The influence of opioids on gastric function:

experimental and clinical studies

Örebro Studies in Medicine 14

Jakob Walldén

The influence of opioids on gastric function:

experimental and clinical studies

© Jakob Walldén, 2008

Title: The influence of opioids on gastric function:experimental and clinical studies

Publisher: Örebro University 2008www.oru.se

Editor: Maria Alsbjer maria.alsbjer@oru.se

Printer: Intellecta DocuSys, V Frölunda 02/2008

issn 1652-4063 isbn 978-91-7668-583-9

14. 66 .

�

�

15

Introduction

Why, as an anesthesiologist, am I writing a thesis about the gastrointestinal tract?

Shouldn’t I be more interested in the brain and the nerves? As a matter of fact,

the stomach and the intestines play a major role during the perioperative period.

The aim of preoperative fasting is an empty stomach at the start of anesthesia in

order to reduce the risk of pulmonary aspiration. Postoperative nausea and

vomiting (PONV), often termed “the big little problem”, is a major concern.

Postoperative ileus (POI) due to impairments in gastrointestinal motility is

common, and it delays the start of oral feeding and the passage of stool. Patients

sometimes rate gastrointestinal symptoms as more severe than postoperative pain,

and impairment of gastrointestinal function often delays discharge from the

hospital. To answer the initial question, the gastrointestinal tract is of central

importance for both perioperative care and the anesthesiologist. Many factors

contribute to the impairment of perioperative gastrointestinal function (1-4), and

the opioid analgesics are one of the major contributors (5, 6).

The overall objective of this work was to acquire more knowledge about the

mechanism and physiology behind opioid effects on the gastrointestinal system.

The current understanding in the field is limited and much more research is

needed (7). In this thesis I have explored the effects of two opioid drugs, fentanyl

and remifentanil, on gastric motility.

Normal gastric motility

The physiological functions of the stomach are to receive ingested food, mix it

with secretions, mechanically break down the contents and finally pass the

contents to the duodenum (2). The proximal stomach, the fundus, functions as a

reservoir and with volume loads, muscles are adapted for maintaining a

continuous contractile tone (8). The distal antral region exhibits phasic and

peristaltic contractile activity and functions both as a pump and a grinding mill

(9). The tone of the pyloric sphincter regulates the outflow to the duodenum (10).

Two patterns of gastric motility can be distinguished – fasting and postprandial

motility. Fasting motility has a housekeeping function and consists of recurrent

contractile activity, sweeping contents distally in the bowel (11). This pattern is

described by the expression “migrating motor complex” (MMC), which is

characterized by three different phases. MMC phase I starts approximately 2-3

hours after a meal, lasts for 1 hour, and during this phase there are only a few

16

contractions every 5 minutes. In MMC phase II, the frequency of contractions

increases, but they are irregular. Phase III starts after a further 30 minutes, lasts

for about 10 minutes, and the coordinated contraction is maximal with clear

propagation of intraluminal contents. The interdigestive pattern terminates

abruptly after ingestion of food. After a meal, the motility pattern is dependent

on the physical state and nutrient content. The stomach exhibits submaximal and

phasic contractile activity, similar to MMC phase II. The contents are mixed,

digested and portioned out to the duodenum through the pylorus.

Myoelectric activity Gastric smooth muscles display a rhythmic electrical activity, slow waves, with a

frequency of approximately 3 cycles per minute. These slow waves originate from

a gastric pacemaker region in the corpus region of the stomach and propagate

towards the pylorus. With influence of the enteric nervous system and other

regulatory mechanisms, the slow waves trigger the onset of spike potentials,

which in turn initiate coordinated contractions of the gastric smooth muscles

(12). Gastric motility and emptying depend on these slow waves.

Neuronal control Neural networks for control of gastric motility are present both on the enteric

and the central levels. The enteric nervous system (ENS) is a separate division of

the autonomic nervous system and has local circuits for integrative functions

independent of extrinsic nervous control. Many of the reflexes, “programs” and

information processing for motility are located within the ENS. However, the

functional physiology of the stomach is dependent on higher levels of control.

(13)

The parasympathetic vagus nerve is a mixed sensory and motor nerve with 90%

afferent fibers, transmitting sensory information to the brainstem, and 10%

efferent fibers with motor functions (14). In the brainstem, the sensory neurons

are located in the nucleus tractus solitarius (NTS) and the motor neurons are

located in the dorsal motor nucleus of vagus (DMV). The two nucleuses are in

proximity to one another and there are dense networks of interneurons between

them, providing sensory information for the motor output, and completing the

vago-vagal reflex loop. The whole complex is also under the influence of higher

centers and circulating hormones (15, 16). There are two subgroups of the

efferent motor nerves, and the neurons are organized separately within the DMV

17

(16). Cholinergic fibers mediate excitations and non-adrenergic non-cholinergic

(NANC) inhibition to the stomach.

The sympathetic innervations to the stomach originate from the thoracic spinal

cord (T6 to T9). They consist of efferent adrenergic fibers and act mainly through

inhibition of the cholinergic transmission in the stomach (17). Together with

sympathetic afferent sensory fibers, the inhibiting gastro-gastric reflex is formed

(18).

The composition of the contents in the intestines also affects motility. Lipids,

carbohydrates, amino acids, low pH and hyperosmolarity in the duodenum

inhibit gastric motility. The “ileal break”, activated by caloric content in the

ileum, inhibits gastric motility and the glucagon-like peptide-1 (GLP-1) is the

proposed mediator (19). Colonic distension decreases gastric tone (20).

Several endogenous substances are involved in the integrative functions.

Cholecystokinin, released in the ileum in the presence of fatty contents, inhibits

gastric emptying mainly through activation of afferent vagal fibers (21). Ghrelin,

a relatively newly discovered gastric peptide, stimulates appetite, food intake and

gastric motility (22). Somatostatin, released from D-cells found throughout the

gastrointestinal tract, has complex actions on motility (21). Motilin, produced in

the duodenum, stimulates stomach motility through direct activation of motilin

receptors on enteric neurons, leading to activation of cholinergic neurons in the

antral region of the stomach (23).

Gastric tone The proximal part of the stomach acts as a reservoir and exhibits a constant

dynamic tone. It adapts for volume loads, and volume waves portion contents to

the distal part of the stomach. The tone is mainly controlled by the autonomous

nervous system. (8, 18)

Tone is not equivalent to pressure. Gastric tone can be expressed as the length of

the muscle fibers in the proximal stomach. As there is an adaptive relaxant reflex,

a volume load might maintain the same intragastric pressure. Therefore, an

almost empty stomach and a full stomach are able to have the same intragastric

pressure, but different tone.

18

Gastric emptying Gastric emptying (GE), the functional outcome of gastric emptying, is dependent

on the character of the stomach contents. The emptying of liquids starts

immediately and follows an exponential profile, meaning that a certain

proportion of the liquids is emptied during each time interval. Usually the time

for emptying half of the liquid contents is about 15-20 minutes. In contrast, any

caloric content or solids in the food causes a change from the liquid pattern of

emptying. For solid food, there is first a delay in emptying, a so-called lag-phase,

where no contents are passed to the duodenum. Contents are then mixed, grinded

and digested. This lag-phase lasts up to 1 hour and is followed by an emptying

that is characterized as linear, as a certain amount of contents is emptied during

each time interval (24). Posture influences gastric emptying, particularly with long

emptying when patients are in a left lateral position (25-27) .

Methods for measuring gastric motility

Gastric emptying rates can be estimated with various methods. The “gold

standard” is scintigraphic methods with radionucelotide labeled test meals (28).

With ultrasound techniques, transpyloric flow and gastric volumes can be

estimated (29-31). Absorption tests, i.e. the paracetamol method, (32-34), and

hydrogen breath tests (35) measure gastric emptying indirectly. Gastric pressures

are studied with manometry catheters and gastric myoelectric activity with

cutaneous electrogastrography (36). The gastric barostat measures proximal

gastric tone (37).

Opioid drugs

Morphine-like alkaloids have been used for centuries for analgesia and sedation.

Morphine was isolated from the opium flower in the beginning of the 19th

century, and some half century later, with introduction of the needle and syringe

in clinical practice, morphine could be administered in a controlled manner.

Opioid drugs are still fundamental in the treatment of severe pain, and morphine

is the reference substance to which other opioids are compared. Numerous

analogues with various pharmacological profiles have been developed. (38)

Opioids mediate their effect via opioid receptors on cell membranes. Currently,

five classes of receptors are identified, but research in humans has focused on the

role of μ-, κ-, and δ-opioid receptors (39). All three receptor classes are expressed

throughout the nervous systems, including the GI tract, and they partly overlap in

distribution and function. The receptors bind to both exogenous opioids, i.e.

morphine, and endogenous opioid peptides. (40, 41). The receptors differ in their

19

pharmacological profiles and have selectivity for the three classes of endogenous

opioid peptides. Analgesia, as well as many of the side effects, are mainly

mediated via activation of the μ-opioid receptor (MOR) (42, 43).

Morphine-analogue drugs exhibit their actions mainly through the MORs (44).

MORs are widely distributed on cell membranes in the body and are present in

the central nervous system (45-47) including at the spinal level (48), on peripheral

nerves (49), and in the gastrointestinal tract (41, 50-52).

On the cellular level, the MOR is coupled to transmembrane G-proteins. The

cellular effects involve hyperpolarizations of the cell membrane via K+ and Ca++

channels and changes in the second messenger systems (i.e. cAMP, IP3). The

physiological result of receptor activation depends on the site of action, but

synaptic transmission is usually inhibited, i.e. via inhibition of presynaptic

excitatory transmitter release or postsynaptic hyperpolarization (38).

Fentanyl Fentanyl is the most common opioid used in daily anesthetic practice. It is a

MOR agonist with analgesic potency 100 times that of morphine. It is highly

lipophilic and has a high volume of distribution. It is usually given as repetitive

bolus injections during anesthesia. Clearance from the body can be long,

especially after repetitive doses. (53)

Remifentanil Remifentanil is a MOR agonist with analgesic potency similar to that of fentanyl.

It has a rapid onset and recovery and is usually administered as a continuous

infusion. Remifentanil is metabolized by unspecific esterases in the body, has a

relatively small volume of distribution, and has a systemic half life of about 10

minutes. The context sensitive halftime (time to reduce the effect site concentra-

tion to 50%) is around 4 minutes and is independent of infusion time. This gives

remifentanil the unique property that after termination of a several hour long

infusion, such as during anesthesia, remifentanil is rapidly cleared and patients

recover within minutes. Remifentanil is used routinely today in anesthetic

management. (54)

Effect of opioids on gastrointestinal motility

In 1917, Trendelenburg was the first to demonstrate the inhibitory effects of

opioids on motility in an experimental setting using isolated animal intestines (55,

56). Since then, effects of opioids on gastrointestinal motility have been studied

20

extensively, but the mechanism is complex and still uncertain (7). Opioid

receptors are widely spread in the gastrointestinal tract and are located on

neurons in the ENS, on secretomotor neurons and on smooth muscles (57-60).

Opioid receptors mediate the action of both exogenous opioids and endogenous

opioid peptides, and generally speaking opioids suppress neural excitability

through the opening of potassium conductance channels, leading to hyperpolari-

zations of cell membranes. Opioids affect a variety of functions including motility

(41, 51, 52, 61-65) and secretion (57), and both μ- and κ-receptors are involved

(40).

Opioid effects on motility can be both excitatory and inhibitory. The stomach

and the intestines are under tonic inhibitory influence from neuronal networks

controlling the coordinated contraction and propulsions. When opioids inhibit

these inhibitory neurons, control from the neuronal network is released and an

uncoordinated non-propulsive contraction occurs (57). This is seen, for example,

when opioids induce a phase III-like activity in the antroduodenal region and

disturb gastric motility (66). The negative effect is seen even with low doses of an

opioid (67).

There is clear evidence that there is both a peripheral and a central mechanism in

the inhibition of gastrointestinal motility (57, 66, 68). Higher centers involved in

the regulation of gastrointestinal motility also express MORs (46, 69, 70). Opioid

receptors are also present on afferent vagal nerve endings projecting to the NTS

(47, 71, 72).

Role of endogenous opioid peptides

Endogenous opioid peptides (enkephalins, β-endorphins and dynorphines) are

located within the GI tract. The function of these peptides is poorly understood,

but they might play a role in the normal control of motility. The distribution of

enkephalinergic neurons is closely matched to neurons expressing MOR (73).

There is evidence that endogenous opioids are released during stress and trauma

and inhibit the normal patterns of motility. After binding to ligands, the MOR

receptor complex is usually internalized into the cell through endocytosis (41).

Using immunhistochemical methods to demonstrate internalized opioid receptors,

the release of endogenous peptides can be studied. In an animal model,

abdominal surgery with and without manipulation of the intestines was

associated with endogenous peptide release, while anesthesia alone was not. (74)

21

Inhibition of endogenous peptides, i.e. through blockage of opioid receptors,

might therefore act prokinetically under conditions of disrupted motility (75).

Anesthetic drugs and gastric motility

Effects of volatile agents on gastrointestinal motility There are only a few published studies on volatile agents and gastrointestinal

motility and no studies regarding sevoflurane. Marshall et al showed (76) that

halothane depressed motility of the stomach, jejunum and colon in dogs and that

activity returned promptly after the agent was withdrawn. In rodents, halothane

and enflurane had profound but different effects on motility (77). Both agents

reduced the frequencies of the slow waves. Halothane reduced phase III activity in

duodenum, intestinal motor activity was slowed after anesthesia, and contractile

activity was affected. Enflurane increased the frequency of MMC during

anesthesia, the frequency slowed to a normal rate after anesthesia, and there were

no major effects on contractile activity. In humans, enflurane and halothane

depress antral motilty and reduce phase II activity (78). To summarize, volatile

anesthetics affect gastric motility, but the effect may cease quickly after

termination of the agents.

Effect of propofol on gastrointestinal motility Propofol in low doses does not influence gastric motility (79), but there is

evidence that propofol may inhibit motility in higher doses. In a laboratory

setting, propofol inhibited spontaneous contractions in human gastric tissue (80).

Prokinetic drugs

Prokinetic drugs can be used to improve and restore gastrointestinal motility.

Available major drug classes with prokinetic properties include antidopaminergic

agents, serotonergic agents and motilin-receptor agonists. However, the drugs

show signs of moderate prokinetic effects with adverse effects, and research on

novel substances is currently intense. (81, 82)

Metoclopramide is a dopamine receptor antagonist that has been used for

decades. As dopamine inhibits gastric motility (83, 84), blockade of the

dopamine-2 receptor (D2) promotes motility. It is also suggested that metoclo-

pramide has effects on serotonergic receptors. Metoclopramide is widely used in

clinical practice, but the prokinetic effects last for only a short time. Also, the side

effects are considerable, as all D2 receptor antagonists might induce extrapyrami-

dal symptoms. Domperidone has properties that are similar to those of

22

metoclopramide, although the most common side effect is hyperprolactemia. The

substance is available in many European countries, but currently not in Sweden.

Tegaserode is a novel serotonergic agent undergoing clinical evaluation. It is a

partial 5-HT4 receptor agonist and 5-HT2b receptor antagonist (85) and

accelerates orocecal transit in volunteers. It has not been associated with serious

side effects. Cisapride is a 5-HT4 receptor agonist with prokinetic actions in

major parts of the gastrointestinal tract, including the stomach. It stimulates

antral and duodenal contraction and improves gastric emptying. However, the

substance was associated with severe cardiac arrhythmias and was withdrawn

from the market in 2001.

The macrolide antibiotic erythromycin is a motilin receptor agonist and it

initiates MMC phase III, and stimulates motility and gastric emptying through

direct effects in the stomach. Compared to other prokinetic drugs, erythromycin

is considered effective. Novel motilin receptor agonists with higher potency and

without antibacterial activity are under development (86, 87).

μ-Opioid receptor antagonists The classical μ-opioidreceptor antagonist naloxone improves opioid induced

bowel dysfunction (88), but as the reversal also antagonizes the analgesic effect of

opioids, the use of naloxone is limited (89).

Research in recent decades has focused on the development of peripheral μ-

receptor antagonists that do not penetrate the blood-brain barrier. Hence,

analgesic effects of opioids are maintained while gastrointestinal effects are

antagonized. Alvimopan is a selective opioid antagonist with extremely limited

oral absorption, and when given orally it does not cross the blood-brain barrier

(39, 90, 91). Clinical phase III trials have showed that Alvimopan accelerates

gastrointestinal recovery after abdominal surgery without compromising opioid

based analgesia (92-96). Metylnaltrexone (MNTX), a derivate of naltrexone, is a

peripheral opioid receptor antagonist that does not cross the blood brain-barrier

and can be administered by both the oral and the intravenous route (97-99).

MNTX is still under investigation in late clinical trials focusing on opioid induced

obstipation (100). In patients treated with opioids, MNTX reduces orocecal

transit time, induces laxation and is well tolerated (100, 101).

23

Gastrointestinal motility during the perioperative period and intensive care

Preoperative fasting One of the most important preparations for patients before anesthesia and

surgery is an empty stomach. Protective reflexes are abundant during anesthesia,

and if regurgitation or vomiting occurs, contents from the stomach might be

aspirated into the lungs, causing fatal aspiration pneumonitis (102). Previously,

NPO (nil per os) after midnight was a rule, but research during the past 20 years

has changed this dogma (103). Current guidelines state a two-hour fast for fluids

and a six-hour fast for solids in healthy patients undergoing elective procedures

(104-107). However, a spectrum of conditions like trauma, pain, emergency

procedures, diabetes and opioid medication are associated with impaired gastric

motility. If the stomach is not considered empty, special procedures are used

routinely for rapidly protecting the airway during the induction of anesthesia

(108).

Early oral intake Today an early start of oral intake after anesthesia and surgery, sometimes within

hours, is common. However, while there is currently no evidence that early intake

diminishes the duration of postoperative ileus, the routine is not associated with

adverse effects, except an increased risk of nausea (109-112). As opioids given

perioperatively might have residual effects during recovery, this might delay the

start of oral intake. Optimizing opioid administration might therefore be

beneficial.

Postoperative ileus Postoperative ileus (POI) is a transient bowel dysmotility that occurs following

abdominal surgery (1, 113). It encompasses delayed gastric and colonic emptying

and failure in the propulsion of the intestinal contents due to atonic bowel (2),

and generally lasts for several days. Inhibitory neural reflexes, neurotransmitters,

inflammatory mediator release and endogenous and exogenous opioids contribute

to the pathogenesis (5). Activation of nociceptive afferent nerves and sympathetic

inhibitory efferent nerves through the spinal reflex is believed to play a major role

(18, 114). Blockade of these nerves with an intra- and postoperative epidural with

local anesthetic reduces gastrointestinal paralysis and enhances recovery by up to

36 hours compared to analgesia with systemic opioids (115). Recent studies have

shown that the prolonged phase of POI is caused by an enteric molecular

inflammatory response in the segments of the intestines manipulated during

surgery (1). Opioid receptors are also up-regulated with the inflammatory

response (49) and might contribute to the impairment.

24

Postoperative nausea and vomiting (PONV) PONV occurs commonly after anesthesia and is described as the “big little

problem”. Patients often recall PONV as their worst experience after undergoing

a surgical procedure. The etiology is multifactorial, and non-smokers, female

gender, history of PONV or motion sickness and the use of opioids are associated

with increased risks (116). The emetic center in the brainstem is in the proximity

of the NTS and DMV. Nausea and vomiting induce changes in gastric motility

through central mechanisms, but there is currently no evidence that motility

changes in the stomach per se induce PONV. However, factors that induce

motility changes also induce PONV. Therefore, it might be difficult to perform

studies in humans where the aim is to study if an intervention with isolated

gastric effects affects nausea and vomiting.

Intensive Care In critical illness, impaired gastric motility is common and is associated with

serious consequences. The underlying mechanisms are tissue ischemia, distur-

bances in fluid-electrolyte balance, abdominal surgery, infections and medications

(i.e. opioids, catecholamines, anticholinergica) (117). The clinical picture includes

enteral feeding intolerance, gastric retention, and paralytic ileus, and about 50%

of intensive care unit (ICU) patients have delayed gastric emptying. As early

enteral administration of nutrition is considered the best practice, with improved

outcome in morbidity and mortality, efforts have been made to promote gastric

motility in the critically ill (118). Erythromycin and metoclopramide are the most

commonly used prokinetics in the ICU (119). With advantage to erythromycin,

(120) both drugs improve gastric emptying (121, 122). Gastric emptying is even

more improved if the two drugs are combined (123). However, rapid tachyphy-

laxia occurs frequently and limits the use of these two drugs (118). The use of

enteral naloxone has been popular in many ICUs, but at this time there is only

weak evidence in the literature. Meissner found that naloxone reduced gastric

residual volumes and the frequency of pneumonia (124), and Mixides showed

that enteral feeding was better tolerated with naloxone (125). Novel prokinetics

are under evaluation for the ICU setting (126).

Genetic variability

In recent years research regarding individual variability in opioid-mediated

analgesia and in side-effects has suggested an association with genetic disposition.

Genetic variations might alter drug effects through changes in metabolizing

enzymes, transport proteins and expression of cellular receptors. Recently, several

25

studies have focused on single nucleotide polymorphism (SNP) in the gene coding

for the μ-opioid receptor (127-131).

Polymorphisms and mutations are variations of the normal ”wildtype” genetic

expression. If the variant is common in the population (>1%), it is termed

polymorphism, and if it is rare (<1%) it is termed mutation. A single nucleotide

polymorphism (SNP) occurs when a position in the DNA strand has two

alternative nucleotides. As all chromosomes exist in pairs, a subject is heterozy-

gote for a SNP if one of the genes carries the variant, and homocygote if both

genes are variants.

One of the most common SNPs in the MOR gene is a change in the nucleotide

base of A>G at position 118 (A118G). This results in an amino acid exchange

from aspargine > aspartate at position 40 (Asn40Asp) in the receptor (127, 131).

The expected frequencies in populations of heterozygous A118G and homozy-

gous subjects are 20 and 2 %, respectively, and there are substantial variations

between ethnic groups (132, 133).

The A118G alteration results in a loss of a putative glycosylation site of the

receptor (134). Investigators report up to 3 times higher affinity to beta-

endorphins for the variant (132), altered signal transduction pathways, and lower

thresholds for morphine in neurone models (135). In contrast, others have

reported no differences in ligand-binding or dose in the cellular response with the

variant (136). The differences might be explained by the use of different cell lines.

Subjects carrying the A118G variant have a diminished pupillary response to the

morphine metabolite morphine-6-glukoronide (M6G) (137). Observations in

patients with renal failure (causes accumulation of M6G) indicate that the variant

decreases side effects and the potency of M6G. There have been speculations

about an M6G toxicity protection by the A118G variant (137). Others report

that analgesic response to M6G is diminished in variants, while respiratory

response (depression) is unchanged (138).

In experimental settings, A118G carriers have a higher threshold for pain (139).

Clinical studies reveal decreased postoperative sensitivity to morphine after knee

arthroplasty (140), carriers of A118G required more morphine for alleviation of

pain caused by malignant disease (141), while there were no differences in opioid

consumption after abdominal hysterectomy (142). After abdominal surgery there

26

was a tendency toward higher morphine consumption with the variant (143).

Interestingly, there is speculation about an association with opioid induced

nausea and vomiting, as carriers of the variant had less symptoms after after

exposure to M6G (144). Compared to a control group, patients switching to

alternative opioids from morphine due to intolerance did not differ in the MOR

gene (145).

The SNP A118G has also been explored in the context of substance addictions.

Regarding alcohol intake, carriers of the variant had a higher sensitivity to

alcohol, became more stimulated and sedated than normal “wildtypes” (146),

and also had a stronger urge to drink more (147). Naltrexone, an opioid receptor

antagonist, blunted the alcohol effect more in subjects with the variant (148).

Some investigators report an association between alcohol dependence and the

variant (146, 149, 150). However, a recent meta analysis concluded that there

was no evidence for such an association (151).

Opioid systems are also believed to inhibit the hypothalamic-pituitary-adrenal

axis (HPA-axis). A blockade of this opioidergic effect releases cortisol. In

response to the MOR antagonist naloxone, the cortisol response was higher

among carriers of the variant (152).

27

Aims of the thesis

- To study effects of the opioid remifentanil on gastric emptying and

evaluate if extreme postures affect gastric emptying.

- To compare postoperative gastric emptying between a remifentanil-

propofol based total intravenous anesthesia and an opioid free se-

voflurane inhalational anesthesia.

- To study effects of remifentanil on proximal gastric tone using a

gastric barostat.

- To study effects of fentanyl on gastric myoelectric activity using a

cutaneous multichannel electrogastrograph (EGG).

- To test the hypothesis that single nucleotide polymorphisms (SNP)

in the μ-opioid receptor gene are associated with the variable effects

on gastric motility caused by opioids.

29

Materials

All studies were approved by the Ethics Committee of Örebro County Council

(prior to 2004) and the Uppsala Regional Ethical Review Board (after 2004).

Study II was also approved by the Swedish Medical Product Agency. All studies

were performed at Örebro University Hospital, Örebro, Sweden, during the

period 2000-2005.

Study I Ten healthy male volunteers (ASA-class I-II) with a mean age of 23.9 years

(range, 21-31) underwent four gastric emptying studies on four separate days.

Study II Fifty patients (ASA-class I-II) undergoing day-case laparoscopic cholecystectomy

were randomly allocated to receive either total intravenous anesthesia with

propofol-remifentanil (TIVA, n=25) or opioid-free inhalational anesthesia with

sevoflurane (GAS, n=25). Five patients (TIVA, n=4, GAS, n=1) were excluded for

perioperative surgical reasons. Postoperative data were analyzed for 21 subjects

in the TIVA group (mean age 45 years, (range 29-64)), females, n= 20) and 24

patients in the GAS group (mean age 46 years, (range 19-69)), females, n= 20).

The gastric emptying study was successful in 18 patients in the TIVA group and

20 patients in the GAS group.

Study III Ten healthy male volunteers (ASA-class I-II) with a mean age 24 years (range, 19-

31) underwent two gastric tone studies on two separate days. Later, analyses of

SNP in the MOR gene were performed. Two subjects did not complete the first

barostat study (glucagon) and 1 subject did not complete the second barostat

study (remifentanil). Genetic analyses were performed in all subjects (n=9) with

successful gastric tone measurements.

Study IV Gastric myoelectric activity was studied with an electrogastrograph in 20 patients

scheduled for elective surgery (ASA-class I-II, mean age 45 years (range, 28-67),

females, n=16) and the effect of a bolus dose of fentanyl 1μg/kg was evaluated.

Later, genetic analyses of SNP in the MOR gene were performed in 18 of the

patients.

31

Methods

Gastric emptying (I-II)

Gastric emptying was studied with the paracetamol method. (Acetaminophen is

the name for paracetamol in North America and in the literature the method is

also called the acetaminophen method). Paracetamol is not absorbed from the

stomach, but is rapidly absorbed from the small intestine. Consequently, the rate

of gastric emptying determines the rate of absorption of paracetamol adminis-

tered into the stomach (32).

Paracetamol 1.5 g dissolved in 200 mL of water (at room temperature) was given

orally (Study I) or through a nasogastric tube (Study II). Blood samples were

taken from an intravenous catheter prior to the administration of paracetamol, at

5, 10, and 15 minutes after administration, and then at 15-minute intervals

during a total period of 120 min. Serum paracetamol was determined by an

immunologic method including fluorescence polarization (TDx acetaminophen®;

Abbott Laboratories, Chicago, IL, USA). Paracetamol concentration curves were

produced and the maximal paracetamol concentration (Cmax), the time taken to

reach the maximal concentration (Tmax), and the area under the serum paraceta-

mol concentration time curves from 0 to 60 minutes (AUC60) and from 0 to 120

minutes (AUC120) were calculated. Tmax was assumed to be 120 minutes if no

paracetamol was detected in any sample. The paracetamol method is a well-

accepted method for studying the liquid phase of gastric emptying, and AUC60

correlates well with measures of gastric emptying performed using isotope

techniques (32, 153)

Gastric tone (III)

Gastric tone was measured by an electronic barostat (SVS®; Synetics AB,

Stockholm, Sweden). The gastric barostat is an instrument with an electronic

control system that maintains a constant preset pressure within an air-filled

flaccid intragastric bag by momentary changing of the volume of air in the bag

(37, 154). When the stomach contracts, the barostat aspirates air to maintain the

constant pressure within the bag, and when the stomach relaxes, air is injected.

The pressure in the bag was set at 2 mmHg above the basal intragastric pressure.

The pressure change at which respiration is perceived on the pressure tracing,

without an increase or decrease in the average volume, is the basal intragastric

pressure.

32

The bag, made of ultrathin polyethylene, has a capacity of 900 ml and is

connected to the barostat by a double-lumen 16 Ch gastric tube. The barostat

measurements followed the recommendations presented in a review article by an

international working team and the barostat instrument fulfilled the criteria

determined by this group (37).

Before the gastric intubation, propofol 0.3 mg/kg was given for sedation. Previous

studies in volunteers have shown that this dose of propofol does not influence

gastric tone (155), and it was given at least 30 min before the study started. The

intragastric bag was folded carefully around the gastric tube and positioned in the

gastric fundus via oral intubation. Thereafter, the gastric bag was unfolded by

being slowly inflated with 300 ml of air under controlled pressure (<20mmHg),

and the correct position of the bag was verified by traction of the gastric tube.

