Pain Control

The most direct beneficial effects of adequate postoperative analgesia include prevention of unnecessary patient discomfort and improved patient satisfaction. In addition, adequate postoperative analgesia may decrease morbidity and postoperative hospital length of stay and, consequently, may decrease cost. Conversely, inadequate postoperative analgesia may increase morbidity by causing hemodynamic, metabolic, immunologic, and hemostatic alterations. Thus aggressive control of postoperative pain may improve some or all of these variables as well as outcomes in high-risk patients after both noncardiac and cardiac surgery. However, achieving optimal pain relief after cardiac surgery may be challenging both because of significant incisional pain (e.g., sternotomy, thoracotomy, and chest tubes) and patient characteristics (e.g., advanced age and pulmonary pathology).

Postoperative analgesia may be attained with a wide variety of techniques (e.g., local anesthetic infiltration, nerve blocks, opioids, nonsteroidal antiinflammatory drugs, or alpha-adrenergic drugs). Traditionally, parenteral opioid administration has been used as first-line therapy for postcardiac surgery pain control. Intravenous opioid use is associated with definite detrimental side effects, most notably respiratory depression and sedation, which may be magnified in the adult cardiac surgical population. Intrathecal and epidural techniques clearly produce reliable analgesia in patients undergoing cardiac surgery and during the last 10 to 15 years have, in fact, been used more often clinically. The use of neuraxial techniques may also help avoid some of the aforementioned side effects encountered with parenteral narcotic administration, particularly if epidural analgesia with local anesthetic is used.

An important potential advantage of the dense analgesia afforded by neuraxial techniques is attenuation of the stress response. Pain, as well as its concomitant sympathetic activation, induces the stress response in surgical patients. Elevated levels of circulating catecholamines can cause unfavorable myocardial oxygen supply and demand profiles by both increasing demand secondary to elevated heart rate and contractility and limiting supply secondary to decreased coronary perfusion time. Increased levels of corticotropin-releasing hormone and other stress hormones and inflammatory mediators may be seen in surgical patients with inadequate analgesia. In addition, these mediators likely play a role in the prothrombotic and immunosuppressed states observed in postsurgical patients. In cardiac surgical patients this stress/inflammatory response may be further compounded and amplified by the humoral response to cardiopulmonary bypass. None of these alterations is desirable in the postcardiac surgery patient, and they have been linked to postoperative organ dysfunction and cardiac events. Theoretically, these deleterious effects of perioperative pain may be attenuated by aggressive use of neuraxial analgesia.

Cardiac Sympathectomy

Intimately related to spinal/epidural analgesia’s attenuation of the pain-induced stress response is the cardiac sympathectomy induced when neuraxial local anesthetics are used. The myocardium and coronary vasculature are densely innervated by thoracic sympathetic nerve fibers that arise from T1 to T5 and profoundly influence total coronary blood flow and distribution. Cardiac sympathetic nerve activation initiates coronary artery vasoconstriction and paradoxic coronary vasoconstriction in response to intrinsic vasodilators. In patients with coronary artery disease, cardiac sympathetic nerve activation disrupts the normal matching of coronary blood flow and myocardial oxygen demand. Furthermore, myocardial ischemia initiates a cardiocardiac reflex mediated by sympathetic nerve fibers, which augments the ischemic process. Cardiac sympathetic nerve activation likely plays a central role in initiating postoperative myocardial ischemia by decreasing myocardial oxygen supply while increasing myocardial oxygen demand. Neuraxial block using local anesthetics produces sympatholysis and prevents many of these undesirable sympathetic effects. In fact, epidural local anesthetics have been shown to attenuate the stress response and improve coronary arteriolar flow secondary to this sympatholytic effect.

