The Effects of Donepezil and Neostigmine in a Patient with Unusual Pseudocholinesterase Activity

Abstract

Donepezil (Aricept, Eisai Inc., Teaneck, NJ), a drug recently approved by the Food and Drug Administration for treatment of Alzheimer's disease, has been reported as safe with no, or only rare, cholinergic side effects [1]. It acts primarily as a reversible acetylcholinesterase inhibitor with an elimination half-life longer than 70 hours [1]. We describe a patient with unrecognized atypical pseudocholinesterase (PsChE) who was receiving long-term donepezil therapy and had received succinylcholine and neostigmine as part of general anesthesia for surgery. After succinylcholine administration, the patient developed a phase II block that was not recognized at the time. Postoperatively, a long neuromuscular block ensued after the administration of neostigmine. We believe that an interaction between the cholinergic effects of neostigmine and donepezil in a patient with atypical PsChE activity resulted in a prolonged neuromuscular block. Case Report An 85-yr-old woman weighing 78 kg with a history of mild Alzheimer's disease and hypertension underwent total abdominal hysterectomy and bilateral oophorectomy. She had undergone no previous operations and had not been exposed to anesthesia. The patient had been taking nifedipine 30 mg three times daily and donepezil 5 mg daily for 5 mo; both medications were taken on the day of surgery. The physical examination and laboratory findings (e.g., aspartate transaminase, lactate dehydrogenase, and alkaline phosphatase levels) were normal. An electrocardiogram obtained before anesthesia showed a normal sinus rhythm of 60 bpm. Anesthesia was induced with 3 mg of d-tubocurarine, 100 [micro sign]g of fentanyl, and 300 mg of thiopental, and endotracheal intubation was facilitated with 100 mg of succinylcholine. Anesthesia was maintained with isoflurane, and muscle paralysis was facilitated with 2 mg of pancuronium that was given when three train-of-four (TOF) twitches were elicited, approximately 35-40 min after succinylcholine administration (in retrospect, only three of four TOF twitches indicated a phase II block). The patient's intraoperative course was uneventful. At the end of the 2-h surgery, the TOF stimuli elicited three twitches, and 5 mg of neostigmine (0.065 mg/kg) and 1 mg of glycopyrrolate were given IV to reverse the neuromuscular block. The patient regained all four twitches in response to TOF stimulation, could follow commands, and was breathing adequately; therefore, the endotracheal tube was removed. However, over the next 5-10 min, her breathing deteriorated and apnea ensued. At this time, no muscle twitches could be elicited. The patient became unresponsive, hypertensive, and tachycardic. Ventilation via a mask with oxygen and sevoflurane was initiated, and her trachea was reintubated. An additional 1 mg of neostigmine and 0.2 mg of glycopyrrolate were again administered, but the neuromuscular block persisted. The patient was transferred to the postoperative care unit and sedated with propofol. Mechanical ventilation was initiated. The patient's body temperature was 35.6[degree sign]C. Over the next 8 h, all four twitches reappeared and adequate spontaneous breathing resumed, but because the patient could not sustain a head lift for 5 s and appeared excessively sedated, we decided to support her ventilation mechanically until the next morning. At 60 min after the second dose of neostigmine, screening tests were performed for PsChE. Her PsChE level was 2.1 U/mL (normal 7.1-19 U/mL), the dibucaine number was 45% (normal 81%-87%), and the fluoride number was 84% (normal 44%-54%). Several days later, we asked for permission to collect a second blood sample to repeat the screening tests for PsChE activity in the absence of neostigmine, but the patient's guardian declined our request. In the absence of repeat testing and because we could not determine whether donepezil or neostigmine may have affected our patient's PsChE activity, we tested their effects on PsChE activity in vitro. Methods Drug effect on PsChE activity was determined by adding known amounts of drug to serum. A solution of donepezil was prepared by pulverizing a 10-mg tablet and dissolving the powder in 50 mL of distilled water. Neostigmine (Gensia, Irvine, CA) was obtained at a concentration of 0.5 mg/mL. Three target donepezil concentrations were selected: 1) the typical serum concentration (25 ng/mL); 2) a supratherapeutic concentration (0.02 mg/mL); and 3) an intermediate concentration (250 ng/mL). For neostigmine, 