During the measurements, the mean gastric volume during each five-minute

interval was calculated.

Electrogastrography (IV)

Electrogastrography (EGG) is the cutaneous recording of gastric myoelectrical

activity, and the activity is closely associated with gastric motility (156). Gastric

smooth muscles display a rhythmic electrical activity, slow waves, with a

frequency of approximately 3 cycles per minute. These slow waves originate from

a gastric pacemaker region in the corpus and propagate towards the pylorus.

With influence of the enteric nervous system and other regulatory mechanisms,

the slow waves trigger the onset of spike potentials, which in turn initiate

coordinated contractions of the gastric smooth muscles (12). Gastric motility and

emptying depend on these slow waves.

Figure 1. Electrode placements in electrogastrographic study: Electrode 3 was placed halfway between the xyphoid process and the umbilicus. Electrode 4 was placed 4 cm to the right of electrode 3. Electrodes 2 and 1 were placed 45 degrees to the upper left of electrode 3, with an interval of 4 to 6 cm. The ground electrode was placed on the left costal margin horizontal to electrode 4. The reference electrode (electrode 0) was placed at the cross point of the horizontal line containing electrode 1 and the vertical line containing electrode 3. (Walldén et al, Acta Anest Scand 2008. In Press.)

33

Six EGG electrodes were placed on the abdomen after skin preparation. The

electrodes consisted of four active electrodes, one reference electrode and one

ground electrode, as illustrated in Figure 1. A motion sensor was also attached to

the abdomen. We used the Medtronic Polygram NET EGG system (Medtronic

A/S, Denmark) for the simultaneous recording of four EGG signals. Our EGG

system was configured to accept an electrode impedance of less than 11 kΩ after

skin preparation. The EGG signal was sampled at ~105 Hz, and this was

downsampled to 1 Hz as part of the acquisition process (157).

All EGG tracings were first examined manually by two of the investigators (JW,

GL). Prior to the analysis, motion artifacts in the EGG signal, indicated by the

motion sensor, were identified and removed manually. For each patient, the EGG

channel with the most typical slow-wave pattern during baseline recording

(before fentanyl) was selected for further analysis.

An overall spectrum analysis was performed on each of the two main 30-minute

segments (before and after fentanyl, respectively) of the selected channel using the

entire time-domain EGG signal (157). Sequential sets of measurement data for

256s with an overlap of 196s were analyzed using fast Fourier transforms and a

Hamming window for the calculation of running power spectra. When the entire

signal was processed, the power spectra for each segment were combined to

arrive at the overall dominant frequency (DF) and power of the dominant

frequency (DP).

The EGG segments and the spectral analysis after fentanyl were further classified

either as 1) Unaffected EGG (no change in DF after fentanyl), 2) Bradygastric

EGG (decrease in DF >=0.3 after fentanyl) or 3) Flatline-EGG (total visual

disappearance of a previously clear sinusoidal 3 cpm EGG-curve after fentanyl)

(see example in figure 3) without any quantifiable DF. When DF was not

quantifiable, DF was set to 0.

Data from the baseline EGG were compared to data from a previous multicenter

study in normal subjects (157) to test if the group in study IV was similar to a

normal population.

Genetic Analysis (III-IV)

Due to the large interindividual variations in the gastric tone response after

remifentanil, we investigated if this variation could be explained by genetic

34

variability, polymorphisms, in the μ-opioid receptor gene. After reviewing the

literature, we decided to analyze polymorphisms with relative high frequencies

and with reports of altered responses. Therefore, we focused on the μ-opioid

receptor gene polymorphisms A118G and G691C (128).

DNA collection and purification. Venous blood (10 ml) was collected from the subjects in EDTA tubes and the

samples were stored frozen at –70°C. Genomic DNA was purified from

peripheral leukocytes in 1 ml of EDTA blood on a MagNA Pure LC DNA

extractor, using the MagNA Pure LC Total Nucleic Acid Isolation Kit – Large

Volume (Roche Diagnostics Corporation, Indianapolis, IN, USA).

Genotyping. Genotyping was performed at CyberGene AB, Huddinge, Sweden. The A118G

SNP in Exon 1 and the IVS2 G691C SNP in Intron 2 were genotyped using

polymerase chain reaction amplification and sequencing. Oligonucleotide primers

(forward: 5'-GCGCTTGGAACCCGAAAAGTC; reverse: 5'-CATTGAG

CCTTGGGAGTT) and (forward: 5'-CTAGCTCATGTTGAGAGGTTC; reverse:

5'-CCAGTACCAGGTTGGATGAG) were used for amplifying gene fragments

containing Exon 1 and Intron 2, respectively. PCR conditions comprised an initial

denaturing step at 95°C for 1 min followed by 30 cycles at 94°C for 1 min,

annealing at 47.2-53.4°C (depending on primer) for 1 min and extension at 68°C

for 3 min, and a final extension at 68°C for 3 min. The amplified fragments were

sequenced using the same primers with the addition of Rev 1-2 5'-

TTAAGCCGCTGAACCCTCCG and the BigDye Terminator v1.1Cycle Sequence

Kit (Applied Biosystems, Foster City, CA, USA). PCR amplification and sequence

reactions were done on ABI GeneAmp 2400 and 9700 (Applied Biosystems).

Sequence analysis was first done on MegaBACE 1000 (Amersham Biosciences,

CA, USA) and then confirmed with ABI 377XL (Applied Biosystems).

Procedure

Study I In a randomized order, gastric emptying was studied on four occasions in each

subject, with at least 1 day between occasions. The subjects were given a

continuous infusion of remifentanil on two occasions while lying either on the

right lateral side with the bed in a 20º head-up position (RHU) or on the left

lateral side with the bed in a 20º head-down position (LHD). On the other two

occasions, no remifentanil infusion was given, and the subjects were studied lying

in the two positions.

35

All subjects fasted for at least 6 h before each study. For the two occasions with

remifentanil, remifentanil was given as a continuous intravenous infusion in a

dose of 0.2 μg· kg-1·min-1 and was started 10 minutes before the ingestion of

paracetamol. The infusion was terminated directly after the last blood sample

(120 min) was drawn.

Study II All patients fasted according to clinical guidelines (107) and were premedicated

with midazolam 1-2 mg I.V. Before induction, all patients received ketorolac 30

mg I.V. In the TIVA group, anesthesia was induced with an infusion of

remifentanil 0.2 μg·kg-1·min –1, followed after 2 minutes by a target-controlled

infusion (TCI) of propofol at 4 μg·mL –1 (induction time 60 seconds). In the GAS

group anesthesia was induced with 8 % sevoflurane via a facial mask. After

attaining an adequate level of anesthesia, muscular relaxation was obtained in

both groups with rocuronium 0.6 mg·kg-1 IV and the trachea was intubated after

90 seconds. In the TIVA group anesthesia was maintained with remifentanil 0.2

μg·kg-1·min –1 and TCI propofol adjusted (2-4 μg·mL –1 ) to maintain a BIS-index

below 50. In the GAS group anesthesia was maintained with sevoflurane,

adjusted to maintain a BIS-below 50. No prophylactic antiemetics were given. A

nasogastric tube was placed in all patients during anesthesia. At the end of

surgery, 20 mL of 0.25% levobupivacaine was infiltrated at the insertion sites of

the laparoscopic instruments, muscular relaxation was reversed with neostigmine

2.5 mg/glycopyrrolate 0.5 mg, and anesthetic agent(s) were terminated. The

patients were extubated in the operating room after return of consciousness and

spontaneous breathing and transferred to the adjacent day-care unit for recovery.

Except for the continuous infusion of remifentanil in the TIVA group, no opioids

were given during anesthesia. The gastric emptying study measurement was

initiated immediately after arrival in the day-care unit.

Patients stayed in the day-care unit for at least 4 hours and PONV and pain

parameters were evaluated every hour. After discharge, patients received a

questionnaire regarding the postoperative 4-24 hour-period, and they rated their

maximal pain and maximal nausea and were questioned about vomiting. In

addition, a telephone interview was performed on the first postoperative day.

Combining the results, we received postoperative data on PONV, maximal VAS-

score for pain, and time to first postoperative opioid analgesic for the time

periods 0-2 hours and 2-24 hours postoperatively.

36

Study III All subjects fasted for at least 6 h before each study. Each subject underwent two

study protocols on two separate days. Before the gastric intubation the subjects

received a bolus dose of propofol (0.3 mg/kg IV). In the first study, the effect of

glucagon on gastric tone was measured. In the second study, gastric tone was

measured during and after a remifentanil infusion and, after a washout-time of 30

minutes, during readministration of remifentanil in combination with glucagon.

The study protocol is illustrated in Figure 2.

During study situations, vital parameters, blood-glucose, nausea and any other

symptoms were recorded.

Later, subjects (n=9) were asked to participate in the genetic analysis of the MOR

gene and we obtained blood samples.

Figure 2 Schematic illustration of the study protocol in study III.

0 min 15 min

Glucagon 1 mg

Propofol0.3 mg ·kg-1

-10 min-40- -80 min

Measurement of Gastric Tone

0 min 75 min15 min 45 min30 min 85 min 95 min

Start Remifentanil Stop Remifentanil Glucagon 1 mg Stop Remifentanil

Start RemifentanilPropofol0.3 mg ·kg-1

-10 min-40- -80 min

Measurement of Gastric Tone

Glucagon Study

Remifentanil Study

0.1�g·kg-1·min-1 0.2 �g·kg-1·min-1 0.3 �g·kg-1·min-10.3 �g·kg-1·min-1Remifentanil

Glucagon

Glucagon

37

Study IV The study was performed in a pre-anesthesia area before the induction of

anesthesia. Patients fasted for at least 6 hours from solid foods and 2 hours from

clear fluids. No premedication was given. While the patient was lying in a

comfortable bed rest position, an intravenous line was inserted and the EGG

recordings were initiated. After achieving a stable EGG signal, a 30-minute

baseline EGG recording was collected. Without discontinuation of the EGG

recording, 1 μgram•kg-1 of fentanyl was given as an intravenous bolus through

the intravenous line and the EGG recording continued for another 30 minutes.

Charts and notes from the recovery unit were reviewed and we collected data

regarding analgesic and antiemetic requirements.

Later, patients were asked to participate in the genetic analysis of the MOR gene

and we obtained blood samples.

Statistics

The significance level was set at 5% in all tests. Data are presented as means (SD)

or medians (ranges).

In study I, repeated-measures analysis of variance was used for overall differences

between the study situations. If the analysis of variance showed differences, a

paired Student’s t-test with Bonferroni Correction was performed between the

study situations.

In study II, the unpaired Student’s t-test was used for comparisons between the

groups of primary outcome variables. For the secondary outcome variables, the

unpaired Student’s t-test, the Mann Whitney U test or Fisher’s exact test were

used.

In study III, repeated-measures analysis of variance was used for overall changes

in gastric tone over time. For comparisons between time periods, Fisher’s PLSD

was used.

In study IV, Wilcoxon’s signed rank test and the 95% confidence interval for the

difference between the medians were used for analysis of the primary EGG

outcome variables. The unpaired t-test was used for the comparison of baseline

EGG data with the historical controls. Fisher’s exact test was used to test

associations between PONV parameters and EGG outcome.

39

Results

Gastric emptying during an infusion of remifentanil and the

influence of posture (I).

Infusion of remifentanil delayed gastric emptying. During the control situations

there were differences in gastric emptying variables between the two extreme

positions, but there were no differences during the infusion of remifentanil (Table

1 and Figure 3). In three subjects, the dose of remifentanil had to be reduced due

to side effects.

Immediate postoperative gastric emptying after total intravenous remifen-

tanil-propofol based anesthesia (II).

There were no differences in postoperative gastric emptying between the TIVA

group and the GAS group. Both groups differed significantly from a pooled

historical control group. However, there was great variability within both study

groups (Table 1 and Figure 4).

Table 1. Gastric emptying variables in study I and study II. AUC60 Cmax Tmax

min μmol mL-1 μmol mL-1 min

Study I (n=10)

Control RHU 5092 (1125) 138 (45) 25 (14)

Control LHD 3793 (1307) 94 (30) 47 (22)

Remifentanil RHU 962 (902) 34 (24) 94 (33)

Remifentanil LHD 197 (128) 16 (14) 109 (10)

Study II

TIVA (n=18) 2458 (2775) 71 (61) 53 (55)

GAS (n=20) 2059 (2633) 81 (37) 83 (41)

Historical controls (n=36) 5988 (1713) 155 (46) 29 (15)

AUC 60 Cmax Tmax Paired T-test with Bonferroni Correction RHU-Remi vs RHU-Control p<0.0001 p< 0.0001 p<0.0001 RHU-Remi vs LHD-Remi NS NS NS RHU-Remi vs LHD-Control p<0.0001 p < 0.0001 p<0.0001 RHU-Control vs LHD-Remi p<0.0001 p < 0.0001 p<0.0001 RHU-Control vs LHD-Control p<0.0083 p <0.0083 NS LHD-Remi vs LHD-Control p < 0.0001 p<0.0001 p<0.0001 Unpaired t-test TIVA vs GAS NS NS NS TIVA vs Historical Controls p<0.001 p<0.001 p<0.001 GAS vs Historical Controls p<0.001 p<0.001 p<0.001

40

Figure 3 Gastric emptying in study I. Mean (SD) concentrations of paracetamol over time.

0

20

40

60

80

100

120

140

160

180

200

0 30 60 90 120

Time (minutes)

Mea

n S-

Para

ceta

mol

con

cent

ratio

n (�

mol

/L) (

SD)

Control- Right lateral side head upControl- Left lateral side head downRemifentanil 0.2�g/kg/min- Right lateral head upRemifentanil 0.2�g/kg/min- Left lateral head down

Figure 4 Gastric emptying in study II. Mean (SD) concentrations of paracetamol over time.

0

50

100

150

200

0 30 60 90 120Minutes

Mea

n (S

D) S

-Par

acet

amol

con

cent

ratio

n (�

mol

/L)

Group TIVA (n=18)

Group GAS (n=20)

Historical Controls (n=36)

41

Gastric tone after injection of glucagon (III)

Glucagon decreased gastric tone in all subjects during the glucagon study. During

the remifentanil study and the ongoing remifentanil infusion, only one subject

had a decrease in gastric tone after the injection of glucagon, while the others

were almost unaffected. (Figure 5).

Gastric tone during an infusion of remifentanil (III)

There were distinct responses in gastric tone during the remifentanil infusion.

However, the responses were variable. Four subjects responded to remifentanil

with a marked increase in gastric tone (decreased volume in bag) that returned to

baseline levels during washout. Four subjects responded to remifentanil with a

marked decrease in gastric tone (increased volume) and maintained a low gastric

tone during the washout period. In one subject (no. 5) gastric tone was almost

unaffected. The mean gastric tone was significantly lower during the washout

period than before starting the infusion. During the readministration of

remifentanil, there were increases in gastric tone among subjects who increased in

tone during the previous remifentanil infusion. The subject with unaffected

gastric tone during the previous infusion increased in gastric tone. The subjects

who maintained a low gastric tone during washout continued to maintain a low

gastric tone. (Figure 5)

Figure 5. Individual gastric volumes in the barostat study (III).

0

200

400

600

800

1000

Bas

elin

e 0

- 5 m

in

5 -

10 m

in

Glu

cago

n

0 -

5 m

in

5 - 1

0 m

in

10 -

15 m

in

Bas

elin

e 0

- 5 m

in

5 - 1

0 m

in

Rem

i 0.1

0

- 5

min

5 - 1

0 m

in

10 -

15 m

in

Rem

i 0.2

0

- 5

min

5 - 1

0 m

in

10 -

15 m

in

Rem

i 0.3

0

- 5

min

5 - 1

0 m

in

10 -

15 m

in

Was

hout

0

- 5

min

5 - 1

0 m

in

10 -

15 m

in

15 -

20 m

in

20 -

25 m

in

25 -

30 m

in

Rem

i 0.3

0

- 5

min

5 - 1

0 m

in

Glu

cago

n

0 -

5 m

in

5 - 1

0 m

in

Was

hout

0

- 5

min

5 - 1

0 m

in

Time

Intr

agas

tric

Bag

Vol

ume

(ml)

Subj 1

Subj 2

Subj 3

Subj 4

Subj 5

Subj 6

Subj 7

Subj 8

Subj 10

Glucagon Studyn=8

Remifentanil Studyn=9

42

Electrogastrography (IV)

Compared to historical controls (157), there were no differences in the baseline

EGG variables.

After the administration of intravenous fentanyl, there was a significant reduction

in both the dominant frequency (DF) and the dominant power (DP) of the EGG

spectra (Figure 7).

Among patients with a flatline-EGG (n=6), the median (range) time from the

administration of intravenous fentanyl to the observed disappearance of the slow

waves was 5 (1-9) minutes. In 5 of these patients, there was reappearance of the 3

cpm slow-wave EGG pattern 30 (29-35) minutes after the administration of

fentanyl.

There was large variation between patients in the response to intravenous

fentanyl. EGG recordings were unaffected in 8 patients, 5 patients developed a

slower DF (bradygastria) and in 6 patients the slow-wave tracings disappeared

totally (flatline-EGG). For an illustration of the effect, see Figure 6.

Figure 6. An individual electrogastrographic response to Fentanyl.

-80

-60

-40

-20

0

20

40

60

80

0:00:00 0:10:00 0:20:00 0:30:00 0:40:00 0:50:00 1:00:00

Time (min)

�V

Fentanyl 1�g/kg I.V.

-80

-60

-40

-20

0

20

40

60

80

00:0400:5300:0300:52Time (min)

�V

Fentanyl 1�g/kg I.V.

Slow-waves 3cpm Disappearance of Slow-waves

-80

-60

-40

-20

0

20

40

60

80

00:0400:5300:0300:52Time (min)

�V

Fentanyl 1�g/kg I.V.

Slow-wavevv s 3cpm Disappearance of Slow-wavevv s

43

Figure 7. Changes in the dominant frequency (DF) and the dominant power (DP) of the electrogastro-graphic spectra after Fentanyl. (Walldén et al. Acta Anest Scand, 2008. In Press)

25

30

35

40

45

50

55

Baseline After Fentanyl 1�g/kg

Dom

inan

t Pow

er (d

B)

*

0

0,5

1

1,5

2

2,5

3

3,5

Baseline After Fentanyl 1�g/kg

Dom

inan

t Fre

quen

cy (c

pm)

*

A B

Genetic study (III-IV)

We found no association between the variable outcome in studies III and IV and

the presence of SNP A118G or G691C in the MOR (Table 2).

Table 2. Results from the determinations of SNPs in the MOR gene with correlations to outcome groups in studies III and IV. 118 A>G genotype IVS2 + 691 G>C genotype Wild Type Hetero- Variant Wild Type Hetero- Variant

zygous zygous (AA) (AG) (GG) (GG) (GC) (CC)

Study III n=7 n=2 n=0 n=5 n=2 n=1

Increased tone (n=4) 4 3 1 Unchanged tone (n=1) 1 1 Decreased tone (n=4) 3 1 1 2 1 Study IV n=15 n=2 n=1 n=0 n=14 n=4

Unaffected EGG (n=6) 5 1 6 Bradygastria (n=5) 4 1 2 3 Flatline (n=6) 5 1 5 1 Excluded from the 1 1 EGG-analysis (n=1)

No associations found between presence of polymorphism and gastric outcome (Chi-Square tests).

PONV (I-IV), other side effects (I-IV) and postoperative pain (II).

In study I, six subjects experienced nausea, three subjects vomited and six subjects

had pruritus during infusion of remifentanil. Seven subjects experienced

44

dysphagia during remifentanil and five subjects complained of headache during

and/or after the infusion of remifentanil.

In study II, the postoperative incidence of nausea and vomiting was high. During

the 0-24 h postoperative period, 16 patients (76%) in the TIVA group and 20

(83%) patients in the GAS group experienced PONV symptoms. However, there

were no significant differences between the groups. There were shorter times to

rescue analgesics in the TIVA group (median 17 minutes) compared to the GAS

group (median 44 minutes).

In study III, 62% (n=5) of the subjects experienced nausea during the glucagon

experiment. During remifentanil, 33% (n=3) experienced nausea and 66% (n=6)

had nausea with the combination of remifentanil and glucagon. Further, during

remifentanil, 77% (n=7) had pruritus, 33% (n=3) had headache and 22% (n=2)

reported dysphagia.

In study IV, the incidence of PONV in the recovery unit was 53% (n=10) and

there was a need for rescue antiemetics in 47% (n=9) of the patients. We found

an association between flatline/bradygastric EGG and the requirement for rescue

analgesics (P=0.02).

45

Discussion

In this thesis I have studied the physiologic effects of opioid drugs on gastric

motility using both standard and novel methods. With the genetic analyses of the

μ-opioid receptor gene, I have introduced new aspects in the field of opioid

induced gastrointestinal motility disturbances.

As expected, opioids had a pronounced effect on gastric motility. Gastric

emptying was delayed, gastric tone altered and there were changes in the EGG

recordings. However, there was great interindividual variability and the

variability could not be explained by genetic variations in the μ-opioid receptor.

Further, we found no difference in postoperative gastric emptying between an

opioid based and opioid free anesthesia, and we suggest that other factors than

opioids contribute to affecting gastric motility.

The Paracetamol method

In studies I-II we used the paracetamol method to study the liquid phase of gastric

emptying. Paracetamol is absorbed in the proximal part of the small intestine,

and as gastric emptying is considered the rate limiting step in the absorption

profile, variables calculated from the paracetamol plasma concentration curve can

be used to describe the emptying rate from the stomach (33). Nimmo et al

showed in 1975 that the area under the concentration curve during the first 60

minutes (AUC60) correlated well with “gold standard” scintigraphic methods

(32). A recent systematic review concluded that the paracetamol method is well

correlated to scintigraphic assessments of gastric emptying (153), and in our

studies we used the validated variables AUC60, AUC120, Tmax and Cmax to

describe gastric emptying. However, it has been suggested that other variables

might be even more accurate, i.e. the ratio between concentrations at two time

points, C(2t)/C(t) or the ratio between two AUC at two time points. Then only

absorption and elimination constants influence the results, and differences

between individuals in volume of distribution, dose and first-passage metabolism

are eliminated (153, 158). It might be valuable to add these variables in future

studies. Also, the use of a salivary instead of a venous sample for the measure-

ment of paracetamol has been proposed, but the method still needs validation

(159). There are also suggestions that studies with the paracetamol method

should be done with crossover designs to reduce the influence of variability

between individuals in pharmacokinetic parameters (158). This might be taken

46

into account in experimental studies, but it would be difficult in clinical

postoperative studies as the surgical procedure cannot be repeated.

The paracetamol method is a simple and cheap bedside method for the evaluation

of gastric emptying, but it is important to remember that it is an indirect

quantification of gastric emptying with limitations regarding interpretation.

Gastric barostat

In study III we used the gastric barostat for the measurement of proximal gastric

tone. It could be difficult to understand the concept of gastric tone, and it is

therefore important to distinguish it from gastric pressure. The smooth muscles in

the proximal stomach have the ability to generate a constant contraction (also

called a tonic contraction) and with that contraction the gastric wall applies a

certain pressure to the intraluminal contents. As the stomach adapts for volume

loads, the smooth muscles are elongated through diminished contraction and the

intraluminal pressure is maintained. This regulation with sustained muscular

activity is referred to as gastric tone (8), and to simplify, changes in gastric tone

are changes in the length of the smooth muscles.

The gastric barostat is the standard method for the evaluation of gastric tone, and

there are currently no other good methods available. The technique is invasive

and involves the introduction of a bag into the stomach (160), and this might

interfere with the response. However, the bag resembles a load of food and we

can consider it as partly physiological. The gastric barostat method is most

commonly used in research regarding the accommodation response, i.e. in the

field of dyspepsia, and usually subjective discomfort and compliance are

evaluated while the bag in the proximal stomach is distended (37, 160). It is

important to point out that we did not perform any distension tests and that we

did not study the accommodation response. We maintained a fixed, relatively low

pressure in the bag and studied effects of an opioid on gastric tone at a specific

pressure level. It might be interesting to perform distension tests with opioids, but

we consider this difficult with remifentanil and other potent opioids as their

analgesic effects blunt the perceptions and might harm the stomach if pressure is

elevated too high.

We found great variability in gastric tone during the remifentanil infusion. We do

not believe this was due to a methodological problem with the gastric barostat.

During the glucagon part of the study all subjects responded with a clear decrease

47

in tone (increased volume). This validates that the gastric barostat was working

properly, since an expected relaxant stimulus, glucagon, decreased the tone in all

subjects. Also, the same barostat equipment and setup were used in previous

studies by our group (84, 155) and we did not observe this kind of variation.

Electrogastrography (EGG)

In study IV we used cutaneous electrogastrography to study gastric myoelectric

activity. This activity is characterized as a constant ongoing fluctuation of the

membrane potential in the syncytium of the gastric smooth muscle cells.

Specialized smooth muscle cells without contractile properties, interstitial cells of

Cajal (ICC), are responsible for the distribution and propagation of the electric

activity. The pace of the fluctuations is normally determined by ICCs in the

corpus region, and the electric potential is propagated distally. These electrical

fluctuations are called gastric slow waves and they usually have a frequency of

about 3 cycles per minute. (156, 157, 161-163) The fluctuations per se do not

initiate muscular contraction, as the electrical potential is below the contraction

threshold. Excitatory stimuli from the controlling enteric network must be

present to initiate spike potentials and contractions (12). With the slow waves,

the pulse and propagation of the propulsive contractions are controlled.

Cutaneous EGG is the summation of electrical potentials from the gastric muscle

in a specific axis. This must be distinguished from electromyographic tracings

with electrodes inserted into the gastric wall; in that case the electrical potential at

a fixed point is measured. After our intervention, we found a lower frequency in

the slow waves and also a disappearance of the waves. The physiological

explanation for the bradygastria might either be a reduction in the frequency of

the pacesetter cells in the corpus or that normal pacesetter ICCs are “knocked-

out” and the slow waves are controlled by more distal ICCs with slower intrinsic

frequency (156). Further, the disappearance of the slow waves might reflect a

disappearance of the oscillations in membrane potential or a total disorganization

of the spontaneous activity. The latter might be more likely, as antral tachygas-

tria, leading to a functional uncoupling of the slow waves, has been observed

after opioid administration (66), and gastric arrhythmias are generally caused by

disruptions of the slow waves(164). Our study is one of the first to use the EGG

in the perioperative setting. We suggest that the method should be used more

frequently, as it measures changes in gastric myoelectric activity, and this might

help us to understand the pathology behind the opioid induced impairments of

gastric myoelectric activity.

48

Genetic testing

Genetic evaluation of the μ-opioid receptor gene was done in studies III and IV.

The findings included major interindividual variability in motility variables in

subjects receiving opioids. Recent reports have suggested that SNPs in the MOR

gene can alter the effects of opioids (129, 165), and to our knowledge there is no

previous work where the issue is explored in the context of gastrointestinal

motility. Therefore, we collected blood samples from subjects who participated in

the studies. Genetic analysis was done by a contracted laboratory using routine

molecular biological techniques. Hence, we are the first to evaluate a possible

association between opioid effects on gastrointestinal motility and genetic

variations in the MOR.

Gastric emptying during a remifentanil infusion and influence of posture (I)

In Study I we evaluated the effect of posture on gastric emptying and the

objectives were in part to evaluate pyloric function. If gastric contents are

passively directed towards the pylorus, gastric emptying would be facilitated in

states of normal motility or when the pyloric sphincter is abnormally relaxed. We

used two extreme body positions in our study protocol – the RHU-position where

contents theoretically are directed towards the pylorus, and the LHD-position

where contents are directed from the pylorus. During the control situations,

gastric emptying was better in the RHU-position. This is in agreement with other

studies where body positions that direct stomach contents towards the pylorus

facilitate gastric emptying (25-27, 166). During the remifentanil infusion, gastric

emptying was delayed in both positions compared to the control situations. This

confirms that remifentanil has the same ability as other MOR agonists to affect

gastric motility and delay gastric emptying (32, 67, 68, 88, 167). However, we

found no significant differences between the positions. This indicates that

remifentanil increases pyloric tone and thereby impairs the flow out to the

duodenum.

Gastric emptying after opioid based vs opioid free anesthesia (II)

In study II we compared gastric emptying in two anesthetic protocols, one with

the opioid remifentanil and the other without opioids. We hypothesized that if

perioperative opioids play a major role in the postoperative inhibition of gastric

motility, there would be differences between the groups. ´The results showed that

gastric emptying was delayed in both groups compared to pooled data from

historical controls. However, we could not find any significant differences

between the groups. This indicates that the use of remifentanil during anesthesia

49

impairs postoperative gastric emptying in the same way as a solely inhalational

anesthesia.

Interestingly, if the figures are compared to the gastric emptying rates during the

remifentanil infusion in Study I, gastric emptying was better in Study II. A

reasonable assumption is that gastric motility during anesthesia with remifentanil

would be affected at least in the same way as during the remifentanil experiments.

As the measure of gastric emptying in study II was done after the cessation of

anesthesia, the results indicate that the inhibitory effect of remifentanil on gastric

emptying was reduced quickly. At the time when we measured gastric emptying

in Study II, the opioid effect might no longer have been present and other factors

might have contributed to the delay. The surgical trauma per se delays gastric

motility (1, 2, 4, 168), and if this factor plays the major role, then there would be

no differences between the groups.

