The N-methyl-d-aspartate (NMDA)-nitric oxide (NO)-cyclic guanosine 3′,5′-monophosphate (cyclic GMP) pathway may have a significant role in modulating the effects of general anesthesia because inhibition of NO synthase (NOS) by means of selective and nonselective isoform inhibitors leads to a reduction of minimum alveolar anesthetic concentration (MAC) and reduction of brain cyclic GMP in animals (1–3). The second messenger, cyclic GMP, regulates a variety of cyclic GMP-dependent kinases, ion channels, and phosphodiesterases (PDE) and may, therefore, interact with other potential anesthetic targets.
Cyclic GMP is hydrolyzed by PDE, and PDE 5, 6, and 9 are specific for cyclic GMP. Clinically approved cyclic GMP-specific PDE 5 inhibitors are available in the form of, for example, sildenafil (Viagra®). PDE 6 is primarily located within photoreceptors and no isoform-specific inhibitors for PDE 9 are licensed in humans.
We hypothesized that modifying cyclic GMP concentrations through selective PDE 5 inhibition would change the sedation requirements for propofol and affect the speed of recovery. We tested this hypothesis by determining the effect of sildenafil on propofol sedation requirements in a double-blind, placebo-controlled, crossover study in healthy volunteers.
After local ethics committee approval and written informed consent, 10 healthy volunteers, aged 18 yr and older and ASA physical status I or II were recruited for the study. All subjects underwent a full medical examination, including drug and allergy history, and were fasted for 6 h for solid food and 2 h for fluids. The study was conducted within the intensive care unit with continuous monitoring of electrocardiogram (ECG), pulse oximetry, and intermittent noninvasive arterial blood pressure.
All volunteers received two IV indwelling cannulae—one for blood sampling and the other for propofol infusion. They were then randomized to receive sildenafil (100 mg) or an identical placebo (gifts from Pfizer, NY) in a double-blind manner, orally 45 min before start of sedation. One week later, the study was repeated with the second (opposite) tablet.
Total IV propofol sedation was started at 0.5 μg/mL predicted plasma concentration using a Fresenius Vial SA Diprifusor® pump (Fresenius Vial, Brezins, France) and increased by 0.1 μg/mL at the end of each 3-min interval if the subject was still able to answer personal questions. Verbal contact was judged to have been lost when the volunteer did not verbally respond to questions about name, date of birth, or address. At this time, the target, calculated, effective propofol concentration as well as the total dose of propofol used and the time were noted. The volunteers were allowed to recover and, on return of spontaneous verbal contact, the calculated and effective propofol concentrations as well as time were again documented. All volunteers were subsequently discharged home and asked not to drive or drink any alcohol and be accompanied by a responsible adult for the following 24 h.
Venous blood was collected into potassium EDTA Vacutainer® tubes before administration of oral sildenafil or placebo, at loss of verbal contact (LVC), and 30 min after LVC. All samples were stored on ice and centrifuged at 2000g at 4°C to separate plasma. Plasma was then snap frozen in liquid nitrogen and stored at –80°C until analysis. Cyclic GMP concentrations were measured using an enzyme immunoassay system (Biotrak, APBiotech, Little Chalfont, UK). Intra- and interassay variability was <10%.
Data were not normally distributed and are presented as median and range. Statistical analysis was performed using Friedman analysis of variance and Wilcoxon's signed rank test as appropriate. A value of P < 0.05 was considered to be statistically significant.
The 6 male and 4 female volunteers had a median age of 28 [range 22–46] yr and median weight of 70 [range 63–102] kg. There were no significant changes in arterial blood pressure, heart rate, or ECG during the study period.
The predicted plasma propofol and the effective brain concentrations at loss and return of verbal contact are given in Table 1. The total dose of propofol was similar for sildenafil and placebo (401.1 [216.4–546.7] mg and 388.8 [236.5–604.0] mg, respectively, P = 0.62).
The total propofol infusion time and time to return of verbal contact are shown in Figure 1. Target and calculated plasma propofol concentration as well as the total dose of propofol were not significantly different between the treatment groups and individuals. The duration of propofol infusion was 62 [45–84] min after placebo pretreatment and was not different from the duration after sildenafil pretreatment, 67 [40–83] min. The median time for return of spontaneous verbal contact, however, was statistically shorter in the sildenafil group compared with placebo (4 [3–5] and 6 [3–8] min, respectively, P = 0.019).
Plasma cyclic GMP concentrations are shown in Table 2. After placebo treatment, plasma cyclic GMP concentrations were decreased significantly at LVC and 30 min thereafter compared with concentrations before sedation (P < 0.004). Cyclic GMP concentrations were not significantly different at LVC and recovery compared with baseline in the sildenafil group. The plasma cyclic GMP concentrations were larger at LVC and 30 min after sedation in the sildenafil group compared with placebo, but this reached statistical significance only at LVC (P = 0.004).
Our results show that selective PDE 5 inhibition does not change propofol sedation requirements but influences recovery time and plasma cyclic GMP concentrations in healthy volunteers.
