John R. Blair

Cardiac Electrophysiologic Effects of Fentanyl and Sufentanil in Canine Cardiac Purkinje Fibers

The electrophysiologic effects of high concentrations of the opioid agonists, fentanyl and sufentanil, on isolated canine cardiac Purkinje fibers were studied. Changes in action potential parameters were examined at the following concentrations: fentanyl 94.6 nM, 0.19 microM, and 0.95 microM; sufentanil 8.6 nM, 86.4 nM, 0.17 microM, and 0.26 microM. Naloxone 5.5 microM was administered after maximal changes were induced by fentanyl in order to explore the possibility of an opioid receptor interaction. Action potential parameters measured were Vmax of phase 0, amplitude, overshoot, maximum diastolic potential, action potential duration at 50%, and 90% repolarization and membrane responsiveness. Fentanyl 0.19 microM and sufentanil-0.17 microM caused a significant lengthening of action potential duration at 50 and 90% repolarization, 6.4% and 7.3%, and 10.2% and 12.4%, respectively, P less than 0.05. Other action potential parameters were not significantly affected by the opioids. Naloxone 5.5 microM alone did not alter action potential characteristics and failed to reverse action potential prolongation produced by fentanyl. The authors suggest that fentanyl and sufentanil prolong action potential duration in canine cardiac Purkinje fibers via direct membrane actions.

Measurement of the low-frequency component of blood pressure variability can assist the interpretation of heart rate variability data.

To the Editor:—Dowd et al. have provided important information about the pharmacokinetics of tranexamic acid (TA) in cardiac surgery with cardiopulmonary bypass. Particularly, they demonstrated the necessity of a continuous infusion of TA to obtain stable therapeutic concentrations. Then, they proposed two different dosage schemes in lowand high-risk patients for bleeding, to obtain TA plasma concentrations of 334 M and 800 M, respectively. Considering the patients with low-risk for bleeding, they recommended a loading dose of 12.5 mg/kg (or greater) over 30 min, a continuous infusion of 6.5 mg · kg 1 · hr , and 1 mg/kg (or greater) added to the priming, whereas in patients with high-risk for bleeding they proposed doses about 2.5 higher. My group published various studies proposing an original pharmacologic protocol for TA that seem very similar to that proposed by Down et al. For patients with low-risk for bleeding; that is, a loading dose of 1 g over 20 min before sternotomy (and not 1 g, 20 min before sternotomy, as erroneously reported in the work of my group cited by Dowd et al.), followed by a continuous infusion of 400 mg/h, and 500 mg added to pump priming. We also applied the same protocol in high-risk patients for bleeding, obtaining a significant reduction of blood loss and allogeneic transfusions. One criticism of the study of Dowd et al. is that the need to increase the doses of TA in this type of patient requires further clinical demonstrations, particularly considering (as the same authors report) that TA plasma concentrations of about 200 M completely inhibit fibrinolysis. Concerning the administration of TA after surgery, I agree with Dowd et al. regarding benefits in the postoperative period depending on intraoperative dosing techniques, but I do not understand why the authors claim that the efficacy of prolonged TA administration in the postoperative period is an open question. In reality, the authors, applying their pharmacokinetic model to our TA protocol, confirmed the conclusions of our study. Our intraoperative TA dosage scheme guarantees therapeutic concentrations for about 12 h, rendering unnecessary postoperative infusion of the drug. It also seems that the potential thrombotic risk intrinsic to antifibrinolytic drugs is underestimated by Down et al. I particularly disagree with their statement, “TA appears to be a very safe medication. . .. Thus our attempt to avoid excessive TA concentrations may not be necessary.” An extensive literature search showed cases of thrombosis following the use of hemostatic drugs such as -aminocaproic acid, aprotinin, and TA, but it would be sufficient to cite a recent case report describing two fatal cases of thrombosis after the use of -aminocaproic acid (very similar to TA with a potency 10 times lower) in cardiac surgical patients operated on with deep hypothermic circulatory arrest, wherein postmortem laboratory analysis revealed the presence of Factor V Leiden. Because it is impossible to identify patients with a preoperative prothrombotic state, it is appropriate to use the minimal effective doses of hemostatic drugs to avoid amplifying these thrombotic complications. Furthermore, the same authors’ group reported in a previous study three cases of stroke in cardiac surgical patients with known peripheral vascular disease, treated intraoperatively with high doses of TA. One would speculate that high concentrations of TA facilitated the formation of a thrombus in the presence of blood flow reductions in a diseased vessel. In conclusion, only large, prospective, blinded studies will establish the real safety and efficacy of the various doses of tranexamic acid in cardiac surgery. Currently, caution is required when using a hemostatic drug.