There is limited knowledge about the effect on gastric motility of the other

anesthetics used in our protocols. Propofol in higher doses might inhibit motility

(80), and volatile agents inhibit motility with an effect that ceases quickly after

termination of the agents (76-78). Remifentanil and volatile agents might

therefore be considered similar regarding the time course of the gastric inhibition,

and that might also explain the finding of no difference. Future studies must

compare remifentanil with other potent opioids and evaluate if postoperative

gastric emptying is enhanced with remifentanil. As an early oral intake is

preferred today, the choice of a perioperative opioid with minimal impact on

postoperative gastric motility could be of importance.

Furthermore, there was great variability in the gastric emptying rates within the

groups, and both groups had patients with normal gastric emptying and patients

with no gastric emptying at all. As patients received IV opioids for severe pain

during recovery, we tested whether there was any association between opioid

analgesia during recovery and gastric emptying rates, but we found no associa-

tion. The variability must be related to other factors.

Effects of remifentanil on gastric tone (III)

In study III we evaluated changes in gastric tone during an IV infusion of

remifentanil. We found that remifentanil had a marked effect on gastric tone, but

there were two distinctly different patterns of reactions, with about half of the

subjects increasing in gastric tone (decreased volume) and about half of the

50

subjects decreasing in gastric tone (increased volume). Due to this variability, we

were not able to statistically prove the response during remifentanil. However,

the gastric tone was statistically lower (higher volume) after the infusion of

remifentanil compared to the baseline period. We believe these are important

findings, as they show that opioid effects on human gastric motility are variable

and complex.

Proximal gastric tone is an important part of gastric motility and it is mainly

controlled by the autonomous nerve system. Vagal cholinergic nerves mediate

excitation (contraction) while vagal non-cholinergic non-adrenergic (NANC)

nerves mediate inhibition (relaxation) (169). Recent studies have identified

nitrous oxide as one of the main transmitters in the NANC pathway. In humans,

the NANC pathway is believed to be silent during fasting conditions and

activated on volume load by the adaptive reflex (170). In addition, there are

sympathetic adrenerigic spinal nerves that inhibit motility mainly through

cholinergic inhibition (17).

Several animal studies have tried to identify targets for the opioid induced

inhibition of gastric motility. It is widely believed that μ-opioid receptor (MOR)

agonists inhibit the release of Ach in the stomach (61), and there is also evidence

that MOR agonists reduce the relaxation induced by the NANC pathway (171).

Opioids might also have direct excitatory effects on gastric smooth muscles (51).

Opioids also act in the central nervous system (CNS). There is evidence that

MORs are present on and inhibit excitatory neurons projecting to gastrointestinal

motor neurons in the dorsal motor complex (DMV) of the medulla (69). In this

way activation of central MORs inhibits the excitatory vagal output, leading to

inhibition of intestinal transit and induction of gastric relaxation in animal

models. In humans, there is evidence that opioids inhibit gastric motility through

a central mechanism (66).

Hence, depending on the current state of autonomous and enteric nerve systems

and the main effect site, opioids have the potential to both increase and decrease

gastric tone.

There are diverging results in the literature regarding the effects of opioids on

gastric tone in humans. Penagini found that morphine increased gastric tone

(172), while Hammas reported a decrease in gastric tone (155). Both studies used

the same dose of intravenous morphine (0.1 mg/kg) and both used a gastric

51

barostat. However, there were important differences between the studies. In the

first study, baseline gastric tone was set to resemble a gastric load of a meal, and

in the second study baseline was set to fasting conditions. The stomach wall was

probably more distended (higher volumes in the intragastric bag) before

morphine in Penagani’s study compared to Hammas’ study, resulting in an

activated adaptive reflex. This leads to completely different baseline conditions.

In Penagini’s subjects there were probably low cholinergic and high NANC vagal

inputs to the stomach, and the reverse baseline conditions were probably present

in Hammas’ subjects. This might explain why a MOR antagonist contracted the

stomach (through NANC inhibition) in one study and relaxed the stomach

(through cholinergic inhibition) in the other study.

An interesting finding in Hammas’ study was that the concurrent administration

of propofol altered the effect of morphine on gastric tone. Propofol per se had no

effect on gastric tone, but after the subsequent administration of morphine,

gastric tone increased (volume decreased), contrary to the response to morphine

alone. We cannot explain the mechanism behind this modulation, but there is

evidence for central interactions and modulations between GABAergic and opioid

pathways (47). Other types of modulations of gastric tone have also been

described; in animals with an intact vagus nerve, noradrenaline relaxed the

proximal stomach while vagotomy reversed this response (169).

Can we explain the variable responses seen in our study within this context?

Remifentanil is a potent MOR agonist and the effect sites are probably both at

the stomach level and in the CNS. We speculate that the “normal” opioid

response during fasting conditions, as seen in Hammas’ study, is a decreased

cholinergic activity resulting in a decrease in gastric tone. However, due to the

high potency of remifentanil, direct smooth muscle effects might predominate in

some subjects, resulting in an increase in tone. Like propofol, remifentanil might

also have properties that modulate the opioid response. The focus of these

speculations is that opioid effects on gastric tone are variable and depend on

factors like the state of the subject and the current status of the neural pathways

and smooth muscles that are involved. This might be an explanation for the

variable results in study III.

Effects of fentanyl on gastric myoelectrical activity (IV)

In study IV we evaluated how fentanyl affected gastric myoelectrical activity.

Before the intervention, all subjects had a 3 cpm slow wave activity, which did

52

not differ from a recent multicenter electrogastrography study in normal subjects

(157). After fentanyl, gastric myoelectrical activity was inhibited, with a decrease

in both the dominant frequency and the dominant power of the electrogastro-

graphic spectra. The electrical activity was disrupted after the administration of

fentanyl, and we observed both bradygastria and disappearance of the slow wave

activity. However, the EGG was unaffected in about half of the subjects.

There are only a few reports in the literature regarding the effects of opioids on

gastric electrical activity. Invasive recordings of gastric myoelectrical activity have

shown that morphine transiently distorts the slow-wave activity and initiates

migrating myoelectric complexes (65, 173). Cutaneous recordings with EGG have

shown that morphine induces tachygastria (66). The shift in the basal EGG

frequency towards bradygastria that we observed in some of the subjects indicates

that opioids inhibit the ICC-network. Bradygastria is believed to be a decrease in

the frequency of the normal pacemaker cells, while other dysrhythmias like

tachygastria have ectopic origins in the stomach (174).

We tried to explain the variability seen in responders and non-responders. One

hypothesis may be a difference between the individuals in the plasma concentra-

tion of fentanyl. Unfortunately, blood samples were not collected during the EGG

study. By using a pharmacokinetic model (53, 175), we calculated the predicted

plasma concentrations of fentanyl for each subject. We were not able to find any

differences in the predicted concentrations between the outcome groups.

However, there is a notable wide variability in the model that may conceal

relevant differences. Further, as body composition affects the pharmacokinetic

profiles of a drug, we tested for differences in body weight and body mass index

between the groups, but found no differences. Also, it cannot be ruled out that

differences between the subjects in pharmacokinetic factors, i.e. distribution

volume, metabolism and clearance, alter the effect-site concentration of fentanyl

and thus the effect on gastric motility.

With the knowledge we have today, we cannot determine the exact mechanism of

the inhibition of myoelectrical activity. Possible locations of opioid receptors are

the interstitial cells of Cajal, interneurons in the enteric nervous system, and nerve

terminals from ascending pathways. There might also be a direct effect on gastric

smooth muscles, but such an effect would probably not affect the slow waves.

53

Our findings confirm that opioids inhibit the electrical activity, but we cannot

explain the variable outcome.

Associations to genetic factors (III, IV).

We hypothesized that genetic variability in the MOR gene was responsible for the

variations seen in the barostat and EGG studies (III and IV), but we did not find

such an association.

There are data indicating that genetic differences are able to alter the gastrointes-

tinal response to opioids. The variable analgesic effect of codeine is related to

genetic variations, leading to different expressions of the enzyme (CYP2D6) that

metabolizes codeine to morphine. Among extensive metabolizers, orocecal transit

time is prolonged compared to poor metabolizers and correlates to higher

morphine concentrations in plasma (176). To our knowledge, there are no studies

regarding the relation of SNP in the μ-opioid receptor to the effect of opioids on

gastrointestinal motility. After reviewing the literature, we decided to analyze two

common SNPs in the μ-opioid receptor gene - A118G and IVS2 G691C (128).

The frequencies of SNP A118G in our material were similar to the frequencies

reported in the literature, and the distributions were in Hardy-Weinberg

equilibrium. There were discrepancies in the distributions of SNP G691C between

studies III and IV. In study III, the distribution was in equilibrium. In study IV, all

investigated subjects were either heterozygote or homozygote to G691C and there

were no normal “wild types” of G691C, and the distribution was not consistent

with the expected distributions in Hardy-Weinberg equilibrium. Our study group

may not represent a normal population, as the majority of subjects were woman

and almost all of them had gallbladder disease. This may introduce a selection

bias. However, with the small sample size it is difficult to draw any conclusions

regarding the distribution.

Our results indicate that pharmacogenetic differences in the opioid receptor gene

may not be a major factor regarding the variable gastric outcome caused by an

opioid. However, due to the small sample size we want to emphasis that our

results are preliminary observations and they must be interpreted with caution.

Genetic variations can still be one co-factor, but not the factor that determined

the outcome in our studies.

54

Side effects of opioids

Nausea and vomiting are known side effects of opioid treatment (177) and we

had a high incidence in our studies. In the studies with volunteers (I and III), one

third to one half of the subjects experienced nausea during the remifentanil

infusion. The incidences of PONV in study II were 48% and 62% (TIVA and

GAS), respectively, and in study II the incidence was around 50%.

Those in Study I who experienced nausea did so during both occasions with

remifentanil. This indicates that there are individual factors that do not change

over time that determine if opioids induce nausea. In study IV we found an

association between opioid induced EGG changes and PONV. We speculate that

in subjects who are sensitive to opioids, both gastric motility changes and nausea

are easily induced. The emetic center and the motor nuclei are located close to

each other in the medulla and neurons influenced by opioids might affect both

systems.

In studies I and III, subjects experienced difficulties swallowing during the

remifentanil infusion. There are reports in the literature showing that potent

opioids can cause dysphagia (178). This side effect provides evidence that potent

opioids inhibit motility patterns through central mechanisms, as swallowing is a

process controlled by neuronal networks in the medulla (179).

Future perspectives

As we still have only small islands of knowledge about the actions of opioids in

the gastrointestinal system and the underlying mechanisms (7), more research is

needed to find out how we can diminish the side effects of the opioids. Novel,

peripheral-acting opioid antagonists are promising and need more evaluation.

However, as opioids also act through central mechanisms in the brain (66), it

might be impossible to antagonize all side effects in the gastrointestinal tract.

Using the results from out studies as a base, we might be able to further explore

the efficiency of the new antagonists. Can we improve gastric emptying during

opioid treatment? How is the dual response we achieved in gastric tone altered,

and can we reveal peripheral and central actions of opioids?

The finding that EGG changes predicted PONV might be useful in helping us

identify subjects at high risk for PONV. Properly designed studies must be

conducted with this issue as the primary hypothesis.

55

Conclusions

- Remifentanil delayed gastric emptying.

- Posture did not influence gastric emptying rates during a remifentanil

infusion.

- There were no differences in postoperative gastric emptying rates between

a remifentanil-propofol based total intravenous anesthesia and an opioid

free sevoflurane inhalational anesthesia.

- Remifentanil both increased and decreased proximal gastric tone and the

responses were individual.

- Fentanyl inhibited gastric myoelectrical activity, although half of the

subjects were “non-responders.”

- “Responders” to fentanyl (EGG changes) had higher incidences of PONV.

- No associations were found between common SNPs in the μ-

opioidreceptor gene and the variable outcomes in the gastric barostat stud-

ies and the EGG studies.

56

Acknowledgments

I wish to express my warm and sincere gratitude to:

All the volunteers and patients who contributed to this thesis.

My friend and tutor Magnus Wattwil, for initiating and guiding me in the field of research, for patiently believing in me despite my remote location and the other projects in my life, and for his incredible knowledge about how-to-get-to-and-survive-a-congress.

My friend and co-tutor Sven-Egron Thörn, for head-hunting me into anesthesia, for invaluable collaboration in my studies, for being a computer-mate, for enthusiasm about everything, and for sharing important things in life.

Greger Lindberg, for collaboration with the electrogastrograph, for the genetic hypothesis, and for constructive and valuable criticism.

Lisbeth Wattwil and Åsa Löfqvist, for all the blood samples and for your unfailing practical support in my projects.

Mathias Sandin, for assistance in the electrogastrography study.

All my fellow colleagues and members of the staff at the Department of Anesthesia, Sundsvall Hospital, for supporting me and being great colleagues and friends.

My boss, Thomas Bohlin, for giving me time for my research.

My former colleagues and members of the staff at ANIVA-kliniken, Örebro, for creating an inspiring research environment.

Hans Malker and FoU-centrum, Landstinget Västernorrland, for believing in my projects and for providing the possibility for me to carry them out.

Margaretha Jurstrand, for deep-freezing my blood samples for the genetic analysis.

The Medical Library at Sundsvall Hospital, for excellent bibliographic service.

Jane Wigertz, for linguistic revision of the text.

My friends and family, hopefully all of you now understand a little of what I have been doing.

Those I have forgotten to mention… many thanks!

Maria, my beloved wife and best friend, for your love and support. If we hadn’t had so much fun together, this thesis would have been defended ages ago…

Andreas, our best gift ever.

The work in this thesis was supported by fundings from Research and Development Center (FoU-Centrum), Västernorrland County Council and Research Committee of Örebro County Council.

57

References

1. Bauer AJ, Boeckxstaens GE. Mechanisms of postoperative ileus. Neurogastroenterol Motil 2004:16 Suppl 2:54-60.

2. Resnick J, Greenwald DA, Brandt LJ. Delayed gastric emptying and postoperative ileus after nongastric abdominal surgery: part I. Am J Gastroenterol 1997:92(5):751-62.

3. Luckey A, Livingston E, Tache Y. Mechanisms and treatment of postoperative ileus. Arch Surg 2003:138(2):206-14.

4. Baig MK, Wexner SD. Postoperative ileus: a review. Dis Colon Rectum 2004:47(4):516-26.

5. Delaney CP. Clinical perspective on postoperative ileus and the effect of opiates. Neurogastroenterol Motil 2004:16 Suppl 2:61-6.

6. Kurz A, Sessler DI. Opioid-induced bowel dysfunction: pathophysiology and potential new therapies. Drugs 2003:63(7):649-71.

7. Hicks GA, DeHaven-Hudkins DL, Camilleri M. Opiates in the control of gastrointestinal tract function: current knowledge and new avenues for research. Neurogastroenterol Motil 2004:16 Suppl 2:67-70.

8. Azpiroz F. Control of gastric emptying by gastric tone. Dig Dis Sci 1994:39(12 Suppl):18S-19S.

9. Read NW, Houghton LA. Physiology of gastric emptying and pathophysiology of gastroparesis. Gastroenterol Clin North Am 1989:18(2):359-73.

10. Ramkumar D, Schulze KS. The pylorus. Neurogastroenterol Motil 2005:17 Suppl 1:22-30.

11. Burks TF, Galligan JJ, Porreca F, Barber WD. Regulation of gastric emptying. Fed Proc 1985:44(14):2897-901.

12. Chang FY. Electrogastrography: Basic knowledge, recording, processing and its clinical applications. J Gastroenterol Hepatol 2005:20(4):502-16.

13. Wood JD, Alpers DH, Andrews PL. Fundamentals of neurogastroenterology. Gut 1999:45 Suppl 2:II6-II16.

14. Chang HY, Mashimo H, Goyal RK. Musings on the wanderer: what's new in our understanding of vago-vagal reflex? IV. Current concepts of vagal efferent projections to the gut. Am J Physiol Gastrointest Liver Physiol 2003:284(3):G357-66.

15. Brookes SJH, Costa M. Functional Histoanatomy of the Enteric Nervous System. In: Johnsson LR, editor. Physiology of the Gastrointestinal Tract. 4th ed: Elsevier Academic Press; 2006. p. 577-602.

16. Travagli RA, Hermann GE, Browning KN, Rogers RC. Brainstem circuits regulating gastric function. Annu Rev Physiol 2006:68:279-305.

17. Abrahamsson H, Glise H. Sympathetic nervous control of gastric motility and interaction with vagal activity. Scand J Gastroenterol Suppl 1984:89:83-7.

18. Glise H, Abrahamsson H. Reflex inhibition of gastric motility pathophysiological aspects. Scand J Gastroenterol Suppl 1984:89:77-82.

19. Naslund E, Bogefors J, Skogar S, Gryback P, Jacobsson H, Holst JJ, Hellstrom PM. GLP-1 slows solid gastric emptying and inhibits insulin, glucagon, and PYY release in humans. Am J Physiol 1999:277(3 Pt 2):R910-6.

20. Bojo L, Cassuto J. Gastric reflex relaxation by colonic distension. J Auton Nerv Syst 1992:38(1):57-64.

58

21. Dockray JD. Gastrointestinal Hormones: Gastrin, Cholecystokinin, Somatostatin, and Ghrelin. In: Johnsson LR, editor. Physiology of the Gastrointestinal Tract. 4th ed: Elsevier Academic Press; 2006. p. 91-120.

22. Masuda Y, Tanaka T, Inomata N, Ohnuma N, Tanaka S, Itoh Z, Hosoda H, Kojima M, Kangawa K. Ghrelin stimulates gastric acid secretion and motility in rats. Biochem Biophys Res Commun 2000:276(3):905-8.

23. Englander EW, Greeley GH. Postpyloric Gastrointetsinal Peptides. In: Johnsson LR, editor. Physiology of the Gastrointestinal Tract. 4th ed: Elsevier Academic Press; 2006. p. 121-159.

24. Hellstrom PM, Gryback P, Jacobsson H. The physiology of gastric emptying. Best Pract Res Clin Anaesthesiol 2006:20(3):397-407.

25. Spiegel TA, Fried H, Hubert CD, Peikin SR, Siegel JA, Zeiger LS. Effects of posture on gastric emptying and satiety ratings after a nutritive liquid and solid meal. Am J Physiol Regul Integr Comp Physiol 2000:279(2):R684-94.

26. Horowitz M, Jones K, Edelbroek MA, Smout AJ, Read NW. The effect of posture on gastric emptying and intragastric distribution of oil and aqueous meal components and appetite. Gastroenterology 1993:105(2):382-90.

27. Burn-Murdoch R, Fisher MA, Hunt JN. Does lying on the right side increase the rate of gastric emptying? J Physiol 1980:302:395-8.

28. Maurer AH, Parkman HP. Update on gastrointestinal scintigraphy. Semin Nucl Med 2006:36(2):110-8.

29. Gentilcore D, Hausken T, Horowitz M, Jones KL. Measurements of gastric emptying of low- and high-nutrient liquids using 3D ultrasonography and scintigraphy in healthy sub-jects. Neurogastroenterol Motil 2006:18(12):1062-8.

30. Hausken T, Mundt M, Samsom M. Low antroduodenal pressure gradients are responsible for gastric emptying of a low-caloric liquid meal in humans. Neurogastroenterol Motil 2002:14(1):97-105.

31. Soreide E, Hausken T, Soreide JA, Steen PA. Gastric emptying of a light hospital breakfast. A study using real time ultrasonography. Acta Anaesthesiol Scand 1996:40(5):549-53.

32. Nimmo WS, Heading RC, Wilson J, Tothill P, Prescott LF. Inhibition of gastric emptying and drug absorption by narcotic analgesics. Br J Clin Pharmacol 1975:2(6):509-13.

33. Clements JA, Heading RC, Nimmo WS, Prescott LF. Kinetics of acetaminophen absorption and gastric emptying in man. Clin Pharmacol Ther 1978:24(4):420-31.

34. Tarling MM, Toner CC, Withington PS, Baxter MK, Whelpton R, Goldhill DR. A model of gastric emptying using paracetamol absorption in intensive care patients [see comments]. Intensive Care Med 1997:23(3):256-60.

35. Jackson SJ, Bluck LJ. Measurement of gastric emptying by octanoate metabolism. Curr Opin Clin Nutr Metab Care 2005:8(5):538-44.

36. Camilleri M, Hasler WL, Parkman HP, Quigley EM, Soffer E. Measurement of gastrointestinal motility in the GI laboratory. Gastroenterology 1998:115(3):747-62.

37. Whitehead WE, Delvaux M. Standardization of barostat procedures for testing smooth muscle tone and sensory thresholds in the gastrointestinal tract. The Working Team of Glaxo-Wellcome Research, UK. Dig Dis Sci 1997:42(2):223-41.

38. Bailey PL, Egan TD, Stanley TH. Intravenous Opioid Anesthestics. In: Miller RD, editor. Anesthesia. 5th ed: Churchill Livingstone; 2000. p. 273-376.

39. Camilleri M. Alvimopan, a selective peripherally acting mu-opioid antagonist. Neurogas-troenterol Motil 2005:17(2):157-65.

40. Greenwood-Van Meerveld B, Gardner CJ, Little PJ, Hicks GA, Dehaven-Hudkins DL. Preclinical studies of opioids and opioid antagonists on gastrointestinal function. Neurogas-troenterol Motil 2004:16 Suppl 2:46-53.

59

41. Sternini C, Patierno S, Selmer IS, Kirchgessner A. The opioid system in the gastrointestinal tract. Neurogastroenterol Motil 2004:16 Suppl 2:3-16.

42. Ferrante FM. Principles of opioid pharmacotherapy: practical implications of basic mechanisms. J Pain Symptom Manage 1996:11(5):265-73.

43. Ganesh A, Maxwell LG. Pathophysiology and Management of Opioid-Induced Pruritus. Drugs 2007:67(16):2323-2333.

44. Sora I, Takahashi N, Funada M, Ujike H, Revay RS, Donovan DM, Miner LL, Uhl GR. Opiate receptor knockout mice define mu receptor roles in endogenous nociceptive re-sponses and morphine-induced analgesia. Proc Natl Acad Sci U S A 1997:94(4):1544-9.

45. Aicher SA, Goldberg A, Sharma S, Pickel VM. mu-opioid receptors are present in vagal afferents and their dendritic targets in the medial nucleus tractus solitarius. J Comp Neurol 2000:422(2):181-90.

46. Tsuchida D, Fukuda H, Koda K, Miyazaki M, Pappas TN, Takahashi T. Central effect of mu-opioid agonists on antral motility in conscious rats. Brain Res 2004:1024(1-2):244-50.

47. Browning KN, Zheng Z, Gettys TW, Travagli RA. Vagal afferent control of opioidergic effects in rat brainstem circuits. J Physiol 2006:575(Pt 3):761-76.

48. Tejwani GA, Rattan AK, McDonald JS. Role of spinal opioid receptors in the antinocicep-tive interactions between intrathecal morphine and bupivacaine. Anesth Analg 1992:74(5):726-34.

49. Pol O, Puig MM. Expression of opioid receptors during peripheral inflammation. Curr Top Med Chem 2004:4(1):51-61.

50. Shahbazian A, Heinemann A, Schmidhammer H, Beubler E, Holzer-Petsche U, Holzer P. Involvement of mu- and kappa-, but not delta-, opioid receptors in the peristaltic motor depression caused by endogenous and exogenous opioids in the guinea-pig intestine. Br J Pharmacol 2002:135(3):741-50.

51. Grider JR, Makhlouf GM. Identification of opioid receptors on gastric muscle cells by selective receptor protection. Am J Physiol 1991:260(1 Pt 1):G103-7.

52. Allescher HD, Ahmad S, Daniel EE, Dent J, Kostolanska F, Fox JE. Inhibitory opioid receptors in canine pylorus. Am J Physiol 1988:255(3 Pt 1):G352-60.

53. Shafer SL, Varvel JR, Aziz N, Scott JC. Pharmacokinetics of fentanyl administered by computer-controlled infusion pump. Anesthesiology 1990:73(6):1091-102.

54. Burkle H, Dunbar S, Van Aken H. Remifentanil: a novel, short-acting, mu-opioid. Anesth Analg 1996:83(3):646-51.

55. Trendelenburg P. Physiological and pharmacological investigations of small intestinal peristalsis. Translation of the article "Physiologische und pharmakologische Versuche uber die Dunndarmperistaltik", Arch. Exp. Pathol. Pharmakol. 81, 55-129, 1917. Naunyn Schmiedebergs Arch Pharmacol 2006:373(2):101-33.

56. Trendelenburg P. Physiologische und pharmakologische Versuche uber die Dunndarmperi-staltik´. Arch Exp Pathol Pharmakol 1917(81):55-129.

57. Wood JD, Galligan JJ. Function of opioids in the enteric nervous system. Neurogastroen-terol Motil 2004:16 Suppl 2:17-28.

58. Kromer W. Endogenous and exogenous opioids in the control of gastrointestinal motility and secretion. Pharmacol Rev 1988:40(2):121-62.

59. Burks TF, Hruby VJ, Galligan JJ, Porreca F. Opioids and CNS control of the gut. Nippon Heikatsukin Gakkai Zasshi 1985:21 Suppl:101-2.

60. Bagnol D, Mansour A, Akil H, Watson SJ. Cellular localization and distribution of the cloned mu and kappa opioid receptors in rat gastrointestinal tract. Neuroscience 1997:81(2):579-91.

60

61. Yokotani K, Osumi Y. Involvement of mu-receptor in endogenous opioid peptide-mediated inhibition of acetylcholine release from the rat stomach. Jpn J Pharmacol 1998:78(1):93-5.

62. Soldani G, Del Tacca M, Bernardini MC, Bardon T, Ruckebusch Y. Peripheral opioid receptors mediate gastrointestinal secretory and motor effects of dermorphin N-terminal tetrapeptide (NTT) in the dog. Neuropeptides 1987:10(1):67-76.

63. Zhang L, Gu ZF, Pradhan T, Jensen RT, Maton PN. Characterization of opioid receptors on smooth muscle cells from guinea pig stomach. Am J Physiol 1992:262(3 Pt 1):G461-9.

64. Bitar KN, Makhlouf GM. Specific opiate receptors on isolated mammalian gastric smooth muscle cells. Nature 1982:297(5861):72-4.

65. Telford GL, Hoshmonai M, Moses AJ, Szurszewski JH. Morphine initiates migrating myoelectric complexes by acting on peripheral opioid receptors. Am J Physiol 1985:249(5 Pt 1):G557-62.

66. Thorn SE, Wattwil M, Lindberg G, Sawe J. Systemic and central effects of morphine on gastroduodenal motility. Acta Anaesthesiol Scand 1996:40(2):177-86.

67. Yuan CS, Foss JF, O'Connor M, Roizen MF, Moss J. Effects of low-dose morphine on gastric emptying in healthy volunteers. J Clin Pharmacol 1998:38(11):1017-20.

68. Murphy DB, Sutton JA, Prescott LF, Murphy MB. Opioid-induced delay in gastric emptying: a peripheral mechanism in humans. Anesthesiology 1997:87(4):765-70.

69. Browning KN, Kalyuzhny AE, Travagli RA. Opioid peptides inhibit excitatory but not inhibitory synaptic transmission in the rat dorsal motor nucleus of the vagus. J Neurosci 2002:22(8):2998-3004.

70. Aicher SA, Mitchell JL, Mendelowitz D. Distribution of mu-opioid receptors in rat visceral premotor neurons. Neuroscience 2002:115(3):851-60.

71. Silverman MB, Hermes SM, Zadina JE, Aicher SA. Mu-opioid receptor is present in dendritic targets of Endomorphin-2 axon terminals in the nuclei of the solitary tract. Neu-roscience 2005:135(3):887-96.

72. Grundy D, Booth CE, Winchester W, Hicks GA. Peripheral opiate action on afferent fibres supplying the rat intestine. Neurogastroenterol Motil 2004:16 Suppl 2:29-37.

73. Patierno S, Zellalem W, Ho A, Parsons CG, Lloyd KC, Tonini M, Sternini C. N-methyl-D-aspartate receptors mediate endogenous opioid release in enteric neurons after abdominal surgery. Gastroenterology 2005:128(7):2009-19.

74. Patierno S, Raybould HE, Sternini C. Abdominal surgery induces mu opioid receptor endocytosis in enteric neurons of the guinea-pig ileum. Neuroscience 2004:123(1):101-9.

75. Sanger GJ, Tuladhar BR. The role of endogenous opioids in the control of gastrointestinal motility: predictions from in vitro modelling. Neurogastroenterol Motil 2004:16 Suppl 2:38-45.

76. Marshall FN, Pittinger CB, Long JP. Effects of halothane on gastrointestinal motility. Anesthesiology 1961:22:363-6.

77. Wright JW, Healy TE, Balfour TW, Hardcastle JD. Effects of inhalation anaesthetic agents on the electrical and mechanical activity of the rat duodenum. Br J Anaesth 1982:54(11):1223-30.

78. Schurizek BA. The effects of general anaesthesia on antroduodenal motility, gastric pH and gastric emptying in man. Dan Med Bull 1991:38(4):347-65.

79. Hammas B, Hvarfner A, Thorn SE, Wattwil M. Propofol sedation and gastric emptying in volunteers. Acta Anaesthesiol Scand 1998:42(1):102-5.

80. Lee TL, Ang SB, Dambisya YM, Adaikan GP, Lau LC. The effect of propofol on human gastric and colonic muscle contractions. Anesth Analg 1999:89(5):1246-9.

