Response to drugs can vary between individuals and between different ethnic populations . Genetic factors represent an important source of inter-individual variation in drug response .
Propofol (2,6-diisopropylphenol) is widely used for total intravenous (i.v.) anaesthesia (TIVA) because of its favourable induction properties and rapid clearance . Propofol, at therapeutic concentrations, exerts its principle pharmacological actions by activation of gamma-aminobutyric acid (GABAA) receptors [4-6]. A new class of human GABAA receptors (subunit epsilon), which confer insensitivity to the potentiating gabaminergic effects of i.v. anaesthetic agents, has been identified . The gene, GABRE, coding for this receptor subunit has the gene map locus Xq28 .
Recovery from general anaesthesia is dependent on factors governing drug sensitivity and drug disposition . Gender has been reported to be a highly significant independent predictor for recovery time . Oxidation of propofol to 4-hydroxypropofol represents a significant pathway in its metabolism. Cytochrome P-450 2B6 (CYP 2B6) is the principal determinant of inter-individual variability of propofol hydroxylation by human liver microsomes in vitro . A significantly reduced CYP2B6 protein expression was documented in carriers of C1459T (R487C) (alleles *5 and *7), found at a frequency of 32.6% in the general population . Thus, compared to the homozygous wild type, CYP2B6 protein expression in individuals homozygous for the R487C variant in exon 9 was reduced by almost eight-fold. CYP2B6 has been identified in brain, and the same polymorphism was associated with lower protein levels . It could be speculated that in vivo, propofol metabolism may be altered by this polymorphism, thereby influencing time to recovery.
We hypothesized that inter-patient variability in the dose of propofol required to achieve bispectral index (BIS) <70 and ‘time to eye opening’ may be influenced by the genotype of the metabolizing enzyme CYP2B6 or that of the drug receptor gene GABRE. If an individual patient's genetic profile could be determined, a better, more cost-effective anaesthetic regimen, with more rapid recovery, could be administered. This would have safety implications in particular for patients resuming activities of daily living soon after day case procedures.
Our study objectives were: (i) to describe the genetic polymorphism of CYP 2B6 and GABRE in patients undergoing day case surgery under TIVA; (ii) to determine if there was a statistically significant correlation between the studied pharmacokinetic (apparent systemic propofol clearance) and pharmacodynamic parameters (dose of propofol required to achieve BIS < 70, time to eye opening) and the described gene polymorphism.
With Institutional Ethics Committee approval and having obtained written informed consent from each, 150 ASA I patients, aged >18 yr, with a body mass index 20-30, scheduled for ‘day case’ procedures of approximately 1-h duration, were studied. Patients with a regular alcohol intake of more than 40 IU week−1 or on concurrent medication were excluded from the study.
A standardized anaesthetic technique consisting of propofol i.v. anaesthesia was used. No preanaesthetic medication was administered. On arrival to the induction room, two 18-G i.v. cannulae were sited (one for blood sampling in the contralateral arm), and 500 mL of Hartmann's solution was infused rapidly (over 20 min). Patient monitoring included electrocardiography, non-invasive blood pressure, pulse oximetry (SPO2), capnography (etCO2) and peripheral temperature. In addition, the BIS was measured and recorded (software version 3.31; Aspect Medical Systems Inc., Natick, MA, USA).
Induction of anaesthesia was achieved by delivering propofol 1% at a rate of 400 mL h−1 (Graseby 3500 infusion pump; Sims Graseby Ltd, Watford, UK) until a BIS of 45 was reached. Propofol requirements for achieving BIS <70 were recorded. After induction, infusions of propofol were supplemented by N2O (66%) in O2 and targeted to maintain BIS values between 40 and 50. No neuromuscular blocking agent was used and a laryngeal mask airway was inserted. Intermittent positive pressure ventilation was employed in order to maintain normocapnia. A remifentanil infusion (0.1-0.3 μg kg−1 min−1 prn) was commenced after induction was achieved and continued until local/regional infiltration (with weight appropriate bupivacaine 0.5%) at the end of surgery. Propofol infusions were abruptly stopped 10-15 min later. Patients were asked repeatedly, at 1-min intervals, in a normal tone voice to open their eyes until an appropriate response was obtained. The following time intervals were recorded: (i) from commencement of propofol infusion to loss of verbal contact (T1) and BIS ≤ 70 (T2, induction time); (ii) from cessation of propofol infusion to eye opening (T3), BIS > 70 (T4), BIS > 90 (T5) and orientation (T6 when the patient states his or her own name, birth date and the name of the hospital, assessed at 2-min intervals).
