Advertisement

Neurosurgery and Spine Procedures in Cancer Patients

  • John WiemersEmail author
  • Claudio E. Tatsui
Living reference work entry

Abstract

Patients with tumors in and around the spinal cord present many challenges to the physicians that care for them. Patient symptoms can range from sensory and motor dysfunction to extreme pain due to nerve impingement and pathological fractures. There is a multitude of factors that make the perioperative course for these patients complicated. Airway management, positioning, blood and fluid management, postoperative pain control, and thromboembolism are some potential problems these patients will encounter. This chapter will help identify and discuss the perioperative management of these patients.

Keywords

Reconstructive spine Cancer spine Perioperative spine care Oncologic spinal tumors Spinal metastasis Intradural tumors Intramedullary tumors Extradural metastasis 

Introduction

Patients with cancer affecting the spine (extradural compartment) or the spinal cord (intradural compartment) present many challenges to the surgeon, anesthesiologist, and critical care physician. Cancer affecting those compartments such as bony spine may result in neurologic dysfunction and/or spinal instability that significantly impact the perioperative management of these patients. Airway management is made more difficult by the limited range of motion mandated in patients with cervical spine instability. Transfer and positioning of patients is more complicated when spinal segments are deemed unstable. If spinal cord or nerve root function is at risk during surgery, then neurophysiologic monitoring may be necessary, adding to the special requirements of the anesthetic management and the overall complexity of patient monitoring. Involvement of the spine with tumor can be extensive with significant extension into paraspinal tissues and body cavities such as the thorax, abdomen, and pelvis. The surgical approach is tailored based on the level of spinal involvement and the extent of disease. Transoral/transmandibular, anterior cervical, transsternal, transthoracic, transabdominal, retroperitoneal, dorsal, and combined approaches, all pose unique perioperative challenges. The hyper-vascularity of many spinal tumors combined with the extensive dissection required to remove them and the complex reconstructive procedures that follow often results in extensive blood loss. Operative time may exceed 10 h, and in some cases a staged approach is used when surgery is expected to require two different routes requiring change in the operative positioning for tumor resection. The successful completion of complex spine tumor surgery depends upon the patient’s ability to tolerate enormous physiologic stresses. Patients must be properly selected and be in optimal medical condition prior to surgery if uncomplicated intraoperative and postoperative courses are to be realized.

Etiology

Tumors in the spinal region are classified into three general categories based on tumor location. These are extradural, intradural extramedullary, and intradural intramedullary lesions.

Extradural Tumors – The bone is the third most common site of metastasis, and the spine is the most common location of bony metastasis. Approximately, 95% of the tumors affecting the spine are metastatic, and the most common types include breast, lung, prostate, renal, melanoma, colon, and thyroid cancers. Primary spine tumors are rare and constitute approximately 5% of all cancer affecting the spine. These lesions can arise anywhere along the vertebral column. Chordomas are the most common primary spine tumors; they arise from remnants of the primitive notochord and may occur at skull base and mobile spine (cervical, thoracic, and lumbar) but most commonly are found in the sacral region [37]. Sarcomas arise from the mesenchymal elements and may erode into the bony spine. Rarely, they can arise directly from bony elements of the spine. Osteosarcoma, chondrosarcoma, leiomyosarcoma, and Ewing sarcoma are all sarcomas that can invade the spinal vertebrae. Benign lesions can arise in the bony spine and depending on neurologic symptoms might need to be surgically removed. These include osteoid osteomas, osteoblastomas, osteochondromas, chondroblastomas, giant cell tumors, vertebral hemangiomas, and aneurysmal bone cysts [19].

Intradural Intramedullary Tumors – The majority of primary spinal cord tumors are gliomas, or bearing a histological resemblance link to normal glial cells. The major types of glial tumors are ependymomas, astrocytomas, and oligodendrogliomas. Intramedullary spinal cord metastases are rare, but increased use of MRI has resulted in more frequent recognition of these lesions.

Ependymomas are the most common intramedullary spinal cord tumor in adults, with a peak age at presentation between 30 and 40 years. These tumors can be located anywhere along the spinal cord, but approximately 50% occur in the lumbosacral spinal cord or filum terminale. Optimal management consists of gross total resection. Although these are infiltrative tumors, a total or near total resection can frequently be achieved without causing neurologic deficits [42].

Myxopapillary ependymoma are slow-growing glial tumors that are typically found in young adults 30–40 years old and are more common in males. The vast majority of these tumors are found in the lumbosacral and thoracolumbar areas of the spine. These tumors can frequently be totally resected, and many patients are cured following gross total resection.

Anaplastic ependymomas are high-grade ependymomas that have anaplastic features on histological examination. Compared to lower-grade ependymomas, anaplastic tumors have a higher recurrence rate and poorer survival rates [36].

Astrocytomas occur throughout the spinal cord. Approximately 50% of spinal cord astrocytomas are pilocytic, and 50% are infiltrative. Pilocytic astrocytomas are well circumscribed and low grade with nonaggressive clinical behavior. Infiltrative astrocytomas of the spinal cord appear as nonencapsulated lesions that enhance minimally on MRI. About one-third are histologically high-grade (anaplastic astrocytoma or glioblastoma). Pilocytic astrocytomas can often be completely or near-completely resected without causing additional neurologic deficits. Diffuse infiltrative astrocytomas are generally much more difficult to resect due to lack of tissue planes and risk of neurologic morbidity.

Intradural Extramedullary Tumors – The most common intradural extramedullary tumors are meningiomas and nerve sheath tumors. Spinal meningiomas can arise from arachnoidal cells anywhere along the neuraxis. Spinal meningiomas most commonly occur within the thoracic spine. These tumors are generally slow growing that adhere to the spinal dura requiring surgical resection. Complete resection can often be achieved, where subtotal resections usually have expectant regrowth with additional resections and/or radiation.

Nerve sheath tumors account for about 25% of tumors arising in the intradural extramedullary space. These tumors consist of neurofibromas and schwannomas. Neurofibromatosis (NF) is a common autosomal-dominant disorder with 100% penetrance in familial lines. Two subtypes have been established: NF-1 and NF-2. The reported prevalence of NF-1 (also known as von Recklinghausen disease or peripheral neurofibromatosis) is 1 in 3,000 to 4,000 individuals. NF-2 (also known as central neurofibromatosis) occurs in 1:40,000 and represents 2.5% of patients with spinal cord tumors. NF-2 is largely associated with ependymomas (intramedullary) and occasionally meningiomas (extramedullary) [23]. Schwannomas are composed of Schwann cells with fibrous tissue. These tumors may show cystic degeneration and hemorrhage. They usually displace nerve roots.

