Cannabis use dates back thousands of years, both for medicinal purposes and as a recreational drug . Its recreational use is very common in many countries, especially the United States, Canada and Australia. In 2000, there were an estimated 150 million users worldwide, representing 3–7% of the world's population older than 15 years . In the United Kingdom, the prevalence of regular use among university students is 20% and 10% of medical students admitted to using it weekly or more .
Elimination of cannabinoids (over 60 different compounds derived from the female plant of Cannabis sativa) from the body is a lengthy process and these compounds may be detectable in the tissues for weeks . Given the high percentage of recreational cannabis users in many societies, it is inevitable that patients are regularly anaesthetized while affected by these compounds. Among the UK day surgical population, the prevalence of use in one survey was 14% .
There are several anecdotes suggesting that patients who had high anaesthetic requirements have subsequently admitted to cannabis use . Animal studies have demonstrated cross-tolerance to barbiturates, opioids, benzodiazepines and phenothiazines [1,5]. Similar interactions are thought possible in humans, but to our knowledge no published data or study has investigated the relationship between cannabis use and propofol, or other anaesthetic drug, requirement. We designed a prospective, randomized, single-blinded clinical trial to investigate whether patients who regularly use cannabis have higher propofol requirements at induction of anaesthesia.
In this prospective, randomized, single-blind study, written informed consent was obtained from 60 male patients aged between 18 and 50 years, ASA I or II, and scheduled for day-case general anaesthesia with a laryngeal mask for airway management. The study protocol was approved by the Institutional Ethics Committee, Royal Perth Hospital, Perth, Australia. Patients using opioids, cocaine, amphetamine, benzodiazepines, selective serotonin reuptake inhibitors, antiretroviral drugs or abusing alcohol (>20 standard drinks per week) were excluded. Other exclusion criteria were weight more than 100 kg, cardiovascular disease, and unstable preinduction bispectral index (BIS) values. Patients were recruited to form two cohorts of 30 patients each: cannabis users (group C, defined as regular use at least once per week for at least the past 6 months) or noncannabis users (group NC, no prior use of cannabis). The attending anaesthetist was blinded to which group the patients belonged. Within each of these groups, each patient was allocated to one of five induction doses of propofol – 1.5, 2, 2.5, 3 or 3.5 mg kg−1. Randomization was achieved using a computer-derived random number sequence and allocation by means of opaque sealed envelopes. Prior to intravenous induction of anaesthesia, 500 ml of crystalloid solution (compound sodium lactate) was administered rapidly to reduce the risk of propofol-induced hypotension. Patients were not premedicated. A BIS XP A 2000 monitor (Aspect Medical Systems Inc., Newton, Massachusetts, USA) was attached and preinduction BIS, pulse and blood pressure values recorded. The allocated propofol dose was injected at a rate of 40 mg every 10 s. Thirty seconds after administration, a clinical assessment was made (looking for loss of consciousness and eyelid reflex and a BIS value <60) to determine whether the depth of anaesthesia was adequate to attempt insertion of a laryngeal mask. If the BIS value was greater than 60, induction was deemed to have failed for this endpoint. If the BIS was less than 60, at 60 s after completion of propofol administration, an attempt was made to insert the laryngeal mask and deemed successful if the laryngeal mask could be inserted without reaction from the patient. It was deemed unsuccessful if any coughing, movement, biting, arousal reactions or resistance from the patient occurred when trying to insert the laryngeal mask. If either BIS was greater than 60 or an attempt to insert the laryngeal mask failed because of an inadequate depth of anaesthesia, increments of propofol 0.5 mg kg−1 were given every 30 s until both a satisfactory BIS endpoint of less than 60 and satisfactory conditions for successful insertion of the laryngeal mask were achieved. Blood pressure was recorded using automated noninvasive oscillotonometry at 1 min intervals. Hypotension was treated at the discretion of the attending anaesthesiologist. Following successful insertion of the laryngeal mask, the study was completed.
