Continuous Spinal Anesthesia with Invasive Hemodynamic Monitoring for Surgical Repair of the Hip in Two Patients with Severe Aortic Stenosis

Abstract

Central neuroaxial blockade has been contraindicated in patients with severe aortic stenosis (AS) [1-4]. Many clinicians avoid using spinal anesthesia in patients with AS because the peripheral sympathetic blockade produced by regional anesthesia can rapidly cause a marked decrease in systemic vascular resistance (SVR) with decreased venous return to the heart and coronary perfusion pressure. In AS, chronic obstruction to left ventricular ejection results in concentric ventricular hypertrophy rendering the heart susceptible to myocardial ischemia, even in the absence of coronary artery disease. Large decreases in SVR, therefore, should be avoided to prevent the catastrophic cycle of hypotension-induced ischemia, subsequent ventricular dysfunction, and worsening hypotension. Indeed, hypotension-induced ischemia with resultant ventricular dysfunction has been described in patients with left ventricular out-flow tract obstruction receiving spinal anesthesia [5]. Recently, several authors have reported greater hemodynamic control achieved with continuous spinal anesthesia (CSA) over epidural or single-dose spinal anesthesia in healthy patients [6,7]. We describe now the use of CSA with invasive hemodynamic monitoring in two patients with critical AS undergoing surgery on the lower extremities. Case Reports Case 1 An 84-yr-old man presented to the operating room for urgent removal and revision of an infected left total hip prosthesis. His past medical history was significant for severe AS complicated by congestive heart failure (New York Heart Association class III). A preoperative echocardiogram revealed an aortic valve area < 0.8 cm (2), an ejection fraction of 25%, severe global hypokinesis, and moderate mitral regurgitation. The patient also had a history of hypertension, atrial fibrillation (AF), and noninsulin-dependent diabetes. His preoperative medications included digoxin, furosemide, enalapril, nitroglycerin patch, and glyburide. He had no known drug allergies. On physical examination, the patient was 180 cm in height and weighed 62 kg. Heart rate (HR) and blood pressure (BP) were 74 bpm and 120/70 mm Hg, respectively. Head and neck evaluation was significant for a Mallampati class II airway and decreased range of motion of the neck. The cardiac examination revealed an irregularly irregular heart beat with a grade 3/6 systolic murmur. The lungs were clear to auscultation. The electrocardiogram (ECG) revealed AF with a ventricular rate of 84 bpm, Q waves in leads V1-V3, and a nonspecific intraventricular conduction delay. Pertinent laboratory values included a hematocrit of 41% and normal coagulation status. Intravenous midazolam (1 mg) and fentanyl (50 micro gram) were given 30 min prior to induction of anesthesia. In addition to standard monitoring, a radial and pulmonary artery (PA) catheter were inserted. While carefully monitoring central venous and pulmonary artery pressures (PAPs), lactated Ringer's solution (1000 mL) was given prior to initiation of spinal anesthesia. A 17-gauge Touhy epidural needle was then inserted at the L3-4 interspace into the subarachnoid space. A 20-gauge catheter was inserted through the needle 3 cm into the subarachnoid space and secured to the back. The patient was repositioned supine, and after confirmation of aspiration of cerebral spinal fluid (CSF) through the catheter, 2.5 mg of plain bupivacaine 0.5% was injected into the subarachnoid space. After reassessment of systemic BP, PAP, systemic vascular resistance (SVR), cardiac output (CO), and sensory level, a second dose of 2.5 mg of plain bupivacaine 0.5% was given 5 min later. A L1 sensory level of anesthesia was obtained 5 min after the second dose, and a third dose of 2.5 mg of plain bupivacaine 0.5% was injected after reassessment of hemodynamic variables again revealed stability. A T8 sensory level of anesthesia was obtained at this time, and the patient was positioned for surgery. (Table 1) displays hemodynamic variables monitored throughout the procedure. Additional doses of 2.5 mg of plain bupivacaine 0.5% were given 90, 135, and 180 min after the initial doses to maintain surgical anesthesia. In addition, the patient was lightly sedated with intermittent intravenous (IV) doses of midazolam 0.5 mg (total of 7 mg for the 4-h procedure). Both the sensory level of analgesia and vital signs remained stable until the conclusion of the operation. The patient received a total of 1900 mL of lactated Ringer's solution and 1 U of packed red blood cells during the procedure. Estimated blood loss was 700 mL.Table 1: Hemodynamic Variables for Patients No. 1 and No. 2The spinal catheter was removed immediately postoperatively, and the patient transported to the recovery room awake, alert, and comfortable. Postoperative pain was managed successfully with patient-controlled analgesia using morphine. After an uneventful postoperative course, the patient was discharged home 5 days later. Specifically, there were no cardiac or postdural puncture headache complications, and the patient was satisfied with the anesthetic management of his case. Case 2 