Epidemiology

Spinal cord tumors are much less prevalent than intracranial tumors, in a ratio of 1:4, but this varies by histology. The intracranial to spinal ratio of astrocytoma is 10:1, and the ratio for ependymomas varies from 3:1 to 20:1, although spinal tumors are the most prevalent location of ependymomas in adults. Spinal ependymomas occur more often in men with a male to female ratio of 1.4 to 2.5:1, whereas meningiomas are more prevalent in women than men. Spinal tumors occur predominantly in young or middle-aged adults and are less common in childhood or after age 60 years. Epidural spinal metastasis is estimated to occur in about 5–10% of all cancer patients. As the survival of patients with systemic cancers improves, the incidence of epidural spinal cord compression from bony metastasis has increased.

Pathophysiology

Extradural Metastasis – Extradural metastasis from primary tumors (lung, breast, GI, prostate, etc.) is spread mainly from the arterial route. Invasion of adjacent vertebral levels often occurs via venous channels in the Batson plexus. Direct invasion through the intervertebral foramina can also occur, especially in the thoracic and lumbar spine. Severe degree of compression of the spinal cord will result in demyelination or axonal destruction. Vascular compromise produces venous congestion and vasogenic edema of the spinal cord resulting in venous infarction and hemorrhage. About 70% of the spinal metastasis occurs in the thoracic spine, 20% in the lumbosacral, and 10% in the cervical region. Involvement of multiple vertebral bodies occurs in up to 50% of cases. Most of the lesions are localized at the anterior portion of the vertebral body (60%), hence the common occurrence of pathologic fractures. In 30% of cases, the lesion infiltrates the pedicle or lamina.

Intradural Intramedullary Tumors – Intradural intramedullary tumors (ependymomas, astrocytomas, hemangioblastomas, etc.) originate from cells within the spinal cord area. Ependymomas are soft, encapsulated, reddish-gray or yellow tumors with modest vascularity. The majority of them are classified as low grade and have a propensity to grow slowly. Originating from ependymal cells, ependymomas are located more centrally (midline) within the spinal cord as compared with other tumors. They appear as focal enlargements within the spinal cord. Various histologic subtypes of ependymomas exist. Astrocytomas are gray, glossy tumors that are characterized by a poorly defined plane. They primarily occur in the cervical spine, and they often involve multiple spinal segments due to its expansive nature. Astrocytomas have an association with neurofibromatosis-1. Adults mainly exhibit higher-grade lesions, whereas low-grade lesions are associated with the younger population [33]. Hemangioblastomas are small benign richly vascularized solitary neoplasms that rarely extend beyond one or two segments of the spinal cord. Within the spinal cord, hemangioblastomas are predominantly located at the posterior or posterolateral region of the spinal canal. Most develop sporadically, but they are associated with von Hippel-Lindau disease, especially when presenting in the spinal cord. Hemangioblastomas are composed of a dense network of vascular capillary channels containing endothelial cells, pericytes, and lipid-laden stromal cells [33].

Clinical Features

A thorough preoperative evaluation often uncovers patient health problems that require attention prior to the proposed surgery or that may alter the perioperative care. A complete history and physical should be obtained, including the history of the present illness, past medical history, and past surgical treatment history. Especial attention needs to be given to the history and physical exam as they are of paramount importance to define the need of further testing and treatment recommendation, which often involves surgery. A good perioperative/anesthetic plan relies heavily on the quality of this information. Although all organ systems are important to the success of an operation, the status of the cardiovascular, pulmonary, and neurologic systems will have a major impact on patient management and outcome. If cardiac or pulmonary disease is suspected, then consultation with specialists in these areas is mandatory as specialized care in the perioperative period will likely be necessary.

Diagnosis

Tumors within or extrinsic to the spinal cord can cause symptoms through disruption of normal neural elements and pathways, producing both local and distal effects. The most frequent local effect is pain that causes nocturnal awakening. Patients often describe this pain as gnawing and unremitting. The site of this may provide an indication of the anatomic location of the tumor [45]. Neurologic dysfunction depends on the level of spinal cord compression, and the resultant symptoms are related to interruption of ascending and descending spinal cord pathways. The most common symptoms are sensory dysfunction, loss of proprioception, muscular weakness, and sphincter dysfunction. Although neurologic manifestations may begin unilaterally, they can progress to involve both sides of the spinal cord and thereby produce bilateral symptoms [45].

The physical exam is necessary to define the exact intensity of neurological dysfunction and correlate it with the probable site of tumor involvement. Documentation of the preoperative neurologic deficits is important to determine if further neurologic deterioration occurred after surgery. An assessment of the patient’s ambulatory status is also necessary since this carries important prognostic significance [45].

Magnetic resonance imaging (MRI) of the spine is currently the diagnostic study of choice, providing excellent delineation of the spinal cord and surrounding structures. Almost all intrinsic spinal cord tumors and metastases enhance with gadolinium [35].

Management

Pharmacologic

Antibiotic Prophylaxis

Postoperative wound infections following spine surgery occur with varying incidence depending on the type of surgery performed. For simple lumbar discectomy, the incidence is approximately 1%, whereas for complex posterior instrumented fusion, the incidence is between 3% and 8% [24]. Malnourished and immunocompromised individuals are at increased risk for developing postoperative wound infection.

Prophylactic antibiotic therapy has been shown to reduce the incidence of infection after spinal surgery [5]. Antibiotic prophylaxis reduces infection rates by 50–70% compared with baseline rates. The optimal choice of antibiotic prophylaxis is based on the most likely organism to be involved in the infection for the type of surgery being performed. In most clean spine cases, resident gram-positive skin flora such as Staphylococcus aureus and S. epidermidis are anticipated. Given their safety, antimicrobial spectra, low incidence of side effects, and pharmacokinetic profile, first-generation cephalosporins are the preferred choice of antibiotic prophylaxis. Cefazolin, in particular, has become the primary choice in clean surgical procedures [10].

In order for antibiotic prophylaxis to be effective, it is critical to administer the drug at a time prior to the skin incision such that the optimal tissue concentration is present when the skin incision is actually made. In general, complete administration of antibiotic prior to but within 1 h. of skin incision is most effective [7]. In addition, adequate antibiotic levels should be maintained throughout the duration of the procedure by timed redosing [10]. Continued postoperative dosing for up to 72 h after surgery has been employed but is no more effective than a single preoperative dose, and it will lead to the proliferation of resistant strains of bacteria [22]. Although vancomycin is more effective than first-generation cephalosporins against gram-positive organisms, use of vancomycin should be limited because of its increased potential for toxicity, cost, and development of microbial resistance [21]. Vancomycin, however, is a reasonable alternative to cefazolin in patients with penicillin/cephalosporin allergy.