We calculated that a sample size of 30 patients per group (six patients to receive each of the five doses of propofol initially) would give adequate data on which to perform logistic regression to estimate an ED50 for successful induction of anaesthesia. Continuous variables were expressed as mean (SD) unless otherwise indicated, and P values were calculated from t-tests. Categorical variables were compared by Fisher's exact test or by the χ2 test, as appropriate. The ED50 and ED95 figures and their confidence intervals were calculated from a probit analysis, using the SAS software package (SAS version 8; SAS Institute Inc., Cary, North Carolina, USA, 1999). A P value less than 0.05 was considered statistically significant.
Data from 60 patients [30 cannabis users (group C), 30 nonusers (group NC)] were analysed. Patient characteristics (Table 1) showed that group C patients were younger than group NC patients (P = 0.043). Alcohol use was not significantly different, with 44% of patients admitting to daily intake of alcohol, whereas 15% of patients of both groups did not take any alcohol at all.
The preanaesthetic BIS value was almost identical (group C 97.7 ± 0.7 vs. group NC 97.3 ± 0.9) and the depression of the BIS at 30 s after the initial dose of propofol did not significantly differ between the groups (C 56.2 ± 26.2 vs. NC 51.0 ± 25.2). A BIS value less than 60 after the initial dose of propofol was achieved in 57% of patients in group C and 73% in group NC (P = 0.18). Placement of a laryngeal mask after the initial propofol dose was successful in 27% of patients in group C and 33% of group NC patients (P = 0.57).
The additional dose of propofol needed to achieve a BIS less than 60 did not differ between groups, but the additional dose required before a laryngeal mask could be inserted in those in whom the initial dose was inadequate was significantly different between groups (group C 117.0 ± 97.5 vs. group NC 65.2 ± 65.7 mg, P < 0.02). The total induction dose of propofol for successful insertion of a laryngeal mask was also significantly greater in group C (314.0 ± 109.3 vs. 263.2 ± 69.5 mg group NC, P < 0.04).
The ED50 and ED95 for successful induction were calculated for both groups (Table 2). The wide scatter of data among group C did not allow calculation of the ED50 for successful insertion of the laryngeal mask. The success rate for a BIS less than 60 after 30 s and successful insertion of a laryngeal mask after 60 s, based on initial induction doses of propofol either up to 2.5 mg kg−1 or above this, is shown in Fig. 1. These values did not significantly differ between groups except for a higher success rate for laryngeal mask insertion in group NC patients who received at least 2.5 mg kg−1 (P < 0.05).
Baseline values for mean arterial pressure (MAP) were not significantly different between groups (group C 139.7 ± 21.2 vs. group NC 136.7 ± 22.9 mmHg), and in both groups the lowest MAP was seen 6 min after induction (group C 108.8 ± 17.6 vs. group NC 104.6 ± 12.0 mmHg).
Our null hypothesis stated that there was no difference in the effective dose of propofol to induce anaesthesia in 50% of patients (ED50), when comparing patients who regularly used cannabis and those who were cannabis naive. Our results do not allow us to reject this; however, there were differences between groups that may be clinically relevant.
Once the randomized induction dose had failed to achieve success for one or both of the endpoints, group C patients needed significantly more propofol in order to allow successful insertion of a laryngeal mask. Second, among those patients who were randomized to an initial propofol dose of at least 2.5 mg kg−1, successful induction was less likely, significantly so for induction to achieve successful insertion of a laryngeal mask. In addition, among cannabis users, we were unable to estimate an ED50 and ED95 because of the wide variability in induction success across the initial doses. This suggests that the study may have been underpowered, and a larger study may have demonstrated a significant difference between the two groups.
Cannabis use is endemic in our society and given that its use has been decriminalized in some countries and some illicit use is tolerated in many, its use is unlikely to abate. Thus, any interaction between cannabis and anaesthetic drugs has important implications for patient care. We are unable to find any studies addressing these matters. It is known that cannabinoids have a number of effects, especially within the central nervous system. Cannabinoid receptors identified to date are designated CB1 and CB2. The receptor CB1 is found in the brain as well as in the peripheral nervous system with high concentrations in the hippocampus, frontal and neocortex, olfactory areas, basal ganglia, cerebellum and spinal cord. CB2 receptors are located more peripherally, mainly associated with the immune system . It is thought that anadamide, palmitoylethanolamide and 2-arachidonylglyderyl ether, all arachidonic acid derivatives, are naturally occurring ligands for these receptors . Well documented effects of cannabis in humans include impairment of short-term memory, analgesia and antiemesis. Effects on memory and nausea are thought to be principally mediated via the CB1 receptor, whereas antinociceptive effects appear to involve other receptors.
