Anesthesia for Spine Surgery and the Prevention of Complications

Summary of Key Points

Preoperative evaluation of the patient scheduled for spine surgery most commonly focuses on functional capacity, neurological assessment, and upper airway examination.
In patients with significant spinal cord compression or an unstable cervical spine undergoing cervical spine surgery, awake fiberoptic intubation may be the safest technique for upper airway management.
Increased intraocular pressure during spine surgery may compromise the ocular perfusion pressure. Maintaining proper mean arterial pressure (MAP) is important to avoid postoperative vision loss.
For anesthetic management during intraoperative electrophysiological monitoring, it is important to avoid muscle relaxants and high concentrations of inhalation anesthetics.
Avoiding hypervolemia is a prerequisite for avoiding endothelial glycocalyx damage.
Goal-directed fluid therapy is the ideal technique for fluid management during spine surgery.
Albumin is the preferred colloid for spine surgery.

Spine surgery has evolved since the 1990s into a multidisciplinary specialty. This chapter presents the most recent evidence-based advances in the perioperative anesthetic management for spine surgery.

Preoperative Assessment

As in other surgeries, preoperative assessment should routinely include a history and physical examination, as well as optimization of the patient for surgery. Preoperative examination of the nontrauma patient’s ability to flex and extend the neck without symptoms while awake is a crucial step in assessing upper airway management during surgery. The anesthesiologist should examine the patient for the presence or absence of Lhermitte sign, sometimes called the “barber chair” phenomenon. Lhermitte sign is an electrical sensation that runs down the back and into the limbs from involvement of the posterior columns and is produced by flexing or extending the neck. This sign suggests compression of the spinal cord in the neck from any cause, such as cervical spondylosis, disc herniation, tumor, or Chiari malformation. The presence of Lhermitte sign should alert the anesthesiologist to use extra caution in maintaining a neutral head position during intubation and throughout the procedure. Awake fiberoptic intubation is often used in patients with significant spinal cord compression or cervical spinal instability. Cervical spondylosis, the commonest indication for cervical spine surgery, can result in anterior cervical osteophyte formation, which may make direct laryngoscopy difficult. Of note, extension at the craniocervical junction is necessary for full mouth opening. Therefore, a small mouth opening should alert the anesthesiologist to the possibility of difficult intubation. Finally, radiological diagnosis of atlantoaxial subluxation, often present in rheumatoid arthritis (RA) patients or Down syndrome patients, should be considered in the preoperative assessment, as direct laryngoscopy in patients with undiagnosed atlantoaxial subluxation has been reported to cause quadriplegia.

The risk for a perioperative cardiovascular event should be properly assessed in the preoperative evaluation. Identification of increased risk provides patients with information that may help them better understand the benefit-to-risk ratio of surgery, especially complex spine surgery, and may lead to interventions that decrease risk. The two most common risk models used in the preoperative assessment are the revised cardiac risk index (RCRI) and the American College of Surgeons’ National Surgical Quality Improvement Program (NSQIP) risk model ( Table 82.1 ). The reported rate of cardiac death or nonfatal myocardial infarction is more than 5% in high-risk procedures, between 1% and 5% in intermediate-risk procedures, and less than 1% in low-risk procedures.

Table 82.1

Comparison Between Revised Cardiac Risk Index and American College of Surgeons’ National Surgical Quality Improvement Program Preoperative Risk Factors Assessment Indexes for Predicting Postoperative Complications

History of ischemic heart	RCRI
History of heart failure	RCRI
History of cerebrovascular disease	RCRI
Insulin dependent diabetes mellitus	RCRI
Preoperative serum creatinine	≥2 mg/dL (RCRI) or >1.5 mg/dL (NSQIP)
Increasing age	NSQIP
American Society of Anesthesiologists class	NSQIP
Preoperative functional status	NSQIP

NSQIP , National Surgical Quality Improvement Program; RCRI, revised cardiac risk index.

