Epinephrine has been known for decades as a potentially useful additive in epidural anesthesia. Because of its vasoconstrictive properties, systemic absorption of other components of the epidural solution—typically local anesthetics and opioids—is reduced ^(1,2). This is expected both to decrease systemic toxicity and to enhance the concentration of local anesthetics and opioids at the neural targets. The hypoalgesic effect of epidural epinephrine in humans has been demonstrated in a placebo-controlled study ⁽³⁾.

Despite its potential usefulness as a component of epidural mixtures, epinephrine is used in very few institutions for postoperative pain management. The likely reasons are twofold: the paucity of high quality clinical trials and the equivocal results in the literature. Interestingly, these factors also may have affected the use of epinephrine as an adjuvant to local anesthetics for intraoperative analgesia. Despite decades of clinical use, the published evidence is so sparse that a systematic review was unable to produce any clear results ⁽⁴⁾.

Clinical Trials

In this context, the article by Niemi and Breivik ⁽⁵⁾ in this issue of Anesthesia & Analgesia is important. A thoracic epidural infusion of ropivacaine 1 mg/mL, fentanyl 2 μg/mL, and epinephrine 2 μg/mL was administered after major thoracic and abdominal surgery. On the first postoperative day, the mixture was replaced by one of two study solutions: the above triple solution or the same one without epinephrine. After 3 h, the study mixture was replaced by the ropivacaine-fentanyl-epinephrine mixture that was initially infused. The same procedure was applied on the second postoperative day, whereby the alternative study solution was infused. The results are impressive. The median pain intensity during cough on the visual analog scale (0–100) remained at approximately 20 during infusion of the triple solution but increased to 80 during infusion of the solution without epinephrine. The median pain score at rest was 0 and more than 50 during infusion of the solution with and without epinephrine, respectively. Patients were more often out of bed and had less nausea during infusion of the triple solution.

Earlier, Niemi and Breivik ⁽⁶⁾ conducted an investigation applying the same design as the present study, whereby bupivacaine instead of ropivacaine was used. The same striking results were obtained. The advantage of adding epinephrine to a bupivacaine-fentanyl solution was also demonstrated in a randomized controlled trial, in which two solutions (with and without epinephrine) were administered in two different groups of patients after abdominal surgery ⁽⁷⁾. Smaller pain scores were found even after 48 h. The incidence of side effects did not differ between the groups. The addition of epinephrine to fentanyl for thoracotomy pain decreased fentanyl consumption and plasma fentanyl concentrations ⁽²⁾. However, pain control and side effects were not affected. When added to a sufentanil bolus after abdominal surgery, epinephrine prolonged analgesia, improved respiratory function, and reduced sedation ⁽⁸⁾.

Although the above studies on thoracic epidural analgesia mostly show a beneficial effect of epinephrine, investigations on lumbar epidural analgesia are less consistent. Cohen et al. ⁽⁹⁾ found no benefit of adding epinephrine 0.5 μg/mL to a bupivacaine-fentanyl infusion after cesarean delivery. The addition of 1.0 μg/mL epinephrine to a bupivacaine-fentanyl combination reduced the infusion rate without affecting pain scores and side effects ⁽¹⁰⁾. When added to pethidine in a patient-controlled epidural regimen after cesarean delivery, epinephrine had no effect on pain scores and increased nausea and pruritus ⁽¹¹⁾. Epinephrine prolonged the duration of analgesia when added to a bolus of diamorphine ⁽¹²⁾, hydromorphone ⁽¹³⁾, morphine ⁽¹⁴⁾, and pethidine ⁽¹⁵⁾. The incidence of side effects was either unaffected ^(13,15) or increased ^(12,14).

Epinephrine per se, at the doses used in epidural mixtures, has not been found to produce side effects. As discussed by Niemi and Breivik ⁽⁵⁾, concerns regarding possible spinal cord ischemia are not supported by the published evidence. Epinephrine-containing solutions are stable for 6 months at 4°C and for 4 days at 22°C ⁽¹⁶⁾.

Pharmacological Aspects

Despite the striking results of the investigation by Niemi and Breivik ⁽⁵⁾, the published data do not consistently show that epinephrine is useful in postoperative epidural mixtures. However, the equivocal results may be explained by analyzing some pharmacological aspects of epidural epinephrine and its interaction with the other components of the epidural mixture.