Pulmonary Function

As less invasive techniques for cardiac surgery have evolved, such as off-pump coronary artery bypass grafting (CABG), heartport, and robotic procedures, there has been a push toward streamlining postoperative care with a focus on anesthetic techniques that permit early postoperative extubation. Early extubation after cardiac surgery decreases the length of intensive care unit (ICU) stay, decreases the length of the hospital stay, and, consequently, decreases the cost of care. Multiple studies have demonstrated that fast-tracking is at least as safe as traditional perioperative management, including a meta-analysis of 10 trials published in 2003. It is important to note, however, that fast-tracking after cardiac surgery has not been shown to improve outcomes. A 2003 meta-analysis showed no difference between fast-track (extubation earlier than 8 hours after surgery) and traditional approaches in mortality rates, myocardial ischemia, or respiratory dysfunction. This study did confirm that fast-tracking decreases length of ICU and hospital stays. Neuraxial anesthesia offers the potential advantage of dense analgesia with less of the sedative and respiratory depressive side effects that can delay extubation. In fact, perhaps the greatest impetus behind the recent interest in neuraxial techniques has been this focus on so-called fast-track anesthesia for cardiac surgery.

Neuraxial analgesia may also help preserve perioperative pulmonary function through opioid sparing, early extubation, and early mobilization. Thoracic epidural anesthesia in patients undergoing major abdominal and noncardiac thoracic surgery has been shown to result in improved postoperative lung function and has been shown to reduce the risk of pulmonary complications.

Hemodynamic Effects

The hemodynamic consequences of neuraxial analgesia vary considerably with the selected technique and agent. When local anesthetics are used, both in the epidural and intrathecal spaces, sympathetic blockade results in decreased heart rate, peripheral vasodilation, venous pooling, and arterial hypotension. These effects are most pronounced with epidural local anesthetic administration (the risk of a total spinal technique precludes the use of intrathecal local anesthetics for patients with advanced cardiac disease) and are mostly absent when narcotics are used alone. Arterial hypotension in CABG patients can cause decreased coronary perfusion and myocardial ischemia. In patients with valvular heart disease or poor ventricular function, both hypotension and bradycardia may cause dramatic drops in cardiac output and, consequently, global hypoperfusion.

Respiratory Effects

High thoracic neuraxial techniques may have negative effects on the respiratory system. When local anesthetics are used in doses sufficient to produce neuraxial anesthesia, intercostal muscle strength may be compromised, which may be clinically significant in patients with pre-existing pulmonary disease.

Neuraxial opioids, on the other hand, may induce respiratory depression. Patients who receive single-shot epidural and intrathecal opioid injections may run the risk of respiratory depression as high as 3% and 7%, respectively. It must be noted, however, that parenteral opioids may carry an even greater risk of causing respiratory depression.

Neurologic Effects

Perhaps the greatest barrier to widespread clinical use of neuraxial analgesia in cardiac surgical patients is the perceived risk of epidural hematoma development. Although results taken directly from cardiac surgery literature are somewhat lacking, there are data regarding the risk of vertebral canal hematoma after neuraxial block placement. A recent prospective study in the United Kingdom estimated the risk of development of epidural hematoma after perioperative (noncardiac) central neuraxial blockade at between 1 : 5700 and 1 : 12,200. This study did not distinguish between lumbar and thoracic block placement, and it is believed by many clinicians that the risk of epidural hematoma is greater with thoracic blocks (although no definitive data are currently available).

In addition, there are several other factors that may raise the risk of epidural hematoma formation in cardiac surgical patients. The majority of patients undergoing cardiac surgery receive full systemic heparinization. Epidural instrumentation in this setting poses a theoretically increased risk of vertebral canal bleeding complications, even when guidelines regarding placement and removal of catheters are adhered to strictly. Furthermore, renal dysfunction, a not infrequently encountered complication of cardiac surgery, as well as the effects of cardiopulmonary bypass, may lead to platelet dysfunction and increased risk of bleeding complications.

Spinal Analgesia

Spinal analgesia for cardiac surgery has several advantages. The use of a small-bore needle may help minimize the risk of a “bloody tap” and epidural hematoma formation. When compared with high thoracic epidural injections, spinal injections are technically easier to perform. In addition, because intrathecal techniques are one-shot injections, postoperative anticoagulation need not be discontinued for catheter removal after surgery.