Intraoperative Neurophysiology Monitoring and Drug Pharmacology

In most intradural spinal procedures, neurophysiology monitoring has become the standard of care to monitor the function of the spinal cord in patients under general anesthesia during resection of tumors. Intraoperative monitoring of motor and sensory evoked potentials allows the surgeon to work around vital neural elements with some degree of confidence that the required surgical manipulations is not damaging important neural pathways within the spinal cord. Decreases in the baseline strength of the evoked potentials give feedback to the surgeon that certain maneuvers are damaging the spinal cord requiring adaptions in the surgeon’s technique, decrease neural tissue manipulation, or reverse previous maneuvers related with retraction or blood supply interruption during surgery. This presents the anesthesiologist with unique anesthetic concerns. Nonparalytic anesthesia can result in patient movement if there is a mismatch between the level of anesthesia and the level of surgical stimulation. A greater degree of vigilance is thus required by the anesthesiologist in order avoid patient bucking, which increases the abdominal pressure, translating in increased venous bleeding, disrupting the surgical flow. The most common monitoring modalities along with their anesthesia implications are discussed below. Somatosensory evoked potentials (SSEPs) are the electrophysiologic responses of the central nervous system to the stimulation of a peripheral nerve. Most commonly, the response to peripheral nerve stimulation is monitored at the level of the cerebral cortex. Peripheral nerves that are commonly used for stimulation are the tibial, peroneal, and median. The monitored neural pathway includes the dorsal root ganglia and the dorsal or posterior column of the spinal cord. Therefore, SSEP monitoring is particularly useful during posterior spinal approaches [31]. The usefulness of SSEPs lies in their ability to demonstrate the functional integrity of neural pathways in an anesthetized patient. Numerous cases have been reported wherein early recognition of changes in SSEPs have changed the surgeon’s technique and apparently prevented neurologic injury [29]. In the operating room, baseline SSEPs are obtained after the induction of anesthesia. Serial recordings are then obtained throughout the surgery and compared to the baseline. Physiologic changes may alter the quality of SSEP recordings. These changes include hypotension, hypothermia, anemia, hypoxia, and changes in pCO2. Changes in the depth of anesthesia can also alter SSEP recordings. Every effort should be made to avoid fluctuations in inhaled gas concentration and bolus injection of hypnotic agents during the period of surgery that is deemed most at risk for neurologic injury (e.g., during decompression and placement of spinal instrumentation). In general, inhaled agents will affect SSEP monitoring more so than intravenous agents. An anesthetic technique that relies on the use of a narcotic agent will generally result in the least amount of interference with SSEPs.

Motor evoked potentials (MEPs) monitor the integrity of motor pathways. The MEPs are a measure of the response of peripheral motor nerves and the muscles they supply from stimulation of the motor cortex. For the upper extremity, the abductor pollicis brevis is usually monitored, while the tibialis anterior, lateral gastrocnemius, and/or abductor hallucis are monitored for the lower extremity. Thus, MEPs monitor the corticospinal tract (i.e., motor cortex, corticospinal tract, nerve root, and peripheral nerve). When SSEPs and MEPs are used together, information about the integrity of both sensory and motor pathways are possible in the anesthetized patient. The MEPs are very sensitive to anesthetic agents. In particular, isoflurane and propofol boluses both cause significant depression of MEPs [40].

Intraoperative Hoffman reflex (H-reflex) monitoring can be used as an adjunct to motor tract assessment. The H-reflex is a true reflex with an afferent arc and an efferent arc that is mediated by alpha motor neurons. It is elicited by electrical stimulation of an afferent mixed peripheral nerve and by then recording the muscle response. The H-reflex can be monitored below the level where the spinal cord ends (where MEPs may be less useful). It monitors the pathway of the reflex and may be suppressed by injury of more cephalad motor tracts. Most commonly, the tibial nerve is stimulated, with recording from the gastrocnemius or soleus muscles [14].

Intraoperative electromyography (EMG) is commonly applied to monitor selective nerve root function during spinal surgery. EMG is “real-time” recording from peripheral vasculature. Common EMG recording sites by spinal levels are as follows: C4, supraspinatus; C5, deltoid and biceps; C6, biceps and wrist extensors; C7, triceps, wrist flexors, and finger extensors; C8, hand intrinsics and finger flexors; T1, hand intrinsics; T6–12, rectus abdominis; L1, iliopsoas; L2, adductor longus and vastus medialis; L4, vastus medialis and vastus lateralis; L5, TA and extensor halluces longus; S1, medial gastrocnemius and peroneus longus; and S2–5, perianal musculature and urethral sphincter. At baseline no muscle activity is recorded from an intact nerve root. Surgical manipulations such as pulling, stretching, or compression of nerves provoke spikes or bursts of activity termed neurotonic discharges, resulting in activity in the corresponding innervated muscle(s). Spontaneous EMG is quite sensitive to irritation of the nerve root, such as retraction of spinal cord or nerve root, saline irrigation, and manipulation during surgery [30].

A decrease in waveform amplitude and an increase in latency may indicate neurologic dysfunction, such as from surgical injury. Specifically, a 50% reduction in amplitude and 10% increase in latency in SSEPs and MEPs are considered pathologic. If worrisome monitoring changes are encountered, then the precise time that they occurred should be noted and correlated with any particular event such as spinal cord retraction, arterial or venous occlusion, fracture or deformity reduction, or placement of spinal segmental stabilization. The surgeon should immediately consider releasing retraction, removing implants, or reversing any corrective reduction that has been made. Correction of hypotension, hypothermia, hypoxia, and other adverse anesthetic events should also be performed rapidly [31].

Anesthetic Effects on Evoked Potentials

Most anesthetics agents alter neural function by producing dose-dependent depression in synaptic activity. In general, inhalational agents (isoflurane, sevoflurane, desflurane) have greater effects on all modes of neuromonitoring than do intravenous (IV) anesthetic agents. Choice and dose of anesthetic agents should be tailored to the modalities used. Every effort should be made to keep the level of anesthesia constant during critical monitoring periods in order to avoid confounding interpretation of changes seen. Volatile inhalational anesthetics with nitrous oxide included cause a dose-dependent decrease in amplitude and an increase in latency of all evoked potentials (SSEPs and MEPs) though motor evoked potentials appear much more sensitive to even low concentrations of inhalational anesthetics.