Epinephrine-induced hypoalgesia ⁽³⁾ is probably explained by absorption into the cerebrospinal fluid and binding to α₂-adrenoreceptors ⁽¹⁷⁾. However, epidural epinephrine is also absorbed systemically ⁽¹⁸⁾ and metabolized by the meninges ⁽¹⁹⁾, which limits its spinal availability. Spinal availability is probably further reduced when epinephrine is administered at a lumbar level as a result of the increased distance between site of administration and spinal cord receptors. This factor is probably less important after thoracic infusion, where the distance between site of administration and spinal cord receptors is decreased. If binding to spinal cord receptors is an important determinant of epinephrine effects, limited spinal availability after lumbar administration could partly explain the lack of effect in some investigations ^(9–11). The effective concentration of epinephrine for lumbar epidural mixtures is not known.

The expected effects of epinephrine-induced vasoconstriction are reduced systemic absorption and increased spinal availability of the other components of the epidural mixture. Hydrophilic opioids, such as morphine, are minimally absorbed systemically, and a large proportion of drug is available for the spinal cord ⁽²⁰⁾. The opposite is true for hydrophobic opioids, such as fentanyl ⁽²⁰⁾. Epinephrine markedly decreases fentanyl plasma concentrations ^(2,6), but has little or no effect on morphine plasma concentrations ⁽²¹⁾. Therefore, the effect of epinephrine-induced vasoconstriction on the ratio spinal/systemic availability may be greater for hydrophobic than for hydrophilic opioids. When added to bupivacaine and epinephrine, fentanyl 2 μg/mL is highly effective and associated with a very infrequent incidence of side effects ⁽²²⁾. Moreover, its plasma concentration remains less than the mean minimum effective analgesic concentration ⁽⁶⁾. This suggests that the presence of epinephrine in a triple mixture renders fentanyl concentrations of 2 μg/mL sufficient to produce effective analgesia. Whether this is the case for other opioids remains unclear.

Opioid-related side effects are caused by both systemic and spinal action. The net impact of epinephrine is probably the result of the balance between induced changes in plasma and cerebrospinal fluid concentrations of the infused opioid, which in turn depends on the chemical characteristics of the opioid administered. This may explain the different influence of epinephrine on side effects in the published literature. It is conceivable that an increase in spinal concentration without significant decrease in plasma concentration could lead to enhanced side effects ^(11,12,14).

The epinephrine-induced changes in systemic and spinal availability of the opioid are likely to be affected also by the opioid concentration in the epidural solution. The relationships between opioid concentration in epinephrine-containing mixtures and concentrations of opioid in plasma and spinal tissues are unclear. The two curves are likely to be very different, so that the resulting dose-response curve in relation to clinical effects is probably complex. It is possible that epinephrine results in changes in opioid plasma and spinal concentrations that may determine improved, unchanged, or even worsened clinical outcomes, depending on the opioid concentration used. The au-thor is not aware of any clinical study investigating the optimal opioid concentration in epinephrine-containing mixtures.

An additional kinetic aspect of epidural opioids needs to be mentioned. One hour after discontinuation of an epinephrine-containing mixture, plasma concentrations of fentanyl increased by 32%⁽⁹⁾. More marked increases in plasma levels after discontinuation were observed for buprenorphine ⁽¹⁰⁾ and sufentanil ⁽²³⁾. This indicates that epinephrine-induced vasoconstriction produces accumulation of opioids at spinal tissues, with subsequent washout and transient increase in systemic absorption when administration is discontinued. The rate of absorption after discontinuation probably depends on the amount of opioid stored in the spinal tissues, which in turn depends on the chemical characteristics and the amount of opioid previously infused. No increase in side effects was observed after discontinuation of the fentanyl-containing mixture ⁽⁹⁾, whereas a large proportion of patients receiving sufentanil experienced lightheadedness and dizziness ⁽²³⁾. There is no report of complications after discontinuation of epinephrine-containing mixtures. However, the possibility of increased side effects should be considered. This drawback is likely to be minimized or eliminated by using small opioid doses and reducing the infusion rate stepwise, rather than by discontinuing the administration abruptly.

Epinephrine is equally effective when added to bupivacaine ⁽⁶⁾ and ropivacaine ⁽⁵⁾, indicating that ropivacaine-induced vasoconstriction does not minimize the benefits of using epinephrine. As for opioids, chemical characteristics and dose and site of administration of the local anesthetic may be important determinants of the effect of epinephrine. Unfortunately, these issues are largely unexplored.

Conclusions

There is now convincing evidence that adding epinephrine to small concentrations of a local anesthetic and fentanyl for thoracic postoperative epidural analgesia is beneficial. However, the results obtained with these mixtures may not be generalized to solutions containing other opioids or administered in the lumbar epidural space. There is a need for studies investigating the dose-effect relationship of epinephrine, particularly in lumbar epidural mixtures, and the kinetics of different opioids and local anesthetics at different doses in epinephrine-containing mixtures.