81. Karamanolis G, Tack J. Promotility medications--now and in the future. Dig Dis 2006:24(3-4):297-307.

61

82. Galligan JJ, Vanner S. Basic and clinical pharmacology of new motility promoting agents. Neurogastroenterol Motil 2005:17(5):643-53.

83. Levein NG, Thorn SE, Wattwil M. Dopamine delays gastric emptying and prolongs orocaecal transit time in volunteers. Eur J Anaesthesiol 1999:16(4):246-50.

84. Levein NG, Thorn SE, Lindberg G, Wattwill M. Dopamine reduces gastric tone in a dose-related manner. Acta Anaesthesiol Scand 1999:43(7):722-5.

85. Beattie DT, Smith JA, Marquess D, Vickery RG, Armstrong SR, Pulido-Rios T, McCullough JL, Sandlund C, Richardson C, Mai N, Humphrey PP. The 5-HT4 receptor agonist, tegaserod, is a potent 5-HT2B receptor antagonist in vitro and in vivo. Br J Phar-macol 2004:143(5):549-60.

86. Park MI, Ferber I, Camilleri M, Allenby K, Trillo R, Burton D, Zinsmeister AR. Effect of atilmotin on gastrointestinal transit in healthy subjects: a randomized, placebo-controlled study. Neurogastroenterol Motil 2006:18(1):28-36.

87. Cowles VE, Nellans HN, Seifert TR, Besecke LM, Segreti JA, Mohning KM, Faghih R, Verlinden MH, Wegner CD. Effect of novel motilide ABT-229 versus erythromycin and cisapride on gastric emptying in dogs. J Pharmacol Exp Ther 2000:293(3):1106-11.

88. Lewis TD. Morphine and gastroduodenal motility. Dig Dis Sci 1999:44(11):2178-86.

89. Foss JF. A review of the potential role of methylnaltrexone in opioid bowel dysfunction. Am J Surg 2001:182(5A Suppl):19S-26S.

90. Taguchi A, Sharma N, Saleem RM, Sessler DI, Carpenter RL, Seyedsadr M, Kurz A. Selective postoperative inhibition of gastrointestinal opioid receptors. N Engl J Med 2001:345(13):935-40.

91. Liu SS, Hodgson PS, Carpenter RL, Fricke JR, Jr. ADL 8-2698, a trans-3,4-dimethyl-4-(3-hydroxyphenyl) piperidine, prevents gastrointestinal effects of intravenous morphine with-out affecting analgesia. Clin Pharmacol Ther 2001:69(1):66-71.

92. Wolff BG, Weese JL, Ludwig KA, Delaney CP, Stamos MJ, Michelassi F, Du W, Techner L. Postoperative ileus-related morbidity profile in patients treated with alvimopan after bowel resection. J Am Coll Surg 2007:204(4):609-16.

93. Herzog TJ, Coleman RL, Guerrieri JP, Jr., Gabriel K, Du W, Techner L, Fort JG, Wallin B. A double-blind, randomized, placebo-controlled phase III study of the safety of alvimopan in patients who undergo simple total abdominal hysterectomy. Am J Obstet Gynecol 2006:195(2):445-53.

94. Tan EK, Cornish J, Darzi AW, Tekkis PP. Meta-analysis: Alvimopan vs. placebo in the treatment of post-operative ileus. Aliment Pharmacol Ther 2007:25(1):47-57.

95. Wolff BG, Michelassi F, Gerkin TM, Techner L, Gabriel K, Du W, Wallin BA. Alvimopan, a novel, peripherally acting mu opioid antagonist: results of a multicenter, randomized, double-blind, placebo-controlled, phase III trial of major abdominal surgery and postopera-tive ileus. Ann Surg 2004:240(4):728-34; discussion 734-5.

96. Senagore AJ, Bauer JJ, Du W, Techner L. Alvimopan accelerates gastrointestinal recovery after bowel resection regardless of age, gender, race, or concomitant medication use. Sur-gery 2007:142(4):478-86.

97. Yuan CS, Foss JF, Moss J. Effects of methylnaltrexone on morphine-induced inhibition of contraction in isolated guinea-pig ileum and human intestine. Eur J Pharmacol 1995:276(1-2):107-11.

98. Foss JF, O'Connor MF, Yuan CS, Murphy M, Moss J, Roizen MF. Safety and tolerance of methylnaltrexone in healthy humans: a randomized, placebo-controlled, intravenous, ascending-dose, pharmacokinetic study. J Clin Pharmacol 1997:37(1):25-30.

99. Yuan CS, Foss JF. Gastric effects of methylnaltrexone on mu, kappa, and delta opioid agonists induced brainstem unitary responses. Neuropharmacology 1999:38(3):425-32.

62

100. Yuan CS. Methylnaltrexone mechanisms of action and effects on opioid bowel dysfunction and other opioid adverse effects. Ann Pharmacother 2007:41(6):984-93.

101. Shaiova L, Rim F, Friedman D, Jahdi M. A review of methylnaltrexone, a peripheral opioid receptor antagonist, and its role in opioid-induced constipation. Palliat Support Care 2007:5(2):161-6.

102. Janda M, Scheeren TW, Noldge-Schomburg GF. Management of pulmonary aspiration. Best Pract Res Clin Anaesthesiol 2006:20(3):409-27.

103. Maltby JR. Fasting from midnight--the history behind the dogma. Best Pract Res Clin Anaesthesiol 2006:20(3):363-78.

104. Ljungqvist O, Soreide E. Preoperative fasting. Br J Surg 2003:90(4):400-6.

105. Stuart PC. The evidence base behind modern fasting guidelines. Best Pract Res Clin Anaesthesiol 2006:20(3):457-69.

106. Soreide E, Eriksson LI, Hirlekar G, Eriksson H, Henneberg SW, Sandin R, Raeder J. Pre-operative fasting guidelines: an update. Acta Anaesthesiol Scand 2005:49(8):1041-7.

107. Soreide E, Ljungqvist O. Modern preoperative fasting guidelines: a summary of the present recommendations and remaining questions. Best Pract Res Clin Anaesthesiol 2006:20(3):483-91.

108. Morris J, Cook TM. Rapid sequence induction: a national survey of practice. Anaesthesia 2001:56(11):1090-7.

109. Andersen HK, Lewis SJ, Thomas S. Early enteral nutrition within 24h of colorectal surgery versus later commencement of feeding for postoperative complications. Cochrane Database Syst Rev 2006(4):CD004080.

110. Han-Geurts IJ, Hop WC, Kok NF, Lim A, Brouwer KJ, Jeekel J. Randomized clinical trial of the impact of early enteral feeding on postoperative ileus and recovery. Br J Surg 2007:94(5):555-61.

111. Charoenkwan K, Phillipson G, Vutyavanich T. Early versus delayed (traditional) oral fluids and food for reducing complications after major abdominal gynaecologic surgery. Cochrane Database Syst Rev 2007(4):CD004508.

112. Mangesi L, Hofmeyr GJ. Early compared with delayed oral fluids and food after caesarean section. Cochrane Database Syst Rev 2002(3):CD003516.

113. Fukuda H, Tsuchida D, Koda K, Miyazaki M, Pappas TN, Takahashi T. Impaired gastric motor activity after abdominal surgery in rats. Neurogastroenterol Motil 2005:17(2):245-50.

114. Liu SS, Carpenter RL, Mackey DC, Thirlby RC, Rupp SM, Shine TS, Feinglass NG, Metzger PP, Fulmer JT, Smith SL. Effects of perioperative analgesic technique on rate of recovery after colon surgery. Anesthesiology 1995:83(4):757-65.

115. Jorgensen H, Wetterslev J, Moiniche S, Dahl JB. Epidural local anaesthetics versus opioid-based analgesic regimens on postoperative gastrointestinal paralysis, PONV and pain after abdominal surgery. Cochrane Database Syst Rev 2001(1):CD001893.

116. Apfel CC, Laara E, Koivuranta M, Greim CA, Roewer N. A simplified risk score for predicting postoperative nausea and vomiting: conclusions from cross-validations between two centers. Anesthesiology 1999:91(3):693-700.

117. Martin B. Prevention of gastrointestinal complications in the critically ill patient. AACN Adv Crit Care 2007:18(2):158-66.

118. Chapman MJ, Nguyen NQ, Fraser RJ. Gastrointestinal motility and prokinetics in the critically ill. Curr Opin Crit Care 2007:13(2):187-94.

119. Rohm KD, Schollhorn T, Boldt J, Wolf M, Papsdorf M, Piper SN. Nutrition support and treatment of motility disorders in critically ill patients - Results of a survey on German intensive care units. Eur J Anaesthesiol 2007:1-9.

63

120. Nguyen NQ, Chapman MJ, Fraser RJ, Bryant LK, Holloway RH. Erythromycin is more effective than metoclopramide in the treatment of feed intolerance in critical illness. Crit Care Med 2007:35(2):483-9.

121. Berne JD, Norwood SH, McAuley CE, Vallina VL, Villareal D, Weston J, McClarty J. Erythromycin reduces delayed gastric emptying in critically ill trauma patients: a random-ized, controlled trial. J Trauma 2002:53(3):422-5.

122. Jooste CA, Mustoe J, Collee G. Metoclopramide improves gastric motility in critically ill patients. Intensive Care Med 1999:25(5):464-8.

123. Nguyen NQ, Chapman M, Fraser RJ, Bryant LK, Burgstad C, Holloway RH. Prokinetic therapy for feed intolerance in critical illness: One drug or two? Crit Care Med 2007.

124. Meissner W, Dohrn B, Reinhart K. Enteral naloxone reduces gastric tube reflux and frequency of pneumonia in critical care patients during opioid analgesia. Crit Care Med 2003:31(3):776-80.

125. Mixides G, Liebl MG, Bloom K. Enteral administration of naloxone for treatment of opioid-associated intragastric feeding intolerance. Pharmacotherapy 2004:24(2):291-4.

126. Roberts DJ, Banh HL, Hall RI. Use of novel prokinetic agents to facilitate return of gastrointestinal motility in adult critically ill patients. Curr Opin Crit Care 2006:12(4):295-302.

127. Mayer P, Hollt V. Pharmacogenetics of opioid receptors and addiction. Pharmacogenet Genomics 2006:16(1):1-7.

128. Ikeda K, Ide S, Han W, Hayashida M, Uhl GR, Sora I. How individual sensitivity to opiates can be predicted by gene analyses. Trends Pharmacol Sci 2005:26(6):311-7.

129. Klepstad P, Dale O, Skorpen F, Borchgrevink PC, Kaasa S. Genetic variability and clinical efficacy of morphine. Acta Anaesthesiol Scand 2005:49(7):902-8.

130. Lotsch J, Geisslinger G. Are mu-opioid receptor polymorphisms important for clinical opioid therapy? Trends Mol Med 2005:11(2):82-9.

131. Palmer SN, Giesecke NM, Body SC, Shernan SK, Fox AA, Collard CD. Pharmacogenetics of anesthetic and analgesic agents. Anesthesiology 2005:102(3):663-71.

132. Bond C, LaForge KS, Tian M, Melia D, Zhang S, Borg L, Gong J, Schluger J, Strong JA, Leal SM, Tischfield JA, Kreek MJ, Yu L. Single-nucleotide polymorphism in the human mu opioid receptor gene alters beta-endorphin binding and activity: possible implications for opiate addiction. Proc Natl Acad Sci U S A 1998:95(16):9608-13.

133. Landau R, Cahana A, Smiley RM, Antonarakis SE, Blouin JL. Genetic variability of mu-opioid receptor in an obstetric population. Anesthesiology 2004:100(4):1030-3.

134. Hollt V. A polymorphism (A118G) in the mu-opioid receptor gene affects the response to morphine-6-glucuronide in humans. Pharmacogenetics 2002:12(1):1-2.

135. Margas W, Zubkoff I, Schuler HG, Janicki PK, Ruiz-Velasco V. Modulation of Ca2+ channels by heterologously expressed wild-type and mutant human micro-opioid receptors (hMORs) containing the A118G single-nucleotide polymorphism. J Neurophysiol 2007:97(2):1058-67.

136. Beyer A, Koch T, Schroder H, Schulz S, Hollt V. Effect of the A118G polymorphism on binding affinity, potency and agonist-mediated endocytosis, desensitization, and resensitiza-tion of the human mu-opioid receptor. J Neurochem 2004:89(3):553-60.

137. Lotsch J, Skarke C, Grosch S, Darimont J, Schmidt H, Geisslinger G. The polymorphism A118G of the human mu-opioid receptor gene decreases the pupil constrictory effect of morphine-6-glucuronide but not that of morphine. Pharmacogenetics 2002:12(1):3-9.

138. Romberg R, Olofsen E, Bijl H. Polymorphism of μ-Opiodreceptor Gene (OPRM1:c.118>G) Does Not Protect Against Opioidinduced Respiratory Depression despite Reduced Analgetic Response? Anesthesiology 2005:102(3):522-30.

64

139. Fillingim RB, Kaplan L, Staud R, Ness TJ, Glover TL, Campbell CM, Mogil JS, Wallace MR. The A118G single nucleotide polymorphism of the mu-opioid receptor gene (OPRM1) is associated with pressure pain sensitivity in humans. J Pain 2005:6(3):159-67.

140. Chou WY, Yang LC, Lu HF, Ko JY, Wang CH, Lin SH, Lee TH, Concejero A, Hsu CJ. Association of mu-opioid receptor gene polymorphism (A118G) with variations in mor-phine consumption for analgesia after total knee arthroplasty. Acta Anaesthesiol Scand 2006:50(7):787-92.

141. Klepstad P, Rakvag TT, Kaasa S, Holthe M, Dale O, Borchgrevink PC, Baar C, Vikan T, Krokan HE, Skorpen F. The 118 A > G polymorphism in the human micro-opioid receptor gene may increase morphine requirements in patients with pain caused by malignant dis-ease. Acta Anaesthesiol Scand 2004:48(10):1232-9.

142. Chou WY, Wang CH, Liu PH, Liu CC, Tseng CC, Jawan B. Human opioid receptor A118G polymorphism affects intravenous patient-controlled analgesia morphine consump-tion after total abdominal hysterectomy. Anesthesiology 2006:105(2):334-7.

143. Coulbault L, Beaussier M, Verstuyft C, Weickmans H, Dubert L, Tregouet D, Descot C, Parc Y, Lienhart A, Jaillon P, Becquemont L. Environmental and genetic factors associated with morphine response in the postoperative period. Clin Pharmacol Ther 2006:79(4):316-324.

144. Skarke C, Darimont J, Schmidt H, Geisslinger G, Lotsch J. Analgesic effects of morphine and morphine-6-glucuronide in a transcutaneous electrical pain model in healthy volun-teers. Clin Pharmacol Ther 2003:73(1):107-21.

145. Ross JR, Rutter D, Welsh K, Joel SP, Goller K, Wells AU, Du Bois R, Riley J. Clinical response to morphine in cancer patients and genetic variation in candidate genes. Pharma-cogenomics J 2005:5(5):324-36.

146. Ray LA, Hutchison KE. A polymorphism of the mu-opioid receptor gene (OPRM1) and sensitivity to the effects of alcohol in humans. Alcohol Clin Exp Res 2004:28(12):1789-95.

147. van den Wildenberg E, Wiers RW, Dessers J, Janssen RG, Lambrichs EH, Smeets HJ, van Breukelen GJ. A functional polymorphism of the mu-opioid receptor gene (OPRM1) influ-ences cue-induced craving for alcohol in male heavy drinkers. Alcohol Clin Exp Res 2007:31(1):1-10.

148. Ray LA, Hutchison KE. Effects of naltrexone on alcohol sensitivity and genetic moderators of medication response: a double-blind placebo-controlled study. Arch Gen Psychiatry 2007:64(9):1069-77.

149. Nishizawa D, Han W, Hasegawa J, Ishida T, Numata Y, Sato T, Kawai A, Ikeda K. Association of micro-opioid receptor gene polymorphism A118G with alcohol dependence in a Japanese population. Neuropsychobiology 2006:53(3):137-41.

150. Bart G, Kreek MJ, Ott J, LaForge KS, Proudnikov D, Pollak L, Heilig M. Increased attributable risk related to a functional mu-opioid receptor gene polymorphism in associa-tion with alcohol dependence in central Sweden. Neuropsychopharmacology 2005:30(2):417-22.

151. Arias A, Feinn R, Kranzler HR. Association of an Asn40Asp (A118G) polymorphism in the mu-opioid receptor gene with substance dependence: a meta-analysis. Drug Alcohol De-pend 2006:83(3):262-8.

152. Chong RY, Oswald L, Yang X, Uhart M, Lin PI, Wand GS. The Mu-Opioid Receptor Polymorphism A118G Predicts Cortisol Responses to Naloxone and Stress. Neuropsycho-pharmacology 2006:31(1):204-11.

153. Willems M, Quartero AO, Numans ME. How useful is paracetamol absorption as a marker of gastric emptying? A systematic literature study. Dig Dis Sci 2001:46(10):2256-62.

154. Azpiroz F, Malagelada JR. Gastric tone measured by an electronic barostat in health and postsurgical gastroparesis. Gastroenterology 1987:92(4):934-43.

65

155. Hammas B, Thorn SE, Wattwil M. Propofol and gastric effects of morphine. Acta Anaesthesiol Scand 2001:45(8):1023-7.

156. Sanders KM, Ördög T, Koh SD, Ward SM. Properties of Electrical Rhytmicity in the Stomach. In: Koch KL, Stern RM, editors. Handbook of Electrogastrography. New York: Oxford University Press; 2004. p. 13-36.

157. Simonian HP, Panganamamula K, Parkman HP, Xu X, Chen JZ, Lindberg G, Xu H, Shao C, Ke MY, Lykke M, Hansen P, Barner B, Buhl H. Multichannel electrogastrography (EGG) in normal subjects: a multicenter study. Dig Dis Sci 2004:49(4):594-601.

158. Sanaka M, Kuyama Y, Yamanaka M. Guide for judicious use of the paracetamol absorption technique in a study of gastric emptying rate of liquids. J Gastroenterol 1998:33(6):785-91.

159. Sanaka M, Kuyama Y, Nishinakagawa S, Mineshita S. Use of salivary acetaminophen concentration to assess gastric emptying rate of liquids. J Gastroenterol 2000:35(6):429-33.

160. Schwizer W, Steingotter A, Fox M, Zur T, Thumshirn M, Bosiger P, Fried M. Non-invasive measurement of gastric accommodation in humans. Gut 2002:51 Suppl 1:i59-62.

161. Parkman HP, Hasler WL, Barnett JL, Eaker EY. Electrogastrography: a document prepared by the gastric section of the American Motility Society Clinical GI Motility Testing Task Force. Neurogastroenterol Motil 2003:15(2):89-102.

162. Simonian HP, Panganamamula K, Chen JZ, Fisher RS, Parkman HP. Multichannel electrogastrography (EGG) in symptomatic patients: a single center study. Am J Gastroen-terol 2004:99(3):478-85.

163. Sanders KM, Koh SD, Ward SM. Interstitial cells of cajal as pacemakers in the gastrointes-tinal tract. Annu Rev Physiol 2006:68:307-43.

164. Wang ZS, Elsenbruch S, Orr WC, Chen JD. Detection of gastric slow wave uncoupling from multi-channel electrogastrogram: validations and applications. Neurogastroenterol Motil 2003:15(5):457-65.

165. Lotsch J, Geisslinger G. Current evidence for a genetic modulation of the response to analgesics. Pain 2006:121(1-2):1-5.

166. Anvari M, Horowitz M, Fraser R, Maddox A, Myers J, Dent J, Jamieson GG. Effects of posture on gastric emptying of nonnutrient liquids and antropyloroduodenal motility. Am J Physiol 1995:268(5 Pt 1):G868-71.

167. Bueno L, Fioramonti J. Action of opiates on gastrointestinal function. Baillieres Clin Gastroenterol 1988:2(1):123-39.

168. Behm B, Stollman N. Postoperative ileus: etiologies and interventions. Clin Gastroenterol Hepatol 2003:1(2):71-80.

169. Jahnberg T. Gastric adaptive relaxation. Effects of vagal activation and vagotomy. An experimental study in dogs and in man. Scand J Gastroenterol Suppl 1977:46:1-32.

170. Tack J, Demedts I, Meulemans A, Schuurkes J, Janssens J. Role of nitric oxide in the gastric accommodation reflex and in meal induced satiety in humans. Gut 2002:51(2):219-24.

171. Storr M, Gaffal E, Schusdziarra V, Allescher HD. Endomorphins 1 and 2 reduce relaxant non-adrenergic, non-cholinergic neurotransmission in rat gastric fundus. Life Sci 2002:71(4):383-9.

172. Penagini R, Allocca M, Cantu P, Mangano M, Savojardo D, Carmagnola S, Bianchi PA. Relationship between motor function of the proximal stomach and transient lower oeso-phageal sphincter relaxation after morphine. Gut 2004:53(9):1227-31.

173. Thorn SE, Wickbom G, Philipson L, Leissner P, Wattwil M. Myoelectric activity in the stomach and duodenum after epidural administration of morphine or bupivacaine. Acta Anaesthesiol Scand 1996:40(7):773-8.

66

174. Qian LW, Pasricha PJ, Chen JD. Origins and patterns of spontaneous and drug-induced canine gastric myoelectrical dysrhythmia. Dig Dis Sci 2003:48(3):508-15.

175. Shibutani K, Inchiosa MA, Jr., Sawada K, Bairamian M. Accuracy of pharmacokinetic models for predicting plasma fentanyl concentrations in lean and obese surgical patients: derivation of dosing weight ("pharmacokinetic mass"). Anesthesiology 2004:101(3):603-13.

176. Mikus G, Trausch B, Rodewald C, Hofmann U, Richter K, Gramatte T, Eichelbaum M. Effect of codeine on gastrointestinal motility in relation to CYP2D6 phenotype. Clin Phar-macol Ther 1997:61(4):459-66.

177. Watcha MF, White PF. Postoperative nausea and vomiting. Its etiology, treatment, and prevention. Anesthesiology 1992:77(1):162-84.

178. Currier DS, Levin KR, Campbell C. Dysphagia with intrathecal fentanyl. Anesthesiology 1997:87(6):1570-1.

179. Martino R, Terrault N, Ezerzer F, Mikulis D, Diamant NE. Dysphagia in a patient with lateral medullary syndrome: insight into the central control of swallowing. Gastroenterol-ogy 2001:121(2):420-6.

115

STUDY I

Klunserna Study 05-10-25, 10.10115

STU

DY

IThe Delay of Gastric Emptying Induced by Remifentanil IsNot Influenced by PostureJakob Wallden, MD*†, Sven-Egron Thorn, MD, PhD*, and Magnus Wattwil, MD, PhD*†

*Department of Anesthesia and Intensive Care, Orebro University Hospital, Orebro, Sweden; and †Department ofMedicine and Care, Faculty of Health Sciences, Linkoping, Sweden

Posture has an effect on gastric emptying. In this study,we investigated whether posture influences the delay ingastric emptying induced by opioid analgesics. Tenhealthymale subjectsunderwent4gastric emptyingstud-ieswith theacetaminophenmethod.Ontwooccasions thesubjectswere given a continuous infusion of remifentanil(0.2 �g · kg�1 · min�1) while lying either on the right lat-eral side ina20°head-uppositionoron the left lateral sidein a 20° head-down position. On two other occasions noinfusionwasgiven, and the subjectswere studied lying inthe two positions. When remifentanil was given, therewere no significant differences between the two posturesin maximal acetaminophen concentration (right side, 34�mol � L�1; versus left side, 16 �mol � L�1), time taken toreach the maximal concentration (94 versus 109 min), or

area under the serum acetaminophen concentration timecurve from 0 to 60 min (962 versus 197 min � �mol � L�1).In the control situation, there were differences betweenthe postures in maximal acetaminophen concentration(138 versus 94 �mol � L�1; P � 0.0001) and area under theserum acetaminophen concentration time curves from 0to 60min (5092 versus 3793min � �mol � L�1; P � 0.0001),but there was no significant difference in time taken toreach themaximalconcentration (25versus47min).Com-pared with the control situation, remifentanil delayedgastric emptying in both postures. We conclude thatremifentanildelaysgastric emptyingand that thisdelay isnot influenced by posture.

(AnesthAnalg 2004;99:429–34)

T he use of IV, epidural, and intrathecal opioids forpostoperative pain relief causes a delay in gastricemptying (1–3). This may delay intake of fluids

and food and influence the absorption of drugs ad-ministered orally. Both systemic and spinal opioidsdelay gastric emptying (2). This delay may be causedby decreased gastric motility and gastric tone or in-creased pyloric tone. The pylorus has a rich enkepha-linergic innervation, and opioids may therefore in-crease pyloric tone (4). Posture influences gastricemptying, particularly with prolonged emptyingwhen patients are in a left lateral position (5–7). Theuse of opioids also increases the risk for postoperativenausea and vomiting (8).

The effects of posture on gastric emptying duringopioid administration have not been studied. If posture

has an effect, then an optimal position may be found thatfacilitates gastric emptying and thereby reduces the neg-ative effects. However, if opioids increase pyloric tone,gastric emptying may not be influenced by posture.

Because postoperative opioids are used in most pa-tients undergoing major surgery, the purpose of thisstudy was to evaluate whether posture can influence thedelayed gastric emptying induced by an opioid. Wecompared the effects on gastric emptying of 2 extremepostures in 10 healthy volunteers with and without theadministration of remifentanil. The objective for usingvolunteers was to eliminate other factors (e.g., surgicalstress and pain) that influence gastric emptying andstudy the pure effect of an opioid. As an opioid, aninfusion of the ultra-short-acting opioid remifentanilwas chosen because of its pharmacological profilewith which a predictable and constant effect could beachieved. The acetaminophen method was used tostudy gastric emptying.

MethodsAfter approval of the study protocol by the ethicscommittee of the Orebro County Council, 10 healthymale volunteers with a mean age of 23.9 yr (range,21–31 yr), a mean weight of 80 kg (range, 71–98 kg),

The study was supported by grants from the Orebro CountyCouncil Research Committee.

Accepted for publication January 20, 2004.Address correspondence to Jakob Wallden, MD, Department of

Anesthesia and Intensive Care, Orebro University Hospital, 701 85Orebro, Sweden. Address e-mail to jakob.wallden@tele2.se. Ad-dress reprint requests to Magnus Wattwil, MD, PhD, Department ofAnesthesia and Intensive Care, Orebro University Hospital, 701 85Orebro, Sweden. Address e-mail to magnus.wattwil@orebroll.se.

DOI: 10.1213/01.ANE.0000121345.58835.93

©2004 by the International Anesthesia Research Society0003-2999/04 Anesth Analg 2004;99:429–34 429

and a mean height of 180 cm (range, 173–188 cm) wererecruited to the study. The subjects gave their in-formed consent to participate after receiving verbaland written information. Only men were recruited,because the menstrual cycle may alter gastric empty-ing (9). None of them was taking any medications, andnone had a history of gastrointestinal disturbances. Ina randomized order, each subject was studied on fouroccasions, with at least 1 day between occasions. Theywere given a continuous infusion of remifentanil ontwo occasions while lying either on the right lateralside with the bed in a 20° head-up position (RHU) oron the left lateral side with the bed in a 20° head-downposition (LHD). On the other two occasions, noremifentanil infusion was given, and the subjects werestudied lying in the two positions (RHU and LHD).

The subjects fasted (both liquids and solids) for atleast 6 h before each study. An indwelling IV catheterwas placed in one arm for the drawing of blood sam-ples. On the occasions when remifentanil was given,an IV line was established in the opposite arm.Remifentanil was administered as a continuous infu-sion in a dose of 0.2 �g · kg�1 · min�1 and was started10 min before the ingestion of acetaminophen. Theinfusion was terminated directly after the last bloodsample (120 min) was drawn.

During the study, the usual monitors were used.Heart rate, arterial blood pressure, oxygen saturation,end-tidal carbon dioxide (CO2), respiratory rate, andsedation level were recorded every fifth minute. Atthe same intervals, the subjects were asked if theywere experiencing nausea or any other symptoms. Thesedation level was recorded as follows: no sedation �1, light sedation � 2, moderate sedation � 3, and deepsedation � 4. The visual analog scale (VAS) 0–10 wasused for nausea, where VAS 0 was no subjectivesymptoms and VAS 10 was the worst nausea thesubjects could imagine.

If the subject showed signs of excessive sedation, re-spiratory depression, severe nausea, or vomiting orshowed signs of other severe symptoms related to theinfusion of remifentanil, the dose was reduced. Duringthe two control situations, heart rate, arterial blood pres-sure, and sedation level were recorded every 15min, andthe subjects were questioned about nausea. Their level ofsedation was checked at the same intervals.

The acetaminophen absorption test was used formeasurement of gastric emptying. Acetaminophen1.5 g dissolved in 200 mL of water was ingested orally,and venous blood samples were taken at 5, 10, and15 min and then at 15-min intervals for 120 min. Whenremifentanil was given, acetaminophen was takenorally 10 min after the start of the infusion. Acetamin-ophen is not absorbed from the stomach but is rapidlyabsorbed from the small intestine. Consequently, therate of gastric emptying determines the rate of absorp-tion of acetaminophen administered into the stomach.