50 ng/mL was selected, which is the serum concentration found 20 min after the IV administration of 0.07 mg/kg [2], a dose similar to that which our patient received. Stock solutions of donepezil (0.2 mg/mL, 2500 ng/mL, and 250 ng/mL) and neostigmine (500 ng/mL) were prepared by serial dilution in distilled water. Ten-fold concentrated stock solutions containing both neostigmine and donepezil were prepared by adding 50 [micro sign]g/mL neostigmine 1:100 (vol/vol) to individual aliquots of the donepezil concentrated stock solutions. The final solutions containing each drug separately or both together in desired concentrations were obtained by adding one volume of a stock solution to nine volumes of serum. Excess serum for testing was obtained in the hematology laboratory from five patients after completion of all ordered laboratory tests and before the blood was to be discarded. These patients were receiving any medications that may have affected PsChE activity. All five serum samples were tested in the presence of three doses of donepezil, neostigmine alone, and three doses of donepezil and neostigmine together. Determinations of PsChE activity were performed on an aca[registered sign] IV analyzer (Dade, Newark, DE) using the manufacturer's protocol (substrate for determination of enzyme activity in this commercial test is butyrylthiocholine). The activity in the presence of neostigmine and/or donepezil was expressed as a percentage of activity in their absence, always using the serum of the same patient as a control. Values are reported as means +/- SD. Statistical analysis was performed with repeated-measures analysis of variance with Bonferroni's adjustment, and differences were considered statistically significant at P < 0.05. Results Donepezil alone at the supratherapeutic concentration (0.02 mg/mL) decreased PsChE activity to 53% of baseline (from 11.6 +/- 3.3 to 6.1 +/- 1.7 U/mL; P < 0.01) (Figure 1). The inhibition of the enzymatic activity by the other two concentrations of donepezil, if any, did not reach statistical significance. Neostigmine 50 ng/mL decreased the PsChE activity to 68% (P = 0.06). The same concentration of neostigmine (50 ng/mL) plus donepezil 25 ng/mL or 0.02 mg/mL reduced the PsChE activity to 44% versus control (P < 0.007) and 25% versus control (P < 0.0001), respectively. Donepezil at therapeutic concentrations (25 ng/mL) magnified the PsChE inhibition produced by 50 ng/mL neostigmine versus neostigmine (P = 0.05). Dibucaine and fluoride numbers were not affected by neostigmine or donepezil, either alone or in combination (data not shown).Figure 1: Average serum pseudocholinesterase (PsChE) activity of five patients in the presence of donepezil (Don; 25 ng/mL, 250 ng/mL, and 0.02 mg/mL), neostigmine (Neo; 50 ng/mL), and three neostigmine (50 ng/mL) plus donepezil (25 ng/mL, 250 ng/mL, and 0.02 mg/mL) concentrations. All values are mean +/- SD. Numbers above respective histograms indicate percentage of PsChE activity remaining based on the control activity equaling 100%. *P < 0.01 versus control. sup [dagger] P < 0.007 versus control. sup [double dagger] P < 0.0001 versus control.beta P < 0.05 versus neostigmine. sup [paragraph sign] P < 0.003 versus neostigmine.Discussion Donepezil, a reversible inhibitor of central acetylcholinesterase, can also inhibit PsChE. Because acetylcholinesterase (at the motor endplates) and PsChE (in plasma) are enzymes involved in hydrolysis of succinylcholine, their inhibition prolongs succinylcholine action [1,3]. In a therapeutic dose (5 mg/d), donepezil inhibits acetylcholinesterase by 64% [4]. However, a literature search did not reveal any clinical report describing the interaction among succinylcholine, neostigmine, donepezil, and PsChE; therefore, we tested these effects in vitro. We demonstrated that neostigmine at a therapeutic dose and/or donepezil at supratherapeutic dose decreased PsChE activity, but dibucaine and fluoride numbers remained unchanged. Our patient declined subsequent serum testing in the absence of neostigmine; therefore, we could not precisely identify her PsChE defect. Because neither donepezil, neostigmine, nor their combination affected dibucaine and fluoride numbers in vitro, the low dibucaine value in our patient must be attributed to an atypical PsChE variant; however, the inhibitor studies were not characteristic of the most commonly described enzyme variants [5]. Her low PsChE activity may, in part, be attributed to the inhibitory effects of neostigmine and donepezil, but because a significant portion of her PsChE was atypical, this