IV anesthetics such as propofol and opioids have less effect on monitoring than inhalational agents, though very deep levels of anesthesia or bolus doses of IV anesthetics can affect neuromonitoring waveforms, especially MEPs. With the exception of ketamine and etomidate, all common IV anesthetics cause a dose-dependent in amplitude and an increase in latency, ultimately producing burst suppression and electrical silence at high doses. Ketamine and etomidate infusions have shown to enhance cortical SSEP amplitude. Ketamine infusions can be beneficial during spine surgery when neuromonitoring is used to enhance the amplitude of the evoked potentials being monitored. Etomidate is not widely used as an infusion to enhance neuromonitoring due to its association with adrenal suppression and worsened outcomes due to sepsis. IV opioids cause small dose-dependent depression of SSEP and MEP responses, though even at very high doses, evoked potentials can be recorded. Infusions of remifentanil, fentanyl, and sufentanil are commonly used as part of the anesthesia during neuromonitoring.

Neuromuscular blocking agents (NMBAs) are needed to facilitate endotracheal intubation at the beginning of the case, but the use of them during the neuromonitoring phase of the surgery depends on what monitoring is needed for the case. MEPs and electromyography (EMG) are affected by neuromuscular blockade. During SSEP monitoring the neuromonitoring team often requests a small infusion of a NMBA to decrease the artifact from large back muscle being stimulated from electrocautery. If NMBAs are used, the degree of paralysis must be monitored, usually with a train-of-four (TOF) peripheral nerve stimulator. Importantly the level of paralysis must be kept constant, which is best accomplished by the use of an infusion of the NMBA.

Blood Loss during the Intraoperative Period

Spinal procedures may result in substantial blood loss. Extradural metastasis of certain tumors are particularly hypervascular and may produce significant blood loss during resection. Renal cell carcinoma, thyroid carcinoma, myeloma, and some sarcomas can be highly vascularized. Preoperative transarterial embolization can greatly reduce the blood supply to these tumors, thereby decreasing the blood loss significantly during resection. There are numerous studies that show the benefit of decreasing intraoperative blood loss with preoperative embolization. Guzman et al. showed that preoperative embolization of various hypervascular spinal metastases greatly reduced the intraoperative blood loss [18]. In a paper by Gottfried et al., the estimated intraoperative blood loss was 4,350 ml in non-embolized patients and 1,800 ml in patients with particle embolization [17].

Surgical technique plays an important role in blood loss. Injecting the skin with a local anesthetic mixture with a 1:100,000 epinephrine solution can dramatically reduce the amount of blood loss during the creation of a long surgical incision. This technique will also limit the excessive use of cautery on the skin. Meticulous attention should be payed to hemostasis and to dissection in avascular planes. Anterior spinal surgery often results in less blood loss than the classic posterior approach and may, therefore, affect surgical decision-making in certain cases. In most cases, maintaining a mean arterial pressure around 65–70 mm Hg is recommended to reduce bleeding.

The use of antifibrinolytics are also an effective means of reducing blood loss, particularly when microvascular oozing from a large open wound is the problem. In orthopedic surgical patients, the use of lysine analogs tranexamic acid (TXA) and epsilon-aminocaproic acid (EACA) is effective when administered as a single dose and when used in multiple-dose regimens (i.e., multiple boluses or bolus plus infusion) [47]. In spine surgery patients, TXA has been shown to consistently reduce estimated blood loss, the need for transfusion, and the total amount of blood transfused when compared with controls [46]. For patients undergoing a vascular anastomosis (e.g., flap coverage of wound) or free fibula grafting or have a hypercoagulable condition, the use of TXA or EACA would be contraindicated.

The suspicion of a hypercoagulable state or disseminated intravascular coagulation (DIC) is heralded by the presence of diffuse oozing without the formation of blood clots. This usually happens in the context of prolonged surgery and excessive bleeding with massive transfusion of blood products (more than one blood volume). The DIC is confirmed by elevated prothrombin and partial thromboplastin times, low platelet count, and elevated fibrin split products. Treatment of DIC is facilitated by early recognition and transfusion of fresh frozen plasma, cryoprecipitate, platelets, and red blood cells. DIC is best prevented by good surgical technique that minimizes blood loss and by the staging of procedures that will require obligatory high blood loss and extended operative time.

Non-pharmacologic

Surgical Approaches and Positioning

Surgery on the spine spans a wide spectrum of interventions. Surgery may involve interventions that are transoral, transmandibular, anterior cervical, transthoracic, transabdominal, or transpelvic. The patient may be positioned supine, prone, lateral, or sitting, and the position of the patient may need to be changed during surgery.

Essentially all patients undergoing spine surgery for treatment of a spinal neoplasm will be under general anesthesia. One exception would be the treatment of some pathologic fractures with either vertebroplasty or kyphoplasty [15]. Inherent to inducing general endotracheal anesthesia comes the task of safely intubating the patient who may have cervical spine instability. Patients with cervical spine instability require awake, fiber-optic intubation. Patients are given a topical anesthetic, and the endotracheal tube is placed under direct fiber-optic visualization while maintaining the neck in a neutral position. The patient’s motor exam is carefully monitored during awake intubation. If any change in the exam is noted, then interventions are either stopped or reversed.

Prone procedures make immediate or emergency access to the airway problematic. Endotracheal tubes and vascular access lines should be tightly secured prior to patient positioning. In case of airway compromise, a stretcher should always be kept nearby so that the patient could be turned supine for emergent airway management. The prone position can also lead to an increase in intra-abdominal and intrathoracic pressure. As a consequence, there is increased ventilatory pressures and decreased cardiac stroke volume. The increased abdominal pressures also increase venous pressures within the epidural vessels, which lead to increased blood loss. For these reasons, patients should be positioned on a Jackson table, four-poster frame, or equivalent device that allows the abdomen to hang free. When placing the patient in the prone position, care must be taken to support the head with the neck in a neutral position while avoiding direct pressure on the eyes. Appropriate padding at the elbows, knees, and feet will prevent nerve and soft-tissue injury. Note that the arms are positioned with enough padding so that they do not hang, and the shoulders are abducted no more than 90° in order to avoid brachial plexus stretch. For upper thoracic and cervical procedures, the head may be fixed in a Mayfield skull clamp, and the arms may be tucked to allow the surgeon unimpeded access to the surgical site. Again, padding of the elbows, wrists, and hands is crucial to decrease the risk of nerve/soft-tissue injury.

The sitting position is an alternative to the prone position for posterior cervical procedures. The main advantage of this approach is less blood loss from the lower venous pressures associated with an upright position. The main disadvantage is the potential for venous air embolism. Intraoperative monitoring for air embolism is mandatory, and a central venous catheter that may be used to aspirate an air embolism should be placed preoperatively.