What is the incidence of inadvertent dural puncture during epidural anesthesia in obstetrics

We thank Harvey and Cave for their comments and appreciate their concerns. As they note, the mechanism(s) by which lipid can reverse local anesthetic systemic toxicity has yet to be fully defined, but it is generally accepted that the predominant effect results from drawing anesthetic from the plasma phase, reducing its effective concentration at the site of action. Consequently, one of the objectives of our studies was to determine the impact of triglyceride chain length on lipid sequestration of anesthetics from human serum in vitro. Although data from similar experiments had been published, these prior experiments examined extraction from buffer, rather than serum. Surprisingly, our results sharply conflicted with these data, as we found greater extraction with Lipofundin (B. Braun Melsungen AG, Melsungen, Germany), a mixed medium-chain triglyceride and long-chain triglyceride (LCT) formulation, when compared with Intralipid (Fresenius Kabi, Uppsala Sweden), an emulsion containing exclusively LCTs. Despite any uncertainty regarding mechanism, lipid resuscitation has become well established as a clinical practice, and promulgated by guidelines published by several authoritative organizations, including the Association of Anaesthetists of Great Britain and Ireland,† the American Society of Regional Anesthesia and Pain Medicine, and the American Heart Association. A unique aspect of the American Heart Association’s guidelines is the association’s explicit recommendation to use an emulsion containing exclusively LCTs. This suggestion was apparently based on the previously published in vitro studies, which our data challenged. However, while we noted the greater extraction by a mixed lipid emulsion, we cautioned, “in vivo studies that confirm [our findings] are obviously required before drawing any confident conclusions.” Despite this limitation, our data had immediate clinical relevance – many (if not most) facilities do not carry more than one lipid formulation, and it was our strong belief that clinicians should not hesitate to use either formulation given both had shown experimental efficacy, and both had been used with apparent success clinically to treat local anesthetic systemic toxicity. Harvey and Cave take objection to our questioning the exclusive use of a LCT emulsion, citing an in vivo study that has been published after the American Heart Association guidelines, and after acceptance of our manuscript. Nevertheless, we agree with their assertion that greater confidence should be placed on data derived in “whole animal” as opposed to “bench-top” experiments, at least as a general principal. However, although the cited study demonstrated superiority of the LCT formulation, there was no significant difference in return of spontaneous circulation, only a higher rate of recurrent asystole with the mixed lipid emulsion. And as the authors note, this difference may reflect the shorter half-life of medium-chain triglycerides. Moreover, return of spontaneous circulation was actually faster with the mixed lipid emulsion, though this difference did not reach statistical significance. Harvey and Cave comment that our results “are insufficient to alter current recommendations for lipid infusion” in local anesthetic systemic toxicity, referencing the American Society of Regional Anesthesia and Pain Medicine guidelines. One of the authors of our paper (KD) was a coauthor of these guidelines, which were deliberately crafted to avoid stipulating a specific lipid, using the generic term “lipid emulsion” throughout. Based on the available literature, we would agree that the scales have tipped in favor of LCT formulations, but in the absence of more definite evidence, clinicians should not hesitate to use a mixed lipid emulsion to treat local anesthetic systemic toxicity. And regardless of formulation, the recent in vivo study by Li et al., as well as clinical experience, emphasize the importance of an adequate continuous lipid infusion following successful response to bolus administration.

Cardiac sympathetic blockade during spinal anesthesia involves both efferent and afferent pathways.

To the Editor-We would like to thank Drs. Thnish and Downs for their interesting case report, “Vagotonia and Cardiac Arrest during Spinal Anesthesia.”’ These authors discuss the contribution of sympathetic and parasympathetic nervous system imbalance during spinal anesthesia as a mechanism of asystole. They state that in a patient with vasovagal syncope, the combination of cardiac sympathetic blockade and vagal stimulation disturbed this autonomic balance even further. It is certainly intuitive to think that because the spinal anesthetic blocks cardiac sympathetic efferent nerves and not vagal efferent nerves, ii relative parasympathetic dominance or vagotonic state would result. However, if this was the case, why would severe bradycardia and asystole not occur more often with high-spinal and epidural anesthesia? As discussed in previous work, another neural pathway is probably Instead of a relative vagal dominance, there was evidence of a decrease in both sympathetic and vagal outflows (efferents) in patients with cardiac sympathectomy after spinal anesthesia. ’The vagal outflow, which was not directly blocked by the anesthetic, was somehow concomitantly reduced or inhibited in parallel with the anesthetic blockade of the sympathetic efferent outflow.’ Therefore, a state of reduced sympathetic and vagal outflow resulted, which fortunately maintained sympathetic and vagal balance and maintained the baseline heart rate. It has been proposed that complete blockade of sympathetic afferent pathways, which have an anatomical distribution similar to the sympathetic efferent pathways, will interrupt necessary visceral communication lines to central neural centers and will result in a selfprotective reduction of parasympathetic activity.’ The activity of sympathetic and parasympathetic (vagal) efferents are normally under the continuous influence of and modulation by sympathetic and parasympathetic afferent input to central control centers.”’ Therefore, blockade of both sympathetic efferent and afferent activity has significance in the overall mechanism of cardiac arrest during spinal anesthesia. We think that sympathectomy of the heart after spinal anesthesia should be thought of as a condition with the potential to develop clinically significant vagal dominance or vagotonia. Bradycardia, asystole, and cardiac arrest from autonomic imbalance during spinal anesthesia are more likely to result if precipitccting events resulting in vagal stimulation occur while cardiac sympathetic blockade exists, as reported by lhrush and Downs.’ We agree with these authors that during spinal anesthesia the clinician should have a high index of suspicion for at-risk scenarios that could result in cardiac arrest and a low threshold for the initiation of prophylactic or resuscitative treatment throughout the perioperativc period.