A venous blood sample (4 mL) was withdrawn for estimation of urea and electrolytes (BUN, creatinine, K, Na, Cl) and liver function (alanine aminotransferase, aspartate aminotransferase, alkaline phosphatase, bilirubine) before induction. Venous blood was further sampled into ethylenediamine tetraacetic acid tubes: 20 mL for genotyping and 4 mL for estimation of propofol plasma concentrations 30 min after induction and at steady state, just before cessation of propofol infusion (after having maintained a constant rate for at least 20 min). The BIS index and infusion rates were recorded at the time of blood sampling. Apparent systemic clearance, calculated from the ratio of infusion rate and plasma concentration at pseudosteady state, was corrected for weight.
Plasma concentrations of propofol
Propofol plasma concentrations were estimated by a high-performance liquid chromatography method with fluorescence detection using thymol as the internal standard (PharmaKinetics, Baltimore, MD, USA). The limit of detection from 0.25 mL of human plasma was 0.1 μg mL−1. During the validation process, the stability of the drug in the matrix was established for storage at −20%, three freeze/thaw cycles and 24 h storage at room temperature. The stability of the drug after extraction was also established.
Leukocyte DNA was isolated by standard methods from blood samples obtained from the 150 patients. The samples were typed for the single-nucleotide polymorphisms (SNPs) using approximately 4 ng of genomic DNA per genotype in a fluorescent-based readout of the assay (Amplifluor, KbioSciences Ltd, Herts, UK). Routine quality control measures were performed (i.e. negative controls, inter-plate testing of a known SNP, intraplate testing of a known SNP, and presence of clear and distinct clusters).
The known C1459T polymorphism from the CYP2B6 gene  was typed, using a primer extension assay based on the sequence context of the genomic SNP (dbSNP reference rs3211371): CCAACATACCAGATC[T/C]GCTTCCTGCCCGCT. In addition, the K262R and Q172H variants (rs2279343 and rs3745274) were genotyped.
Four GABRE variants (Table 1) were identified from the Leelab database (http://www.bioinformatics.ucla.edu/snp/) , as SNPs contained with a reasonably high frequency within expressed GABRE sequences from Genbank.
Efficient statistical approaches should simultaneously estimate haplotypes and differences in distribution of a phenotype among haplotypes. We implemented the method of Schaid and colleagues , which permits such an analysis for a quantitative trait.
Effects of genotypes on various times and drug clearance were analysed using the Wilcoxon signed rank sum test. Only three phenotypes, time to BIS <70, apparent systemic clearance and time to eye opening, were chosen for genotypic analysis to avoid loss of power that results from extensive multiple testing of multiple end-points. A likelihood analysis determined association of major haplotypic groupings (pooling those with <10% frequency) with qualitative and quantitative outcomes (time to BIS < 70) and quantitative outcomes (rank of clearance and rank of time to eye opening) using the haplo.score software (www.mayo.edu/hsr/people/schaid.html) . Haplotype frequencies were estimated simultaneously using the hapipf program http://ideas.respec.org/c/boc/bocode/5425501.html) . Since the GABRE gene is X-linked, the likelihood model was estimated separately for males and females and the χ2-test (3 degrees of freedom (d.f.)) likelihood differences from the male and female models were summed. This sample size has sufficient power to detect a strong effect, of the order of a three- or four-fold increase in risk of having a slow propofol clearance.
A total of 150 (60 female and 90 male) patients were studied. Mean age (range) was 37.8 (18, 74) yr. Preoperative urea, electrolytes and liver function tests were within normal range in all patients. Mean (±SD) duration of propofol infusion was 65.3 (±25.8) min (Table 2). No adverse effect was documented and no readjustment of the propofol infusion rate had to be made as a consequence of hypotension. Types of surgery included inguinal hernia repair, removal of metal from upper and lower limbs, varicose veins, and hand and foot surgery.
The induction dose of propofol required to lower BIS from a baseline value of 97 (±1.5) to <70 varied from 1.4 to 4.8 mg kg−1, thus showing a 3.2-fold inter-patient variability (2.5 ± 0.5 mg kg−1; Table 2). Propofol plasma concentrations estimated at 30 min after induction and at pseudosteady state (at discontinuation of the infusion) varied from 0.7 to 7.8 μg kg−1 (11.2-fold) and from 0.9 to 6.9 μg kg−1 (seven-fold), respectively, while BIS was maintained constant between 40 and 50 (Table 3). Apparent systemic clearance estimated at pseudosteady state varied from 9.1 to 55.8 mL min−1 kg−1 (6.1-fold variability).