Patients are positioned supine for anterior cervical, transsternal, and midline transabdominal/pelvic procedures. The anterior cervical approach requires that the patient’s arms be tucked at their side, limiting the anesthesiologist’s access to intravenous lines and monitoring equipment. Airway interventions are impeded by the proximity of the surgical field to the endotracheal tube. The lower thoracic, lumbar, and sacral spine may be accessed in a supine position. Inherent in this approach are the anesthetic risks that accompany the increased fluid requirements that are inherent during intra-abdominal surgery and the surgeon’s need to retract on visceral organs, aorta, and vena cava. Excessive retraction on the intra-abdominal and intrapelvic great vessels (aorta, vena cava, iliac arteries, and veins) can adversely affect circulation to the lower extremities. Frequent pedal pulse checks ensure lower extremity perfusion during lengthy procedures.

The lateral position is used to approach the anterior aspects of the thoracic or lumbar spine or when access to both the anterior and posterior spine is desired. The lateral position is inherently unstable, so steps must be taken to secure the patient. A beanbag combined with a strap over the hips is usually sufficient. The head is cushioned with the neck in neutral position, and the arms are placed at an angle of 90° with respect to the torso so as not to stretch the brachial plexus. The elbows are padded to protect the ulnar nerves. An axillary roll must be placed to protect the neurovascular structures in the axilla from injury. Lastly, the hips and knees are flexed, and cushions are placed under the fibular head of the lower leg and between the legs. A double-lumen endotracheal tube allows selective ventilation of the dependent lung so that the nondependent lung can be easily retracted from the surgical field. During intrathoracic or intra-abdominal approaches, sudden hypotension without bleeding may occur. In most cases, this is related to decreased blood return to the right side of the heart. The common causes are retractors placed on the inferior vena cava. The solution is to remove the retractors, especially those retracting the inferior vena cava, liver, or spleen and to wait for the blood pressure to normalize. Failure of the blood pressure to normalize after removal of retractors should raise suspicion for other systemic problems.

Airway Management

Indications for an awake intubation include the risk of delayed gastric emptying, the need to assess neurologic function after intubation is complete (in cases such as an unstable cervical spine), or the presence of a neck stabilization device (such as a halo vest or cervical traction), which prevents adequate airway maintenance in an unconscious patient. Direct laryngoscopy with minimal in-line stabilization or a hard collar is an accepted means of intubation for many patients provided this can be achieved without any neck movement. Awake fiber-optic intubation will be required in patients wearing devices such as halo vests, or hard collars, which make conventional airway access impossible, and in those patients where difficulty is anticipated because of anatomical reasons such as micrognathia, limited mouth opening, etc.

Perioperative Vision Loss

Postoperative vision loss (POVL) is a rare but devastating complication that can occur with spinal surgeries. Ophthalmic complications have been reported to occur in less than 0.2% of spine surgeries. The major causes of visual loss in this patient population include ischemic optic neuropathy (ION), central retinal artery occlusion (CRAO), and retinal vein occlusion (RVO). CRAO and RVO are attributed to embolic load and/or direct globe compression, emphasizing the need to protect the eyes from direct pressure while in the prone position [44]. In these cases, visual loss is usually unilateral and associated with other signs of pressure (e.g., ophthalmoplegia, ptosis, or altered sensation in the territory of the supraorbital nerve). Horseshoe-shaped headrests should be avoided in prone patients as they have been implicated in cases of CRAO [27].

ION almost always results in permanent visual loss. Independent risk factors for the development of ION include male sex, obesity, use of the Wilson frame (head lower than the heart), longer anesthetic time, greater estimated blood loss, and lower percent of colloid in the non-blood fluid replacement [4]. Although the final common pathway is thought to be hypoperfusion of the optic nerve, there is no clear association with either intraoperative systemic hypotension or with the presence of peripheral vascular disease or diabetes [27].

Procedures requiring a combined approach (anterior-posterior) to the spine are commonly longer in duration and are associated with significant blood loss and increased postoperative complications when compared with procedures requiring an isolated anterior or posterior approach. Consideration should be given to staging these combined approaches to decrease the risk of postoperative vision loss in these patients [4].

Because of the devastating nature of this complication, patients should be informed of an increase incidence of visual loss after spinal operations that are expected to be of prolonged duration and associated with significant blood loss.

Management Algorithm

Morbidity and mortality after spinal fusion can approach 23% and 0.5%, respectively, and up to 10% of lumbar spine fusion patients will require care in an ICU. Factors independently associated with increased morbidity after spine surgery include advanced age, male gender, and increased comorbidity burden. Specific comorbid conditions highly associated with perioperative adverse events include pulmonary hypertension, congestive heart failure, renal failure, and pre-surgical coagulopathy [16].

Postoperative Airway Management

Special attention is paid to the patient’s airway presurgery, but a significant number of problems can present themselves at the completion of surgery. The most common indication for intensive care unit (ICU) admission after spine surgery is the need for ventilatory support. Airway and facial edema commonly occur during long procedures in the prone position and with the administration of large volumes of intravenous fluid. The patient with significant airway edema is at risk for airway obstruction after extubation.

The incidence of airway compromise requiring reintubation after anterior procedures on the cervical spine has been reported as up to 1.9% [27]. Anterior cervical spinal surgery may result in recurrent laryngeal nerve injury or hematoma, causing airway obstruction after extubation. The most common cause of vocal cord paralysis is compression of the recurrent laryngeal nerve within the endolarynx. Monitoring ETT cuff pressure and release after retractor placement may prevent injury to the recurrent laryngeal nerve [28].

Tracheoesophageal edema or perforation may also develop in the tissue of the neck because the esophagus and trachea are retracted during these procedures to obtain access to the cervical spine. Risk factors include multiple level surgery, blood loss of more than 300 ml, duration of >5 h, a combined anterior and posterior operation, and previous cervical surgery. High-risk patients should be monitored closely in the ICU, and some consideration should be made for a staged extubation utilizing an airway exchange catheter once a leak is confirmed around the endotracheal tube [27].

Postoperative Neurologic Monitoring

Every attempt should be made to awaken the patient for neurologic assessment prior to leaving the operating room. Patients not suitable for extubation should still be allowed to awaken for neurologic exam prior to leaving the operating room. If the patient demonstrates a neurologic deficit, then a CT scan through the operated levels should be obtained immediately to rule out the possibility of malpositioned instrumentation or wound hematoma. A CT study will help the surgeon localize the exact level of the malpositioned spinal hardware, allowing removal or repositioning of the device maintaining the spinal stabilization.

