15 international institutions were enrolled. They reported changes in the American Spinal Injury Association (ASIA) Lower Extremity Motor Scores (LEMS) at hospital discharge, 6 weeks, and 6 months postoperatively. At hospital discharge, a decline in LEMS was seen in 22.18% of patients, whereas 12.78% showed improvement in their scores. This percentage improved at 6-month follow-up where 10.82% of patients showed a persistent decline in preoperative LEMS, while 20.52% showed improvement and 68.66% showed maintenance of their preoperative LEMS. In patients with abnormal preoperative LEMS, 51.5% showed improvement upon discharge whereas 83.3% showed improvement of their LEMS by 6 months. This multicenter study helped clarify immediate and short-term postoperative expectations on neurologic complication after complex spinal deformity correction. Given this inherent higher risk of neurologic complications associated with spinal deformity correction, the goal of this chapter is to provide an overview on the causes, prevention, and management of catastrophic intraoperative neurologic events in spine deformity surgery.

PREOPERATIVE EVALUATION

History and Physical Examination

Clinical evaluation should begin with a comprehensive medical history and physical examination. Patients should be evaluated for preoperative medical comorbidities and other conditions that may pose an increased risk of intraoperative neurologic deficit. A complete physical examination should include height, weight, skin evaluation, gait assessment, as well as hip, spine, and extremity range of motion evaluation. Deep tendon reflexes and sensory and motor function evaluation are crucial to obtain a baseline neurologic status. Long tract signs, such as sustained ankle clonus, positive Babinski, hyperreflexia, or gait instability, should be assessed and documented. Advanced imaging studies are pursued if any signs of myelopathy are present. Having the patient lie prone on the examination table is a helpful method to evaluate the flexibility, or lack thereof, of the spinal deformity and may provide a preliminary impression of how he or she will look on the operating table. Having the patient assume this prone position several times a day prior to surgery may be beneficial in high-risk cases to help acclimatize the body and spinal cord to the intraoperative position.

Imaging Studies

Imaging studies are essential for preoperative planning and optimal surgical outcomes. Initial imaging should include standing full-spine posteroanterior (PA) and lateral radiographs with visualization of the hip joints. Full-spine radiographs allow evaluation of the global coronal and sagittal alignment, whereas supine and bending radiographs allow for flexibility assessment of the deformity. Lateral dynamic radiographs may help identify any focal instability.

If available, previous radiographs may provide information on the natural history and progression of the deformity. In severe spine deformities, two-dimensional radiographs are not sufficient to fully appreciate the anatomy. Obtaining a computed tomography (CT) scan of the spine with three-dimensional (3D) reconstructions is helpful on delineating the extent of the deformity as well as for evaluation of the rotation and structural anomalies not fully perceived on radiographs (Fig. 1A-C). In severe deformities, a full-sized 3D printed model of the spine may serve as a

preoperative planning tool in addition to being a tangible intraoperative guide for exposure, instrumentation, and deformity correction (Fig. 1D).

Figure 1. PA and lateral radiographs (A) of a 13-year-old female with severe kyphoscoliotic deformity and multiple prior spinal surgeries. Posterior and lateral 3D reconstructions (B) and axial views of the CT scan demonstrate diastematomyelia as indicated in the arrow (C) as well as multiple lumbosacral vertebras with posterior element defects (B). A full-sized 3D printed model of the spine (D) may serve as a guide for surgical dissection, instrumentation, and deformity correction. MRI showing a split spinal cord (E). Advanced imaging studies, such as CT scans and MRI, help provide more detail about the rotation and structural anomalies associated with the spine deformity.

Figure 2. Spinal cord risk classification. T2-weighted MRI images of the thoracic apical level in spine deformity. Type 1: normal cord morphology and visible CSF between the cord and concave pedicle (A). Type 2: normal cord morphology with no intervening CSF between the cord and concave pedicle (B). Type 3: deformed spinal cord with no intervening CSF between pedicle and cord (C).

As part of the preoperative planning, magnetic resonance imaging (MRI) is used to visualize areas of spinal cord or nerve root compression as well as pathologies of the neural axis, such as tethered cord, syringomyelia, and Chiari malformations (Fig. 1E). Spinal cord compression causing abnormal signal intensity changes in T1- and T2-weighted MRI images has been associated with histopathologic changes of the spinal cord.⁸ The degree of spinal cord compression observed in MRI can also help assess the risk of IONM data loss. Siekatycki et al.⁹ used T2-weighted MRI images to categorize the morphology of the spinal cord at the apical level of the thoracic spine deformity. Type 1 is defined as a round-appearing cord with visible cerebrospinal fluid (CSF) between the spinal cord and the apical concave pedicle. Type 2 is a round spinal cord with no intervening CSF between the cord and apical pedicle. A type 3 is a deformed spinal cord with no intervening CSF between the pedicle (Fig. 2). In their study, patients with a type 3 spinal cord had a 28 times greater odds of IONM data loss with deformity correction as compared to patients with normal spinal cord morphology.