At induction, T1 and T2 varied from 50 to 330 s (6.6-fold) and from 80 to 350 s (4.4-fold), respectively. At emergence, T3, T4, T5, T6 varied from 25 to 1725 s (69-fold), from 99 to 3390 s (34-fold), from 15 to 1665 s (111-fold) and from 120 to 1860 s (15.5-fold), respectively.
No significant correlation could be established between the apparent systemic propofol clearance and either of the emergence times (T3-6, r = −0.11, −0.23, −0.20 and −0.20, respectively) or between age and clearance (r = −0.16) or plasma concentration of propofol (r = 0.26). However, there was positive correlation between plasma concentrations of propofol estimated at 30 min and before cessation (r = 0.79, P < 0.01), as well as between the rate of propofol infusion and plasma concentrations of propofol (r = 0.79, P < 0.01).
CYP2B6 C1459T (R487C) polymorphism had the following distribution in the population studied: T : T 1%, T : C 25% and C : C 74%. These are similar to those described by Lang and colleagues  (1%, 13% and 86%, respectively) in a Caucasian population. Figure 1 indicates the time to eye opening and apparent systemic propofol clearance for patients carrying CYP2B6T, compared to the CC genotype. No difference in apparent systemic propofol clearance was found between carriers of the T allele and the C : C genotype (25.5 (SE 0.96) vs. 27.6 (SE 1.52) mL min−1kg−1, P = 0.08, Wilcoxon signed rank sum test). Similarly, no difference in time to eye opening was found between carriers of the T allele and the C : C genotype (T-carrier mean 619, SE 41) vs. CC genotype (mean 643, SE 29; P = 0.73). Figure 1 suggests that the variance in both time to eye opening and apparent systemic clearance may be slightly greater in males with the CC genotype than among the T carriers. Carrying the T allele is unlikely to contribute to a very large risk of being in the slower recovery half of the patient group (upper 95% CI = 1.68). Among the CYP2B6 haplotypes, *1 (172H-262K-487R), *6B (172Q-262R-487R) and *5 (172H-262K-487C) had an incidence of 0.63, 0.22 and 0.14, respectively. There was no significant haplotypic association with apparent systemic clearance (P = 0.48, 3 d.f.) or time to eye opening (P = 0.55). The only suggestive effect was that the HKC haplotype was associated with a slower clearance (P = 0.05; non-significant P = 0.2 after correction for multiple testing).
The haplotypic distribution of the four allelic variants in the GABRE locus was determined (Table 4a and b). This indicates that at the four variant positions [mRNA358]G/T, 20118C/T, 20326C/T and 20502 A/T, there are four common haplotypes TTCT, CTCT, TGCT and TTTA, with respective frequencies of 0.49, 0.22, 0.15 and 0.09. These haplotypes can be largely inferred from the genotypes of three of the variants, namely [mRNA358]G/T, 20118C/T, and then interchangeably 20326C/T or 20502 A/T.
Time to BIS < 70 was strongly skewed (Fig. 2) and we defined as ‘propofol induction resistant’, a small subset of 24 predominantly male patients taking >170 s to achieve BIS < 70 (Fig. 2). We sought to determine firstly if the distribution of genotypes was markedly skewed in this small group of 24 patients. Statistical analysis focused on the three commonest haplotypes (TTCT, CTCT and TGCT), pooling other haplotypes (see Table 5). The significance of association in females of the four different haplotypic categories was assessed using the haplo.score software . The significance of the male and female data was then subsequently combined as a likelihood ratio with 3 d.f. Calculating the probability from the summed effects across the male and female haplotype association analyses, we observed no significant association (P = 0.69 for BIS < 70 of ≥170 s; see Table 5).
As secondary hypotheses, GABRE associations with other clinical features were explored. Individual genotypes were investigated in relation to apparent systemic clearance and time to eye opening. There were no significant effects of carrier status for mRNA358G, 20118C and 20326T on apparent systemic clearance (P = 0.82, 0.92 and 0.60, respectively) or on time to eye opening (P = 0.70, 0.82 and 0.93, respectively).
