MITOCHONDRIA are the principal source of cellular energy metabolism. The cellular machinery necessary for the Krebs cycle, fatty acid oxidation, and, most importantly, oxidative phosphorylation, reside within mitochondria. Mutations in mitochondrial proteins cause striking clinical features in the two tissues types requiring the most energy, muscle and nerve. These features include myopathies, cardiomyopathies, encephalopathies, seizures, and cerebellar ataxia. It is increasingly common for children with documented mitochondrial disease to undergo surgery. Often, a muscle biopsy is required for diagnosis, but a wide array of surgical procedures is possible. Such procedures usually involve a general anesthetic for these young children. Although many different anesthetic techniques have been used successfully for patients with mitochondrial disease, 1,2there are reports of serious complications occurring during and after anesthetic exposure. 3,4As a result, there is a general belief among anesthetists that these patients are at increased risk from the stress of surgery and anesthesia. 5
Mitochondrial dysfunction most commonly affects function of the nervous system and of the muscular system. Because volatile anesthetics exert much of their effects through these same tissues, it is possible that patients with mitochondrial disease may have an abnormal sensitivity to volatile anesthetics. General anesthetics have their main clinical effect in tissues such as brain and heart, tissues that are highly dependent on oxidative metabolism. General anesthetics have been shown to depress oxidative phosphorylation in cell-free studies of isolated mitochondria. 6,7These reports indicate that anesthetics depress mitochondria only at concentrations higher than those used clinically. However, even at doses commonly used in the operating room, general anesthetics may cause significant depression of mitochondria in normal patients. 8An investigation in a model organism has also shown that a mutation in a mitochondrial protein leads to increased sensitivity to volatile anesthetics. 9
To limit the anesthetic dose for surgery to only that absolutely necessary for patients with mitochondrial disease, we have been using a Bispectral Index®monitor (BIS®; Aspect Medical Systems, Inc., Newton, MA) during surgery for patients with putative mitochondrial disorders. We induced anesthesia in 16 children who had tentative diagnoses of mitochondrial disease and who presented to our operating room for general anesthesia. Most of the scheduled procedures were skeletal muscle biopsies to assess oxidative phosphorylation and electron transport. We report herein our observations of BIS values noted during routine induction of these 16 children with mitochondrial myopathies. The sensitivities of the patients to sevoflurane were related to their diagnoses of mitochondrial function.
Case Reports
Institutional Review Board approval was provided by the University Hospitals of Cleveland. We induced anesthesia in 16 children undergoing surgery for treatment relating to or diagnosis of defects in mitochondrial function. These patients were receiving their normal care and were not part of a prospective study. Preoperatively, all children had clinical findings consistent with mitochondrial disease. Their symptoms included hypotonia, developmental delay, failure to thrive, and lactic acidosis. In most cases, the surgeries were for muscle biopsies to determine the specific aspect of mitochondrial function, which was affected. Because patients with mitochondrial dysfunction are believed to be at increased risk during anesthesia, 5we elected to induce anesthesia in these patients slowly using a BIS®monitor to indicate their depth of anesthesia. No preoperative sedation was given, and sevoflurane (in 100% oxygen) was the sole agent for induction of anesthesia in all cases. All inductions were performed by the same anesthesiologist (P. G. M.) using a tight-fitting mask with continuous monitoring of end-tidal gas concentrations. During induction, inspired sevoflurane was slowly increased by 0.5% every 2 min until a BIS value of 60 or less was reached. Inspired sevoflurane was increased only after end-tidal concentration of sevoflurane was constant for at least 1 min. Each induction (except in the patients requiring very low doses of sevoflurane) took approximately 10 min. When a patient reached a BIS value of 60, loss of consciousness was confirmed by lack of response to voice and by loss of eyelash reflexes. This approach was not intended to reach a true steady state, but to approximate one; we did not alter the induction to a great degree from normal, and we did not subject these patients to an unusual practice. When a BIS value of 60 or less was reached, nitrous oxide (70%) and 1–2 μg/kg fentanyl were added, and the surgeons prepared the patient and proceeded with the case. Maintenance of anesthesia for surgery was conducted using sevoflurane, nitrous oxide, and fentanyl when indicated. Local analgesia with lidocaine at the surgical site was used in all cases. Maintenance, emergence, and recovery from anesthesia were normal in all cases.