Serum acetaminophen was determined by an immu-nologic method including fluorescence polarization(TDx® acetaminophen; Abbott Laboratories; NorthChicago, IL). Acetaminophen concentration curveswere produced, and the maximal acetaminophen con-centration (Cmax), the time taken to reach the maximalconcentration (Tmax), and the area under the serumacetaminophen concentration time curves from 0 to60 min (AUC60) were calculated. Tmax was assumed tobe 120 min if no acetaminophen was detected in anysample. The acetaminophen method is a well acceptedmethod for studying the liquid phase of gastric emp-tying, and AUC60 correlates very well with measuresof gastric emptying performed with isotope tech-niques (10,11).

A prior power calculation was performed and de-signed to detect differences in AUC60 between the 2postures when remifentanil was given. On the basis ofdata from previous studies, the estimated sample sizewas 10 volunteers with a power of 80% at the 5%significance level.

The results are presented as means with standarddeviations. Repeated-measures analysis of variancewas used for overall differences between the studysituations. If the analysis of variance showed differ-ences, a paired Student’s t-test with Bonferroni’s cor-rection was used for comparisons between the situa-tions. The significance level was set at 5% in all tests.

ResultsThe acetaminophen concentration curves are pre-sented in Figure 1. There were significant differencesin AUC60 (P � 0.001), Cmax (P � 0.001), and Tmax (P �0.001) among the 4 study situations. During theremifentanil infusion, AUC60 was lower, Cmax wassmaller, and Tmax was longer in both postures com-pared with the control situations. During the controlsituations, there were statistically significant differ-ences, with a higher AUC60 and a larger Cmax in theRHU position. There were no statistically significantdifferences in AUC60, Cmax, or Tmax between the 2postures when remifentanil was given (Table 1).

In 3 subjects (30%), the dose of remifentanil had tobe reduced during the study because of side effects(Table 2). Six subjects (60%) experienced nausea, threesubjects (30%) vomited, and six subjects (60%) hadpruritus during at least one of the remifentanil situa-tions. There was no nausea, vomiting, or pruritusduring the control situations (Table 3).

Systolic blood pressure and heart rate were stable inall situations during the study. Arterial blood pressuredecreased slightly compared with the initial pressurein all situations, but no further changes were detected.Heart rate decreased in the LHD position when noinfusion was given (Table 4).

430 ANESTHETIC PHARMACOLOGY WALLDEN ET AL. ANESTH ANALGREMIFENTANIL AND GASTRIC EMPTYING 2004;99:429–34

STU

DY

I

Respiratory rate decreased in the RHU position, andend-tidal CO2 increased in both positions during theremifentanil infusion (Table 4). There was no changein oxygen saturation. In 7 subjects (70%), the respira-tory rate decreased during the remifentanil infusion to�5 breaths/min. After verbally reminding the volun-teers to breathe and, for one subject, reducing theinfusion of remifentanil, the respiratory rate immedi-ately returned to an acceptable level.

Seven subjects experienced difficulty swallowingduring the remifentanil infusion, but the symptomceased within minutes after the infusion was termi-nated. Five subjects complained of headache duringand after the infusion of remifentanil, and in somesubjects the headache persisted for several hours.

DiscussionThis study has demonstrated that body position influ-ences gastric emptying of fluids, that remifentanil insmall doses delays gastric emptying of fluids, and thata change in body position does not influence the delayin gastric emptying induced by remifentanil. Duringthe control situation, the RHU position resulted infaster gastric emptying than the LHD position.

Gastric emptying is influenced by at least threemechanisms—gastric tone, gastric motility, and pylo-ric tone. The proximal fundus of the stomach func-tions as a reservoir, and the muscles are adapted formaintaining a continuous contractile tone. The distalantrum/pyloric area of the stomach exhibits phasicand peristaltic contractile activity and functions bothas a pump and a grinding mill (12). The tone of thepyloric sphincter regulates the outflow to the duode-num. Consequently, changes in any of these factorswill affect the rate of gastric emptying.

There are limited reports on the effects of posture ongastric emptying. Anvari et al. (13) found that gastricemptying of nonnutrient liquids was faster in the sit-ting position compared with the left lateral position

and that even after a delay in gastric emptying in-duced by atropine there were differences between thepositions. The faster emptying in the sitting positionbefore atropine was associated with increased antralperistaltic activity and increased pyloric pressure, butafter atropine, no differences in antropyloroduodenalmotility could be observed. The mechanism for thechange in motility was thought to be due to effects ofgravity rather than primarily to changes in motility.Other authors also report that the left lateral positionis associated with a delay in gastric emptying (5–7),and our findings are in accordance with these results.

The effect of gravity on gastric emptying is depen-dent on pyloric tone. Even if posture moves the gastriccontents toward the pylorus and there is a high pylorictone, gastric emptying will be difficult. In the controlsituation in this study, emptying time was fast in bothpostures. This indicates that passage through the py-loric region was easy. Opioids decrease gastric tone(14), but even if gastric tone was decreased, gastricemptying would have been facilitated by the RHUposition. Because our study showed no significantdifferences in gastric emptying between the RHU andLHD positions during remifentanil infusion, these re-sults indicate that remifentanil increases pyloric toneand thereby impairs the flow into the duodenum. Ithas been clearly shown that the pylorus has a richenkephalinergic innervation (4), which may explainthe effect of opioids on pyloric obstruction. No con-clusions about gastric motility can be drawn on thebasis of the results of our study.

Several studies have demonstrated that both systemicand spinal opioids delay gastric emptying (1,15,16), andthese effects are both peripherally and centrally medi-ated (2). Opioid receptors are present in the gastric tract,and recently developed opioid antagonists such asmeth-ylnaltrexone and alvimopan, which do not pass theblood-brain barrier, reverse the opioid-induced inhibi-tion of gastrointestinal motility (17,18).

Opioids pass the blood-brain barrier and have thepotential to regulate motility through a central mecha-nism. The dorsal vagal complex, located in the medullarbrainstem, receives sensory information from the gastro-intestinal tract through afferent vagal fibers and is alsothe origin of efferent vagal fibers projecting to the gas-trointestinal tract. �-Opioid receptors have been identi-fied in the synaptic connections within this region, andopioid agonists given locally inhibit gastric motility anddecrease gastric tone (19). The role of opioids in thebrainstem’s normal physiological control of gastrointes-tinal motility is controversial, because local injection ofthe opioid antagonist naloxone has not been found toinfluence motility per se (20). However, intrathecal mor-phine has been shown to inhibit motility and delay gas-tric emptying (2), so clinical studies consequently sup-port the findings that opioids can inhibit motilitythrough a central mechanism.

Figure 1. Mean (sd) serum acetaminophen concentrations duringthe four study situations.

ANESTH ANALG ANESTHETIC PHARMACOLOGY WALLDEN ET AL. 4312004;99:429–34 REMIFENTANIL AND GASTRIC EMPTYING

The respiratory rate decreased and end-tidal CO2increased during the infusion of remifentanil. CO2induces relaxation of smooth muscle in vascular beds,but we are not aware of any reports concerning theeffects of CO2 on gastrointestinal motility. However,hypercapnia also induces sympathetic stimulation,and the increased sympathetic activity may influencegastric function, with delayed gastric emptying. Soonafter the infusion of remifentanil was terminated, therespiratory rate and end-tidal CO2 were normalized.

Half of the volunteers experienced nausea duringthe remifentanil infusion. Five subjects vomited, butthis was late in the study and therefore had no majoreffect on the acetaminophen study. Opioids are

known to cause nausea and vomiting, but the mecha-nisms are complex. The action is believed to be medi-ated through activation of the chemoreceptor triggerzone (located in the area postrema) (8).

Sixty percent of the subjects experienced pruritus.Pruritus is often seen after the administration of opi-oids, particularly after spinal administration, in whichthere are reports of incidences up to 50% (21).

Remifentanil induced difficulties in swallowing inmost volunteers. Swallowing is a complex motor be-havior controlled by neuronal networks in the brain-stem, and after the administration of intrathecal fent-anyl, there are reports of dysphagia (22). This suggeststhat the mechanism is mediated by a central action.

Table 1. Mean (sd) of AUC60, Tmax, and Cmax in Two Different Body Postures With and Without Infusion ofRemifentanil 0.2 �g � kg�1 � min�1

Variable

Right lateral side head-up position Left lateral side head-down position

Remifentanil Control Remifentanil Control

AUC60 (min � �mol � L�1) 962 (902) 5092 (1125) 197 (128) 3793 (1307)Cmax (�mol � L�1) 34 (24) 138 (45) 16 (14) 94 (30)Tmax (min) 94 (33) 25 (14) 109 (10) 47 (22)

AUC60 Cmax Tmax

RHU-Remi vs RHU-Control P � 0.0001 P � 0.0001 P � 0.0001

RHU-Remi vs LHD-Remi NS NS NS

RHU-Remi vs LHD-Control P � 0.0001 P � 0.0001 P � 0.0001

RHU-Control vs LHD-Remi P � 0.0001 P � 0.0001 P � 0.0001

RHU-Control vs LHD-Control P � 0.0083 P � 0.0083 NS

LHD-Remi vs LHD-Control P � 0.0001 P � 0.0001 P � 0.0001

Paired Student’s t-tests with Bonferroni’s correction after the analysis of variance detected differences. The significance level was set at 5%, with P � 0.0083considered significant.

AUC60 � area under the serum acetaminophen concentration curve from 0 to 60 min during the study; Cmax � maximum acetaminophen concentration;Tmax � time taken to reach the maximum acetaminophen concentration.

Table 2. Adjustments of the Initial Dose of Remifentanil 0.2 �g � kg�1 � min�1 in Three Subjects Because of SideEffects

SubjectNo. Position

Time after startof remifentanila

(min) Side effect

Adjusted dose ofremifentanil

(�g � kg�1 � min�1)

4 RHU 86 Nausea (VAS � 9) 0.196 Nausea (VAS � 7) 0.05

105 Vomited 0LHD 120 Vomited 0

5 RHU 120 Vomited 0LHD 83 Nausea (VAS � 6) 0.1

103 Nausea (VAS � 6) 0.05113 Headache 0.025

9 RHU 31 Respiratory rate 3 breaths/min 0.15LHD 0 Dose reduced from start because of

respiratory depression duringRHU position

0.15

36 Respiratory rate 3 breaths/min 0.1

RHU � right lateral head-up position; LHD � left lateral head-down position; VAS � visual analog scale (scale 0–10; no subjective symptom � 0, worstsubjective symptom � 10).

a Infusion of remifentanil was started 10 min before the ingestion of acetaminophen. The infusion was normally terminated after 130 min, when the last bloodsample was taken.

432 ANESTHETIC PHARMACOLOGY WALLDEN ET AL. ANESTH ANALGREMIFENTANIL AND GASTRIC EMPTYING 2004;99:429–34

STU

DY

I

Even though remifentanil is an ultra-short-actingdrug, the subjects complained of headache and latenausea after the infusion was stopped. This is proba-bly not a direct effect of remifentanil and may insteadbe an effect of the increased CO2 level, with a possiblecerebral influence.

Because of the extensive side effects found in vol-unteers, we do not consider remifentanil a suitabledrug for postoperative analgesia. A multicenter eval-uation of the use of remifentanil for early postopera-tive analgesia found an occurrence of adverse respira-tory events in 29% of patients (23) and concluded thatthe technique is probably not practical for routineclinical use.

However, with adequate monitoring, remifentanil isa valuable drug for studying the effects of opioids inexperimental setups with volunteers. With this drug,

it is possible to study the dose-response effects ofopioids.

In conclusion, remifentanil delays gastric emptying,and this delay is not influenced by changes in bodyposture. During the control situation, the RHU posi-tion facilitated gastric emptying.

References1. Murphy DB, Sutton JA, Prescott LF, Murphy MB. Opioid-

induced delay in gastric emptying: a peripheral mechanism inhumans. Anesthesiology 1997;87:765–70.

2. Thorn SE, Wattwil M, Lindberg G, Sawe J. Systemic and centraleffects of morphine on gastroduodenal motility. Acta Anaesthe-siol Scand 1996;40:177–86.

3. Thoren T, Wattwil M. Effects on gastric emptying of thoracicepidural analgesia with morphine or bupivacaine. AnesthAnalg 1988;67:687–94.

Table 3. Incidence of Nausea, Vomiting, and Pruritus for Each Posture With and Without Infusion of Remifentanil 0.2�g � kg�1 � min�1 (n � 10)

Variable

Right lateral side head-up position Left lateral side head-down position

Remifentanil Control Remifentanil Control

Nausea 5/10 0/10 6/10 0/10Vomiting 3/10 0/10 2/10 0/10Pruritus 4/10 0/10 4/10 0/10

Five subjects had nausea in both remifentanil situations. The maximal nausea VAS score for those subjects who experienced nausea was in the right lateralhead-up position (median, 6; range, 3–10) and in the left lateral head-down position (median, 4.5; range, 2–9).

Two subjects vomited in both remifentanil situations. There was prior nausea in all subjects who vomited.Two subjects had pruritus in both remifentanil situations.VAS � visual analog scale (Scale 0–10; no subjective symptom � 0, worst subjective symptom � 10).

Table 4. Vital Variables During the Study

VariableBeforestart �10 to 0 min 0–30 min 31–60 min 61–90 min 91–120 min P value

Mean systolic blood pressure (mm Hg)RHU with remifentanil infusion 120 (10.8) 105 (9.2) 109 (10.9) 106 (10.2) 105 (7.3) 105 (8.8) �0.0001a

RHU with no infusion (control) 126 (15.8) — 111 (12.7) 106 (9.6) 104 (8.3) 107 (10.6) �0.0001a

LHD with remifentanil infusion 129 (12.2) 111 (10.5) 109 (14.7) 106 (10.8) 105 (9.5) 106 (11.1) �0.0001a

LHD with no infusion (control) 121 (12.8) — 104 (11.4) 99 (6.9) 100 (8) 106 (7.7) �0.0001a

Mean heart rate (bpm)RHU with remifentanil infusion 66 (12) 66 (15) 68 (18) 70 (14) 70 (13) 70 (14) NSRHU with no infusion (control) 68 (11) — 65 (12) 63 (10) 63 (7) 63 (7) NSLHD with remifentanil infusion 68 (10) 64 (10) 63 (10) 66 (10) 66 (10) 67 (12) NSLHD with no infusion (control) 66 (12) — 60 (9) 59 (6) 59 (5) 58 (6) 0.0055a

Mean respiratory rate (breaths/min)RHU with remifentanil infusion 13.4 (3.3) 9.6 (3.4) 8.8 (3) 8.5 (4) 8.7 (3.6) 9 (3.6) �0.0001a

LHD with remifentanil infusion 10.9 (4.4) 8.5 (2.2) 9.2 (2.4) 9.6 (2.7) 9.2 (3.4) 9.9 (2.1) NSMean end-tidal CO2 (%)

RHU with remifentanil infusion 5.3 (0.4) 6.0 (0.7) 6.8 (0.9) 6.9 (1.1) 6.8 (1.1) 6.8 (0.9) �0.0001b

LHD with remifentanil infusion 5.5 (0.5) 6.0 (0.6) 7.0 (0.6) 6.9 (0.8) 6.9 (1) 6.5 (1.2) �0.0001b

Mean oxygen saturation (%)RHU with remifentanil infusion 98 (1) 98 (1) 98 (1) 98 (1) 98 (1) 98 (1) NSLHD with remifentanil infusion 98 (1) 99 (2) 98 (1) 98 (1) 98 (1) 98 (1) NS

Values are mean (sd).RHU � right lateral side head-up position; LHD � left lateral side head-down position; NS � not significant.Repeated-measurement analysis of variance was used to evaluate differences over time in the monitored variables.a Significant changes in values between before start and during the infusion, but no detectable changes during the infusion of remifentanil.b Significant increase in end-tidal CO2 during the infusion of remifentanil.

ANESTH ANALG ANESTHETIC PHARMACOLOGY WALLDEN ET AL. 4332004;99:429–34 REMIFENTANIL AND GASTRIC EMPTYING

4. Edin R, Lundberg J, Terenius L, et al. Evidence for vagal en-kephalinergic neural control of the feline pylorus and stomach.Gastroenterology 1980;78:492–7.

5. Horowitz M, Jones K, Edelbroek MA, et al. The effect of postureon gastric emptying and intragastric distribution of oil andaqueous meal components and appetite. Gastroenterology 1993;105:382–90.

6. Burn-Murdoch R, Fisher MA, Hunt JN. Does lying on the rightside increase the rate of gastric emptying? J Physiol 1980;302:395–8.

7. Spiegel TA, Fried H, Hubert CD, et al. Effects of posture ongastric emptying and satiety ratings after a nutritive liquid andsolid meal. Am J Physiol Regul Integr Comp Physiol 2000;279:R684–94.

8. Apfel CC, Roewer N. Risk assessment of postoperative nauseaand vomiting. Int Anesthesiol Clin 2003;41:13–32.

9. Notivol R, Carrio I, Cano L, et al. Gastric emptying of solid andliquid meals in healthy young subjects. Scand J Gastroenterol1984;19:1107–13.

10. Nimmo WS, Heading RC, Wilson J, et al. Inhibition of gastricemptying and drug absorption by narcotic analgesics. Br J ClinPharmacol 1975;2:509–13.

11. Medhus AW, Lofthus CM, Bredesen J, Husebye E. Gastricemptying: the validity of the paracetamol absorption test ad-justed for individual pharmacokinetics. NeurogastroenterolMotil 2001;13:179–85.

12. Read NW, Houghton LA. Physiology of gastric emptying andpathophysiology of gastroparesis. Gastroenterol Clin North Am1989;18:359–73.

13. Anvari M, Horowitz M, Fraser R, et al. Effects of posture ongastric emptying of nonnutrient liquids and antropyloroduode-nal motility. Am J Physiol 1995;268:G868–71.

14. Hammas B, Thorn SE, Wattwil M. Propofol and gastric effects ofmorphine. Acta Anaesthesiol Scand 2001;45:1023–7.

15. Yuan CS, Foss JF, O’Connor M, et al. Effects of low-dose mor-phine on gastric emptying in healthy volunteers. J Clin Phar-macol 1998;38:1017–20.

16. Lydon AM, Cooke T, Duggan F, Shorten GD. Delayed postop-erative gastric emptying following intrathecal morphine andintrathecal bupivacaine. Can J Anaesth 1999;46:544–9.

17. Yuan CS, Foss JF. Gastric effects of methylnaltrexone on mu,kappa, and delta opioid agonists induced brainstem unitaryresponses. Neuropharmacology 1999;38:425–32.

18. Taguchi A, Sharma N, Saleem RM, et al. Selective postoperativeinhibition of gastrointestinal opioid receptors. N Engl J Med2001;345:935–40.

19. Browning KN, Kalyuzhny AE, Travagli RA. Opioid peptidesinhibit excitatory but not inhibitory synaptic transmission in therat dorsal motor nucleus of the vagus. J Neurosci 2002;22:2998–3004.

20. Gue M, Junien JL, Bueno L. Central and peripheral opioidmodulation of gastric relaxation induced by feeding in dogs.J Pharmacol Exp Ther 1989;250:1006–10.

21. Rathmell JP, Pino CA, Taylor R, et al. Intrathecal morphine forpostoperative analgesia: a randomized, controlled, dose-ranging study after hip and knee arthroplasty. Anesth Analg2003;97:1452–7.

22. Currier DS, Levin KR, Campbell C. Dysphagia with intrathecalfentanyl. Anesthesiology 1997;87:1570–1.

23. Bowdle TA, Camporesi EM, Maysick L, et al. A multicenterevaluation of remifentanil for early postoperative analgesia.Anesth Analg 1996;83:1292–7.

434 ANESTHETIC PHARMACOLOGY WALLDEN ET AL. ANESTH ANALGREMIFENTANIL AND GASTRIC EMPTYING 2004;99:429–34

STUDY II

Klunserna Study 05-10-25, 10.12123

STU

DY

II

J Anesth (2006) 20:261–267DOI 10.1007/s00540-006-0436-3

Original articles

The effect of anesthetic technique on early postoperative gastricemptying: comparison of propofol-remifentanil and opioid-freesevoflurane anesthesia

Jakob Walldén1, Sven-Egron Thörn2, Åsa Lövqvist2, Lisbeth Wattwil2, and Magnus Wattwil2

1 Department of Anesthesia, Sundsvall Hospital, 851 86 Sundsvall, Sweden2 Departments of Anesthesia and Intensive Care, Örebro University Hospital, Örebro, Sweden

Introduction

Gastric emptying is an essential part of gastrointestinalmotility, and a postoperative delay may postpone theearly start of oral feeding and alter the bioavailability oforally given drugs [1]. Today a majority of our patientsundergo surgery on an ambulatory basis and an impor-tant part of the care is to have them tolerate oral nutri-tion and per-oral analgesics as soon as possible. A delayin gastric emptying may therefore postpone a patient’sdischarge.

Activation of inhibitory neural pathways by the surgi-cal trauma, a local inflammatory response in the gas-trointestinal tract, and the drugs used perioperativelycontribute to the impairment of gastric motility [2], and,of the drugs used, opioids are thought to constitute themost important factor.

The extent to which anesthetic technique contributesto the early postoperative inhibition of gastric motilityis uncertain. With an inhalation technique withoutopioids, the effect of inhalation agents on gastric motil-ity may cease quickly after discontinuation of the agent[3]. An intravenous technique with an ultra-short-acting opioid, to minimize the negative opioid effect onmotility, combined with propofol, which has antiemeticproperties, and to some degree, antagonizes the opioideffect on gastric motility [4], may favor motility.Both methods are, theoretically, optimal for gastricmotility. However, when these anesthetic techniquesare used in major surgery there may be a needfor opioid analgesics in the early postoperative period,as the residual analgesic properties of the anestheticscease quickly. If one of the techniques proves to havea faster gastric emptying rate, this may have animpact on the choice of anesthesia to optimize gastricmotility.

The aim of this study was to compare the effect onearly gastric emptying between two anesthetic methods,an inhalation opioid-free sevoflurane-based anesthesia

AbstractPurpose. A postoperative decrease in the gastric emptying(GE) rate may delay the early start of oral feeding and alterthe bioavailability of orally administered drugs. The aim ofthis study was to compare the effect on early gastric emptyingbetween two anesthetic techniques.Methods. Fifty patients (age, 19–69 years) undergoingday-case laparascopic cholecystectomy were randomlyassigned to received either total intravenous anesthesia withpropofol/remifentanil/rocuronium (TIVA; n = 25) or inhala-tional opioid-free anesthesia with sevoflurane/rocuronium(mask induction; GAS; n = 25). Postoperative gastric empty-ing was evaluated by the acetaminophen method. Afterarrival in the recovery unit, acetaminophen (paracetamol)1.5g was given through a nasogastric tube, and blood sampleswere drawn during a 2-h period. The area under the serum-acetaminophen concentration curve from 0–60min (AUC60),the maximal concentration (Cmax), and the time to reach C-max (Tmax) were calculated.Results. Twelve patients were excluded due to surgical com-plications (e.g., conversion to open surgery) and difficulty indrawing blood samples (TIVA, n = 7; GAS, n = 5). Gastricemptying parameters were (mean ± SD): TIVA, AUC60,2458 ± 2775 min·μmol·l−1; Cmax, 71 ± 61 μmol·l−1; and Tmax,81 ± 37min; and GAS, AUC60, 2059 ± 2633min·μmol·l−1;Cmax, 53 ± 53 μmol·l−1; and Tmax, 83 ± 41 min. There were nosignificant differences between groups.Conclusion. There was no major difference in early postop-erative gastric emptying between inhalation anesthesia withsevoflurane versus total intravenous anesthesia with propofol-remifentanil. Both groups showed a pattern of delayed gastricemptying, and the variability in gastric emptying was high.Perioperative factors other than anesthetic technique mayhave more influence on gastric emptying.

Key words Gastrointestinal motility · Gastric emptying ·Anesthesia, inhalation · Anesthesia, intravenous · Analgesics,opioid · Cholecystectomy, laparoscopic

Address correspondence to: J. WalldénReceived: March 20, 2006 / Accepted: July 28, 2006

262 J. Walldén et al.: Anesthetic technique and gastric emptying

and an intravenous propofol-remifentanil basedanesthesia.

Patients, materials, and methods

Fifty patients (American Society of Anesthesiologists[ASA] physical status I and II) undergoing day-caselaparoscopic cholecystectomy at Örebro UniversityHospital, Sweden, were included in this study. Thestudy protocol was approved by the Ethics Committeeof the Örebro County Council and by the SwedishMedical Product Agency. The patients entered thestudy after giving verbal and written consent. Patientswere randomly allocated (by the use of sealed envel-opes) to receive either total intravenous anesthesia(TIVA group; n = 25) or total inhalation anesthesia(GAS group; n = 25). An independent nurse preparedall the sealed envelopes from of a computer-generatedtable before the study started. Investigators (J.W.,M.W., S.E.T.) enrolled patients to the study. The envel-opes were opened by the investigators just before theinduction of anesthesia. There was no blinding in thestudy.

Patients were excluded from the study if the proce-dure was converted to open cholecystectomy, or if theduration of surgery exceeded 150 min.

The gastric emptying study was started immediatelyafter the patient’s arrival at the recovery unit. Duringthe first 24 h after surgery, the incidence of postopera-tive nausea and vomiting (PONV) and pain, and theneed for opioid analgesics were evaluated by means ofobservations in the recovery unit, a telephone inter-view, and a questionnaire.

The primary endpoints in the study were the gastricemptying parameters, and we tested the hypothesis thatthere would be a difference in gastric emptying betweenthe study groups.

For the secondary outcome variables (PONV, pain,opioid need) we were aware that the number of patientsmight be too small to detect differences.

The patients fasted for 6 h but were allowed to drinkclear fluids up to 2 h before premedication. All patientsreceived premedication with midazolam 1–2 mg IV atthe day-care unit, 20–30min before the induction ofanesthesia. In the operating room, patients underwentroutine monitoring, including continuous processedelectroencephalography (Bispectral index [BIS]-monitor; Aspect Medical Systems, Newton, MA, USA).Before induction, all patients received ketorolac 30mgIV. In the TIVA group, anesthesia was induced with aninfusion of remifentanil 0.2μg·kg−1·min−1, followed, after2 min, by a target-controlled infusion (TCI) of propofolat 4μg·ml−1 (induction time, 60 s). In the GAS group,anesthesia was induced with 8% sevoflurane via a facial

mask. After an adequate level of anesthesia was at-tained, muscular relaxation was obtained in both groupswith rocuronium 0.6 mg·kg−1 IV, and the trachea wasintubated after 90 s. In the TIVA group, anesthesia wasmaintained with remifentanil 0.2 μg·kg−1·min−1 and TCIpropofol, adjusted (2–4 μg·ml−1) to maintain a BISindex below 50. In the GAS group, anesthesia wasmaintained with sevoflurane, with concentrations ad-justed to maintain a BIS index below 50. No prophylac-tic antiemetics were given. A nasogastric tube wasplaced in all patients during anesthesia. At the end ofsurgery, 20 ml of 0.25% levobupivacaine was infiltratedat the insertion sites of the laparoscopic instruments,muscular relaxation was reversed with neostigmine2.5 mg/glycopyrrolate 0.5mg, and anesthetic agent(s)were terminated. The patients were extubated in theoperating room after return of consciousness and spon-taneous breathing and transferred to the adjacentday-care unit for recovery. Except for the continuousinfusion of remifentanil in the TIVA group, no opioidswere given during anesthesia.

Acetaminophen absorption was used as an indirectmeasure of gastric emptying [5]. Acetaminophen is notabsorbed from the stomach, but is rapidly absorbedfrom the small intestine. Consequently, the rate of gas-tric emptying determines the rate of absorption ofacetaminophen administered into the stomach.Immediately after patients’ arrival at the day-care unit,acetaminophen 1.5 g, dissolved in 200ml of water (atroom temperature), was given through the nasogastrictube. Prior to administration, correct placement of thetube was verified by auscultation over the stomach areaduring the injection of 20ml of air into the tube. Thetube was removed after acetaminophen was given.Blood samples were taken from an intravenous catheterprior to the administration of acetaminophen and then5, 10, and 15min after the administration, and then at15-min intervals during a period of 120min. Serumacetaminophen was determined by an immunologicmethod, including fluorescence polarization (TDx ac-etaminophen; Abbott Laboratories, Chicago, IL, USA).Acetaminophen concentration curves were produced,and the maximal acetaminophen concentration (Cmax),the time taken to reach the maximal concentration(Tmax), and the area under the serum-acetaminophenconcentration time curves from 0 to 60min (AUC60) and0 to 120 min (AUC120) were calculated. Tmax was as-sumed to be 120min if no acetaminophen was detectedin any sample. The acetaminophen method is a well-accepted method for studying the liquid phase of gastricemptying, and the AUC60 correlates well with measuresof gastric emptying performed using isotope techniques[5].

The patients stayed in the day-care unit for at least4 h. During this period, nausea, vomiting, and pain were

STU

DY

II

J. Walldén et al.: Anesthetic technique and gastric emptying 263

evaluated every hour. Nausea and pain were evaluatedwith a visual analogue scale (VAS), and occurrences ofvomiting were recorded. Droperidol 0.5–1mg IV wasgiven on request as the first rescue antiemetic accordingto the routines of the department. If not sufficient,ondansetron 2–4mg IV was given as the second drug.If patients scored more than 3 on the VAS for pain,ketobemidone 1–2mg IV was given. Ketobemidoneis an opioid analgesic with properties similar to thoseof morphine and is widely used in the Scandinaviancountries.