National Surgical Quality Improvement Program Database Risk Model

The NSQIP database is used to determine risk factors associated with intraoperative/postoperative myocardial infarction or cardiac arrest (MICA). The NSQIP is composed of risk factors such as type of surgery, dependent functional status, abnormal creatinine, American Society of Anesthesiologists (ASA) class, and increased age as predictors of MICA. The NSQIP risk model has higher predictive ability than the RCRI (0.884 vs. 0.747).

Revised Cardiac Risk Index

The RCRI performs well in distinguishing patients at low risk compared with high risk for all types of noncardiac surgery, but it is less accurate in patients undergoing only vascular noncardiac surgery. Moreover, it does not capture risk factors for noncardiac causes of perioperative mortality. Of note, only one-third of perioperative deaths are because of cardiac causes ( Table 82.2 ).

Table 82.2

Incidence of Postoperative Complications as Measured by the Revised Cardiac Risk Index

No risk factors	0.4%
One risk factor	1.0%
Two risk factors	2.4%
Three or more risk factors	5.4%

Patients whose estimated risk is less than 1% are labeled as being low-risk and require no additional cardiovascular testing. However, patients whose risk is 1% or higher may require additional cardiovascular evaluation. Often, these are patients with known or suspected coronary artery or valvular heart disease. Further evaluation, in addition to consultation with a cardiologist, may include stress testing, echocardiography, or 24-hour ambulatory monitoring. In addition, no further tests usually are required in patients who can perform four or more metabolic equivalents of activity. However, for those whose functional capacity is lower or unknown, additional testing is indicated if it will influence perioperative care.

Biomarkers

The biomarkers brain natriuretic peptide (BNP) and N-terminal pro-BNP (NT-proBNP) have been used for perioperative evaluation. These biomarkers are produced by myocytes in response to stress and are important prognostic indicators of heart failure. Cuthbertson and associates demonstrated that a BNP greater than 40 pg.mL ^–1 has a sensitivity of 75% and a specific of 70% for predicting perioperative death or myocardial injury; it also performed better than the RCRI. The combination of C-reactive protein and NT-proBNP had better predictive ability for perioperative major cardiovascular events in noncardiac surgery than other risk models.

Ankle Brachial Index

The ankle brachial index (ABI) is valuable for cardiovascular risk qualification. The ABI is easily determined by measuring the systolic blood pressure (SBP) with a portable Doppler ultrasound machine on each arm and on the dorsalis pedis and posterior tibial arteries of each ankle. The highest of the two arm pressures is selected, as is the highest of the two pressures of each ankle. The ABI is obtained by dividing the highest ankle SBP in each leg by the highest arm pressure. The normal ABI would be greater than 0.9 and 1.3 or lower. Values between 0.41 and 0.9 are associated with peripheral arterial disease (PAD), and values 0.4 or lower with severe PAD. Therefore, the addition of ABI to traditional measures of cardiovascular risk factors will improve the sensitivity, specificity, and predictive values for a future cardiovascular event.

Movement of the Cervical Spine With Intubation

Placement of the endotracheal tube requires a complex series of movements. The primary force applied by the laryngoscope is upward lift, which results in extension of the atlantooccipital interspace. The lift also results in flexion at lower vertebrae. There is evidence that insertion of the laryngoscope results in maximal extension at the occiput and atlas, with flexion below C2. External stability methods may reduce movement during direct laryngoscopy (DL), but will also make glottis visualization more difficult.

Atlantoaxial instability, observed in patients with RA and Down syndrome, holds important clinical significance for anesthesiologists. In atlantoaxial instability, the odontoid process is no longer firmly held against the back of the anterior arch of C1, owing either to disruption of the transverse ligament or damage to the odontoid process itself.

Roughly 30% of patients with severe RA may exhibit some instability at C1‒C2, although few patients require surgical correction. Therefore, it is advisable for all patients with severe RA to have periodic flexion and extension radiographs, certainly before undergoing any surgical procedure. Similarly, roughly 15% of patients with Down syndrome patients may exhibit laxity in the transverse ligament. It may be advisable that Down syndrome patients have cervical dynamic radiographs before any surgical procedure that requires DL or extensive neck manipulation.