There are also disadvantages to the intrathecal approach. Select centers have been using a “total spinal” technique for adult cardiac surgical patients, but the administration of sufficient local anesthetic doses to achieve this can result in dramatic drops in both heart rate and blood pressure and has precluded wider clinical use. Given this limitation, spinal administration of opioids has become the more widespread intrathecal approach. Although intrathecal opioids given before surgery provide analgesia that persists into the postoperative period, they have several disadvantages when compared with local anesthetics. Unlike local anesthetics, neuraxial opioids do not produce sympathetic blockade. Thus many of the theoretical advantages already listed do not apply when intrathecal opioids are selected for neuraxial analgesia. Furthermore, as already described, intrathecal opioids can cause significant respiratory depression in up to 7% of patients. Another disadvantage of spinal analgesia when compared with epidural techniques is the inability to continue to provide supplemental analgesia after the effects of the initial injection dissipate.

Epidural Analgesia

Although in theory a single-shot epidural injection for analgesia is feasible for cardiac surgical patients, placement of an indwelling epidural catheter and intermittent dosing or continuous infusion of the drug is the most common approach in both the literature and clinical practice. Epidural catheter placement offers the advantage of incremental dosing and, consequently, tighter hemodynamic control. Safe administration of local anesthetics through an epidural catheter allows modulation of sympathetic tone and the stress response in addition to providing excellent analgesia. Epidural catheters can be left in place after surgery, which allows continued administration of local anesthetics and opioids to help control postoperative pain. The risk of respiratory depression after epidural narcotic administration is significantly less than the risk observed after intrathecal opioid injection.

On the other hand, high thoracic epidural catheters can be challenging to place. The risk of a bloody tap and epidural hematoma formation may be significantly higher after thoracic epidural placement than after lumbar spinal injection. In addition, removal of the indwelling catheter after surgery may increase the risk of hematoma formation. Finally, removal of epidural catheters requires cessation of anticoagulants and antiplatelet medications and may put patients taking these medications at risk of thrombotic or embolic events.

Spinal Analgesia

The overwhelming majority of studies in the literature have used intrathecal morphine (administered before induction of general anesthesia) to provide a prolonged tail of postoperative analgesia. Some investigators have used intrathecal fentanyl and sufentanil in conjunction with morphine to provide enhanced intraoperative analgesia. As already mentioned, other investigators have used intrathecal local anesthetics in attempts to provide dense intraoperative analgesia and induce cardiac sympathectomy. Still others have studied the use of intrathecal clonidine in addition to neuraxial opioids. Regardless of the agent and regimen used, these studies demonstrate the efficacy of intrathecal injections in providing postoperative analgesia.

Two early randomized, blinded, placebo-controlled clinical studies underscore the ability of intrathecal morphine to induce significant postoperative analgesia after cardiac surgery. Vanstrum and colleagues prospectively randomly assigned 30 patients to receive either intrathecal morphine (0.5 mg) or intrathecal placebo before induction of anesthesia. Patients who received intrathecal morphine required significantly less intravenous morphine than control subjects receiving placebo during the initial 30 hours after intrathecal injection. Associated with this enhanced analgesia was a substantially decreased need for antihypertensive medications during the immediate postoperative period. Chaney and colleagues prospectively randomly assigned 60 patients to receive either intrathecal morphine (4.0 mg) or intrathecal placebo before induction of anesthesia. The tracheal extubation time was similar in all patients. Patients who received intrathecal morphine required significantly less intravenous morphine than control subjects receiving placebo during the initial postoperative period.

Numerous other nonrandomized clinical investigations (e.g., retrospective and observational) attest to the ability of intrathecal morphine to produce substantial postoperative analgesia in patients after cardiac surgery, the quality of which depends not only on the intrathecal dose administered but also on the type and amount of intravenous drugs used for the intraoperative baseline anesthetic. The optimal dose of intrathecal morphine or other agents for achieving maximum postoperative analgesia with minimum undesirable drug effects is uncertain.