After discharge from the day-care unit, the patientsthemselves completed a questionnaire about PONVand pain during the time period 4–24h postoperatively.The patients scored the maximal pain and maximal nau-sea on a VAS and were questioned as to whether theyhad vomited or not. A nurse or doctor also performed atelephone interview on the first postoperative day,during which patients were questioned about events ofpain, nausea, or vomiting after discharge. Combiningthe observations from the recovery unit, the question-naire, and the telephone interview, we acquired vari-ables regarding the incidence of PONV during 0–2hand 2–24h, the need for antiemetics in the day-care unit,the maximal VAS score for pain during the periods 0–2h and 2–24h, the time to first dose of opioid analgesics,and the total dose of opioids given. These variableswere regarded as secondary outcome variables in thestudy.

Sample size was calculated based on the AUC60 as theprimary outcome variable. A difference of at leastone-third of AUC60 under normal conditions wasconsidered clinically significant. Based on previousstudies [6], we estimated the minimal difference to be2000min·μmol·l−1 and the within-group SD for theAUC60 to be 2000 min·μmol·l−1. For a power of 0.8

and α = 0.05, a sample size of 17 patients in each groupwas calculated to be appropriate. From previous studieswith the acetaminophen method, we had the experiencethat, in some patients, it might be difficult to drawvenous blood samples due to a constricted venous sys-tem. For this reason, we increased the study populationto 25 patients in each group.

To be able to compare our gastric-emptying resultswith a normal gastric-emptying profile (in our contextwithout any influence from anesthesia, surgery, pain,drugs, etc) we used a pooled dataset of controlgastric-emptying measurements from three previousstudies by our group. In the first study [6] the controlswere taken 4–5 weeks after an open cholecystectomy (n= 17; ASA, I–II; mean (±SD) age, 49 ± 15 years; male, n= 4; female, n = 13); in the second study (unpublisheddata), 4 weeks after abdominal surgery (n = 9; ASA,I–II; mean age, 69 ± 10 years; male, n = 7; female, n = 2);and in the third study, the controls were young healthymale volunteers in an experimental setting [7] (n = 10;ASA, I; mean age, 24 ± 3.4 years). In all control mea-surements, 1.5g acetaminophen dissolved in 200ml ofwater was given orally after a period of fasting andblood samples were taken every 15min during 2 h. Thehandling and laboratory analysis of the samples werethe same as in the current study, as described above.The mean serum-acetaminophen concentration curveof the pooled data is presented in Fig. 1, and thegastric emptying parameters were (mean ± SD): AUC60,5988 ± 1713 min·μmol·l−1; Cmax, 145 ± μmol·l−1; and Tmax,29 ± 15min.

The primary outcome variables AUC60, AUC120, Cmax,and Tmax, are presented as means with SDs. The second-ary outcome variables are presented as events, num-bers, or medians with ranges. Unpaired Student’s t-test,Mann-Whitney U-test, or Fisher’s exact test was used

Fig. 1. Mean (+SD) serum (S)-acetaminophen concentrations during thegastric emptying study after propofol-remifentanil total intravenous anesthesia(TIVA) or opioid-free sevoflurane (GAS)anesthesia. As a reference for normal gas-tric emptying, a group of historical con-trols, pooled from control groups in threeprevious studies (see the Methods sectionfor description), is included in the graph

264 J. Walldén et al.: Anesthetic technique and gastric emptying

for statistical analysis, and P < 0.05 was considered sta-tistically significant.

Results

Fifty patients were included in the study from April2002 to January 2003. Five patients (TIVA, n = 4; GAS,n = 1) were excluded due to conversion to open chole-cystectomy or prolonged duration of surgery (>150 min)due to choledochal stones. In 7 patients (TIVA, n = 3;GAS, n = 4) there were difficulties in drawing bloodsamples for the acetaminophen concentration analysis.Hence, a total of 12 patients (TIVA, n = 7; GAS, n = 5)were excluded from the analysis of the primary outcomevariable.

Patient characteristics are presented in Table 1. Sur-gery and anesthesia were uneventful in all patients.There were no differences between the groups in dura-tion of surgery or duration from end of surgery to startof the gastric emptying studies.

Acetaminophen concentration curves are presentedin Fig. 1. There were no differences between the groupsin the primary outcome variables, AUC60, AUC120, Cmax,or Tmax (Table 2). Both groups differed significantly (P <0.01) from the pooled historical control group. Of the 38patients eligible for the primary outcome analysis, only1 patient had no detectable acetaminophen in any of theblood samples (i.e., no gastric emptying at all); seeTable 3.

Table 1. Patient characteristics and time variables before the start of the gastricemptying study

TIVA group GAS group(n = 24) (n = 21) P valuea

Age (years) 45 (29–64) 46 (19–69) NSHeight (cm) 168 (152–189) 169 (158–187) NSWeight (kg) 80 (56–112) 75 (56–100) NSFemales 20 16 NSMales 4 5 NSSmokers 4 4 NSASA Class I 19 17 NSASA Class II 5 4Duration of surgery (min) 74 (25–148) 70 (65–108) NSDuration from end of surgery 8 (2–17) 9 (2–22) NS

to tracheal extubation (min)Duration from end of surgery 19 (10–30) 22 (8–45) NS

to arrival at recovery unit (min)Duration from end of surgery 24 (13–35) 26 (17–45) NS

to start of GE study (min)

Values are given as means with ranges or numbersTIVA, total intravenous anesthesia with remifentanil and propofol; GAS, total inhalationanesthesia with sevoflurane; GE, gastric emptyinga Unpaired Student’s t-test or Fisher’s exact test

Table 2. Mean and SD of AUC60, AUC120, Cmax, and Tmax in the two study groups

TIVA group GAS group 95% CI for the differenceVariable (n = 18) (n = 20) between the means P valuea

AUC60 (min·μmol−1·l−1) 2458 ± 2775 2059 ± 2633 −1390 to 2188 NS (P = 0.65)AUC120 (min·μmol−1·l−1) 5889 ± 5750 4288 ± 4820 −1877 to 5079 NS (P = 0.36)Cmax(μmol·l−1) 71 ± 61 53 ± 55 −20 to 56 NS (P = 0.35)Tmax(min) 81 ± 37 83 ± 41 −28 to 24 NS (P = 0.85)

TIVA, total intravenous anesthesia with remifentanil and propofol; GAS, total inhalation anesthesia with sevoflurane; AUC60, AUC120, areaunder the serum-acetaminophen concentration curve at 0–60min and 0–120 min; Cmax, maximum acetaminophen concentration; Tmax, time takento reach the maximum acetaminophen concentration; CI, confidence interval; NS, not significanta Unpaired Student’s t-test

Table 3. Number of patients without detectable serum aceta-minophen (no gastric emptying at all) at different time periods

TIVA group GAS group(n = 18) (n = 20) P valuea

0–60 Min 3 1 NS0–120 Min 1 0 NS

TIVA, total intravenous anesthesia with remifentanil and propofol;GAS, total inhalation anesthesia with sevoflurane; NS, not significanta Fisher’s exact test

STU

DY

II

J. Walldén et al.: Anesthetic technique and gastric emptying 265

Secondary outcome variables were obtained in 45patients (TIVA, n = 21; GAS, n = 24). The questionnairewas completed by 20 patients (95%) in the TIVA groupand 23 patients (96%) in the GAS group. The telephoneinterview was performed in 20 patients (95%) in theTIVA group and 22 patients (92%) in the GAS group.For the period 2–24h postoperatively, secondary out-come variables could be obtained in all patients.

There were no statistically significant differencesbetween the groups in the incidence of nausea,vomiting, or PONV (Table 4). Twelve (57%) patients inthe TIVA group and 10 (42%) patients in the GASgroup were given rescue antiemetics in the recoveryunit.

There were no differences between the groupsin maximal VAS scores for pain, the need for opioid

analgesics, or the dose of opioid analgesics. The time tothe first administration of opioids in the recovery unitwas significantly longer in the GAS group (Table 5).

Discussion

This study demonstrates that patients anesthetized withan inhalational, opioid-free regimen with sevofluranehad a gastric emptying pattern in the early postopera-tive period (0–2h) similar to that in patients anes-thetized with an intravenous propofol-remifentanilregimen. When our results were compared with the gas-tric emptying pattern seen in a normal state (no anes-thesia and no surgery), gastric emptying could beconsidered to be delayed in both groups.

Table 5. Pain variables

Variable TIVA group GAS group P valuea

n = 21 n = 24Median (range) for the highest VAS score for pain 0–2h 5 (0–9) 4 (0–9) NSMedian (range) for the highest VAS score for pain 2–24h 4 (0–10) 4 (0–7) NSNumber of patients with need for opioid analgesics in recovery unit 17 (81%) 20 (83%) NS

n = 17 n = 20Median (range) total dose of ketobemidone IV (mg) in patients 5.9 (1.5–11) 5.0 (2.0–11) NS

who received opioid analgesicsMedian (range) time from arrival at recovery unit to first dose of 17 (0–45) 44 (0–155) <0.01

ketobemidone (min) in patients who received opioid analgesics

TIVA, total intravenous anesthesia with remifentanil and propofol; GAS, total inhalation anesthesia with sevoflurane; NS, not significant; VAS,100-mm visual analogue scalea Mann-Whitney U-test or Fisher’s exact test

Table 4. Numbers (%) of patients with events of postoperative nausea and/or vomiting(PONV) during the study

TIVA group GAS groupVariable (n = 21) (n = 24) P valuea

Postoperative 0–2hNausea 10 (48%) 15 (62%) NSVomiting 2 (10%) 4 (17%) NSNausea or vomiting 10 (48%) 16 (67%) NS

Postoperative 2–24hNausea 11 (52%) 16 (67%) NSVomiting 5 (24%) 8 (33%) NSNausea or vomiting 12 (57%) 16 (67%) NS

Postoperative 0–24hNausea 15 (71%) 20 (83%) NSVomiting 6 (29%) 8 (33%) NSNausea or vomiting 16 (76%) 20 (83%) NS

TIVA, total intravenous anesthesia with remifentanil and propofol; GAS, total inhalationanesthesia with sevoflurane; NS, not significantEvent of nausea 0–2h, VAS for nausea >10mm at day-care unit; event of nausea 2–24 h, VAS fornausea >10mm at day-care unit or VAS for nausea >10mm on questionnarie, or nausea reportedat telephone interview; VAS, 100-mm visual analogue scalea Fisher’s exact test

266 J. Walldén et al.: Anesthetic technique and gastric emptying

Our study was powered to detect major differences ingastric emptying rate, and the results indicate that theremight be a small difference, with faster gastric emptyingin the total intravenous anesthesia group. However,gastric emptying was greatly delayed in both groups,and we do not consider a potential difference of thissmall magnitude as clinically relevant.

There was great variability in the gastric emptyingrate within the groups. We tested the hypothesis of acorrelation between opioid administration in the earlypostoperative period and gastric emptying rate, but wefound no relation (data not shown). There was both fastand slow emptying among patients who received opioidanalgesics during the gastric emptying study, as well asamong those who did not receive any opioid analgesicsor those who received opioid analgesics after the gastricemptying study was completed. The use of opioid anal-gesics and antiemetics in the recovery period is part ofthe overall perioperative care of the patients and ispartly a consequence of the anesthetic technique. Thesefactors cannot be eliminated and should be consideredas part of the anesthetic technique.

It is always doubtful to include historical data as acontrol. However, we thought it would be valuable torelate the gastric emptying profile seen in the groups inthe present study to a normal gastric emptying profile,which, in our context, means under no influence of anes-thesia, surgery, drugs, pain etc. To create a reference,we pooled data from control situations in three previousstudies performed under different conditions. The gas-tric emptying profiles for these data, both the individualcontrol groups and the pooled group, are similar tothose in other control situations published in the litera-ture [8–10]. We consider our control dataset as an ac-ceptable estimate of a normal gastric emptying profile.It would have been ideal to have control values for eachpatient included in the study, but, unfortunately, thatwas not the study design.

We were aware that the number of patients might betoo small to detect any differences in postoperative nau-sea and vomiting (PONV) [11], and we could not detectany statistically significant differences in PONV be-tween the groups. PONV was not a primary endpoint inthis study, but we considered it valuable to have thePONV recordings. There was a tendency in our studytoward a higher incidence of PONV in the GAS group,and it has been reported that volatile agents may bea main cause of vomiting in the early postoperativeperiod [12]. To draw any conclusions about differencesin PONV between the anesthetic techniques, a largernumber of patients must be studied.

The incidence of PONV was high in both groups. Themajority of patients were non-smoking women, andopioids were given as analgesics in the recovery unit. IfApfel’s simplified risk score [13] were to be applied, the

predicted incidence of PONV would be high in patientswith these characteristics. As there are no data on howantiemetics affect gastric emptying, no prophylacticantiemetics were given.

There is probably no direct relation between gastricemptying and PONV. We have previously shown thatthe perioperative gastric emptying rate is not a predic-tor for PONV [14], and gastric decompression duringanesthesia does not reduce the incidence of PONV [15].

There was a shorter time to the first dose of post-operative opioid analgesics in the group receiving theintravenous anesthesia. This may be explained either bya residual effect of the inhalation agent [16] or by hype-ralgesia caused by remifentanil [17].

Previous studies comparing the effects on gastrointes-tinal motility exerted by different general anesthetictechniques in the clinical situation are limited, and thesehave not shown any differences between different tech-niques [9,18,19]. The results from our study are in accor-dance with these study results, as we found no majordifferences between the groups. Nothing can beconcluded as to what extent the anesthetics used areinvolved in the postoperative impairment of gas-trointestinal motility. Other factors, such as the surgicaltrauma or individual sensitivity to the drugs used maybe more important.

There are several experimental studies addressing theeffects of anesthetic drugs on gastrointestinal motility.The inhibitory effect of opioids on gastrointestinalmotility has been studied extensively. This effect ismainly mediated via opioid receptors, but the mecha-nism and understanding are complex and still uncertain[20]. Opioids inhibit motility even at low doses [21], andthe mechanism is both peripherially and centrally medi-ated [22]. Propofol at low doses does not influence gas-tric motility [23], but there is evidence that propofolmay inhibit motility at higher doses. In a laboratorysetting, propofol inhibited spontaneous contractions inhuman gastric tissue [24]. There are only a few studieson volatile agents and gastrointestinal motility. Volatileanesthetics have inhibitory effects on gastric motility,but the effect may cease quickly after termination of theagents [3, 25].

The anesthetic techniques used in this study, oneopioid-free and one with an ultra-short-acting opioid,would, theoretically, be ideal for optimizing gastricemptying. However, the majority of patients haddelayed gastric emptying with both of these methods.This indicates that it may be difficult to further improveearly gastric emptying by further altering the methodsof general anesthesia. We cannot exclude the possibilitythat all general anesthetic methods have inhibitory ef-fects on early postoperative gastric emptying. Otherperioperative factors may also have main impacts onearly gastric emptying, and it is difficult to distinguish

STU

DY

II

J. Walldén et al.: Anesthetic technique and gastric emptying 267

between all the factors involved. However, intra-operative and postoperative intravenous fluid restric-tion promotes the return of gastrointestinal motility andreduces complications after abdominal surgery [26].Minimizing the surgical trauma during the laparoscopicprocedure reduces pain and nausea [27].

The weakness in our study is that the variability ofgastric emptying was higher than expected, whichresulted in loss of power. However, we believe that ourstudy indicates that, even after optimizing the anes-thetic regimen, gastric emptying is delayed for themajority of patients. In both groups there were severalpatients with fast gastric emptying and there may alsohave been a small difference between the groups thatwas not detected in our study. The high variability mayhave been due to factors other than the anestheticsused, and must be addressed in future studies.

In summary, there were no major differences in earlypostoperative gastric emptying between opioid-freesevoflurane anesthesia and intravenous propofol-remifentanil anesthesia. The variability was high in bothgroups, and perioperative factors other than the anes-thetics used may have greater influence on early postop-erative gastric emptying.

Acknowledgments. This study was supported by grants fromÖrebro County Council, Örebro, and Emil Andersson’s Fundfor Medical Research, Sundsvall.

References

1. Watcha MF, White PF (1992) Postoperative nausea and vomiting.Its etiology, treatment, and prevention. Anesthesiology 77:162–184

2. Bauer AJ, Boeckxstaens GE (2004) Mechanisms of postoperativeileus. Neurogastroenterol Motil 16 (Suppl 2):54–60

3. Schurizek BA (1991) The effects of general anaesthesia onantroduodenal motility, gastric pH and gastric emptying in man.Dan Med Bull 38:347–365

4. Hammas B, Thorn SE, Wattwil M (2001) Propofol and gastriceffects of morphine. Acta Anaesthesiol Scand 45:1023–1027

5. Nimmo WS, Heading RC, Wilson J, Tothill P, Prescott LF (1975)Inhibition of gastric emptying and drug absorption by narcoticanalgesics. Br J Clin Pharmacol 2:509–513

6. Thorn SE, Wattwil M, Naslund I (1992) Postoperative epiduralmorphine, but not epidural bupivacaine, delays gastric emptyingon the first day after cholecystectomy. Reg Anesth 17:91–94

7. Wallden J, Thorn SE, Wattwil M (2004) The delay of gastricemptying induced by remifentanil is not influenced by posture.Anesth Analg 99:429–434

8. Nimmo WS, Littlewood DG, Scott DB, Prescott LF (1978) Gas-tric emptying following hysterectomy with extradural analgesia.Br J Anaesth 50:559–561

9. Mushambi MC, Rowbotham DJ, Bailey SM (1992) Gastric emp-tying after minor gynaecological surgery. The effect of anaesthetictechnique. Anaesthesia 47:297–299

10. Kennedy JM, van Rij AM (2006) Drug absorption from the smallintestine in immediate postoperative patients. Br J Anaesth97:171–180

11. Apfel CC, Roewer N, Korttila K (2002) How to study post-operative nausea and vomiting. Acta Anaesthesiol Scand 46:921–928

12. Apfel CC, Kranke P, Katz MH, Goepfert C, Papenfuss T, RauchS, Heineck R, Greim CA, Roewer N (2002) Volatile anaestheticsmay be the main cause of early but not delayed postoperativevomiting: a randomized controlled trial of factorial design. Br JAnaesth 88:659–668

13. Apfel CC, Laara E, Koivuranta M, Greim CA, Roewer N (1999)A simplified risk score for predicting postoperative nausea andvomiting: conclusions from cross-validations between two cen-ters. Anesthesiology 91:693–700

14. Wattwil M, Thorn SE, Lovqvist A, Wattwil L, Klockhoff H,Larsson LG, Naslund I (2002) Perioperative gastric emptying isnot a predictor of early postoperative nausea and vomiting inpatients undergoing laparoscopic cholecystectomy. Anesth Analg95:476–479

15. Burlacu CL, Healy D, Buggy DJ, Twomey C, Veerasingam D,Tierney A, Moriarty DC (2005) Continuous gastric decompres-sion for postoperative nausea and vomiting after coronaryrevascularization surgery. Anesth Analg 100:321–326

16. Matute E, Rivera-Arconada I, Lopez-Garcia JA (2004) Effects ofpropofol and sevoflurane on the excitability of rat spinalmotoneurones and nociceptive reflexes in vitro. Br J Anaesth93:422–427

17. Hood DD, Curry R, Eisenach JC (2003) Intravenous remifentanilproduces withdrawal hyperalgesia in volunteers with capsaicin-induced hyperalgesia. Anesth Analg 97:810–815

18. Freye E, Sundermann S, Wilder-Smith OH (1998) No inhibitionof gastro-intestinal propulsion after propofol- or propofol/ketamine-N2O/O2 anaesthesia. A comparison of gastro-caecaltransit after isoflurane anaesthesia. Acta Anaesthesiol Scand 42:664–669

19. Jensen AG, Kalman SH, Nystrom PO, Eintrei C (1992) Anaes-thetic technique does not influence postoperative bowel function:a comparison of propofol, nitrous oxide and isoflurane. Can JAnaesth 39:938–943

20. Hicks GA, DeHaven-Hudkins DL, Camilleri M (2004) Opiates inthe control of gastrointestinal tract function: current knowledgeand new avenues for research. Neurogastroenterol Motil 16(Suppl 2):67–70

21. Yuan CS, Foss JF, O’Connor M, Roizen MF, Moss J (1998)Effects of low-dose morphine on gastric emptying in healthy vol-unteers. J Clin Pharmacol 38:1017–1020

22. Thorn SE, Wattwil M, Lindberg G, Sawe J (1996) Systemic andcentral effects of morphine on gastroduodenal motility. ActaAnaesthesiol Scand 40:177–186

23. Hammas B, Hvarfner A, Thorn SE, Wattwil M (1998) Propofolsedation and gastric emptying in volunteers. Acta AnaesthesiolScand 42:102–105

24. Lee TL, Ang SB, Dambisya YM, Adaikan GP, Lau LC (1999)The effect of propofol on human gastric and colonic muscle con-tractions. Anesth Analg 89:1246–1249

25. Marshall FN, Pittinger CB, Long JP (1961) Effects of halothaneon gastrointestinal motility. Anesthesiology 22 363–366

26. Brandstrup B, Tonnesen H, Beier-Holgersen R, Hjortso E,Ording H, Lindorff-Larsen K, Rasmussen MS, Lanng C, Wallin L,Iversen LH, Gramkow CS, Okholm M, Blemmer T, Svendsen PE,Rottensten HH, Thage B, Riis J, Jeppesen IS, Teilum D,Christensen AM, Graungaard B, Pott F (2003) Effects of intrave-nous fluid restriction on postoperative complications: comparisonof two perioperative fluid regimens: a randomized assessor-blinded multicenter trial. Ann Surg 238:641–648

27. Cengiz Y, Janes A, Grehn A, Israelsson LA (2005) Randomizedtrial of traditional dissection with electrocautery versus ultrasonicfundus-first dissection in patients undergoing laparoscopic chole-cystectomy. Br J Surg 92:810–813

STUDY III

Klunserna Study 05-10-25, 10.13135

STU

DY

III

Effects of Remifentanil on Gastric Tone Jakob Walldén, MD * †; Sven-Egron Thörn, MD,PhD §†; Greger Lindberg, MD,

PhD ‡; Magnus Wattwil, MD, PhD §†

* Department of Anesthesia, Sundsvall Hospital, Sundsvall; † School of Health and Medical Sciences, Örebro University, Örebro ‡ Karolinska Institutet, Department of

Medicine, Karolinska University Hospital Huddinge; Huddinge § Department of Anes-thesia and Intensive Care, Örebro University Hospital, Örebro;

SWEDEN

Objectives: Opioids are well known for impairing gastric motility. The mechanism is far from clear and there is wide interindividual variability. The purpose of this study was to evaluate the effect of remifentanil on proximal gastric tone. Materials and Methods: Healthy volunteers were studied on two occasions and proximal gastric tone was measured by a gastric barostat. On the first occasion (n=8) glucagon 1 mg IV was given as a reference for a maximal relaxation of the stomach. On the second occasion (n=9) remifentanil was given in incremental doses (0.1, 0.2 and 0.3 μg•kg-

1•min-1) for 15 min each, followed by a washout period of 30 minutes. Thereafter re-mifentanil was readministered, and 10 minutes later glucagon 1 mg was given. Mean in-tragastric bag volumes were calculated for each 5-minute interval. Analyses of single nu-cleotide polymorphisms (SNP) A118G and G691C in the μ-opioid receptor (MOR) gene were done in all subjects. Results: Glucagon decreased gastric tone in all subjects. Remifentanil had a marked effect on gastric tone; we found two distinct patterns of reactions with both increases and de-creases in gastric tone, and during the remifentanil infusion glucagon did not affect gas-tric tone. We found no association between SNPs A118G and G691C and the two pat-terns of gastric tone reactions to remifentanil. Conclusions: Remifentanil induced changes in gastric tone with both increases and de-creases in tone. As a preliminary observation, the variation between individuals could not be explained by SNPs in the MOR gene. Keywords: Gastrointestinal motility; Gastric tone; Analgesics, Opioid; Polymorphism, Single Nucleotide; Receptors, Opioid, mu/*genetics; Genotype This study was supported by grants from FoU-centrum, Sundsvall, Emil Andersson’s Fund for Medical Research, Sundsvall and Örebro County Council Research Committee. The manuscript is submitted.

2 Jakob Wallden

Introduction Preoperative fasting, bioavailability of drugs given orally (i.e. premedication), gastric re-tention with the associated risk of aspiration, and the postoperative start of oral intake are examples of issues that are highly dependent on gastric motility. Gastric motility is often impaired during and after surgery/anesthesia as a result of many contributing fac-tors. Opioids, given as part of the anesthesia and the postoperative analgesic regimes, play a major role in this impairment. Gastric emptying, the functional goal of gastric motility, is determined by an integrative motor activity in the stomach. The proximal part of the stomach acts as a reservoir and exhibits a constant dynamic tone that adapts to the volume load. The distal part of the stomach exhibits a distinct peristaltic activity and acts both as a pump towards the duo-denum and a grinding mill. Gastric tone can be expressed as the length of the muscle fibers in the proximal stomach. The tone is not equivalent to pressure. As there is an adaptive relaxant reflex, a volume load can maintain the intragastric pressure. Therefore, an almost empty stomach and a full stomach are able to have the same intragastric pressure, but different tone. The gas-tric barostat, which maintain a constant pressure in an air-filled intragastric bag, meas-ures gastric tone as isobaric volume variations (1). Opioids are well known for impairing gastric motility and emptying (2-4). However, knowledge about the effects of opioids on proximal gastric tone is limited and results from published research are divergent (5, 6). There is also little knowledge about how the highly potent μ-opioid receptor agonist remifentanil affects gastric motility. The primary aim of this study was to evaluate the effect of the μ-opioid receptor agonist remifentanil on proximal gastric tone during fasting conditions. The study was per-formed in healthy volunteers and the gastric barostat was used to measure gastric tone. There are substantial inter-individual differences in the general response to opioids (7), and recent studies have suggested that polymorphism of the μ-opioid receptor (MOR) gene, with an altering of the receptor-function, may be a cause of the variation (8-10). Since we found large variations in gastric tone in response to remifentanil in this study, we also investigated if the variation was correlated to the presence of two different poly-morphisms in the μ-opioid receptor gene.

Methods Following approval of the study protocol by the Ethics Committee of the Örebro County Council, 10 healthy male volunteers with a mean age of 24 years (range, 19-31 years), a mean weight of 75 kg (range, 60-84 kg) and a mean height of 182 cm (range, 171-185 cm) were recruited to the study. The subjects gave their informed consent to participate after receiving verbal and written information. Only men were recruited, since the men-strual cycle may alter gastric motility (11). None of them were taking any medications and there was no history of gastrointestinal disturbances. Each subject underwent two study protocols on two separate days. In the first study, the effect of glucagon on gastric tone was measured. Glucagon is a potent inhibitor of gas-trointestinal motility and induces a powerful relaxation of the stomach, resulting in an increase in gastric volume (12). The objectives were to study the effect of glucagon in or-der to obtain an estimate of maximal stomach relaxation and to test the performance of the gastric barostat system.

STU

DY

IIIRemifentanil and Gastric Tone 3

In the second study, gastric tone was measured during and after a remifentanil infusion and, after a washout time of 30 minutes, also during readministration of remifentanil in combination with glucagon. Remifentanil is an ultra-short-acting opioid (μ-opioid recep-tor agonist) with a predictable and constant effect. Measurement of gastric tone Gastric tone was measured by an electronic barostat (SVS®; Synetics AB, Stockholm, Sweden). The gastric barostat is an instrument with an electronic control system that maintains a constant preset pressure within an air-filled flaccid intragastric bag by mo-mentary changes in the volume of air in the bag. When the stomach contracts, the baro-stat aspirates air to maintain the constant pressure within the bag, and when the stomach relaxes, air is injected. The pressure in the bag was set at 2 mmHg above the basal intra-gastric pressure. The pressure change at which respiration is perceived on the pressure tracing- without an increase or decrease in the average volume- is the basal intragastric pressure. The bag, made of ultrathin polyethylene, has a capacity of 900 ml and is con-nected to the barostat by a double-lumen 16 Ch gastric tube. The barostat measurements were performed following the recommendations presented in a review article by an inter-national working team, and the barostat instrument fulfilled the criteria determined by this group (13). Procedure The subjects fasted for at least six hours before each study. An IV line was established in one arm and an IV-infusion of 5% buffered glucose 100ml/hour was given. Before the gastric intubation the subjects received a bolus dose of propofol (0.3 mg/kg IV). Previous studies in volunteers have shown that this dose of propofol, which was given at least 30 min before the study started, does not influence gastric tone (6). The intragastric bag was folded carefully around the gastric tube and positioned in the gastric fundus via oral in-tubation. Thereafter, the gastric bag was unfolded by being slowly inflated with 300 ml of air under controlled pressure (<20mmHg), and the correct position of the bag was verified by traction of the gastric tube. The gastric bag was completely deflated and thereafter inflated with air to a pressure 2 mmHg above the intragastric pressure. During the study the participants were lying down, positioned on their right side, and were asked to relax comfortably. Volume and pressure in the gastric bag were continuously recorded by the electronic barostat and sampled in the computer. The mean gastric bag volume during each 5-min interval was calculated. Glucagon study After 10 min of stable basal gastric tone recording the subjects were given an intravenous bolus dose of 1 mg glucagon. Mean gastric volumes before the injection, and during the time intervals 0 – 5 min, 5 –10 min, and 10-15 min after the injection, were calculated. For a schematic illustration of the study protocol, see Figure 1. Remifentanil Study After 10 minutes of stable basal gastric tone recordings, a continuous intravenous infu-sion of remifentanil was started. The initial dose was 0.1 μg•kg-1•min-1, after 15 minutes the dose was increased to 0.2 μg•kg-1•min-1, and after a further 15 minutes the dose was increased to 0.3 μg•kg-1•min-1. The infusion was discontinued after 45 minutes, follow-ing which there was a washout period of 30 minutes. Thereafter, remifentanil was read-ministered in a dose of 0.3 μg·kg-1·min-1, and 10 minutes later glucagon 1 mg was given intravenously and the remifentanil infusion was continued for a further 10 minutes. For a schematic illustration of the study protocol, see Figure 1.