The NMDA-NO-cyclic GMP pathway is potentially a major target for general anesthetics (2,4,5). It is interlinked with a host of other intracellular pathways, receptor regulation, and intracellular calcium homoeostasis. Cyclic GMP is a ubiquitous second messenger and is described in almost every tissue. The volatile anesthetics halothane (3,6), isoflurane (7), enflurane (8), and sevoflurane (9,10), all reduce cerebral cyclic GMP concentration in animal studies. This is at least partially attributed to a decrease in NOS activity and occurs in the presence of chronic NOS inhibition with the NOS inhibitors NG-nitro-l-arginine-methyl-ester and 7-niroindazole (7,11,12). Unlike the reports from studies of volatile anesthetics, the results of cyclic GMP measurements from in vivo studies are incomplete and conflicting for IV anesthetics. Pentobarbital (80 mg/kg) was reported to decrease cerebral cyclic GMP concentration in ventilated rats (13), whereas a smaller dose (50 mg/kg) did not change cyclic GMP concentrations (3). Ketamine is reported to increase cyclic GMP in ventilated rats, which is consistent with the results of an increase in NO oxidation products as measured by microdialysis (3,14). Cerebral cyclic GMP concentrations from in vivo studies have not been reported for propofol because of its short duration of action after IV injection and difficult experimental control in animals. We selected a target-controlled sedation protocol for its ease of use and control of sedation in humans.
The duration of cyclic GMP signaling is dependent not only on the rate of cyclic GMP formation, but also on its rate of breakdown. The cyclic nucleotides adenosine 3′,5′-cyclic monophosphate (AMP) and GMP are hydrolyzed by PDE, of which there are currently 11 isoforms described. Most PDE catalyze the breakdown of both cyclic nucleotides (PDE 1, 2, 3, 10, and 11), whereas PDE 5, 6, and 9 are specific for cyclic GMP and PDE 4, 7, and 8 are specific for cyclic AMP. The enzyme activity of the PDE is regulated by cyclic AMP and cyclic GMP concentrations and is subject to feedback activation and inhibition (15–17). Sildenafil has an acceptable degree of selectivity for PDE 5 with minor side effects, such as visual disturbances and headaches, which are probably the result of partial inhibition of PDE 6 and PDE 1, respectively (17). Sildenafil is also thought to cross the blood-brain barrier (18). We selected sildenafil for this study while accepting that cyclic GMP could also to some extent be hydrolyzed by other nonspecific PDE isoforms.
Very little is known about the effects of anesthesia on cyclic GMP in humans, although we have previously demonstrated that saliva cyclic GMP increases during inhaled anesthesia in patients undergoing minor obstetric procedures (19). Reports of cyclic GMP concentrations and variations in plasma or human brain in relation to anesthesia are not available. Plasma cyclic GMP seems to be tightly regulated and displays nocturnal variation (20), but it is conceivable that variables affecting daily life such as diet, fluid intake, and stress situations may also influence plasma cyclic GMP concentration. The volunteers in the present study underwent the two parts of the study exactly one week apart in order to minimize this potential effect.
Target and calculated plasma propofol concentration as well as the total dose of propofol were not significantly different between the groups and individuals, indicating that sildenafil pretreatment did not increase propofol sedation requirements and possibly did not affect the pharmacokinetic properties of propofol in our study. Power calculations based on the current results indicate that >250 volunteers would be required to demonstrate a statistically significant effect of sildenafil on calculated propofol plasma concentration. However, the predicted propofol concentrations are based on an algorithm from a different population and a potential influence of sildenafil on pharmacokinetics of propofol will require detailed pharmacokinetic studies (21–23). The ability of other PDE to hydrolyze cyclic GMP within the brain may account for the observed lack of increased propofol sedation requirements in the placebo group. A combination of PDE inhibitors may elucidate this but would produce unacceptable cardiovascular side effects in healthy volunteers.
Return of verbal contact from propofol sedation was significantly faster after sildenafil pretreatment compared with placebo. This was not attributed to an increased propofol infusion time or the total amount of propofol used and indicates that changes in plasma cyclic GMP concentration may correlate with anesthesia in humans, providing evidence that the NMDA-NO-cyclic GMP pathway is a potential target for propofol sedation in humans. However, no direct link to brain cyclic GMP concentrations can be made and further studies are required. Plasma samples obtained from the jugular venous bulb may allow for more detailed studies from cyclic GMP originating within the brain but this was outside the scope of this study.
The difference in recovery time between the sildenafil and placebo groups is small and unlikely to be of direct clinical significance in the dose (1–2 mg/kg) often used for erectile dysfunction. It exceeds the starting dose for treatment of pulmonary hypertension (0.3 mg/kg) but is less than the maximal dose of 2–3 mg/kg at which systemic hypotension occurs. A combination with other PDE inhibitors may theoretically lead to a greater difference in recovery time and, perhaps, in propofol sedation requirements. Changes in other clinical variables such as cognitive function may help to further elucidate any important clinical interactions.