Following major reconstructive surgery for spine tumors, patients should be observed in the intensive care unit for 24–48 h. Neurologic examination is performed every 1 h. Motor power in all four extremities and sensory exam are most important to check. Rarely, patients will have a delayed onset of neurologic deficit. In these cases, hypotension and spinal cord ischemia may be the culprit. The immediate response to a delayed postoperative deficit should be to aggressively treat any hypotension if present and to then obtain a CT scan of the operated site to rule out the development of a hematoma or instrumentation failure. If CT myelography is obtained, then cerebrospinal fluid (CSF) should be removed prior to instillation of the intrathecal contrast to avoid the potential for increasing extramural spinal cord pressure, which could further compromise spinal cord perfusion. A lumbar drain may be left in place to divert CSF and reduce extramural spinal cord pressure for up to 72 h. if spinal cord ischemia is suspected [1]. Sacrifice of thoracic nerve roots and corresponding radicular arteries during tumor removal has the potential to cause relative spinal cord ischemia. In case of nerve root/radicular artery sacrifice, intraoperative and postoperative hypotension should be avoided and reversed aggressively if encountered.

Postoperative Vision Assessment

In high-risk cases, assessment of vision should be performed as soon as possible in the ICU and an early ophthalmic opinion sought if there is a suggestion of visual compromise. Initial management should include optimization of arterial pressure, oxygenation, and correction of anemia. Treatment with agents such as acetazolamide has not been beneficial, and there is rarely any useful improvement in vision, so attention should be focused on preventative measures. Other possible causes of POVL include cortical ischemia and hemorrhage from a cerebral tumor [27].

Coagulopathies

The suspicion of a hypercoagulable state or disseminated intravascular coagulation (DIC) is heralded by the presence of diffuse oozing without the formation of blood clots. This usually happens in the context of prolonged surgery and excessive bleeding with massive transfusion of blood products (more than one blood volume). The DIC is confirmed by elevated prothrombin and partial thromboplastin times, low platelet count, and elevated fibrin split products. Treatment of DIC is facilitated by early recognition and transfusion of fresh frozen plasma, cryoprecipitate, platelets, and red blood cells. DIC is best prevented by good surgical technique that minimizes blood loss and by the staging of procedures that will require obligatory high blood loss and extended operative time.

Pain Management

Patients undergoing major reconstructive spine surgery after tumor resection are at high risk for experiencing significant postoperative pain as a result of the extent of surgery that has been performed often involving extensive muscle dissection both anteriorly over the abdomen and thorax and dorsally over the paraspinal support muscles. Poorly controlled pain has a major impact on the development of other postoperative complications, as the patient’s ability to cough, breathe, ambulate, and participate actively in their postoperative recovery will be extremely limited if pain is severe. Immediate- and long-term patient outcome may be improved as a result of optimized postoperative analgesia.

Although opioids remain the primary analgesic agent for the management of acute postoperative pain after major reconstructive spine surgery, opioid-related adverse effects inhibit rapid recovery and rehabilitation. The concept of multimodal analgesia was introduced more than a decade ago as a technique to improve analgesia and reduce the incidence of opioid-related adverse events [6]. Opioids can be administered IV (continuous infusion and patient-controlled analgesia devices with or without background infusions), epidural, and intrathecal routes. Their use, via the IV route in particular, is associated with side effects such as respiratory depression, nausea and vomiting, sedation, and gastrointestinal ileus. The latter may be especially disadvantageous after major spinal surgery, when some degree of paralytic ileus is common [31].

In addition to intravenous opioids, components of multimodal perioperative pain management may include the following: ketamine, gabapentinoids, acetaminophen, and nonsteroidal anti-inflammatory drugs (NSAIDs). The use of ketamine as an analgesic adjunct has been shown to reduce postoperative opioid requirements. A randomized, prospective study of 102 opioid-tolerant patients with chronic pain having spine surgery found that ketamine administration resulted in short- and long-term (up to 6 weeks) opioid consumption without an increase in side effects [40]. Gabapentinoids (gabapentin, pregabalin) reduce postoperative pain, opioid consumption, and some opioid-related adverse effects when used as a preoperative adjunctive analgesic [39]. Continuing administration in the postoperative period is likely to be more effective than a single preoperative dose of either of these medications [34]. Acetaminophen can be used as part of a multimodal pain control regimen after spine surgery. A meta-analysis of randomized trials found that the addition of acetaminophen (intravenous or oral) to morphine following major surgery resulted in a small but statistically significant decrease in morphine use postoperatively [25]. NSAIDs are controversial as an adjunct pain control modality as some studies have shown that NSAIDs may adversely affect bone healing [11]. When used in the postoperative setting, special attention to the renal function should be given; the use of NSAIDs can decrease renal filtration rates. Also, the platelet function can be affected, and the risk of postoperative bleeding can be increased.

Malnutrition

Patients undergoing major spine surgery are at significant risk of malnutrition postoperatively. Extensive surgical wounds significantly increase caloric requirements, and optimal wound healing depends on adequately meeting these needs. A normal active adult may require between 2500 and 3000 calories/day, whereas the surgical patient may need as much as 6000 calories/day [26]. Two of the most common complications following major reconstructive spine surgery are infection and wound healing problems, and malnutrition may play a significant role in this.

Many patients requiring surgery for spine tumors may already be malnourished prior to surgery. A postoperative ileus, lasting 2–5 days, is often present after major spinal surgery, particularly after anterior approaches to the spine. For these reasons, consideration should be given to aggressive enteral or total parenteral nutrition before and after surgery. Patients should not be allowed to maintain a catabolic state in the immediate postoperative period. For patients who cannot tolerate adequate oral intake on the first or second postoperative day, then a nasogastric feeding tube should be placed and tube feeds given at caloric goal [8]. Because depletion of nutritional parameters appears to correlate with an increased incidence of wound complications, the maintenance of adequate nutrition may result in a decrease in these complications [38].

Prevention of Thromboembolism and Early Mobilization

Venous thromboembolism remains a significant cause of morbidity and mortality in surgical patients. In adult patients undergoing major reconstructive spine surgery, the incidence of deep venous thrombosis has been reported to vary from 0.3% to 25% using Doppler imaging techniques as a screening tool [32]. Patients with spinal cord injury are known to have a significant incidence of pulmonary embolism [43]. Immobilization increases the incidence of deep vein thrombosis. In one study, the incidence of symptomatic pulmonary embolism following anterior/posterior spine surgery was 6% [9]. Factors that may contribute to the development of deep vein thrombosis in postoperative spine patients include prolonged immobilization, extended operative times, and prone positioning that may lead to compression of femoral veins. Also, despite modern internal fixation methods, patients may be slow to mobilize after major spine surgery because of postoperative pain and critical care needs.