Atlantoaxial instability arises from the fact that C1 is a rigid ring affixed firmly to the base of the skull. If the transverse ligament is damaged, lifting the skull and C1 will result in an increase in the anterior atlantodental interval, and hence a decrease in the posterior atlantodental interval. In other words, C2 remains fixed whereas C1 slides anteriorly, with the cord potentially becoming compressed or trapped in the space behind the odontoid.

Dynamic factors in the cervical spinal column affect the degree of cord compression. Hyperextension narrows the spinal canal by shingling the laminae and buckling the ligamentum flavum. Translation or angulation between vertebral bodies in flexion or extension can narrow the space available for the cord. Patients who lack cord compression statically may compress the cord dynamically, leading to the development of myelopathic symptoms during intubation with DL. Translation of adjacent vertebral bodies in flexion or extension may compress the spinal cord ( Fig. 82.1 ). Therefore, maintaining the neutral neck position during intubation is of utmost importance to avoid cervical spinal cord injury, especially during cervical spine surgery.

Fig. 82.1, Dynamic mechanical factors, such as normal or abnormal motion, can create the so-called pioneer phenomenon, which may be associated with different patterns of encroachment of the cord.

The primary force applied during DL is upward lift with a little angular force. This force can be as high as 50 to 70 Newton (N) (45 N is sufficient to lift 4.5 kg or 10 lb). The more difficult the exposure, as may occur with cervical spondylosis, the greater the force that may need to be applied, leading to a greater likelihood of cervical spinal cord injury during intubation. DL with the Macintosh 3 blade results in near maximum extension at the occiput and C1. ^, Use of the new videoscope Airtraq resulted in a 23% reduction in cervical spine motion in a cadaveric study; it exerted only 20% of the force typically exerted by DL.

Fiberoptic intubation is an ideal method for intubation, resulting in the smallest degree of upper cervical spine motion (ideal especially when cervical spine movement is not feasible). The use of the laryngeal mask airway or intubating laryngeal mask airway is not recommended for upper airway management in the patient with an unstable cervical spine, as these exert high pressure against the upper cervical vertebrae during insertion, during inflation, and while in situ.

Manual In-Line Stabilization and Cricoid Pressure

The goal in manual in-line stabilization (MLS) is to apply force to the head and neck equal in magnitude and opposite in direction to those generated by DL so as to limit movement that might result during airway management. However, MLS not only failed to reduce movement at the site of instability in cadaver models but also limited the anesthesiologists’ ability to visualize the vocal cords. ^, Cricoid pressure, as long as it was not excessive, did not result in movement in a cadaver model of an injured cervical spine.

Central Cord Syndrome and Direct Laryngoscopy

Central cord syndrome (CCS) was first described by Schneider in 1954. Classic CCS presents as a spinal cord injury with weakness in the upper extremities greater than that in the lower extremities in patients with underlying cervical spondylosis. The pathological mechanism often involves hyperextension of the cervical spine with compression of the cord, by osteophytes anteriorly, and by enfolded ligamentum flavum posteriorly, impinging on the central core of the spinal cord and leading to ischemia, edema, or hematomyelia. Hyperextension may seem mild, as with direct intubation, but in the setting of cervical spondylosis it can result in marked neurological injury. Younger patients with congenital cervical stenosis also are at increased risk of sustaining CCS as a result of hyperextension injury, as during DL.

Patients with chronic renal failure are prone to spinal degenerative disease. The term destructive spondyloarthropathy is used to describe a process occurring in hemodialysis patients, which can affect the cervical spine. Therefore, the head of a patient with chronic renal failure should be kept in a neutral position during endotracheal intubation and during surgery.