4 Jakob Wallden

Figure 1 (opposite page) Schematic illustration of the study design

Monitoring and safety During both studies, the usual monitors were used. Heart rate, blood pressure, oxygen saturation, end-tidal carbon-dioxide (CO2), respiratory rate and sedation level were re-corded every fifth minute. At the same intervals, the subjects were asked if they were ex-periencing nausea or any other symptoms. The sedation level was recorded as follows: No sedation = 1, Light sedation =2, Moderate sedation = 3 and Deep sedation = 4. A vis-ual analog scale (VAS) ranging from 0-10 was used for nausea, where VAS 0 was no sub-jective symptoms and VAS 10 was the worst nausea the subjects could imagine. Blood glucose was followed during both studies. In the glucagon study, blood glucose was measured just before and 15 min after the administration of glucagon. In the remifentanil study, blood glucose was measured during the baseline period and just before and 15 minutes after the administration of glucagon. If the subject showed signs of excessive sedation, respiratory depression, severe nausea or vomiting, or showed signs of other severe symptoms related to the infusion of remifen-tanil, the dose was reduced or discontinued. Genetic analyses Due to the large inter-individual variations in the gastric tone response after remifentanil, we investigated if this variation could be explained by genetic variability, polymor-phisms, in the μ-opioid receptor gene. After reviewing the literature, we decided to ana-lyze polymorphisms with relative high frequencies and with reports of altered responses. Therefore, we focused on the μ-opioid receptor gene polymorphisms A118G and G691C (14). As ethnicity has impact on genetic expressions, we reviewed patient data and found that all of the subjects were Caucasians. DNA collection and purification. Venous blood (10 ml) was collected from the subjects in EDTA tubes and the samples were stored frozen at –70°C. Genomic DNA was puri-fied from peripheral leukocytes in 1 ml of EDTA blood on a MagNA Pure LC DNA ex-tractor, using the MagNA Pure LC Total Nucleic Acid Isolation Kit – Large Volume (Roche Diagnostics Corporation, Indianapolis, IN, USA). Genotyping. Genotyping was performed at CyberGene AB, Huddinge, Sweden. The A118G SNP in Exon 1 and the IVS2 G691C SNP in Intron 2 were genotyped using po-lymerase chain reaction amplification and sequencing. Oligonucleotide primers (forward: 5'-GCGCTTGGAACCCGAAAAGTC; reverse: 5'-CATTGAGCCTTGGGAGTT) and (forward: 5'-CTAGCTCATGTTGAGAGGTTC; reverse: 5'-CCAGTACCAGGTTGGATGAG) were used for amplifying gene fragments contain-ing Exon 1 and Intron 2, respectively. PCR conditions comprised an initial denaturing step at 95°C for 1 min followed by 30 cycles at 94°C for 1 min, annealing at 47.2-53.4°C (depending on primer) for 1 min and extension at 68°C for 3 min, and a final extension at 68°C for 3 min. The amplified fragments were sequenced using the same primers with the addition of Rev 1-2 5'-TTAAGCCGCTGAACCCTCCG and the BigDye Terminator v1.1Cycle Sequence Kit (Applied Biosystems, Foster City, CA, USA). PCR amplification and sequence reactions were done on ABI GeneAmp 2400 and 9700 (Applied Biosys-tems). Sequence analysis was first done on MegaBACE 1000 (Amersham Biosciences) and then confirmed with ABI 377XL (Applied Biosystems).

0 min

15 min

Glucagon 1 m

gP

ropofol0.3 m

g ·kg-1

-10 min

-40--80 min

Measurem

entof Gastric

Tone

0 min

75 min

15 min

45 min

30 min

85 min

95 min

Start R

emifentanil

Stop R

emifentanil

Glucagon 1 m

gS

top Rem

ifentanil

Start R

emifentanil

Propofol

0.3 mg ·kg-1

-10 min

-40--80 min

Measurem

entof Gastric

Tone

Glucagon Study

Rem

ifentanil Study

0.1μg·kg-1·m

in-1

0.2 μg·kg-1·m

in-1

0.3 μg·kg-1·m

in-1

0.3 μg·kg-1·m

in-1

Rem

ifentanil

Glucagon

Glucagon

STU

DY

III

6 Jakob Wallden

Statistics The results are presented as means with standard deviations and medians with ranges. Repeated measures ANOVA was used for evaluating overall differences between the study situations. If the statistical analysis showed differences, Fisher’s PLSD was used for comparisons between the situations. For the analysis, the remifentanil study was split into two parts. The Chi-square test was used for analysis of the genetic variations. The significance level was set at 5% in all tests.

Results Eight subjects completed the glucagon study and nine subjects completed the remifentanil study. One subject (no. 8) did not tolerate the gastric tube during the glucagon study and terminated participation in both study protocols. One subject refused to participate in the glucagon study after completing the remifentanil study. Glucagon study Glucagon induced a significant decrease in gastric tone (increase in volume) in all sub-jects (n=8) (Table 1 and Fig. 2). There was a temporary increase in heart rate after the injection of glucagon (Before: 70 (6.1) min-1; 0-5 min: 87 (8.7) min-1; p<0.001), other vital variables were normal and sta-ble. Blood glucose increased after glucagon (Before: 5.4 (1.4) mmol L-1; After: 11.1 (2.2) mmol L-1; p<0.001). 5 subjects experienced nausea (VAS 4 (2-8)) after receiving gluca-gon. Table 1 Gastric tone in healthy volunteers (n=8) studied with a barostat. Intragastric bag volumes (ml) after intravenous glucagon 1 mg. Mean (SD) ml ANOVA Median (range) ml Before Glucagon -10 to -5 minutes 138 (16) 168 (65-224) -5 to 0 minutes 156 (20) 158 (68-192) After Glucagon 1mg P <0.0001 0 to 5 minutes 362 (40)* 329 (230-454) 5 to 10 minutes 456 (47)* 410 (299-701) 10 to 15 minutes 448 (50)* 387 (330-714)

Change over time evaluated with repeated measures ANOVA. Pairwise comparison between the periods with Fisher’s PLSD. * = Significant difference (p<0.05) compared to “Before Glucagon -5 to 0 min” Figure 2 (opposite page) Gastric tone measured with a gastric barostat. The curves represent individual intragastric bag volumes during the studies. In the first part, glucagon 1 mg was given as an intravenous bolus in-jection. In the second part, remifentanil was given at the doses of 0.1, 0.2 and 0.3 μg•kg-1•min-1

STU

DY

III

Figure 2

0

200

400

600

800

1000

Baseline 0 - 5 min

5 - 10 min

Glucagon 0 - 5 min

5 - 10 min

10 - 15 min

Baseline 0 - 5 min

5 - 10 min

Remi 0.1 0 - 5 min

5 - 10 min

10 - 15 min

Remi 0.2 0 - 5 min

5 - 10 min

10 - 15 min

Remi 0.3 0 - 5 min

5 - 10 min

10 - 15 min

Washout 0 - 5 min

5 - 10 min

10 - 15 min

15 - 20 min

20 - 25 min

25 - 30 min

Remi 0.3 0 - 5 min

5 - 10 min

Glucagon 0 - 5 min

5 - 10 min

Washout 0 - 5 min

5 - 10 min

Time

Intragastric Bag Volume (ml)S

ubj 1

Subj 2

Subj 3

Subj 4

Subj 5

Subj 6

Subj 7

Subj 8

Subj 10

Glucagon S

tudyn=8

Rem

ifentanil Study

n=9

8 Jakob Wallden

Remifentanil study There were variable responses in gastric tone during the initial 45-minute infusion of re-mifentanil and the subsequent washout period of 30 minutes (Table 2 and Fig. 2). Four subjects (no. 1, 2, 3, 7) responded to remifentanil with a marked increase in gastric tone (decreased volume) that decreased during washout. Four subjects (no. 4, 6, 8, 10) re-sponded to remifentanil with a marked decrease in gastric tone (increased volume) and maintained a low gastric tone during the washout period. In one subject (no. 5) gastric tone was almost unaffected. The mean gastric tone was significantly lower during the washout period than before starting the infusion. Table 2 Gastric tone in healthy volunteers (n=9) studied with a gastric barostat. Intragastric bag volumes (ml) during infusion of remifentanil and in combination with intravenous glucagon 1 mg. Intragastric bag volumes (ml) Mean (SD) ml Median (range) Interval Period for volume Repeated Measure

during the study measurement ANOVA

Before Remifentanil 117 (44) 107 (62-192) -10 to 0 min -5 to 0 min During Remifentanil 0 to 45 min 0.1 μg•kg-1•min-1 156 (170) 114(1 - 473) 0 to 15 min 10 to 15 min 0.2 μg•kg-1•min-1 219 (240) 70 (1-542) 15 to 30 min 25 to 30 min P=0.0012 0.3 μg•kg-1•min-1 250 (291) 59 (0-722) 30 to 45 min 40 to 45 min Washout period 1 320 (276)* 304 (25–785) 45 to 75 min 55 to 60 min 394 (237)* 379 (90 – 820) 70 to 75 min Readmin Remifentanil 0.3 μg•kg-1•min-1 342 (314) 367 (0-856) 75 to 95 min 80 to 85min P=0.6 + Glucagon 1 mg 308 (316) 339 (0-879) at 85 min 90 to 95 min Washout period 2 347 (310) 242 (1-839) 95 to 105 min 100 to 105 min Change over time evaluated with repeated measures ANOVA. Pairwise comparison between the periods with Fisher’s PLSD. * = Significant difference (p<0.05) compared to “Before Remifentanil”.

During the initial remifentanil infusion there were significant decreases in heart rate (Be-fore: 67 (4.9 min-1; Minimum during Remi 0.1: 61 (4.6) min-1; p<0.001) and respiratory rate (Before: 12 (1.8) min-1; Minimum during Remi 0.2: 8 (2.3) min-1; p<0.001) and sig-nificant increases in end-tidal CO2 (Before: 5.4 (0.3) %; Maximum during Remi 0.3: 7.4 (1.3) %; p<0.001) and sedation level (Before: 1 (0); Maximum during Remi 0.3: 2 (0.7); p<0.05). The administration of glucagon at the end of the study induced a significant in-crease in systolic blood pressure (Before: 122 (9) mmHg; After: 137 (22) mmHg; p<0.001), heart rate (Before: 61 (4.1) min-1; After: 85 (22) min-1; p<0.001) and blood glucose (Before: 6.2 (1.1) mmol L-1; After: 10.3 (1.1) mmol L-1; p<0.001). One subject (no. 3) became too sedated during the highest dose of remifentanil and thin-fusion was discontinued. During readministration this subject received remifentanil 0.2 μg kg-1min-1.

STU

DY

IIIRemifentanil and Gastric Tone 9

Subjects experienced pruritus (n=7), nausea (n=3, VAS 1 (1-3)), headache (n=3) and diffi-culties swallowing (n=2) during the remifentanil infusion. After glucagon, the incidence of nausea increased (n=6; VAS 4.5 (2-7)). During the readministration of remifentanil there were increases in gastric tone among subjects with increased tone during the previous remifentanil infusion. The subject with unaffected tone during the previous infusion had an increase in gastric tone. The subjects who maintained a low gastric tone during washout continued to maintain a low gastric tone. Only one subject (no. 5) responded with a decrease in gastric tone after the injection of glucagon during the readministration of remifentanil. Genotype study Blood samples for genetic analysis were obtained from all 9 subjects in the study. We found no correlation between the gastric response to remifentanil and the polymorphisms A118G and G691C (Table 3). Table 3 Gastric tone response to remifentanil and correlation to genotype groups (n=9). 118 A>G genotype IVS2 + 691 G>C genotype Wild Type Hetero Variant Wild Type Hetero- Variant zygous zygous (AA) (AG) (GG) (GG) (GC) (CC) n=7 n=2 n=0 n=5 n=2 n=1 78% 22% 0 % 56% 22% 11% Increased tone (n=4) 4 3 1 Unchanged tone (n=1) 1 1 Decreased tone (n=4) 3 1 1 2 1 No association found between gastric tone response to remifentanil and presence of polymorphism A118G [Wild Type vs (Heterozygous OR Variant)] ; Chi-square test, P=0.097 No association found between gastric tone response to remifentanil and presence of polymorphism in G691C [Wild Type vs (Heterozygous OR Variant)]; Chi-square test , P =0.23

Discussion The major finding in this study is the marked effect of remifentanil on gastric tone. We found two distinctly different patterns of reactions, with about half of the subjects in-creasing in gastric tone (decreased volume) and about half of the subjects decreasing in gastric tone (increased volume). Due to this variability, we were not able to statistically prove a response during remifentanil. However, the gastric tone was significantly lower (higher volume) after the infusion of remifentanil compared to the baseline period. We believe these are important findings, as they show that opioid effects on human gastric motility are variable and complex. As expected, we found that glucagon decreased gastric tone in all subjects. In addition, we evaluated if the variable response in gastric tone to remifentanil could be explained by the single nucleotide polymorphisms A118G and G691C in the μ-opioid receptor gene, but we found no association.

10 Jakob Wallden

We have tried to explain the variability in gastric tone during the remifentanil infusion. We do not believe this is due to a methodological problem with the gastric barostat. Dur-ing the glucagon part of the study all subjects responded with a clear decrease in tone (increased volume). This validates that the gastric barostat was working properly, as an expected relaxant stimulus, glucagon, decreased tone in all subjects. Also, the same baro-stat equipment and setup have been used in previous studies by our group (6, 15) and we have not observed this kind of variation. There are several limitations in our study. There was no control group, and we cannot completely rule out that there was a time effect involved for the change in gastric tone. However, there was a stable baseline level in gastric tone before remifentanil and the dis-tinct changes in gastric tone after start of the infusion, as well as the changes after dis-continuation, are in agreement with the timing of the pharmacodynamic properties of remifentanil (16). This provides us with evidence that the effects are mediated by re-mifentanil. The number of subjects in this study was small. We expected a similar response to re-mifentanil in all subjects, but instead there were two kinds of divergent responses. As this is the first study to describe this dual effect, we consider our observations as important despite the lack of statistical power. Future studies may evaluate the quantitative relation between the responses, and the mechanisms behind them, in a larger group of subjects. Basic knowledge about the regulation of gastric tone is needed to explain the effects of opioids. Proximal gastric tone is an important part of gastric motility and is mainly con-trolled by the autonomous nerve system. Vagal cholinergic nerves mediate excitation (contraction) while vagal non-cholinergic non-adrenergic (NANC) nerves mediate inhibi-tion (relaxation) (17). Recent studies have identified nitrous oxide as one of the main transmitters in the NANC pathway. In humans, the NANC pathway is believed to be silent during fasting conditions and to be activated by the adaptive reflex (18). In addi-tion, there are sympathetic adrenerigic spinal nerves that inhibit motility mainly through cholinergic inhibition (19). Several animal studies have tried to identify targets for the opioid induced inhibition of gastric motility. It is widely believed that μ-opioid receptor (MOR) agonists inhibit the release of Ach in the stomach (20), and there is also evidence that MOR agonists reduce the relaxation induced by the NANC pathway (21). Opioids might also have a direct ex-citatory effect on gastric smooth muscles (22). Hence, depending on the current state of autonomous and enteric nerve systems and the main effect site, opioids have the potential to both relax and contract the stomach. Opioids also act in the central nervous system (CNS). There is evidence that MORs are present on and inhibit excitatory neurons projecting to gastrointestinal motor neurons in the dorsal motor complex (DMV) of the medulla (23). In this way activation of central MORs inhibits the excitatory vagal output, leading to inhibition of intestinal transit and induction of gastric relaxation in animal models. In humans, there is evidence that opioids inhibit gastric motility through a central mechanism (24). There are diverging results in the literature regarding the effects of opioids on gastric tone in humans. Penagini found that morphine increased gastric tone (5) while Hammas reported a decrease in gastric tone (6). Both studies used the same dose of intravenous morphine (0.1 mg/kg) and both used a gastric barostat. However, there were important differences between the studies. In the first study, baseline gastric tone was set to resem-ble a gastric load of a meal and in the second study, baseline was set to fasting condi-tions. The stomach wall was probably more distended (higher volumes in intragastric

STU

DY

IIIRemifentanil and Gastric Tone 11

bag) before morphine in Penagani’s study compared to Hammas’ study, resulting in an activated adaptive reflex. This leads to completely different baseline conditions. In Penagini’s subjects there were probably low cholinergic and high NANC vagal inputs to the stomach and the reverse baseline conditions were probably present in Hammas’ sub-jects. This might explain why a MOR antagonist contracted the stomach (through NANC inhibition) in one study and relaxed the stomach (through cholinergic inhibition) in the other study. An interesting finding in Hammas’ study was that the concurrent administration of pro-pofol altered the effect of morphine on gastric tone. Propofol per se had no effect on gas-tric tone, but after the subsequent administration of morphine, gastric tone increased (volume decreased), contrary to the response of morphine alone. We cannot explain the mechanism behind this modulation, but there is evidence for central interactions and modulations between GABAergic and opioid pathways (25). Other types of modulations of gastric tone have also been described; in animals with an intact vagus nerve, noradrenaline relaxed the proximal stomach while vagotomy reversed this response (17). Can we explain the variable responses seen in our study within this context? Remifen-tanil is a potent MOR agonist and the effect sites are probably both at the stomach level and in the CNS. We speculate that the “normal” opioid response during fasting condi-tions, as seen in Hammas’ study, is a decreased cholinergic activity resulting in a decrease in gastric tone. However, due to the high potency of remifentanil, direct smooth muscle effects might predominate in some subjects, resulting in an increase in tone. Like propo-fol, remifentanil might also have properties that modulate the opioid response. The fo-cus of these speculations is that opioid effects on gastric tone are variable and depend on factors like the state of the subject and the current status of the neural pathways and smooth muscles involved. This might be an explanation for the variable results in our study. As expected, glucagon decreased gastric tone in all subjects. The effect of glucagon is be-lieved to be an indirect inhibition of cholinergic activity (26). Among those subjects who already had low gastric tone a further decrease was not expected. With the exception of one subject, the administration of glucagon during the remifentanil infusion did not re-sult in a change in gastric tone. As the opioids might act on the smooth muscle level, glucagon might not have the ability to modulate the opioidergic effects on gastric motil-ity. We tested the hypothesis that pharmacogenetic differences in the μ-opioid receptor gene were responsible for the variable outcome. Investigators recently reported that the occur-rence of single nucleotide polymorphisms (SNP) in the μ-opioid receptor gene is associ-ated with altered responses to an opioid (8-10). We found no association between the presence of SNPs and the response in gastric tone after remifentanil. The results are in agreement with a recent published study by our group where we evaluated a variable ef-fect of fentanyl on electrogastrography (EGG) recordings (27). However, the results from the genetic analysis must be interpretated with care. The study was designed and pow-ered for the barostat variables and to investigate associations to genetic factors, a larger sample size is needed (28). The result of no association must be regarded as preliminary observations and has to be confirmed in properly designed studies. The observation might be an indication that the SNP A118G and G691C are not major factors for the observed variability, but we can neither confirm nor rule out that the presence of SNPs in the MOR alter opioid effects on motility. Do our results have any implications for the clinical situation? The main message is that gastric effects of opioids are variable, and it is not possible today to predict the response

12 Jakob Wallden

in the individual patient. An example of the variability is seen in the clinic, where opioids induce nausea and vomiting in some patients while other patients are totally unaffected. In future studies we need to evaluate whether the different kinds of gastric responses are of clinical significance. In summary, remifentanil induced distinct changes in gastric tone with both increases and decreases in tone. The effect of remifentanil on gastric tone is probably dependent on the current state of the systems involved. As a preliminary observation, the variations be-tween individuals could not be explained by SNPs in the MOR gene.

References 1. Azpiroz F. Control of gastric emptying by gastric tone. Dig Dis Sci 1994:39(12 Suppl):18S-19S. 2. Nimmo WS, Heading RC, Wilson J, Tothill P, Prescott LF. Inhibition of gastric emptying and drug

absorption by narcotic analgesics. Br J Clin Pharmacol 1975:2(6):509-13. 3. Lewis TD. Morphine and gastroduodenal motility. Dig Dis Sci 1999:44(11):2178-86. 4. Wallden J, Thorn SE, Wattwil M. The delay of gastric emptying induced by remifentanil is not influ-

enced by posture. Anesth Analg 2004:99(2):429-34. 5. Penagini R, Allocca M, Cantu P, Mangano M, Savojardo D, Carmagnola S, Bianchi PA. Relationship

between motor function of the proximal stomach and transient lower oesophageal sphincter relaxa-tion after morphine. Gut 2004:53(9):1227-31.

6. Hammas B, Thorn SE, Wattwil M. Propofol and gastric effects of morphine. Acta Anaesthesiol Scand 2001:45(8):1023-7.

7. Coulbault L, Beaussier M, Verstuyft C, Weickmans H, Dubert L, Tregouet D, Descot C, Parc Y, Lien-hart A, Jaillon P, Becquemont L. Environmental and genetic factors associated with morphine re-sponse in the postoperative period. Clin Pharmacol Ther 2006:79(4):316-324.

8. Klepstad P, Rakvag TT, Kaasa S, Holthe M, Dale O, Borchgrevink PC, Baar C, Vikan T, Krokan HE, Skorpen F. The 118 A > G polymorphism in the human micro-opioid receptor gene may increase morphine requirements in patients with pain caused by malignant disease. Acta Anaesthesiol Scand 2004:48(10):1232-9.

9. Chou WY, Yang LC, Lu HF, Ko JY, Wang CH, Lin SH, Lee TH, Concejero A, Hsu CJ. Association of mu-opioid receptor gene polymorphism (A118G) with variations in morphine consumption for anal-gesia after total knee arthroplasty. Acta Anaesthesiol Scand 2006:50(7):787-92.

10. Klepstad P, Dale O, Skorpen F, Borchgrevink PC, Kaasa S. Genetic variability and clinical efficacy of morphine. Acta Anaesthesiol Scand 2005:49(7):902-8.

11. Notivol R, Carrio I, Cano L, Estorch M, Vilardell F. Gastric emptying of solid and liquid meals in healthy young subjects. Scand J Gastroenterol 1984:19(8):1107-13.

12. Notivol R, Coffin B, Azpiroz F, Mearin F, Serra J, Malagelada JR. Gastric tone determines the sensi-tivity of the stomach to distention. Gastroenterology 1995:108(2):330-6.

13. Whitehead WE, Delvaux M. Standardization of barostat procedures for testing smooth muscle tone and sensory thresholds in the gastrointestinal tract. The Working Team of Glaxo-Wellcome Research, UK. Dig Dis Sci 1997:42(2):223-41.

14. Ikeda K, Ide S, Han W, Hayashida M, Uhl GR, Sora I. How individual sensitivity to opiates can be predicted by gene analyses. Trends Pharmacol Sci 2005:26(6):311-7.

15. Levein NG, Thorn SE, Lindberg G, Wattwill M. Dopamine reduces gastric tone in a dose-related manner. Acta Anaesthesiol Scand 1999:43(7):722-5.

16. Burkle H, Dunbar S, Van Aken H. Remifentanil: a novel, short-acting, mu-opioid. Anesth Analg 1996:83(3):646-51.

17. Jahnberg T. Gastric adaptive relaxation. Effects of vagal activation and vagotomy. An experimental study in dogs and in man. Scand J Gastroenterol Suppl 1977:46:1-32.

18. Tack J, Demedts I, Meulemans A, Schuurkes J, Janssens J. Role of nitric oxide in the gastric accom-modation reflex and in meal induced satiety in humans. Gut 2002:51(2):219-24.

19. Abrahamsson H, Glise H. Sympathetic nervous control of gastric motility and interaction with vagal activity. Scand J Gastroenterol Suppl 1984:89:83-7.

20. Yokotani K, Osumi Y. Involvement of mu-receptor in endogenous opioid peptide-mediated inhibition of acetylcholine release from the rat stomach. Jpn J Pharmacol 1998:78(1):93-5.

STU

DY

IIIRemifentanil and Gastric Tone 13

21. Storr M, Gaffal E, Schusdziarra V, Allescher HD. Endomorphins 1 and 2 reduce relaxant non-adrenergic, non-cholinergic neurotransmission in rat gastric fundus. Life Sci 2002:71(4):383-9.

22. Grider JR, Makhlouf GM. Identification of opioid receptors on gastric muscle cells by selective recep-tor protection. Am J Physiol 1991:260(1 Pt 1):G103-7.

23. Browning KN, Kalyuzhny AE, Travagli RA. Opioid peptides inhibit excitatory but not inhibitory syn-aptic transmission in the rat dorsal motor nucleus of the vagus. J Neurosci 2002:22(8):2998-3004.

24. Thorn SE, Wattwil M, Lindberg G, Sawe J. Systemic and central effects of morphine on gastroduode-nal motility. Acta Anaesthesiol Scand 1996:40(2):177-86.

25. Browning KN, Zheng Z, Gettys TW, Travagli RA. Vagal afferent control of opioidergic effects in rat brainstem circuits. J Physiol 2006:575(Pt 3):761-76.

26. Shimatani T. Involvement of cholinergic motor neurons in pharmacological regulation of gastrointes-tinal motility by glucagon in conscious dogs. J Smooth Muscle Res 1997:33(4-5):145-62.

27. Wallden J, Lindberg G, Sandin M, Thorn S-E, Wattwil M. Effects of fentanyl on gastric myoelectrical activity - a possible association to polymorphisms of the μ-opioid receptor gene? Acta Anaesthesiol Scand 2008:In Press.

28. Belfer I, Wu T, Kingman A, Krishnaraju RK, Goldman D, Max MB. Candidate gene studies of human pain mechanisms: methods for optimizing choice of polymorphisms and sample size. Anesthesiology 2004:100(6):1562-72.

STUDY IV

Klunserna Study 05-10-25, 10.15161

STU

DY

IV

Effects of fentanyl on gastric myoelectrical activity – a possible association to polymorphisms of the

μ-opioid receptor gene? Jakob Walldén, MD * †; Greger Lindberg, MD, PhD ‡ ;Mathias Sandin, MD §;

Sven-Egron Thörn, MD,PhD §†;Magnus Wattwil, MD, PhD §†

* Department of Anesthesia, Sundsvall Hospital, Sundsvall; † Department of Clinical Medicine, Örebro University, Örebro; ‡ Karolinska Institutet, Department of Medicine, Karolinska University Hospital Huddinge; Stockholm; § Department of Anesthesia and

Intensive Care, Örebro University Hospital, Örebro; SWEDEN Background: Opioids have inhibitory effects on gastric motility, but the mechanism is far from clear. Electrical slow waves in the stomach determine the frequency and the peri-staltic nature of gastric contractions. The primary aim of this study was to investigate the effects of the opioid fentanyl on gastric myoelectric activity. As there were large varia-tions between the subjects we investigated if the variation was correlated to single nu-cleotide polymorphisms (SNP) of the μ-opioidreceptor gene. Methods: We used cutaneous multichannel electrogastrography (EGG) to study myoelec-trical activity in 20 patients scheduled for elective surgery. Fasting EGG was recorded for 30 minutes, followed by intravenous administration of fentanyl 1μg•kg-1 and subsequent EGG recording for 30 minutes. Spectral analysis of the two recording periods was per-formed and variables assessed were dominant frequency (DF) of the EGG and its power (DP). Genetic analysis of the SNP A118G and G691C of the μ-opioidreceptor gene were performed with PCR-technique. Results: There was a significant reduction in DF and DP after intravenous fentanyl. However, there was a large variation between the patients. In eight subjects EGG was unaffected, five subjects had a slower DF (bradygastria) and in six subjects the slow waves disappeared. We found no correlation between the EGG outcome and presence of A118G or G691C in the μ-opioidreceptor gene. Conclusions: Fentanyl inhibited gastric myoelectrical activity in about half of the sub-jects. The variation could not be explained by SNP in the μ-opioid receptor gene. Due to small sample size, results must be regarded as preliminary observations. Keywords: Gastrointestinal motility; Gastric myoelectrical activity; Electrogastrography; Analgesics, Opioid; Polymorphism, Single Nucleotide; Receptors, Opioid, mu/*genetics; Genotype This study was supported by grants from FoU-centrum, Sundsvall and Emil Andersson’s Fund for Medical Research, Sundsvall. Accepted for publication in Acta Anaesthesiologica Scandinavica on January 2, 2008.