Unfortunately, the true incidence of thromboembolic complications in spine surgery patients remains unknown, and the need for postoperative prophylaxis is not yet widely accepted. Use of intermittent compression boots appears to provide a reasonable level of primary prevention [13]. One study, however, found an unacceptably high rate of symptomatic pulmonary embolism (6%) in patients undergoing anterior and posterior spine surgery treated with compression boots alone [9].

Hesitancy to use prophylactic pharmacologic anticoagulation in spine surgery patients is due to the theoretic risk of bleeding complications, which may result in epidural hematoma and spinal cord/cauda equina compression. Currently, limited evidence suggests that bleeding complications associated with the use of low molecular weight heparin are not significant in patients felt to be at greatest risk – those undergoing craniotomy [20]. Also, the combination of compression stockings plus treatment with low-dose heparin decreases the incidence of symptomatic deep vein thrombosis [3]. Limited reporting indicates that the use of low molecular weight heparin is safe and effective in preventing deep vein thrombosis and its complications in patients undergoing spine surgery [41]. Based on the available evidence, some feel that it is safe and effective to treat postoperative spine patients with compression boots and subcutaneous enoxaparin, 30 mg every 12 h. Enoxaparin may be given just prior to surgery or started the morning after surgery.

Early mobilization is reported to significantly reduce the incidence of perioperative complication, shorten duration of in-hospital stay, and contribute to improved perioperative functional status in spinal surgery patients [12]. Delaying ambulation by just 24 h may contribute to higher complication rates and inferior functional outcomes [2].

Conclusion/Summary

The care of the cancer patient undergoing spinal procedures requires close monitoring of possibly multiple systems postoperatively. Spinal surgery often involves multiple specialties/services that are involved in the care and the ultimate success of the surgical procedure involved. From neurologic monitoring to postoperative acute pain to possible correction of coagulopathies, the patient postoperatively will need close specialized care in the immediate postsurgical period.