In each patient, mitochondrial function was evaluated via assays of oxidative phosphorylation and electron transport. Oxidative phosphorylation was evaluated by supplying the isolated mitochondria with substrates specific for each complex of the electron transport chain. 10For example, glutamate and pyruvate plus malate were used to evaluate complex I, whereas succinate was used to evaluate complex II. Oxidative phosphorylation was measured using the rate of oxygen uptake in the presence and absence of adenosine diphosphate with the specific substrates. Failure to oxidize with use of a particular substrate was indicative of a problem with the associated complex or an enzyme or complex interacting with the associated complex.
Electron transport chain enzymes were evaluated by solubilizing the mitochondria. When solubilized, the enzyme complexes were evaluated by their ability to accept electrons from electron donors specific for the isolated complexes and enzymes and transfer them to specific electron acceptors. 10These studies and the oxidative phosphorylation studies were part of the planned care for all of these patients and were performed in the laboratory of Dr. Charles Hoppel (Professor, Departments of Pharmacology and Medicine, Case Western Reserve University, Cleveland, Ohio).
No specific defects in oxidative phosphorylation were found in six children. Six other patients were diagnosed with complex III dysfunction. Three patients had mitochondrial defects associated with complex I function. The remaining child had morphologic and metabolic features of Leigh disease but was not tested for complex I function. Remarkably, these last four patients required a very low dose of sevoflurane to reach an anesthetic depth (BIS value of 60) associated with loss of consciousness (table 1).
Discussion
The oxidative phosphorylation pathway consists of five protein complexes (I–V). Complexes I and II independently transfer electrons to coenzyme Q and then sequentially to complexes III, IV, and V. In isolated mitochondria, complex I is capable of using several carbon sources as fuel, among them pyruvate (when coupled with malate) and glutamate. Complex II is restricted to the use of succinate as an energy source. The remaining complexes can also be independently analyzed with specific electron donors. The constellation of presenting symptoms in these patients did not distinguish their underlying defect in these complexes (table 1).
Surprisingly, four patients required very low doses of sevoflurane to reach a BIS value of 60. Each of these patients had changes that potentially involve the function of complex I, the first complex of the electron transport chain. Our observations are consistent with reports of the effects of volatile anesthetics on in vitro assays of mitochondrial function. These studies showed that complex I was the most sensitive of any step in oxidative phosphorylation to inhibition by volatile anesthetics. 6,7However, the in vitro changes were relatively modest, even on exposure of mitochondria to large doses of volatile anesthetics. What is striking to us is that these observations in humans match those seen in a simple animal model, Caenorhabditis elegans . In this animal, a nematode, a defect in complex I function reduces the EC50of volatile anesthetics to approximately one fifth that of wild type. A defect in complex II function had no effect on anesthetic sensitivity in C. elegans . With either defect, the animals move normally but are short-lived and very sensitive to hyperoxia. 3
It is important to note that the end point measured herein does not correspond to the classic minimum anesthetic concentration (MAC) end point of immobility to a painful stimulus, nor do the values represent equilibrium values or EC50s for sevoflurane concentrations. Determining such an end point in these patients would increase difficult ethical questions and necessitate altering the induction. However, the current end point used was determined during a normal anesthetic induction and only required proceeding somewhat slower than usual to obtain an end-tidal sevoflurane concentration that was not changing quickly. Such an approach may represent a method for approximating anesthetic sensitivity during pilot studies without exposure of patients to increased risk or discomfort. We did not obtain these data during a prospective study. Therefore, we do not have matched controls of patients without mitochondrial disease. However, we have induced anesthesia in more than 25 presumably healthy children of the same age range by the same method and found that a typical value of 3–3.5% sevoflurane is needed to reach a BIS value of 60.
Because mitochondria provide most of the energy used by mammalian cells, it is possible that anesthetics affect patients with mitochondrial disease by causing decreased energy supplies. Because the central nervous system has a high demand for energy, changes in level of consciousness might be highly susceptible to anesthetics in patients with mitochondrial dysfunction. Complex I is the primary entry point for electrons into the electron transport chain in normal mitochondria, so changes affecting complex I may intensely affect central nervous system metabolism. However, the profound hypersensitivity of a subset of patients with mitochondrial disease to volatile anesthetics was unexpected, even in light of our studies with a model organism. If this observation can be corroborated in a large study, it would represent a change in sensitivity to volatile anesthetics in humans due to a well-characterized genetic defect. The mechanisms of action of volatile anesthetics are not yet known; alterations in anesthetic sensitivity may indicate important molecular species involved in anesthetic action. These findings also indicate that patients with mitochondrial disease resulting from a variety of molecular changes are varied in their response to anesthetics and possibly in their risk during surgery.