Chin lift, jaw thrust, and DL can cumulatively cause movement of the cervical spine and dynamic hyperextension injury to the spinal cord, thereby inducing CCS. Therefore, the use of either asleep or awake fiberoptic intubation, while keeping the patient’s head in a neutral position, especially during cervical spine surgery, may be the best way to avoid cervical spinal cord injury.

Pathophysiological Changes With Prone Position

Cardiovascular Changes

Using a noninvasive cardiac output monitor, both cardiac index (CI) and venous return decreased in unanesthetized, healthy volunteers in the prone position. CI decreased compared with the supine position as follows: knee–chest position (20%), on pelvic props from a modified Relton–Hall frame under the anterior superior iliac spines and padded support under the chest (17%), on an evacuatable mattress (11%), and on pillows (3%; one pillow under the thorax and one under the abdomen, leaving the abdomen free to move). Toyota and Amaki studied transesophageal echocardiograms in 15 healthy patients undergoing prone-position lumbar laminectomy. The prone position caused left ventricular volume and compliance to decrease. These changes were attributed to a decrease in the venous return because of inferior vena caval compression, and decreased left ventricular compliance because of increased intrathoracic pressure in the prone position. These results have been confirmed by other studies, using thermodilution pulmonary artery catheters to measure the CI when transferring from the supine to the prone position. Cardiac output in these studies decreased by 17% to 24%. The reduction in cardiac output in the prone position also leads to a decrease in the metabolism of propofol. A reduction in propofol metabolism while in the prone position could also explain the results of Sudheer and colleagues, who showed a significant reduction in cardiac output in the prone position during maintenance of anesthesia using propofol compared with isoflurane. Pearce observed vena caval pressures to be 0 to 40 mm H ₂ O in patients in the prone position with the abdomen hanging free. In contrast, patients with abdominal compression had vena caval pressures greater than 300 mm H ₂ O. Increased venous pressure not only increases bleeding during spine surgery, owing to congestion of vertebral veins, but also can impair spinal cord perfusion.

The use of the prone position with abdominal compression was identified as a plausible cause of spinal cord ischemia, leading to neurological deficits after cervical laminectomy. The authors of this case series recommended the avoidance of abdominal compression and hypotension, especially in myelopathic patients for whom maintenance of spinal cord perfusion pressure is of paramount importance.

Hemodynamic and Fluid Management during Spine Surgery in the Prone Position

The aim of fluid management is to maintain normovolemia and adequate tissue perfusion while avoiding tissue edema. The discovery of the endothelial glycocalyx (EG) has changed our understanding of tissue perfusion. The EG consists of membrane-bound proteoglycans and glycoproteins that form a network in which plasma proteins are retained. The main constituents of the glycocalyx are syndecan, heparan sulfate, and hyaluronan. EG plus bound fluids and plasma proteins forms the endothelial surface layer (ESL), which has a thickness of about 1 μm. In humans, the ESL contains approximately 700 to 1000 mL of noncirculating plasma.

An intact EG is the key factor in maintaining an intact vascular barrier, maintaining a proper filtration rate, and avoiding tissue edema despite high extravascular colloid osmotic pressure (COP) in the interstitial tissues. EG retains plasma and generates the endothelial surface layer, which also has a high COP. In a small gap below the EG, the concentration of proteins is lower than in the interstitial space, allowing small net fluid filtration into the interstitial tissue. The EG structure makes the arteriolar and capillary domains relatively impermeable. However, venules represent a suitable site for fluid filtration through their gaps and pores. Because colloids are able to escape through venular pores, there are low osmotic pressure differences, as well as low hydrostatic differences. The result is a slow net filtration through venular pores. The latter property is in accordance with the newly appreciated fact that there is no net reabsorption of fluid in the venular segments of the microcirculation.

In summary, a small amount of fluid and protein exits the blood vessels at all times, but it is removed from the interstitial space in a timely manner via the lymphatic system under normal physiological conditions.