In examining the safety of spinal analgesia for cardiac surgery, a mathematical model for predicting this outcome concluded that the risk is approximately 1 : 10,000. Therefore none of the available studies is sufficiently powered to assess the risk of epidural hematoma. It is important to note that no cases of epidural hematoma have been reported in cardiac surgical patients who have received spinal analgesia. In addition, no epidural hematoma adverse events have been reported in any of the studies investigating intrathecal analgesia for cardiac surgery.

None of the available investigations alone allows us to reach conclusions regarding the clinically relevant advantages (if any) of spinal analgesia over parenteral opioid administration. There have been, however, two meta-analyses of prospective, randomized controlled trials. In 2004, a meta-analysis by Liu and colleagues of 17 trials encompassing 668 patients showed no significant impact of intrathecal analgesia on major clinical endpoints such as mortality, myocardial infarction, time to extubation, and arrhythmias. More recently, Zangrillo and colleagues performing a meta-analysis of 24 trials totaling 1106 patients confirmed these findings and also demonstrated no decrease in hospital length of stay in the spinal analgesia group.

In summary, clinical investigations involving intrathecal techniques indicate that administration of intrathecal morphine before induction of general anesthesia produces reliable postoperative analgesia after cardiac surgery. To date, no known cases of epidural hematoma have developed after spinal analgesia for cardiac surgery, which suggests that this technique is safe. However, two separate meta-analyses have failed to demonstrate any significant clinical benefit to the use of spinal analgesia for cardiac surgery.

Thoracic Epidural Analgesia

Thoracic epidural analgesia with local anesthetics alone or with epidural narcotics has been shown to provide effective postoperative pain control in patients undergoing CABG surgery. The use of thoracic epidural analgesia for pain control has been consistently shown to be narcotic sparing in the postoperative period. Numerous clinical studies attest to this fact.

An early mathematical analysis by Ho and colleagues in patients subjected to systemic heparinization required for cardiopulmonary bypass (without a single episode of hematoma formation reported in the literature as of the year 2000) estimated that the maximum risk may be as frequent as 1 : 2400. As use of neuraxial techniques in cardiac surgical patients has become more prevalent, reports of epidural hematoma in cardiac surgical patients have begun to surface. In 2004, the first report of hematoma formation associated with epidural instrumentation the day before scheduled cardiac surgery was published. The first report of hematoma formation during the immediate postoperative period after cardiac surgery (catheter inserted immediately before surgery after induction of general anesthesia) occurred the same year. A letter to the editor in 2006 details permanent paraplegia in two patients undergoing cardiac surgery with thoracic epidural supplementation and hints at two additional patients who experienced hematoma formation associated with catheter insertion the day before scheduled cardiac surgery. However, the most recent estimation of epidural hematoma risk in the cardiac surgical population is 1 : 12,000, which is consistent with the upper range of risk seen in the general surgical population. In addition, others have argued that the addition of large series reporting neuraxial analgesia for cardiac surgery to the literature without any cases of epidural hematoma within these series suggests that the risk in this population may not be appreciably greater than the risk observed in general surgery patients. It is important to note that, in much of the thoracic epidural literature, the catheters were placed the day before surgery. There are insufficient data to evaluate whether same-day placement of epidural catheters raises the risk of epidural hematoma occurrence.

Whereas hematoma formation is clearly a major concern, thromboembolic complications when normalization of coagulation variables is achieved for epidural removal may also be a consideration. Chaney and Labovsky report a case in which a patient receiving postoperative anticoagulation for atrial fibrillation and a mechanical aortic valve experienced an embolic stroke after normalization of coagulation variables to remove a thoracic epidural catheter placed electively before cardiac surgery. In light of this report, the risk of thromboembolic events in patients requiring anticoagulants or antiplatelet medication postoperatively also needs to be taken into consideration before a decision is made to place epidural catheters in these subjects.