In this study, we characterized the inter-patient variability in the dose of propofol required to achieve BIS < 70 as well as in the estimated apparent systemic clearance and ‘time to eye opening’ following TIVA. Although it seemed possible that the cause of this variability would be genetically determined, the analysed genetic variants in the CYP2B6 and GABAA(ε) genes did not account for the majority of this variation in vivo.
A strength of this study is the rigorous standardization of the total i.v. anaesthetic technique. Depth of anaesthesia is best measured by a processed electroencephalographic monitor. The BIS is a dimensionless parameter (0-100) derived from the cortical electroencephalogram, which has been shown to correlate with measures of drug-induced sedation and hypnosis. Awake values of BIS generally range from 90 to 100, and values <60 indicate a very low probability of response to verbal command [17,18]. Similar to other studies, the target BIS of 40-50 was selected in order to achieve a comparable depth of anaesthesia .
Remifentanil was used in conjunction with propofol because of its short context sensitive half-life. In addition, due to its unique esterase metabolism it does not interact with the hydroxylation of propofol when compared to other opioids . On the other hand, the local infiltration/regional block at the end of surgery served the purpose of providing a relatively pain-free recovery from anaesthesia.
There are a number of confounding variables that may have influenced our results. Clearance is the chief descriptive tool for both metabolic and excretory elimination of drugs. Plasma clearance refers to the volume of distribution cleared of drug in unit time and is independent of the dose or volume of distribution. Clearance of anaesthetic drugs by the liver is influenced by the rate of liver blood flow and specific enzyme activity. These vary widely with age and health . In our study, although patient age varied between 18 and 74 yr, all were healthy and were on no concurrent medication. Plasma clearance can be estimated from the plasma concentration at steady state, when the rate of infusion equals the rate of drug elimination . Morgan and colleagues  reported that plasma propofol concentrations approach steady state within 20 min following the start of infusion. We found no difference between the plasma concentrations of propofol at 30 min and at the end of surgery (assumed to be steady state), probably explained by the administration of an initial bolus, hence steady state was achieved immediately. The lack of relationship between clearance and recovery times in this study indicates that drug distribution may still be the major factor influencing recovery after a 1-h infusion of propofol.
In a large number of human liver samples studied in vitro, a significantly reduced CYP2B6 protein expression and S-mephenytoin N-demethylase activity was found in carriers of the C1459T (R487C) variant . Thus, compared to the homozygous wild type, CYP2B6 protein expression in individuals homozygous for the R487C variant in exon 9 was reduced by almost eight-fold. Although in vitro CYP2B6 was the principal determinant of inter-individual variability in the hydroxylation of propofol by liver microsomes , our results did not demonstrate this in vivo. Likely explanations for this discrepancy include the possibility that the contribution of a particular CYP isoform to a biotransformation reaction in vivo is also dependent on substrate affinity and concentration. Propofol is a very high clearance drug; consequently, clearance depends primarily on hepatic blood flow and only minimally on metabolism. Any potential decrease in CYP-mediated oxidative metabolism would have to be extremely large to have an effect on clearance. Moreover, a 20- to 100-fold variation in CYP2B6 expression is likely to result from a variety of extrinsic factors (diet, drug administration). CYP2B6 is a highly polymorphic gene, and any single polymorphism may not be the primary determinant of variability. A recent in vitro study  found no effect of C1459T or any other single polymorphism, but found an effect of the *6B haplotype in individuals with significant alcohol exposure history. Exclusion of subjects with a history of alcohol exposure in the present study, while reducing this as a confounding variable, may have limited the ability to detect a CYP2B6 genetic effect. Finally, besides CYP2B6, additional enzymes such as CYP2A6, 2C8, 2C9, 2C18, 2C19 and 1A2 may play a role, especially when substrate concentrations are high . Glucuronidation and sulphation also contribute to propofol metabolism.
Our results are in line with previous studies [11,25] regarding the observation that women emerge from anaesthesia faster than men. However, this phenomenon cannot be explained by genetic differences at the four major haplotypes at the X-linked GABRE gene. While it remains possible that the one-sixth of patients with prolonged recovery time from propofol are enriched for a genetic variant in GABRE, the chance of such a hypothetical variant not being in disequilibrium with one of the four major haplotypes may be fairly low. It remains of interest to determine what combination of genetic and environmental factors may contribute to the variation in time of emergence from anaesthesia.