Important Functions of the Glycocalyx

The EG plays a very important role in maintaining the proper functions of the immune and coagulation systems. Normally, the EG contains small endothelial adhesion molecules. Degradation of EG by cytokines, especially by tumor necrosis factor, or by ischemic reperfusion injury, exposes the adhesion molecules to immunocompetent cells, which enhances leukocyte and platelet adhesion. After shedding of the EG, circulating glycocalyx components like heparan sulfates have a direct chemotactic effect on leukocytes and increase their presence at the site of inflammation. Consequently, the destruction of EG can trigger the inflammatory cascade. Therefore, maintaining the integrity of the EG might represent a promising therapy for inflammation and ischemic/reperfusion injury. EG has an important mechanosensory role, as it translates intravascular shear stress into biochemical activation of endothelial cells to release nitric oxide (NO). EG is a crucial component for binding and regulating enzymes involved in the coagulation cascade. Furthermore, the most important inhibitor of thrombin and factor Xa, antithrombin III, is firmly attached to the EG. It is therefore not surprising that hyperglycemia-induced loss of EG is accompanied by activation of coagulation and vascular dysfunction in diabetic patients ( Fig. 82.2 ).

Fig. 82.2, Physiological role of the glycocalyx. The endothelial glycocalyx regulates nitric oxide synthase activity, harbors superoxide dismutase, and serves as a physical barrier for macromolecules, including plasma proteins and lipoproteins. In addition, the glycocalyx attenuates platelet and leukocyte adhesion.

Perioperative Fluid Management and Glycocalyx

Perioperative fluid management is one of the key factors in maintaining the integrity of the EG. It is well known that iatrogenic acute hypervolemia can lead to the release of atrial natriuretic peptide (ANP). ANP induces shedding of EG components, mainly syndecan-1, thereby increasing shifts of fluid and macromolecules into the interstitial space. Thus the ability of ANP to increase capillary permeability to water, solutes, and macromolecules might be at least partially explained by its capacity to disturb EG structure. The average insensible fluid loss is only about 0.5 mL/kg/h via skin and airways in the awake adult. During abdominal surgery, insensible fluid loss increases to only 1 mL/kg/h. Avoiding hypovolemia and hypervolemia through careful perioperative fluid management is an important element in maintaining a healthy EG, and thereby limiting perioperative fluid and protein shifts into the interstitial space. An intact ESL is essential to avoid excessive tissue edema. Therefore a goal-directed fluid approach is essential to maintain normovolemia and ESL integrity, reducing postoperative complications like anastomotic leaks, nausea and vomiting, infections, and pulmonary complications.

Many of these complications may result from excessive use of crystalloids. Of note, 80% of the infused crystalloids are distributed in interstitial tissues under all conditions. Therefore, the use of crystalloid at a rate of 1 to 2 mL/kg/h for maintenance and the use of isooncotic colloid like albumin for the replacement of blood loss may be ideal for fluid management during spine surgery.

Suitable and Practical Techniques for Goal-Directed Fluid Therapy and Perfusion Pressure During Spine Surgery

Fluid management during spine surgery in the prone position represents a challenge because of decreased right ventricle preload induced by intrathoracic pressure, and consequently decreased stroke volume (SV). The aim of fluid management is to maintain the mean systemic filling pressure (MSFP) within normal levels (7–11 mm Hg) to ensure normal venous return and cardiac output. Of note, venous return equals MSFP minus central venous pressure. Goal-directed fluid therapy (GDFT) is considered the preferred method for maintaining proper intravascular filling without fluid overload. Fluid overload is a major problem that results in increased facial edema, delayed postoperative extubation, and even increased length of hospital stay.