too may have resulted in the unusually long neuromuscular block. What may be more relevant is that we do not know the magnitude of inhibition of the acetylcholinesterase at the motor endplate by donepezil. Typically, an individual heterozygous for atypical PsChE has a moderately delayed recovery from succinylcholine [5]. At 35 min after the administration of 1 mg/kg succinylcholine, our patient had three TOF twitches, which suggests a phase II block; however, this block was only recognized retrospectively. This type of block indicates a prolonged interaction between neuromuscular receptors and succinylcholine (slow hydrolysis), a situation that can be encountered in succinylcholine overdose, in patients with drug-induced PsChE deficiency, and in patients with an atypical enzyme (in whom succinylcholine is not hydrolyzed). The duration of succinylcholine action is directly related to the absolute PsChE activity [6]. However, only when the normal PsChE concentration decreases below 75% does prolongation of the succinylcholine block become clinically evident [7]. Our in vitro experiments showed that a supratherapeutic dose of donepezil reduced PsChE by 47%; therefore, in patients with normal PsChE activity, donepezil would not typically prolong the action of succinylcholine; however, theoretically, it may do so in patients with already low PsChE activity. Although our patient had very low postoperative PsChE activity, this in itself cannot explain the lengthy neuromuscular block in the absence of succinylcholine. It is possible that the phase II block was induced by neostigmine in this patient with atypical PsChE who was receiving donepezil. If excessive neostigmine is administered to a patient, it may cause a paradoxical neuromuscular block [8,9], even more so to an individual with low PsChE activity. This may explain the block in our patient because at the end of surgery, neostigmine administration first reversed the block: the patient temporarily regained all TOF twitches and had good muscle strength, allowing tracheal extubation. However, just several minutes later, neuromuscular block recurred and our patient required endotracheal intubation. Such an event would occur if nonhydrolyzed acetylcholine had accumulated at the neuromuscular junction. It is possible that donepezil enhanced the inhibitory effect of neostigmine on the enzyme acetylcholinesterase and that an excessive inhibition of acetylcholine hydrolysis resulted and a block by acetylcholine ensued. In addition, the administration of pancuronium bromide, through its inhibition of cholinesterase activity, may have contributed to prolonged neuromuscular block in our patient. The activity of cholinesterase is 40% inhibited 45 min after an injection of 0.1 mg/kg pancuronium [10]. Although donepezil has an excellent clinical safety record, adverse effects such as muscle weakness and fasciculations (which are the symptoms typically seen with acetylcholine-induced depolarization) have been reported, albeit rarely. Such rare adverse neuromuscular effects during donepezil therapy may indicate that there are susceptible individuals who develop neuromuscular cholinergic symptoms. At least some of these individuals may be those rare patients with an acquired PsChE deficiency that does not become clinically evident unless they receive drugs that inhibit PsChE (such as donepezil). Although adverse neuromuscular symptoms or signs during donepezil therapy are rare, the manufacturer recognizes the potential for interaction between donepezil and certain drugs used during anesthesia and recommends its discontinuation 3 wk before surgery. The manufacturer also warns of potential synergistic effects with cholinergic agonists, such as bethanechol and other cholinesterase inhibitors [1]. Our present study of PsChE activity in vitro suggests that administering neostigmine to patients who are taking donepezil has an additive effect on the decrease in PsChE activity. Although donepezil at a therapeutic concentration did not reduce the PsChE activity, the same dose in the presence of neostigmine induced an additional reduction, resulting in a residual PsChE activity of 44% (Figure 1). In conclusion, the additive inhibitory effects of neostigmine and donepezil on PsChE activity (and presumably on motor endplate acetylcholinesterase activity) can explain the long-lasting neuromuscular block in our patient with atypical PsChE variant. However, we believe that such a reduction in a patient with normal enzyme activity would not result in a clinically significant alteration of neuromuscular transmission.