PATIENTS with myasthenia gravis (MG), an autoimmune disorder characterized by reduction in functional acetylcholine receptors at the neuromuscular junction, 1show increased sensitivity to nondepolarizing muscle relaxants. 2–5During surgery necessitating muscle relaxation, paralytic techniques must be chosen carefully based on neuromuscular function. Many reports have described neuromuscular responses to nondepolarizing muscle relaxants during general anesthesia in MG patients, but little information is available regarding responses during sevoflurane or propofol anesthesia. We describe an MG patient anesthetized with each of these agents given singly in procedures separated by 7 months.
A 42-yr-old man was scheduled for fenestration of a mediastinal abscess. The patient, 174 cm tall and weighing 71 kg, had an 8-yr history of MG. When thymectomy was performed shortly after onset of MG, his symptoms decreased but did not resolve. Long-term therapy with prednisolone was required, which may have contributed to abscess formation in the mediastinum. The patient showed a positive response to intravenous injection of edrophonium (10 mg) and also showed a decremental response preoperatively after electromyography with 3-Hz stimulation. Antibody to acetylcholine receptor was highly increased in serum (24.6 nm; normal < 0.37 nm). He was classified as having type IIb MG by the Osserman classification. 6Results of preoperative examination were unremarkable except for the limb weakness. Prednisolone administration (7.5 mg/day) was continued until the morning of surgery. Hydroxyzine (50 mg) and atropine (0.5 mg) were administered intramuscularly as premedication 1 h before induction of anesthesia. Anesthesia was induced with 4 mg/kg thiopental and 1.5 μg/kg fentanyl followed by sevoflurane (2%) and nitrous oxide (60%) in oxygen. The trachea was intubated after topical anesthesia, without administration of muscle relaxants. The lungs were ventilated so that end-tidal carbon dioxide pressure was maintained at approximately 40 mmHg. Skin temperature over the forehead and the thenar region was maintained between 32 and 33°C.
Neuromuscular function was assessed intraoperatively by acceleration measurements in the orbicularis oculi and adductor pollicis muscles. Electrodes were placed 2 cm anterior to the earlobe and at the wrist. Piezoelectric transducers (TOF Guard; Biometer International, Odense, Denmark) were affixed to the upper eyelid using adhesive tape, as well as to the distal phalanx of the thumb. The temporal branch of the facial nerve and the ulnar nerve were stimulated with train-of-four supramaximal square pulses of 0.2 ms duration (60 mA). The nerves were stimulated every 15 s. At the eyelid, the transducer measured acceleration generated by circumferential contraction of the orbicularis oculi muscle. 7After 10 min to allow the response to stabilize, 10 μg/kg vecuronium was administered intravenously, followed by increments of 5 μg/kg until blockade in the adductor pollicis muscle exceeded 90%. Responses of the orbicularis oculi and adductor pollicis muscles were monitored continuously. After blockade exceeding 90% was obtained, inspired concentration of sevoflurane was decreased to 1.0%. Neuromuscular data for the orbicularis oculi and adductor pollicis muscles were recorded on a memory card of the TOF Guard. All graphic and numerical neuromuscular data were obtained using TOF Guard Reader software (Biometer International).
Blockades induced in the orbicularis oculi and adductor pollicis muscles with 10 μg/kg vecuronium were 91% and 4%, respectively. Effective doses for 50% and 90% blockade (ED50and ED90) were measured in the orbicularis oculi and adductor pollicis muscles using linear regression for a dose–response curve plotting the probit transformation of the block (%) versus the logarithm of the dose (μg/kg). We could not obtain these data for the orbicularis oculi because the first incremental dose of vecuronium produced 100% block. In the adductor pollicis muscle, ED50and ED90were 16.6 and 25.6 μg/kg, respectively. The patient was extubated the next day, with no residual neuromuscular effects of the drug.