2 Jakob Wallden

Introduction Electrogastrography (EGG) is the cutaneous recording of gastric myoelectrical activity and the activity is closely associated to gastric motility (1). Gastric smooth muscles dis-play a rythmic electrical activity, slow-waves, with a frequency of approximately 3 cycles per minute. These slow-waves originate from a gastric pacemaker region in the corpus and propagate towards the pylorus. With influence of the enteric nervous system and other regulatory mechanisms, the slow-waves trigger the onset of spike potentials, which in turn initiate coordinated contractions of the gastric smooth muscles (2). Gastric motil-ity and emptying depend on these slow waves. Opiate drugs are well known to impair gastric motility. The mechanistic understanding how this impairment is mediated is far from clear (3) although opioid receptors are dis-tributed all over the gastrointestinal tract. To our knowledge there is only one study in the literature where cutaneous EGG was used for studying gastric effects of opiates (4) and in that study morphine induced tachygastria. The primary aim of our study was to investigate how the short acting opiate fentanyl affects gastric myoelectrical activity as recorded with cutaneous EGG. There are substantial inter-individual differences in the general response to opioids (5) and recent studies have suggested that polymorphism of the μ-opioid receptor (MOR) gene with an altering of the receptor-function may be a cause of the variation (6). Since we found large variations in the EGG response to an opioid, we also investigated if this variation was correlated to the presence of two different polymorphisms in the μ-opioid receptor gene. Methods After approval of the study protocol by the ethics committee of the Örebro County Council, 20 patients undergoing surgery on an ambulatory basis were recruited to the study. The subjects gave their informed consent to participate after receiving verbal and written information. Patient characteristics are presented in Table 1. The study was done before the induction of anesthesia in a pre-anesthetic area. Patients had fasted for at least 6 hours from solid foods and 2 hours from clear fluids. No pre-medication was given. While the patient was lying in a comfortable bed rest position, an intravenous line was inserted and the EGG recordings were initiated. After achieving a stable EGG signal, a 30-minute baseline EGG recording was collected. Without discon-tinuation of the EGG recording, 1 μgram•kg-1 of fentanyl was given as an intravenous bolus through the intravenous line and the EGG recording continued for another 30 minutes. Multichannel Electrogastrography Six EGG electrodes were placed on the abdomen after skin preparation. The electrodes consisted of four active electrodes, one reference electrode and one ground electrode as illustrated in Figure 1. A motion sensor was also attached to the abdomen. We used Medtronic Polygram NET EGG system (Medtronic A/S, Denmark) for the simultaneous recording of four EGG signals. Our EGG system was configured to accept an electrode impedance of less than 11 kΩ after skin preparation. The EGG signal was sampled at ~105 Hz, and this was downsampled to 1 Hz as part of the acquisition process (7). EGG analysis All EGG tracings were first examined manually by two of the authors (JW, GL). Prior the analysis motion artifact in the EGG signal, indicated by the motion sensor, were iden-

STU

DY

IVEffects of Fentanyl on EGG 3

tified and removed manually. For each patient, the EGG channel with the most typical slow-wave pattern during baseline recording (before fentanyl) was selected for further analysis. An overall spectrum analysis was performed on each of the two main 30 minute seg-ments (before and after fentanyl respectively) of the selected channel using the entire time-domain EGG signal (7). Sequential sets of measurement data for 256s with an over-lap of 196s were analyzed using fast Fourier transforms and a Hamming window for the calculation of running power spectra. When the entire signal was processed, the power spectra for each segment were combined to arrive at the overall dominant frequency (DF) and power of the dominant frequency (DP).

Table 1 Patient characteristics Age (yr) 45 (28-67) Height (cm) 169 (155-180) Weight (kg) 77 (54-124) Body Mass Index 27 (18-39) Females 16 Males 4 Smokers 3 ASA I 16 ASA II 4 Values are given as means with ranges or numbers. Figure 1 Electrogastrography electrode placement: -Electrode 3 was placed halfway between the xy-phoid process and the umbilicus. -Electrode 4 was placed 4 cm to the right of elec-trode 3. -Electrode 2 and 1 were placed 45 degree to the upper left of electrode 3, with an interval of 4 to 6 cm. -The ground electrode was placed on the left costal margin horizontal to electrode 4. -The reference electrode (electrode 0) was placed at the cross point of the horizontal line containing electrode 1 and the vertical line containing elec-trode 3.

4 Jakob Wallden

The EGG segments and the spectral analysis after fentanyl were further classified either as 1) Unaffected EGG (no change in DF after fentanyl), 2) Bradygastric EGG (decrease in DF >=0.3 after fentanyl) or 3) Flatline-EGG (total visual disappearance of a previously clear sinusoidal 3 cpm EGG-curve after fentanyl) (see example in figure 3) without any quantifiable DF. When DF was not quantifiable, DF was set to 0. Data from the baseline EGG were compared to data from a previous multicenter study in normal subjects (7) to test if our study group was similar to a normal population. Predicted fentanyl concentrations in blood was calculated using Shibutani’s modification of Shafer’s formula (8, 9). Genetic analyses Due to the large interindividual variation in the EGG pattern after fentanyl, we decided to investigate if this variation could be explained by genetic variability, polymorphisms, in the μ-opioid receptor gene. We decided to analyze polymorphisms with a relative high frequency and after reviewing the literature, we focused on the μ-opioid receptor gene polymorphisms A118G and G691C (10). As ethnicity has impact on genetic expressions, we reviewed patient data and found that all of the subjects were Caucasians. DNA collection and purification. Venous blood (10 ml) was collected from the subjects in EDTA tubes and the samples were stored frozen at –70°C. Genomic DNA was puri-fied from peripheral leukocytes in 1 ml of EDTA blood on a MagNA Pure LC DNA ex-tractor, using the MagNA Pure LC Total Nucleic Acid Isolation Kit – Large Volume (Roche Diagnostics Corporation, Indianapolis, USA). Genotyping. Genotyping was performed at CyberGene AB, Huddinge, Sweden. The A118G SNP in Exon 1 and the IVS2 G691C SNP in Intron 2 were genotyped using po-lymerase chain reaction amplification and sequencing. Oligonucleotide primers (forward: 5'-GCGCTTGGAACCCGAAAAGTC; reverse: 5'-CATTGAGCCTTGGGAGTT) and (forward: 5'-CTAGCTCATGTTGAGAGGTTC; reverse: 5'-CCAGTACCAGGTTGGATGAG) were used for amplifying gene fragments contain-ing Exon 1 and Intron 2, respectively. PCR conditions comprised an initial denaturing step at 95°C for 1 min followed by 30 cycles at 94°C for 1 min, annealing at 47.2-53.4°C (depending on primer) for 1 min and extension 68°C for 3 min, and a final extension at 68°C for 3 min. The amplified fragments were sequenced using the same primers with the addition of Rev 1-2 5'-TTAAGCCGCTGAACCCTCCG and the BigDye Terminator v1.1 Cycle Sequence Kit (Applied Biosystems, Foster City, CA, USA). PCR amplification and sequence reactions were done on ABI GeneAmp 2400 and 9700 (Applied Biosys-tems). Sequence analysis was first done on MegaBACE 1000 (Amersham Biosciences, CA, USA) and then confirmed with ABI 377XL (Applied Biosystems). Postoperative data Charts and notes from the recovery unit were reviewed and we collected data regarding analgesic and antiemetic requirements. The decision to include these data was done after the initial study was terminated. Statistics In order to detect a mean intraindividual difference of 1 cpm in the dominant frequency (DF) with 1 cpm as the expected standard deviation of the difference, a sample size of 12 was calculated (alpha=0.05, beta = 0.2). To further increase power and compensate for possible exclusions sample size was set to 20. The results are presented as medians with interquartile ranges. Wilcoxon's signed rank test and the 95% confidence interval of the difference between the medians were used for analysis of the primary EGG outcome

STU

DY

IVEffects of Fentanyl on EGG 5

variables. The unpaired t-test was used for the comparisons of baseline EGG data to the historical controls and for the comparisons of predicted fentanyl concentrations and body composition between the outcome groups. Analysis of change over time for the vi-tal variables (blood pressure, heart rate, oxygen saturation) was performed using a gen-eral linear model of variance for repeated measures. The Chi-squared test was used for the analysis of associations between the EGG outcome and the genetic variations. The significance level was set to 5% in all statistical tests.

Results All patients (n=20) tolerated the administration of fentanyl well and there were no ad-verse events. One patient was excluded from the EGG analysis due to major artifacts in the EGG recording (both before and after fentanyl). We interpreted the artifacts as elec-tromagnetical interference in the ambience. Blood pressure, heart rate and oxygen saturation were normal during the whole study period with small statistically significant decreases in systolic and diastolic blood pressure after administration of fentanyl (data not shown). Compared to historical controls (7), there were no differences in the baseline EGG vari-ables, see Table 2. After the administration of intravenous fentanyl, there was a significant reduction in both dominant frequency (DF) and dominant power (DP) of the EGG spectra, see Table 3. Individual changes in DF and DP are presented in Figure 2. There was large variation between patients in the response to intravenous fentanyl. EGG recordings were unaffected in 8 patients, 5 patients developed a slower DF (bradygastria) and in 6 patients the slow-wave tracings disappeared totally (flatline-EGG). For an illus-tration of the effect, see Figure 3. Among patients with a flatline-EGG (n=6), the median (range) time from the administra-tion of intravenous fentanyl to the observed disappearance of the slow waves was 5 (1-9) minutes. In 5 of these patients, there was a reapperance of the 3 cpm slow-wave EGG pattern 30 (29-35) minutes after the administration of fentanyl. We found no difference between the outcome groups in predicted fentanyl concentrations derived from a pharmacokinetic model (unaffected vs affected group (ng·mL-1): 10 min: 0.45 (0.061) vs 0.43 (0.065) (p=0.6); 20 min: 0.34 (0.046) vs 0.32 (0.049) (p=0.6); 30 min: 0.26 (0.035) vs 0.25 (0.037) (p=0.6)). Further, there were no difference between the groups in body weight (unaffected vs affected group (kg): 73.8 (13.5) vs 79 (19.3) (p=0.29)) or BMI (unaffected vs affected group (kg·m-2): 25.3 (4.8) vs 27.9 (5.1) (p=0.50)). Blood samples for genetic analysis were obtained from 18 subjects in the study. We found no correlation between the gastric response to fentanyl and the polymorphisms A118G or G691C (Table 4). We found an association between requirement for postoperative antiemetic and the gas-tric response to fentanyl (Table 5).

6 Jakob Wallden

25

30

35

40

45

50

55

Baseline After Fentanyl 1μg/kg

Dom

inan

t Pow

er (d

B)

*

0

0,5

1

1,5

2

2,5

3

3,5

Baseline After Fentanyl 1μg/kg

Dom

inan

t Fre

quen

cy (c

pm)

*

Table 2 EGG parameters from the baseline recordings compared to a multicenter study in nor-mal subjects (7).

Baseline recording

(n=19)

Historical control (n=60).

Difference between means with 95% C.I. P-value

Dominant Frequency (cpm) 2.92 ± 0.17 2.89 ± 0.26 0.03 (-0.1 to 0.16) p=0.64

Dominant Power (dB) 42.1 ± 3.8 42.4 ± 6.3 -0.30 (-3.5 to 2.7) p=0.85

Values are means (SD). Paired students t-test.

Table 3 Changes in EGG variables before (-30 to 0 min) and after (0 to 30 min) the administra-tion of 1μg•kg-1 intravenous fentanyl.

Baseline

recording After fentanyl 1μg•kg-1 I.V.

Difference between the medians with 95% C.I. P-value

Dominant Frequency (cpm) 2.9 ( 2.8 - 3.0 ) 2.5 ( 0 – 2.9 ) 0.4 (0-2.9) p=0.002

Dominant Power (dB) 41 ( 39 – 45 ) 38 ( 35 – 40 ) 3 (0.6-3.5) p=0.002

Values are medians with interquartile ranges. Wilcoxons signed rank test. C.I. = Confidence Inter-val. Figure 2 A: Individual values for the dominant frequency (DF) in the electrogastrographic (EGG) spectra before and after intravenous fentanyl 1μg•kg-1. If there was no dominant fre-quency, DF was set to zero. B: Individual values for the dominant power (DP) in the EGG spectra before and after intravenous fentanyl 1μg·kg-1. (* = Wilcoxons signed rank test, p<0.05) A B

STU

DY

IVEffects of Fentanyl on EGG 7

Figure 3 Electrogastrographic (EGG) tracing in a patient where the gastric slow waves disap-peared after intravenous fentanyl 1μg•kg-1 with a close-up of the period were fentanyl was administered.

-80

-60

-40

-20

0

20

40

60

80

25:00 30:00 35:00 40:00Time (min)

μV

Fentanyl 1μg/kg I.V.

Slow-waves 3 cpm Disappearance of Slow-waves

Table 4 EGG classification and genotype groups (n=18). 118 A>G genotype IVS2 + 691 G>C genotype

Wild Type (AA) n=15 83%

Hetero-zygous (AG) n=2 11%

Variant (GG) n=1 6%

Wild Type(GG) (n=0) 0%

Hetero-zygous (GC)

(n=14) 78%

Variant (CC) (n=4) 22%

Unaffected EGG (n=6)

5 1 6

Bradygastria (n=5) 4 1 2 3

Flatline (n=6) 5 1 5 1

Excluded from the EGG-analysis (n=1)

1 1

No association found between EGG-classification and prescence of polymorphism A118G [Wild Type vs (Heterozygous OR Variant)] ; Chi-square test, P=0.99. No association found between EGG-classification and prescence of polymorphism in G691C [Wild Type vs (Heterozygous OR Variant)]; Chi-square test was not possible to perform as there were no cases in “Wild type”. Two subjects, both classified as “Unaffected EGG”, did not participate in the genotype analysis.

8 Jakob Wallden

Table 5 PONV, Postopertive antiemetic and the correlation to the EGG outcome after fentanyl.

EGG after fentanyl

Unaffected (n=8)

Bradygastric or flatline (n=11)

PONV at the recovery unit No (n=9) 6 3 Yes (n=10) 2 8

Antiemetic administration at recovery unit No (n=10) 7 3 Yes (n=9) 1 8

Fishers exact test (2-sided): PONV vs EGG-classification; P=0.070; Antiemetics vs EGG-classification; P=0.020* PONV= Postopertive Nausea and Vomiting; EGG = Electrogastrography.

Discussion In this study, we found that intravenous administration of the opioid fentanyl 1μg·kg-1

inhibited gastric myoelectric activity with a decrease in both the dominant frequency and the dominant power of the electrogastrographic spectra. The responses were highly indi-vidual with responders and non-responders. We tested if this variability could be ex-plained by the single nucleotide polymorphisms A118G and G691C of the μ-opioid re-ceptor gene, but we found no association. We found that the EGG response to fentanyl predicted the need for postoperative antiemetic treatment. The results from the genetic analysis must be interpretated with care. The study was de-signed and powered for the EGG variables and to investigate associations to genetic fac-tors, a larger sample size is needed (11). The result of no association must be regarded as preliminary observations and has to be confirmed in properly designed studies. The ob-servation might be an indication that the SNP A118G and G691C are not major factors for the observed variability, but we can neither confirm nor rule out that the presence of SNPs in the MOR alter opioid effects on motility. In this study, we hypothesized that opioids would impair gastric electrical activity. Coor-dinated peristaltic contractions of gastric smooth muscles, initiated by electrical depolari-zation, are the bases for gastric emptying of solids. The stomach displays a rhythmic de-polarization that is characterized by slow waves with a frequency of about 3 cycles per minute (cpm). The electrical activity originates in the corpus region of the stomach and propagates towards the pylorus. Specialized pacemaker cells, the interstitial cells of Cajal (ICC), are localized as a network around the myenteric plexus of the stomach and are responsible for the generation and conduction of the slow waves (1). We used cutaneous electrogastrography, which revealed 3 cpm slow wave activity 30 minutes before the in-tervention with fentanyl in all subjects. Data from the baseline period did not differ from a recent multicenter electrogastrography study in normal subjects (7) and we consider the baseline period in our subjects as normal electrical activity. The electrical activity was disrupted after the administration of fentanyl and we observed both bradygastria and disappearance of the slow wave activity. However, in about half of the subjects, EGG was unaffected.

STU

DY

IVEffects of Fentanyl on EGG 9

Opioids are known inhibitors of gastrointestinal motility and there are numerous studies in the literature regarding the effect of opioids on gastroduodenal function. Gastric emp-tying is delayed (12-14), antroduodenal motility increased (15), pyloric spasm induced (16) and there are reports of both relaxation (17) and contraction (18) of the gastric fun-dus. The understanding how these effects are mediated is far from clear (3) and the opioids might act on opioid receptors at different levels, both within the stomach (19, 20) and in the central nervous system (4, 21, 22). There are only a few reports in the literature about the effects of opioids on gastric elec-trical activity. Invasive recordings of gastric myoelectric activity have shown that mor-phine transiently distort the slow-wave activity and initiate migrating myoelectric com-plexes (23, 24). Cutaneous recordings with EGG showed that morphine induced tachy-gastria (4). The shift in the basal EGG frequency towards bradygastria that we observed in some of the subjects indicates that opioids inhibit the ICC-network. Bradygastria is believed to be a decrease in the frequency of the normal pacemaker cells while other dys-rhythmias like tachygastria have ectopic origins in the stomach (25). There was no randomization or blinding in this study. We compared the EGG before and after an intervention and the subjects acted as their own control. We were aware that there are risks for bias and other errors with this study design. Other factors like time-effects, emotions or other stimuli may have contributed to the outcome. However, the onset-time of the effect among the responders was consistent with the pharmacological properties of fentanyl and we suggest that fentanyl is responsible for the EGG-changes. To further investigate and verify our results, a double-blinded randomized control trial is needed. We tried to explain the variability seen with responders and non-responders. One hy-pothesis may be a difference between the individuals in the plasma concentration of fen-tanyl. Unfortunately we did not collect blood samples during the EGG study. By using a pharmacokinetic model (8, 9), we calculated the predicted plasma concentrations of fen-tanyl for each subject. We were not able to find any differences in the predicted concen-trations between the outcome groups. However, there is a notable wide variability in the model that may conceal relevant differences. Further, as body composition affects the pharmacokinetic profiles of a drug, we tested for differences in body weight and body mass index between the groups, but found no differences. Also, we cannot rule out that differences between the subjects in pharmacokinetic factors, i.e. distribution volume, me-tabolism or clearance, alter the effect-site concentration of fentanyl and thus the effect on gastric motility. We tested the hypothesis if pharmacogenetic differences in the μ-opioid receptor were responsible for the variable outcome. Recently investigators have reported that occur-rence of single nucleotide polymorphisms (SNP) in the μ-opioid receptor gene are associ-ated with an altered responses to an opioid (26, 27). There are data supporting that ge-netic differences are able to alter gastrointestinal response to opioids. The variable anal-gesic effect of codein is related to genetic variations, leading to different expression of the enzyme (CYP2D6) that metabolizes codein to morphine. Among extensive metabolizers, oro-cecal transit time is prolonged compared to poor metabolizers and correlates to higher morphine concentrations in plasma (28). To our knowledge there are no studies on the relation of SNP in the μ-opioid receptor to the outcome of opioids on gastrointes-tinal motility. After reviewing the literature, we decided to analyze two SNPs in the μ-opioid receptor gene with a high reported frequency (10). The SNP A118G is also the polymorphism that is believed to have clinical significance.

10 Jakob Wallden

We found no association between the presence of SNPs and EGG changes after the inter-vention with fentanyl. Two subjects, both classified as “unaffected EGG”, refused to par-ticipate in the genotyping study. By using simulated genotype data with all hypothetical outcomes for the two subjects, we found that there were no changes in the results if they have been participating. The frequencies of A118G in our material were similar to the reported frequencies in the literature. All investigated subjects were either heterozygote or homozygote to G691C and there were no normal “wild types” of G691C. This SNP is reported in a high frequency, but the distribution in our study is not consistent with the expected distributions from the Hardy-Weinberg equilibrium. Our study group may not represent a normal population, as the majority of the subjects are woman and almost all of them had a gallbladder disease and this may introduce a selection bias. However, with the small sample size it is difficult to draw any conclusions regarding the distribution. Our results indicate that pharmacogenetic differences in the opioid receptor gene may not be a major factor for the variable gastric outcome by an opioid. However, due to the small sample size, we want to emphasis that our results are preliminary observations and the interpretation of the results have to be cautious. Retrospectively we reviewed the anesthetic and postoperative records regarding intraop-erative and postoperative opioid requirements, pain assessments, postoperative nausea or vomiting (PONV) and antiemetic treatments. We correlated the data to the EGG and pharmacogenetic results. We found a higher requirement of antiemetic treatments among the subjects classified as responders to fentanyl. There was also a tendency towards a higher incidence of PONV in this group. As we investigated many factors, these results may have resulted by chance, so other explanations are possible. Those subjects who re-sponded with gastric effects of fentanyl may somehow have a higher sensitivity to opioids, which may also results in higher emetic response to opioids. There is also a pos-sibility that the changes in gastric motility per se induce emesis. As there are great limita-tions in the retrospective review, we want to emphasis that our results are only an indica-tion for a possible association between PONV and opioid induced changes in gastric mo-tility. The pattern of responders and non-responders on the EGG after fentanyl raises the ques-tion if opioid-side effects follow a present or non-present pattern. In daily clinical rou-tine, we also observe that some patients experience PONV on opioid treatment, while others are totally unaffected. If there is such “switch”, the “trigger” must be identified. The issue is probably complex and might i.e. include pharmacokinetic, pharmacody-namic and pharmacogenetic factors. In summary, the opioid fentanyl induced changes in the gastric slow waves with brady-gastria and disappearance of the slow waves. Pharmacogenetic differences in the μ-opioid receptor gene could not explain the variability with responders and non-responders, but as our sample size was small the findings must be regarded as preliminary observations. Further studies are needed to survey the mechanism of gastric effects of opioids and the source of the great variability.

References 1. Sanders KM, Ördög T, Koh SD, Ward SM. Properties of Electrical Rhytmicity in the Stomach. In: Koch

KL, Stern RM, editors. Handbook of Electrogastrography. New York: Oxford University Press; 2004. p. 13-36.

2. Chang FY. Electrogastrography: Basic knowledge, recording, processing and its clinical applications. J Gastroenterol Hepatol 2005:20(4):502-16.

STU

DY

IVEffects of Fentanyl on EGG 11

3. Hicks GA, DeHaven-Hudkins DL, Camilleri M. Opiates in the control of gastrointestinal tract function: current knowledge and new avenues for research. Neurogastroenterol Motil 2004:16 Suppl 2:67-70.

4. Thorn SE, Wattwil M, Lindberg G, Sawe J. Systemic and central effects of morphine on gastroduodenal motility. Acta Anaesthesiol Scand 1996:40(2):177-86.

5. Coulbault L, Beaussier M, Verstuyft C, Weickmans H, Dubert L, Tregouet D, Descot C, Parc Y, Lien-hart A, Jaillon P, Becquemont L. Environmental and genetic factors associated with morphine response in the postoperative period. Clin Pharmacol Ther 2006:79(4):316-324.

6. Klepstad P, Rakvag TT, Kaasa S, Holthe M, Dale O, Borchgrevink PC, Baar C, Vikan T, Krokan HE, Skorpen F. The 118 A > G polymorphism in the human micro-opioid receptor gene may increase mor-phine requirements in patients with pain caused by malignant disease. Acta Anaesthesiol Scand 2004:48(10):1232-9.

7. Simonian HP, Panganamamula K, Parkman HP, Xu X, Chen JZ, Lindberg G, Xu H, Shao C, Ke MY, Lykke M, Hansen P, Barner B, Buhl H. Multichannel electrogastrography (EGG) in normal subjects: a multicenter study. Dig Dis Sci 2004:49(4):594-601.

8. Shibutani K, Inchiosa MA, Jr., Sawada K, Bairamian M. Accuracy of pharmacokinetic models for pre-dicting plasma fentanyl concentrations in lean and obese surgical patients: derivation of dosing weight ("pharmacokinetic mass"). Anesthesiology 2004:101(3):603-13.

9. Shafer SL, Varvel JR, Aziz N, Scott JC. Pharmacokinetics of fentanyl administered by computer-controlled infusion pump. Anesthesiology 1990:73(6):1091-102.

10. Ikeda K, Ide S, Han W, Hayashida M, Uhl GR, Sora I. How individual sensitivity to opiates can be pre-dicted by gene analyses. Trends Pharmacol Sci 2005:26(6):311-7.

11. Belfer I, Wu T, Kingman A, Krishnaraju RK, Goldman D, Max MB. Candidate gene studies of human pain mechanisms: methods for optimizing choice of polymorphisms and sample size. Anesthesiology 2004:100(6):1562-72.

12. Wallden J, Thorn SE, Wattwil M. The delay of gastric emptying induced by remifentanil is not influ-enced by posture. Anesth Analg 2004:99(2):429-34.

13. Nimmo WS, Heading RC, Wilson J, Tothill P, Prescott LF. Inhibition of gastric emptying and drug ab-sorption by narcotic analgesics. Br J Clin Pharmacol 1975:2(6):509-13.

14. Yuan CS, Foss JF, O'Connor M, Roizen MF, Moss J. Effects of low-dose morphine on gastric emptying in healthy volunteers. J Clin Pharmacol 1998:38(11):1017-20.

15. Lewis TD. Morphine and gastroduodenal motility. Dig Dis Sci 1999:44(11):2178-86. 16. Edin R, Lundberg J, Terenius L, Dahlstrom A, Hokfelt T, Kewenter J, Ahlman H. Evidence for vagal

enkephalinergic neural control of the feline pylorus and stomach. Gastroenterology 1980:78(3):492-7. 17. Hammas B, Thorn SE, Wattwil M. Propofol and gastric effects of morphine. Acta Anaesthesiol Scand

2001:45(8):1023-7. 18. Penagini R, Allocca M, Cantu P, Mangano M, Savojardo D, Carmagnola S, Bianchi PA. Relationship

between motor function of the proximal stomach and transient lower oesophageal sphincter relaxation after morphine. Gut 2004:53(9):1227-31.

19. Bitar KN, Makhlouf GM. Specific opiate receptors on isolated mammalian gastric smooth muscle cells. Nature 1982:297(5861):72-4.

20. De Schepper HU, Cremonini F, Park MI, Camilleri M. Opioids and the gut: pharmacology and current clinical experience. Neurogastroenterol Motil 2004:16(4):383-94.

21. Aicher SA, Mitchell JL, Mendelowitz D. Distribution of mu-opioid receptors in rat visceral premotor neurons. Neuroscience 2002:115(3):851-60.

22. Browning KN, Kalyuzhny AE, Travagli RA. Opioid peptides inhibit excitatory but not inhibitory synap-tic transmission in the rat dorsal motor nucleus of the vagus. J Neurosci 2002:22(8):2998-3004.

23. Telford GL, Hoshmonai M, Moses AJ, Szurszewski JH. Morphine initiates migrating myoelectric com-plexes by acting on peripheral opioid receptors. Am J Physiol 1985:249(5 Pt 1):G557-62.

24. Thorn SE, Wickbom G, Philipson L, Leissner P, Wattwil M. Myoelectric activity in the stomach and duodenum after epidural administration of morphine or bupivacaine. Acta Anaesthesiol Scand 1996:40(7):773-8.

25. Qian LW, Pasricha PJ, Chen JD. Origins and patterns of spontaneous and drug-induced canine gastric myoelectrical dysrhythmia. Dig Dis Sci 2003:48(3):508-15.

26. Klepstad P, Dale O, Skorpen F, Borchgrevink PC, Kaasa S. Genetic variability and clinical efficacy of morphine. Acta Anaesthesiol Scand 2005:49(7):902-8.

27. Chou WY, Yang LC, Lu HF, Ko JY, Wang CH, Lin SH, Lee TH, Concejero A, Hsu CJ. Association of mu-opioid receptor gene polymorphism (A118G) with variations in morphine consumption for analgesia after total knee arthroplasty. Acta Anaesthesiol Scand 2006:50(7):787-92.

28. Mikus G, Trausch B, Rodewald C, Hofmann U, Richter K, Gramatte T, Eichelbaum M. Effect of co-deine on gastrointestinal motility in relation to CYP2D6 phenotype. Clin Pharmacol Ther 1997:61(4):459-66.

Publikationer i serien Örebro Studies in Medicine

1. Bergemalm, Per-Olof (2004). Audiologic and cognitive long-term sequelae from closed head injury.

2. Jansson, Kjell (2004). Intraperitoneal Microdialysis. Technique and Results.

3. Windahl, Torgny (2004). Clinical aspects of laser treatment of lichen sclerosus and squamous cell carcinoma of the penis.

4. Carlsson, Per-Inge (2004). Hearing impairment and deafness. Genetic and environmental factors – interactions – consequences. A clinical audiological approach.

5. Wågsäter, Dick (2005). CXCL16 and CD137 in Atherosclerosis.

6. Jatta, Ken (2006). Inflammation in Atherosclerosis.

7. Dreifaldt, Ann Charlotte (2006). Epidemiological Aspects on Malig-nant Diseases in Childhood.

8. Jurstrand, Margaretha (2006). Detection of Chlamydia trachomatis and Mycoplasma genitalium by genetic and serological methods.

9. Norén, Torbjörn (2006). Clostridium difficile, epidemiology and antibiotic resistance.

10. Anderzén Carlsson, Agneta (2007). Children with Cancer – Focusing on their Fear and on how their Fear is Handled.

11. Ocaya, Pauline (2007). Retinoid metabolism and signalling in vas-cular smooth muscle cells.

12. Nilsson, Andreas (2008). Physical activity assessed by accelerometry in children.

13. Eliasson, Henrik (2008). Tularemia – epidemiological, clinical and diagnostic aspects.

14. Walldén, Jakob (2008). The influence of opioids on gastric function:

experimental and clinical studies.