An 84-yr-old woman presented to the operating room for urgent open reduction and internal fixation of a left intertrochanteric femur fracture. Her past medical history was significant for severe AS complicated by congestive heart failure (New York Heart Association class III). Preoperative echocardiogram and cardiac catheterization revealed a peak aortic transvalvular gradient of 84 mm Hg, left ventricular hypertrophy, an ejection fraction of 50%, and moderate mitral regurgitation. The coronary arteries were normal. In addition, the patient had a history of angina and recurrent supraventricular tachycardia. The pulmonary history was significant for long-standing tobacco use (>50 pack/yr) and severe chronic obstructive pulmonary disease. Preoperative pulmonary function tests revealed a forced vital capacity of 0.72 L (28% predicted) and forced expiratory volume in 1 s of 0.65 L (34% predicted). Further medical problems included a history of breast cancer, chronic lymphocytic leukemia, anemia, and peptic ulcer disease. The patient's medications were digoxin and diltiazem. She was allergic to penicillin. On physical examination, the patient weighed 43.6 kg and was 162.5 cm in height. Preoperative BP and HR were 150/70 mm Hg and 80 bpm, respectively. Examination of the head and neck revealed a Mallampati class II airway. Cardiovascular examination was notable for jugular venous distention 10 cm above the sternal notch. The heart had a regular rate and rhythm with a S3 gallop and Grade 3/6 systolic ejection murmur radiating to the axilla. Auscultation of the lungs was notable for bilateral basilar rales with distant breath sounds. Pertinent preoperative laboratory values included a hematocrit of 32.7% and normal coagulation status. ECG revealed normal sinus rhythm with left atrial enlargement, left ventricular hypertrophy, and an age undetermined septal infarct present on previous ECGs. IV fentanyl (50 micro gram) was given 30 min prior to induction of anesthesia. Monitors were applied as described in Case 1. While carefully monitoring central venous pressures and PAPs, lactated Ringer's solution (500 mL) was given prior to initiation of spinal anesthesia. Next, a continuous spinal catheter was placed at the L3-4 interspace in the same manner as described in Case 1. After aspiration of CSF to confirm correct catheter placement, 20 mg of plain 2% lidocaine was injected intrathecally. After reassessment of systemic BP, PAP, SVR, CO, and sensory level, an additional 20 mg of plain 2% lidocaine was injected 5 min after the initial dose. A T10 sensory level of analgesia was obtained at this time and vital signs remained stable. Hemodynamic variables throughout the operation are presented in Table 1. In addition to the continuous spinal catheter, the patient received a total of 1 mg of IV midazolam during the case for sedation. Both the sensory level of analgesia and vital signs remained stable until the conclusion of the operation. The patient received a total of 1200 mL of lactated Ringer's solution during the procedure. Estimated blood loss was 200 mL. Prior to transfer from the operating room bed to the intensive care unit (ICU), the patient had an episode of supraventricular tachycardia which was treated with 6 mg of IV adenosine. This hemodynamically unstable rhythm resolved immediately and converted promptly to sinus rhythm. The patient was admitted to the ICU as preoperatively planned for overnight monitoring. While in the ICU, postoperative pain was well controlled via the continuous spinal catheter with the patient receiving a total of 0.1 mg of Duramorph Registered Trademark (Elkins-Sinn, Inc., Cherry Hill, NJ) in two divided doses. The spinal catheter was removed the following morning. The patient was subsequently discharged 7 days later from the hospital without complications and reported being satisfied with her anesthetic management. Discussion General anesthesia is often preferred over central neuroaxial blockade for the anesthetic management of patients with AS. This choice is often made because peripheral sympathetic nervous system block produced by regional anesthesia can lead to an undesirable decrease in SVR. Indeed, when regional anesthesia is selected, epidural rather than spinal anesthesia is often recommended due to the more likely gradual onset of peripheral sympathetic nervous system blockade [8-10]. We contend that CSA offers many of the advantages of epidural anesthesia. With the appropriate invasive monitoring, the onset of peripheral sympathetic block develops in a gradual and controlled fashion using CSA. Indeed, Klimscha et al. [7] showed that CSA in healthy elderly patients resulted in a significantly lesser decrease in BP and lower incidence of vasopressor use when compared to age-matched controls receiving continuous epidural anesthesia. However, both Schnider et al. [6] and Klimscha et al. [7] point out the maximum hemodynamic effects produced by local anesthetic injection may not occur for up to 20 min after CSA or epidural injection. In the two previous reports, Schnider et al. [6] waited 6 min before reinjection of local anesthetic, whereas Klimscha et al. [7] waited 25 min to reinject. Although we used a shorter dosing interval (5 min) and a smaller