Seven months later, the patient was scheduled to undergo left lung resection to treat the abscess. His myasthenic condition had worsened; the dose of prednisolone had been increased to 25 mg/day, and azathioprine (50 mg/day) had been added. Antibodies to acetylcholine receptors in serum were 77.2 nm. The Osserman class remained IIb. Premedication was the same as before. Anesthesia was induced with 1.5 mg/kg propofol and 1.5 μg/kg fentanyl, followed by continuous intravenous infusion of propofol at 8 mg · kg−1· h−1during administration of oxygen. Neuromuscular function was monitored as before. After baseline measurement, 10 μg/kg vecuronium was administered, followed by additional increments of the same size. Blockades in the orbicularis oculi and the adductor pollicis muscles with 10 μg/kg vecuronium were 81% and 0%, respectively. ED50and ED90were calculated as previously. Ventilation with a bag and mask was continued as greater than 90% blockade of the adductor pollicis muscle was induced for double-lumen tracheal tube intubation. ED50and ED90were 3.0 and 17.7 μg/kg in the orbicularis oculi muscle, and 27.7 and 42.5 μg/kg in the adductor pollicis muscle, respectively. Anesthesia was maintained with propofol and fentanyl, and recovery was uneventful.
As sensitivity of nondepolarizing muscle relaxants is increased in patients with MG, 1–5monitoring of their neuromuscular function is of particular importance. Neuromuscular function is influenced not only by severity of MG, but by anesthetics. In healthy patients, inhaled anesthetics accelerate neuromuscular blockade by nondepolarizing muscle relaxants, but intravenous anesthetics like propofol have only a minor effect. 8–10In healthy preparations, propofol reduces acetylcholine release from nerve terminals in brain and vascular smooth muscle, 11–13whereas in MG patients, little information is available about the effects of propofol. Moreover, the backgrounds of MG patients differ greatly in titers of antibodies to acetylcholine receptors, MG symptoms, muscles affected, and treatments. Therefore, estimation of the actual influence of anesthetics on neuromuscular function in patients with MG is considerably difficult. We describe one patient with MG who underwent anesthesia twice. MG was worse at the time of the second operation than at the first. Anesthetics used were sevoflurane for the first operation and propofol for the second.
Anesthetic potency between sevoflurane (first procedure) and propofol (second procedure) is difficult to compare: first, the former is inhalation, whereas the latter is intravenous anesthetics; second, corticosteroid inhibits the synthesis of GABAergic steroids 14and may lead to antagonistic interference with propofol. Therefore, comparison of their neuromuscular effects requires a deliberate interpretation.
We used the orbicularis oculi as a measure of neuromuscular blockade of forehead muscles. The orbicularis oculi and corrugator supercilii muscles are representative of forehead muscles and can be monitored by piezoelectric transducers, 15but the orbicularis oculi has been proposed as an informative index in patients with myasthenia gravis. 5
We tried to calculate ED50and ED90in both muscles, but the orbicularis oculi muscle was so sensitive at the time of the first operation that the first incremental administration of vecuronium caused 100% blockade, so a dose–response curve for the orbicularis muscle during sevoflurane anesthesia could not be constructed. However, in the orbicularis muscle, administration of 10 μg/kg vecuronium caused 91% blockade during sevoflurane anesthesia and 81% during propofol anesthesia. Sensitivity to vecuronium was greater during sevoflurane anesthesia than during propofol anesthesia in both muscles, even though MG was worse at the time of surgery during propofol anesthesia.
Antagonistic interactions between aminosteroid relaxant (vecuronium) and corticosteroid were shown by a previous study. 16The dose of prednisolone was increased from 7.5 (first procedure) to 25 mg/day (second procedure). Therefore, in the second procedure, neuromuscular effect had been expected to be greater if the myasthenic condition was the same. These results indicated that the difference in effective doses between sevoflurane and propofol anesthesia may be more marked.
As in a previous report of MG patients, sensitivity to vecuronium was greater in the orbicularis oculi than in the adductor pollicis muscle, 5unlike the situation in healthy patients. 7,17The pattern of increased sensitivity to vecuronium in both muscles in this patient was consistent with MG. The differential response in the orbicularis oculi and the adductor pollicis muscle can be attributed to a greater decrease in margin of safety for neuromuscular transmission in the ocular muscles than in the adductor pollicis, reflecting the preferential involvement of ocular muscles in MG. 1
In the current MG patient, sevoflurane had a more marked effect than did propofol on neuromuscular function, as is true for healthy patients. Residual neuromuscular blockade would be more frequent during sevoflurane anesthesia than during propofol anesthesia. However, the effective dose of nondepolarizing muscle relaxants varied considerably in the MG patients described previously. Inhaled anesthetics, such as sevoflurane, remain important when muscle relaxation is necessary; blockade need not be kept light in MG patients, but neuromuscular function must be closely monitored.
In conclusion, we described two separate instances of anesthetic management with different agents in the same patient with MG. As is true in general, propofol had less neuromuscular effect than sevoflurane in the current MG patient.