Cross-References

References

  1. 1.
    Ackerman L, Traynelis V. Treatment of delayed onset neurological deficit after aortic surgery with lumbar cerebrospinal fluid drainage. Neurosurgery. 2002; 51:1414–21.CrossRefGoogle Scholar
  2. 2.
    Adogwa O, Elsamadicy A, Fialkoff J, et al. Early ambulation decreases length of hospital stay, peri-operative complications and improves functional outcomes in elderly patients undergoing surgery for correction of adult degenerative scoliosis. Spine. 2017;42:1429–5.CrossRefGoogle Scholar
  3. 3.
    Agnelli G, Piovella F, Buoncristiani P, et al. Enoxaparin plus compression stockings compared with compression stockings alone in the prevention of venous thromboembolism after elective neurosurgery. N Engl J Med. 1998;339:80–5.CrossRefGoogle Scholar
  4. 4.
    American Society of Anesthesiologists Task Force of Perioperative Visual Loss. Practice advisory for perioperative visual loss associated with spine surgery. Anesthesiology. 2012;116:274–85.CrossRefGoogle Scholar
  5. 5.
    Barker F II. Efficacy of prophylactic antibiotic therapy in spinal surgery: a meta-analysis. Neurosurgery. 2002;51:391–400.CrossRefGoogle Scholar
  6. 6.
    Buvanendran A, Kroin J. Multimodal analgesia for controlling acute postoperative pain. Curr Opin Anaesthesiol. 2009;22:588–93.CrossRefGoogle Scholar
  7. 7.
    Classen D, Evans R, Pestotnik S, et al. The timing of prophylactic antibiotics and the risk of surgical wound infection. N Engl J Med. 1992;326:281–6.CrossRefGoogle Scholar
  8. 8.
    Daly J, Redmond H, Gallagher H. Perioperative nutrition in cancer patients. J Parenter Enter Nutr. 1992;16:100S–5S.CrossRefGoogle Scholar
  9. 9.
    Dearborn J, Hu S, Tribus C, et al. Thromboembolic complications after major thoracolumbar spine surgery. Spine. 1999;24:1471–6.CrossRefGoogle Scholar
  10. 10.
    DeWald R, Arlet V, Carl A, et al., editors. Spinal deformities: the comprehensive text. New York: Thieme; 2003.Google Scholar
  11. 11.
    Dodwell E, Latorre J, Parisini E, et al. NSAID exposure and risk of nonunion: a meta-analysis of case-control and cohort studies. Calcif Tissue Int. 2010;87:193.CrossRefGoogle Scholar
  12. 12.
    Epstein N. A review article on the benefits of early mobilization following spinal surgery and other medical/surgical procedures. Surg Neurol Int. 2014; 5(suppl 3):S66–73.CrossRefGoogle Scholar
  13. 13.
    Ferree B, Wright A. Deep venous thrombosis following posterior lumbar spinal surgery. Spine. 1993; 18:1079–82.CrossRefGoogle Scholar
  14. 14.
    Feyissa A, Tummala S. Intraoperative neurophysiologic monitoring with Hoffmann reflex during thoracic spine surgery. J Clin Neurosci. 2015;22:990.CrossRefGoogle Scholar
  15. 15.
    Fourney D, Schomer D, Nadar R, et al. Percutaneous vertebroplasty and kyphoplasty for painful vertebral body fractures in cancer patients. J Neurosurg. 2003;98:21–30.CrossRefGoogle Scholar
  16. 16.
    Fu K, Smith J, Polly D, et al. Correlation of higher preoperative American society of anesthesiology grade and increased morbidity and mortality rates in patients undergoing spine surgery. J Neurosurg Spine. 2011;14:470.CrossRefGoogle Scholar
  17. 17.
    Gottfried O, Schloesser P, et al. Embolization of metastatic spinal tumors. Neurosurg Clin. 2004;15(4): 391–9.CrossRefGoogle Scholar
  18. 18.
    Guzman R, Dubach-Schwizer S, et al. Preoperative transarterial embolization of vertebral metastases. Eur Spine J. 2005;14(3):263–8.CrossRefGoogle Scholar
  19. 19.
    Hansen M, Nellissery M, Bhatia P. Common mechanisms of osteosarcoma and paget’s disease. J Bone Miner Res. 1999;14(Suppl 2):39.CrossRefGoogle Scholar
  20. 20.
    Iorio A, Agnelli G. Low-molecular weight and unfractionated heparin for prevention of venous thromboembolism in neurosurgery: a meta-analysis. Arch Intern Med. 2000;160:2327–32.CrossRefGoogle Scholar
  21. 21.
    Janning S, Rybak M. Antimicrobial prophylaxis in surgery. In: Dipro J, Talbert R, Yee G, Matzke G, Wells B, Posey L, editors. Pharmacotherapy: a pathophysiologic approach. Stamford: Appleton & Lange; 1999. 1888–1889.Google Scholar
  22. 22.
    Kernodle D, Barg N, Kaiser A, et al. Low-level colonization of hospitalized patients with methicillin-resistant coagulase-negative staphylococci and emergence of the organisms during surgical antimicrobial prophylaxis. Antimicrob Agents Chemother. 1988;32:202–8.CrossRefGoogle Scholar
  23. 23.
    Lee M, Rezai A, Freed D, et al. Intramedullary spinal cord tumors in neurofibromatosis. Neurosurgery. 1996;38(1):32–7.CrossRefGoogle Scholar
  24. 24.
    Levi A, Dickman C, Sonntag V, et al. Management of postoperative infections after spinal instrumentation. J Neurosurg. 1997;86:975–80.CrossRefGoogle Scholar
  25. 25.
    McDaid C, Maund E, Rice S, et al. Paracetamol and selective and non-selective non-steroidal anti-inflammatory drugs (NSAIDs) for the reduction of morphine-related side effects after major surgery: a systematic review. Health Technol Assess. 2010;14:1.CrossRefGoogle Scholar
  26. 26.
    McPhee I, Williams R, Swanson C. Factors influencing wound healing after surgery for metastatic disease of the spine. Spine. 1998;23:726–32.CrossRefGoogle Scholar
  27. 27.
    Nowicki R. Contin Educ Anaesth Crit Care Pain. 2014;14(4):147–52.CrossRefGoogle Scholar
  28. 28.
    Osborn I, Naruse R, Ferrario L. Airway management in neurosurgical patients. In: Hagberg C, Artime C, Aziz M, editors. Hagberg and benumof’s airway management. 4th ed. Philadelphia: Elsevier; 2018. p. 729–30.Google Scholar
  29. 29.
    Owen J. The application of intraoperative monitoring during surgery for spinal deformity. Spine. 1999;24:2649–62.CrossRefGoogle Scholar
  30. 30.
    Park J, Hyun S. Intraoperative neurophysiological monitoring in spinal surgery. World J Clin Cases. 2015;3(9):765–73.CrossRefGoogle Scholar
  31. 31.
    Raw D, Beattie J, et al. Anaesthesia for spinal surgery in adults. Br J Anaesth. 2003;91(6):886–904.CrossRefGoogle Scholar
  32. 32.
    Rokito S, Schwartz M, Neuwirth M. Deep vein thrombosis after major reconstructive spinal surgery. Spine. 1996;21:853–8.CrossRefGoogle Scholar
  33. 33.
    Samartzis D, Gillis C, et al. Intramedullary spinal cord tumors: part I-epidemiology, pathophysiology, and diagnosis. Global Spine J. 2015;5(5):425–35.CrossRefGoogle Scholar
  34. 34.
    Schmidt P, Ruchelli G, Mackey S, et al. Perioperative gabapentinoids: choice of agent, dose, timing, and effects on chronic postsurgical pain. Anesthesiology. 2013;119:1215.CrossRefGoogle Scholar
  35. 35.
    Sevick R, Wallace C. MR imaging of neoplasms of the lumbar spine. Magn Reson Imaging Clin N Am. 1999;7:539.PubMedGoogle Scholar
  36. 36.
    Shrivastava R, Epstein F, Perin N, et al. Intramedullary spinal cord tumors in patients older than 50 years of age. Management and outcome analysis. J Neurosurg Spine. 2005;2:249.CrossRefGoogle Scholar
  37. 37.
    Stacchiotti S, Casali P, Lo Vullo S, et al. Chordoma of the mobile spine and sacrum: a retrospective analysis of a series of patients surgically treated at two referral centers. Ann Surg Oncol. 2010;17:211.CrossRefGoogle Scholar
  38. 38.
    Stambough J, Beringer D. Postoperative wound infections complicating adult spine surgery. J Spinal Disord. 1992;5:277–85.CrossRefGoogle Scholar
  39. 39.
    Tiippana E, Hamunen K, Kontinen V, et al. Do surgical patients benefit from perioperative gabapentin/pregabalin? A systematic review of efficacy and safety. Anes Analg. 2007;104:1545.CrossRefGoogle Scholar
  40. 40.
    Ubags L, Kalkman C, Been H, et al. The use of ketamine or etomidate to supplement sufentanil/N2O anesthesia does not disrupt monitoring of myogenic transcranial motor evoked responses. J Neurosurg Anesthesiol. 1997;9:228–33.CrossRefGoogle Scholar
  41. 41.
    Voth D, Schwarz M, Hahn K, et al. Prevention of deep vein thrombosis in neurosurgical patients: a prospective double-blind comparison of two prophylactic regimen. Neurosurg Rev. 1992;15:289–94.CrossRefGoogle Scholar
  42. 42.
    Waldron J, Laperriere N, Jaakkimainen L, et al. Spinal cord ependymomas: a retrospective analysis of 59 cases. Int J Radiat Oncol Biol Phys. 1993;27:223.CrossRefGoogle Scholar
  43. 43.
    Waring W, Karunas R. Acute spinal cord injuries and the incidence of clinically occurring thromboembolic disease. Paraplegia. 1991;29:8–16.PubMedGoogle Scholar
  44. 44.
    Warner M. Postoperative visual loss: experts, data, and practice. Anesthesiology. 2006;105:641.CrossRefGoogle Scholar
  45. 45.
    Welch W, Jacobs G. Surgery for metastatic spinal disease. J Neuro-Oncol. 1995;23:163.CrossRefGoogle Scholar
  46. 46.
    Zufferey P, Merquiol F, Laporte S, et al. Do antifibrinolytics reduce allogeneic blood transfusion in orthopedic surgery? Anesthesiology. 2006;105:1034.CrossRefGoogle Scholar
  47. 47.
    Zufferey P, Lanoiselee J, Chapelle C, et al. Tranexamic acid bolus plus infusion is not more effective than a single bolus in primary hip arthroplasty: a randomized controlled trial. Anesthesiology. 2017;127:413.CrossRefGoogle Scholar

Copyright information

© Springer Nature Switzerland AG 2019

Authors and Affiliations

  1. 1.MD Anderson Cancer CenterThe University of TexasHoustonUSA
  2. 2.MD Anderson Cancer CenterHoustonUSA

Section editors and affiliations

  • Garry Brydges
    • 1
  1. 1.Department of Anesthesiology Division of Anesthesia, Critical Care and Pain MedicineThe University of Texas MD Anderson Cancer CenterHoustonUSA

Personalised recommendations