Some studies suggest that the use of thoracic epidural anesthesia may increase the incidence of hypotension requiring vasopressor administration. The ultimate clinical significance, if any, of this hypotension when appropriately managed is unknown.

In the last decade several meta-analyses have suggested that the use of thoracic epidural analgesia for cardiac surgery may be superior to general anesthesia alone. In 2004, Liu pooled data from 15 studies including 1178 patients comparing general anesthesia with thoracic epidural analgesia to general anesthesia alone in patients undergoing CABG surgery. No differences were seen between groups in mortality or myocardial infarction rates, but the meta-analysis did find statistically significant decreases in pulmonary complications, dysrhythmias, pain scores, and time to extubation in the patients who received thoracic epidural analgesia. In 2010, Bignami and colleagues analyzed 33 trials (2336 patients) comparing general anesthesia with thoracic epidural analgesia to general anesthesia alone in patients undergoing cardiac surgery. In their analysis, they found that thoracic epidural analgesia reduced the time to extubation, the risk of renal failure, and the composite endpoint myocardial infarction/death (although when assessed independently there were no differences in myocardial infarction and mortality). In 2011, Svircevic and colleagues published a meta-analysis of 28 articles including a total of 2731 patients: 1416 patients received general anesthesia alone and 1315 patients received general anesthesia with thoracic epidural analgesia. They concluded that thoracic epidural analgesia in patients undergoing cardiac surgery reduces the risk of postoperative supraventricular arrhythmias and respiratory complications. In addition, their data suggested that thoracic epidural analgesia may reduce the risk of mortality, myocardial infarction, and stroke, although the study was insufficiently powered to achieve statistical significance for these endpoints.

In assessing these meta-analyses, several important factors must be considered. First, the control arms of the analyzed studies included patients who received high-dose narcotic regimens. These subjects may have had an exaggerated effect on extubation times and, by extension, the risk of pulmonary complications and may account for the differences in these endpoints detected by the meta-analysis. In addition, all three of the meta-analysis included data from a large study (420 patients) published by Scott in 2001. In this study, 420 patients undergoing cardiac surgery were prospectively randomly assigned (nonblinded) to receive either thoracic epidural bupivacaine/clonidine and general anesthesia or general anesthesia alone (control group). Epidural infusions were continued for 96 hours after surgery (titrated according to need). In control patients, postoperative analgesia was obtained with intravenous opioids. After surgery, striking clinical differences were observed between the two groups. Postoperative supraventricular arrhythmia, respiratory tract infection, renal failure, and acute confusion were all decreased in thoracic epidural patients when compared with control patients. However, several limitations in this study may account for the observed differences. First and foremost, the clinical protocol dictated that beta-adrenergic blockers could not be used during or after surgery for the 5 days of the study period. Because approximately 90% of this study’s patients were taking beta-adrenergic blockers before surgery, this unique perioperative management (discontinuation of beta-adrenergic blockers) clouds interpretation of postoperative supraventricular arrhythmia data. Patients in the epidural group received epidural clonidine, a cardioprotective drug that may also account for the decreased risk of tachyarrhythmias observed in the experimental subjects. Also, despite prospective randomization, substantially fewer patients receiving thoracic epidural catheters were active smokers before surgery when compared with control subjects, which clouds interpretation of postoperative extubation times and respiratory tract infection data. Given the large size of this study relative to the meta-analyses performed, it is possible that some of the favorable outcomes demonstrated by the meta-analyses may be disproportionately due to the included Scott data.

A recent large (654-subject) and well-designed randomized controlled trial by Svircevic and colleagues comparing general anesthesia with epidural analgesia to fast-track (remifentanil-based) general anesthesia showed similar extubation times and composite pain scores and risk of pulmonary complications, cardiac arrhythmias, and myocardial infarction in both groups. The only observed benefit of thoracic epidural analgesia was lower early postoperative pain scores. These data suggest that a regimented fast-track general anesthetic may offer many of the benefits of thoracic epidural analgesia without the additional risk associated with epidural placement.