In summary, we have shown that selective PDE 5 inhibition with sildenafil influenced recovery time and plasma cyclic GMP concentrations in humans but did not influence propofol sedation requirements. This is further evidence that the NMDA-NO-cyclic GMP pathway has a role in modulating the effects of propofol sedation in humans.
1. Johns RA, Moscicki JC, DiFazio CA. Nitric oxide synthase inhibitor dose-dependently and reversibly reduces the threshold for halothane anesthesia: a role for nitric oxide in mediating consciousness? Anesthesiology 1992;77:779–84.
2. Tonner PH, Scholz J. [The NO/cGMP signal transduction system: a central target for anesthetics?] Anasthesiol Intensivmed Notfallmed Schmerzther 1999;34:78–89.
3. Galley HF, Le Cras AE, Logan SD, Webster NR. Differential nitric oxide synthase activity, co-factor availability and cGMP accumulation in the central nervous system during anaesthesia. Br J Anaesth 2001;86:388–94.
4. Johns RA. Nitric oxide, cyclic guanosine monophosphate, and the anesthetic state. Anesthesiology 1996;85:457–9.
5. Sonner JM, Antognini JF, Dutton RC, et al. Inhaled anesthetics and immobility: mechanisms, mysteries, and minimum alveolar anesthetic concentration. Anesth Analg 2003;97:718–40.
6. Nahrwold ML, Lust WD, Passonneau JV. Halothane-induced alterations of cyclic nucleotide concentrations in three regions of the mouse nervous system. Anesthesiology 1977;47:423–7.
7. Adachi T, Shinomura T, Nakao S, et al. Chronic treatment with nitric oxide synthase (NOS) inhibitor profoundly reduces cerebellar NOS activity and cyclic guanosine monophosphate but does not modify minimum alveolar anesthetic concentration. Anesth Analg 1995;81:862–5.
8. Vulliemoz Y, Verosky M, Alpert M, Triner L. Effect of enflurane on cerebellar cGMP and on motor activity in the mouse. Br J Anaesth 1983;55:79–84.
9. Ichinose F, Huang PL, Zapol WM. Effects of targeted neuronal nitric oxide synthase gene disruption and nitroG-L-arginine methylester on the threshold for isoflurane anesthesia. Anesthesiology 1995;83:101–8.
10. Masaki E, Kondo I. Methylene blue, a soluble guanylyl cyclase inhibitor, reduces the sevoflurane minimum alveolar anesthetic concentration and decreases the brain cyclic guanosine monophosphate content in rats. Anesth Analg 1999;89:484–9.
11. Ichinose F, Mi WD, Miyazaki M, et al. Lack of correlation between the reduction of sevoflurane MAC and the cerebellar cyclic GMP concentrations in mice treated with 7-nitroindazole. Anesthesiology 1998;89:143–8.
12. Fukuda T, Saito S, Sato S, et al. Halothane minimum alveolar anesthetic concentration and neuronal nitric oxide synthase activity of the dorsal horn and the locus ceruleus in rats. Anesth Analg 1999;89:1035–9.
13. Kant GJ, Muller TW, Lenox RH, Meyerhoff JL. In vivo effects of pentobarbital and halothane anesthesia on levels of adenosine 3′,5′-monophosphate and guanosine 3′,5′-monophosphate in rat brain regions and pituitary. Biochem Pharmacol 1980;29:1891–6.
14. Wu J, Kikuchi T, Wang Y, et al. NOx-concentrations in the rat hippocampus and striatum have no direct relationship to anaesthesia induced by ketamine. Br J Anaesth 2000;84:183–9.
15. Pyne NJ, Arshavsky V, Lochhead A. cGMP signal termination. Biochem Soc Trans 1996;24:1019–22.
16. Juilfs DM, Soderling S, Burns F, Beavo JA. Cyclic GMP as substrate and regulator of cyclic nucleotide phosphodiesterases (PDEs). Rev Physiol Biochem Pharmacol 1999;135:66–104.
17. Soderling SH, Beavo JA. Regulation of cAMP and cGMP signaling: new phosphodiesterases and new functions. Curr Opin Cell Biol 2000;12:174–9.
18. Milman HA, Arnold SB. Neurologic, psychological, and aggressive disturbances with sildenafil. Ann Pharmacother 2002;36:1129–34.
19. Englehardt T, Galley HF, MacLennan FM, Webster NR. Saliva cyclic GMP increases during anaesthesia. Br J Anaesth 2002;89:635–7.
20. Zhdanova IV, Simmons M, Marcus JN, et al. Nocturnal increase in plasma cGMP levels in humans. J Biol Rhythms 1999;14:307–13.
21. Flezzani P, Alvis MJ, Jacobs JR, et al. Sufentanil disposition during cardiopulmonary bypass. Can J Anaesth 1987;34:566–9.
22. Coetzee JF, Glen JB, Wium CA, Boshoff L. Pharmacokinetic model selection for target controlled infusions of propofol: assessment of three parameter sets. Anesthesiology 1995;82:1328–45.
23. Yeganeh MH, Ramzan I. Determination of propofol in rat whole blood and plasma by high-performance liquid chromatography. J Chromatogr B Biomed Sci Appl 1997;691:478–82.