MECHANISMS OF NEUROLOGIC DEFICIT

There are numerous etiologies of neurologic deficit during spine deformity surgery including spinal cord stretching or compression during correction, direct trauma to the cord, and vascular compromise, among others.²,¹⁰,¹¹,¹²,¹³,¹⁴ Risk factors associated with IONM alerts and neurologic deficits include patients with preexisting neurologic deficit, myelopathy, previous intradural or anterior thoracic surgery, severe kyphosis, persistent intraoperative hypotension, significant intraoperative blood loss, increased deformity angular ratio (DAR), and patients with unobtainable IONM signals.²,⁵,¹⁵,¹⁶,¹⁷,¹⁸,¹⁹

Patient Positioning

Prior to surgical incision, careful attention should be directed to patient positioning. Special considerations are taken in patients with severe myelopathy, and in certain cases, baseline IONM signals may be needed prior to positioning. Fiberoptic intubation should be considered in patients with inflammatory arthropathies or cervical instability. Before prone positioning, all bony prominences and areas at risk of compressive neuropathy (knees, anterior superior iliac spine, and elbows) should be well padded and protected. Chest pads are positioned leaving sufficient space between the patient’s axilla and chest pad as to minimize brachial plexus compression. Avoiding excessive shoulder abduction and hyperextension while assuming a 90/90-degree shoulder/elbow position is crucial to decrease the risk of brachial plexopathy.

Instrumentation

Although the use of pedicle screws has been shown to be a safe and reliable method of instrumentation, direct spinal cord trauma or nerve root deficit during instrumentation may still occur.²⁰,²¹,²²,²³,²⁴ A systematic review on pedicle screw-related complications in scoliosis surgery revealed a 4.2% rate of screw malposition.²³ This percentage increased to 15.7% in studies were postoperative CT scans were obtained. Out of 1666 patients and a total of 4570 pedicle screws placed, only one temporary neurologic complication associated with screw placement was seen. The reported durotomy rate due to screw malposition was found to be 0.35% per screw inserted. Despite the relatively high rate of screw malposition, the neurologic risk associated with pedicle screw placement remains low. If any medial pedicle perforation is suspected, the pedicle screw is removed, screw trajectory is probed and interrogated, and the decision to use different instrumentation (hooks, wires), redirect the pedicle screw, or avoid instrumenting that level is made.

Intraoperative Hypotension

Even though spine deformity patients may benefit from relative hypotension during surgical exposure, persistent hypotension throughout the case can be detrimental. Adequate cord perfusion should be maintained during deformity correction, as hypotension is a known cause of IONM data loss and neurologic deficit.²,¹⁰,¹¹,¹²,¹³,²⁵ The spinal cord is more prone to vascular insult during surgeries around the “critical zone” of T4-T9.²⁶ Mean arterial pressure is maintained at 60-70 mm Hg during exposure to minimize blood loss, whereas during correction of the deformity, it is elevated to 80-90 mm Hg to maintain cord perfusion. This is particularly important in cases of severe kyphosis in which stretching of the anterior spinal artery during deformity correction may furthermore preclude perfusion to the spinal cord.

Othman and Lenke established the role of hypotension as a cause of IONM data loss during kyphosis correction.¹³ They reported an acute loss of IONM data associated with a sudden decrease in mean arterial pressure before any corrective maneuver, which resolved after elevating the mean arterial pressure with the use of intravenous vasopressors. In addition to prolonged hypotension, spinal cord ischemia may result secondary to hypoxia caused by decreased hemoglobin level or cord hypoperfusion after ligation of segmental blood vessels during an
anterior-based procedure.¹²,²⁷,²⁸ In a retrospective review of more than 12 000 adult and pediatric spinal surgeries, poor spinal cord perfusion was the second most common cause of IONM alerts, causing ˜12% of the total alerts and second only to screw placement at 30%.³ Raising the mean arterial pressure as a first response to an average of 86 mm Hg after an IONM alert was successful in restoring the spinal cord monitoring data in 20% of pediatric spine deformity patients.²⁹

Deformity Correction

Direct and indirect spinal cord deficit may occur during deformity correction. Surgical instruments may cause direct injury to the spinal cord, especially in cases where a significant portion of the dura is exposed, such as in three-column osteotomies. Surgeons should be aware of the exposed neural elements at all times. A “no-fly zone,” where no instruments are passed, is established above the neural elements in cases where a significant amount of dura is exposed, as to not let any instrument fall in the operative field and inadvertently injure the spinal cord.

Special attention should be paid to neuromonitoring changes during deformity correction. Releasing the tension on the spinal cord by reversing the correction should be considered if neuromonitoring data remain at warning criteria once other potential causes of spinal cord deficit have been addressed (Case #1). The stability of the spinal column is taken into consideration before removing and reversing the correction, especially during three-column osteotomies. If removal of the rod involves further destabilization of the spine, partial stability should be maintained to avoid further spinal cord compromise. If there is improvement in IONM data after release of the correction, the surgeon should consider attempting a more moderate correction vs in situ fusion with minimal correction. If no improvement in IONM data is seen after correction release and elevation of the mean arterial pressure, other causes of spinal cord compression are contemplated. Osteotomy sites should be inspected prior and after corrective maneuvers. The relationship between bony structures and neural elements may change after deformity correction, and areas of the spinal cord that seemed free may now be compressed by overhanging lamina, pedicles, or other bony structures.

The DAR is a radiographic measure used to assess the segmental angularity of a deformity. It is calculated by dividing the preoperative maximum sagittal or coronal Cobb angle by the number of vertebras spanning the deformity (Fig. 3). A higher DAR implies a more acute angle of the deformity, thereby a higher risk of compression of the spinal cord during deformity correction. In patients undergoing a vertebral column resection (VCR) through a posterior approach, a total DAR ≥ 25 or a sagittal DAR ≥ 15 represents a much higher risk or IONM events and postoperative neurologic deficits.¹⁸ In pediatric patients, a total DAR > 45 degrees per level and a sagittal DAR > 22 degrees per level have been associated with a 75% incidence of IONM alert during a posterior-based VCR.¹⁹

Only gold members can continue reading. Log In or Register to continue