The Site of Action of Epidural Fentanyl

Abstract

In Response: We thank Candan and Arslan for their interest in our study (1), which addressed a long-standing controversy in the scientific literature. Twenty-two published clinical and human volunteer studies are divided into approximately equal camps supporting either systemic or spinal mechanisms for epidural fentanyl analgesia. When we analyzed these studies according to their modes of drug administration, we found that almost all the studies that administered epidural fentanyl as a continuous infusion demonstrated systemic analgesia and almost all the studies that administered epidural fentanyl as a bolus demonstrated segmental or spinal analgesia. This finding was independent of the volume of drug dilution. We agree with Candan and Arslan that bolus administration followed by an infusion is often employed in a clinical setting. However, the objective of this study was to examine the hypothesis that epidural fentanyl administered as a bolus elicits segmental analgesia by predominantly acting at the level of the spinal cord, whereas epidural fentanyl administered as a continuous infusion elicits nonsegmental analgesia by predominantly acting at supraspinal sites. The results from studies that administered mixed bolus and infusion protocols were variable and did not fall into a predictable pattern. The simultaneous administration of a bolus and an infusion is not as well suited to address this question, as is the administration of a bolus or an infusion regime alone. The major finding of our study was that epidural fentanyl at a dilution commonly used in clinical practice resulted in segmental analgesia when injected as a bolus but caused nonsegmental analgesia when administered as a continuous infusion. Our study did not attempt to address the question why the site of action may be different when comparing bolus injection with infusion techniques. We commented that “we can only speculate about the underlying mechanism explaining our findings” and offered the different concentration gradients of fentanyl between the epidural and intrathecal space after bolus injection and during infusion as a conceivable one. Candan and Arslan, offer an alternative plausible explanation. They suggest that the total amount of drug administered per unit of time may be more important than the concentration gradient for determining whether epidural fentanyl acts predominantly at spinal or supraspinal sites. While we agree that there may have been merit in using identical dilutions of fentanyl when comparing bolus and infusion administration, this is likely to be of minor importance. They point out, as we did in the Discussion in our article, that several earlier studies administered epidural fentanyl bolus in dilutions ranging between 5–25μg/mL and demonstrated segmental analgesia in all cases; in our infusion group, we administered drug in 3–10 μg/mL dilutions and demonstrated equal spinal and supraspinal analgesia. While it may be interesting to speculate as to whether the concentration gradient or the total amount of drug injected per unit of time determine the predominant site of action of epidurally injected fentanyl, it is methodologically more relevant to study an epidural bolus versus infusion regime that results in similar fentanyl plasma concentrations. In the editorial accompanying our article, Mather and Cousins (2) raised doubts about one of our conclusions, i.e., that an epidural infusion of fentanyl resulted in a predominant supraspinal drug effect. They pointed out that the plasma concentrations resulting from the infusion regime were significantly larger than those resulting from the bolus regime and attributed this to a threefold difference in total fentanyl dose over the study period in the infusion group. However, this was only true when comparing the small-dose bolus with the small-dose infusion and the large-dose bolus with the large-dose infusion regime used in our study (30 μg bolus versus 30 μg/h over 210 min and 100 μg bolus versus 100 μg/h over 200 min). We designed the dosing protocol such that the large-dose bolus and the small-dose infusion received an identical dose of fentanyl over the same period of time. Interestingly, higher plasma concentrations were observed following the large-dose bolus (AUC 60.8 ± 40.8 ng/mL/min, Cmax 0.47 ± 0.39 ng/mL) rather than during the small-dose infusion regime (AUC 32.0 ± 14.6 ng/mL/min, Cmax 0.26 ± 0.11 ng/mL). Notwithstanding, in this comparison, bolus administration still resulted in segmental analgesia, whereas the infusion regime resulted in nonsegmental analgesia. This finding, discussed in the original study, provides further support for our hypothesis, yet was obtained under conditions addressing the methodological concerns raised in accompanying editorial. Yehuda Ginosar, MBBS Ed Riley, MD Martin Angst, MD Department of Anesthesiology and Critical Care Medicine Hebrew University Hadassah School of Medicine Jerusalem, Israel [email protected]