1. Meyer UA. Overview of enzymes of drug metabolism. J Pharmacokinet Biopharm
2. Ingelman-Sundberg M. Functional consequences of polymorphism of xenobiotic metabolizing enzymes. Toxicol Lett
3. Watson KR, Shah MV. Clinical comparison of ‘single agent’ anaesthesia with sevoflurane versus target controlled infusion of propofol
. Br J Anaesth
4. Patten D, Foxon GR, Martin KF, Halliwell RF. An electrophysiological study of the effects of propofol
on native neuronal ligand-gated ion channels. Clin Exp Pharmacol Physiol
5. Krasowki MD, Jenkins A, Flood P, Kung AY, Hopfinger A, Harrison NL. General anaesthetic potencies of a series of propofol
analogs correlate with potency for potentiation of gamma-aminobutyric acid (GABA) current at the GABA(A) receptor but not with lipid solubility. J Pharmacol Exp Ther
6. Buggy DJ, Nicol B, Rowbotham DJ, Lambert DG. Effects of intravenous anaesthetic agents on glutamate release: a role for GABAA receptor-mediated inhibition. Anaesthesiology
7. Davies PA, Hanna MC, Hales TG, Kirkness EF. Insensitivity to anaesthetic agents conferred by a class of GABA(A) receptor subunit. Nature
8. Wilke K, Gaul R, Klauck SM, Poustka A. A gene in human chromosome band Xq28 (GABRE) defines a putative new subunit class of the GABA(A) neurotransmitter receptor. Genomics
9. Hachenberg T. Perioperative management with short-acting intravenous anaesthetics. Anaesthesiol Reanim
10. Gan TJ, Glass PS, Sigl J et al
. Women emerge from general anesthesia with propofol
/alfentanil/nitrous oxide faster than men. Anesthesiology
11. Court MH, Duan SX, Hesse LM, Venkatakrishnan K, Greenblatt DJ. Cytochrome P
-450 2B6 is responsible for interindividual variability of propofol
hydroxylation by human liver microsomes. Anesthesiology
12. Lang T, Klein K, Fischer J et al
. Extensive genetic polymorphism in the human CYP2B6
gene with impact on expression and function in human liver. Pharmacogenetics
13. Miksys S, Lerman C, Shields PG, Mash DC, Tyndale RF. Smoking, alcoholism and genetic polymorphisms alter CYP2B6
levels in human brain. Neuropharmacology
14. Irizarry K, Kustanovich V, Li C et al
. Genome-wide analysis of single nucleotide polymorphisms in human expressed sequences. Nat Genet
15. Schaid DJ, Rowland CM, Tines DE, Jacobson RM, Poland GA. Score tests for association between traits and haplotypes when linkage phase is ambiguous. Am J Hum Genet
: 425-434. Epub 2001 Dec 27.
16. Mander AP. Haplotype analysis in population-based association studies. Stata J
17. Sigl JC, Chamoun NG. An introduction to bispectral analysis for the electroencephalogram. J Clin Monit
18. Glass PSA, Bloom M, Kearse L, Rosow C, Sebel P, Manberg P. Bispectral analysis measures sedation and memory effects of propofol
, midazolam, isoflurane and alfentanil in healthy volunteers. Anesthesiology
19. Johansen JW, Sebel PS. Development and clinical application of electroencephalographic bispectrum monitoring. Anesthesiology
20. Hoke JF, Cunningham F, James MK, Muir KT, Hoffman WE. Comparative pharmacokinetics and pharmacodynamics of remifentanil, its principle metabolite (GR90291) and alfentanil in dogs. Pharmacology
21. Smith BE. Pharmacokinetics without calculus - an introduction. Int Anesthesiol Clin
22. Morgan DJ, Campbell GA, Crankshaw DP. Pharmacokinetics of propofol
when given by intravenous infusion. Br J Clin Pharmacol
23. Hesse LM, He P, Krishnaswamy S et al
. Pharmacogenetic determinants of interindividual variability in bupropion hydroxylation by cytochrome P
450 2B6 in human liver microsomes. Pharmacogenetics
24. Guitton J, Buronfosse T, Desage M et al
. Possible involvement of multiple human cytochrome P
450 isoforms in the liver metabolism of propofol
. Br J Anaesth
25. Hoymork SC, Raeder J, Grimsmo B, Steen PA. Bispectral index, predicted and measured drug levels of target-controlled infusions of remifentanil and propofol
during laparoscopic cholecystectomy and emergence. Acta Anaesthesiol Scand