initial dose of local anesthetic successfully without complication, further study is required to determine the best dosing criteria and interval. Instead of using a fixed dose or time interval, systemic BP, PAP, SVR, CO, and sensory level should be assessed prior to reinjection of local anesthetic. CSA offers the additional advantage over epidural anesthesia in that catheter placement is technically easier and aspiration of CSF provides confirmation of correct catheter placement. The continuous spinal catheter also has the similar advantage to an epidural catheter in that it may be left in place postoperatively for pain management, minimizing the need for systemic opiates and their attendant risks. However, the potential higher incidence of respiratory depression with spinal versus epidural opiate administration should not be over-looked [11]. CSA avoids many of the disadvantages of general anesthesia. In contrast to general anesthesia, use of a continuous spinal catheter allows patient communication of subjective feelings of distress throughout the operation. In addition, the hemodynamic perturbations of direct laryngoscopy and intubation are avoided with CSA. Moreover, the use of volatile anesthetics in patients with AS may lead to myocardial depression, peripheral vasodilation, and loss of normal atrial systole. Likewise, CSA obviates the need for neuromuscular blockade, which may lead to undesirable fluctuations in HR. In addition, a regional anesthetic technique, such as CSA, in a patient with coexisting chronic obstructive pulmonary disease avoids complications associated with positive pressure ventilation and possible infectious risks associated with intubation [12]. However, CSA does have potential complications. First, CSA should be used with caution in patients in whom a difficult endotracheal intubation is anticipated. The potential need for emergent control of the airway must be balanced with the risks of awake, sedated fiberoptic intubation under controlled circumstances prior to induction of anesthesia. Second, peripheral sympathetic nervous system block produced by CSA may be deleterious in situations of profound blood loss. This is especially true in the setting of AS where precipitous decreases in SVR can lead to the catastrophic cycle of hypotension-induced ischemia, subsequent ventricular dysfunction, and worsening hypotension. Third, AS is often complicated by global ventricular hypokinesis and AF. These patients are often anticoagulated and CSA would be contraindicated. Finally, many of the complications associated with single-dose spinal anesthesia including postdural puncture headache, persistent paresthesia, low back pain, and risk of infection also apply to CSA. We think the performance of CSA in patients with severe AS is best guided using a PA catheter. In addition to being able to monitor volume status, a PA catheter allows early diagnosis of pulmonary hypertension, ischemia, and right or left ventricular failure [13]. Due to the decreased left ventricular (LV) compliance as well as the increased LV end-diastolic pressure and volume, adequate preload is essential in the patient with AS to maintain CO. It is our practice in patients with normal cardiac function to hydrate with 10-15 mL/kg of crystalloid prior to induction of central neuroaxial blockade. In patients with severe LV dysfunction, we use invasive hemodynamic monitoring to guide prehydration. If while attempting to hydrate these patients with administration of 10-15 mL/kg of crystalloid, the pulmonary capillary wedge pressure (PCWP) increases by more than 3-4 mm Hg, the infusion is stopped. The patient in Case 1 received 1000 mL of crystalloid with no change in PCWP, while the crystalloid administered in Case 2 was terminated after 500 mL due to an increase in PCWP. Furthermore, a PA catheter may be extremely helpful in guiding treatment of hypotension in the patient with AS. If the hypotension is associated with decreased central venous pressure and PCWP, we would first administer crystalloid or colloid in an attempt to maintain the PCWP similar to that obtained preoperatively. If no immediate improvement in the cardiac filling pressures is obtained, a vasoconstrictor such as phenylephrine might be added to increase SVR, preload, and coronary perfusion. Ephedrine should be used cautiously in the patient with AS, as the resultant tachycardia may precipitate myocardial ischemia. On the other hand, if hypotension occurs in a nonischemic patient with a PCWP unchanged from that obtained preoperatively, or in association with an increased PCWP, then inotropic drugs should be given. With the aide of invasive hemodynamic monitoring, we were able to successfully induce and maintain CSA in controlled fashion while maintaining control of the cardiac filling pressures. In this manner, we were successful in avoiding administration of any cardiovascular inotropes or vasopressors. Our experience shows CSA to be a safe anesthetic technique with minimal hemodynamic disturbances. Although general anesthesia is historically considered the anesthetic of choice for patients with AS, CSA is an attractive alternative for the management of surgery on the lower extremities when used with appropriate invasive monitoring.