During spine surgery, GDFT can be easily conducted using either stroke volume variation (SVV) or pulse pressure variation (PPV) and can be measured by pulse contour analysis using the Flo Trac/Vigileo system. The other technique for measuring SVV is measuring dynamic changes in descending aortic blood flow and SV by esophageal Doppler probe. The Doppler probe is used to guide the fluid boluses to maintain a corrected flow time greater than 0.35 s or to keep giving fluid boluses as long as SV continues to increase by more than 10%. However, in the PPV method for GDFT the fluid boluses are usually given when PPV is greater than 15% (sensitivity = 100%, specificity = 80%). The ratio of PPV/SVV is called dynamic arterial elastance and is used as indication for the vascular resistance. If the dynamic arterial elastance is 0.9 or over, which indicates intact vascular resistance, a fluid bolus could be used to increase the blood pressure. However, if the dynamic arterial elastance is 0.9 or under, which indicates vasodilation or vasoplegia, then the use of vasopressors is the preferred solution to increase the blood pressure.

In our practice, we prefer to use an esophageal Doppler probe or PPV to guide fluid management during spine surgery, especially in the prone position. It should be remembered that tidal volume should be 8 to 10 mL/kg during the measurement period. If the blood pressure remains lower than the required target even after fluid supplementation, we administer vasopressors or inotropes to reach the target blood pressure. Of note, the use of vasopressors constricts the splanchnic circulation and thereby enhances the MSFP. In addition, vasopressors constrict the precapillary sphincter and decrease the perfusion pressure in the capillaries, thereby reducing filtration pressure and tissue edema ( Table 82.3 ).

Table 82.3

Comparison Between Pulse Pressure Variation and Stroke Volume Variation

Pulse pressure variation (PPV)	PPV is calculated from the mean values of four minimum and maximum pulse pressures (PP) averaged during the previous 30 s	$PPV = {PP}_{max} - {PP}_{\min} / \frac{1}{2} ({PP}_{\max} + {PP}_{\min}) \times 100$ $PPV = {PP}_{max} - {PP}_{\min} / \frac{1}{2} ({PP}_{\max} + {PP}_{\min}) \times 100$
Stroke volume variation (SVV)	SVV is calculated from the mean values of four minimum and four maximum stroke volumes (SVs) averaged during the previous 30 s	$SSV + {SSV}_{max} - {SS}_{\min} / \frac{1}{2} ({SSV}_{\max} + {SS}_{\min}) \times 100$ $SSV + {SSV}_{max} - {SS}_{\min} / \frac{1}{2} ({SSV}_{\max} + {SS}_{\min}) \times 100$

Collectively, the aim of fluid management is to maintain normovolemia and proper tissue perfusion, while avoiding fluid overload.

Commonly Used Colloids During Spine Surgeries

Albumin

Albumin, a natural plasma protein with a molecular weight of 69,000 KDa, accounts for the greatest proportion of plasma COP. For preparations commonly used in clinical practice, albumin is isolated from pooled human plasma and has been considered to be the gold standard solution for fluid resuscitation in the critically ill population.

This intrinsic effect of albumin is most likely based on its electrostatic binding properties. The charges exposed by the molecules forming the EG are mainly negative, whereas an albumin molecule carries not only negative charges (carboxylate groups), but also positive charges (arginine, lysine) at physiological pH. Therefore, the presence of positive charges in albumin enable it to attach to the EG and promote ESL integrity. It has been shown in an isolated perfused heart model that providing albumin to the endothelium before and after ischemia maintained vascular integrity during reperfusion and alleviated the development of tissue edema. In contrast, other synthetic colloid molecules, which expose only negative charges on their surfaces, are not able to maintain the integrity of the ESL and the vascular barrier like albumin. Therefore, using albumin instead of synthetic colloids better reduces interstitial tissue edema. The EG prefers albumin for evoking NO-mediated coronary dilatation. Therefore, albumin helps to reverse stagnation, thrombosis, and corpuscular adherence within cortical venules in the reperfusion phase after focal brain ischemia. ^, One metaanalysis included 17 studies that randomized 1977 participants with sepsis; the use of albumin to resuscitate patients with sepsis was associated with lower mortality compared with other fluid resuscitation regimens.

You're Reading a Preview

Become a Clinical Tree membership for Full access and enjoy Unlimited articles

Become membership

If you are a member. Log in here