There have also been reports of coronary ischaemia, myocardial infarction and transitory cerebral ischaemia in healthy young adult cannabis users with normal coronary arteries . The mechanism is not known but several explanations have been proposed, including an increase in heart rate mediated by sympathetic stimulation and reduced parasympathetic activity.
A limitation of the study is that the use of alcohol might have been a confounding factor; however, the evidence that ethanol and propofol show cross-tolerance is far weaker than that for benzodiazepines and ethanol . To our knowledge, there is only one clinical study that has shown differences in propofol requirements as a function of drinking habit . In a recent study by Servin et al., they concluded that chronic alcoholism induces only mild changes in the pharmacokinetics of propofol. An animal study  revealed that chronic ethanol consumption in rats did not affect the action of propofol. Although the groups reported similar alcohol use and we attempted to exclude heavy users, the self-reporting of alcohol use and the level of consumption is notoriously unreliable . Ideally, in a study such as ours, all users of alcohol would have been excluded and this was confirmed by testing, but we did not consider the trial would be feasible in our patient population under those conditions. Nevertheless, the groups were well matched for any potential effect of alcohol on propofol requirement.
It would also have been of interest to have measured the levels of cannabinoids among participants. It is possible that the level of delta9-tetrahydrocannabinol d(delta9-THC) plays a role in the response to propofol and we speculate that variation in patient levels might explain the variability of response we noted among the users of cannabis. We did not specifically question group C individuals on recent use of the drug. This cannabinoid and its metabolites are highly fat-soluble and are retained in the lipid-rich tissues for substantial periods. There are also large interindividual differences in rates of metabolism, which would make any direct relationship even more complex . Further studies focusing on whether a dose-related interaction occurs between delta9-THC and propofol would be interesting.
Our study suggests that regular cannabis users show a more variable response to induction of anaesthesia with propofol and that higher doses of propofol may be required to achieve both loss of consciousness and adequate jaw relaxation and depression of airway reflexes for insertion of a laryngeal mask. Until more information is obtained, it would seem prudent to at least seek a history about cannabis use during routine preoperative assessment.
The present study was funded by the Royal Perth Hospital Department of Anaesthesia and Pain Medicine Research Fund. We thank Aspect Medical Systems, USA, for the loan of the BIS monitor.
1 Ashton CH. Adverse effects of cannabis and cannabinoids. Br J Anaesth 1999; 83:637–649.
2 United Nations Office for Drug Control and Crime Prevention. Global illicit drug trends 2003
. New York: United Nations office for Drug Control and Crime Prevention; 2003.
3 Mills PM, Penfold N. Cannabis abuse and anaesthesia. Anaesthesia 2003; 58:1125.
4 Symons IE. Cannabis smoking and anaesthesia. Anaesthesia 2002; 57:1134–1145.
5 Kumar RN, Chambers WA, Pertwee RG. Pharmacological actions and therapeutic uses of cannabis and cannabinoids. Anaesthesia 2001; 56:1059–1068.
6 Iverson L. Cannabis and the brain. Brain 2003; 126:1252–1270.
7 Culic V. Acute risk factors for myocardial infarction. Int J Cardiol 2007; 117:260–269.
8 McCowan C, Marik P. Refractory delirium tremens treated with propofol
: a case series. Crit Care Med 2000; 28:1781–1784.
9 Fassoulaki A, Farinotti R, Servin F, Desmonts JM. Chronic alcoholism increases the induction dose of propofol
in humans. Anesth Analg 1993; 77:553–556.
10 Servin FS, Bougeois B, Gomeni R, et al
. Pharmacokinetics of propofol
administered by target controlled infusion to alcoholic patients. Anesthesiology 2003; 99:576–585.
11 Inagawa G, Sato K, Kikuchi T, et al
. Chronic ethanol consumption does not affect action of propofol
on rat hippocampal acetylcholine release in vivo. Br J Anaesth 2004; 93:737–739.
12 Dawson DA. Methodological issues in measuring alcohol use. Alcohol Res Health 2003; 27:18–29.