Anesthesiology 2005; 103:845–54 © 2005 American Society of Anesthesiologists, Inc. Lippincott Williams & Wilkins, Inc. Nitrous Oxide Revisited Evidence for Potent Antihyperalgesic Properties Philippe Richebe´ , M.D., Ph.D.,* Cyril Rivat, Ph.D.,† Cyril Creton, M.Sc.,† Jean-Paul Laulin, Ph.D.,‡ Pierre Maurette, M.D., Ph.D.,§ Marc Lemaire, M.D.,࿣ Guy Simonnet, Ph.D.#

Background: Although opioids are unsurpassed analgesics NITROUS oxide is commonly used in humans for pain for surgery, they also induce an N-methyl-D-aspartate–depen- relief. Experimental animal studies have revealed that dent enhancement of postoperative hyperalgesia. Because ni- nitrous oxide induces opioid peptide release in the peri- trous oxide (N2O) has anti–N-methyl-D-aspartate properties, the purpose of this study was to evaluate nitrous oxide ability to aqueductal brainstem, leading to disinhibition (activa- prevent such an opioid-induced hyperalgesia in rats. tion) of the descending noradrenergic inhibitory path- Downloaded from http://pubs.asahq.org/anesthesiology/article-pdf/103/4/845/360312/0000542-200510000-00024.pdf by guest on 01 October 2021 ␥ Methods: First, preventive effects of 50/50% N2O–O2 on the ways via inhibition of -aminobutyric acid–mediated development of delayed hyperalgesia observed after inflamma- interneurons. This results in a negative modulation of the 1–3 tory pain (hind paw carrageenan injection on D0) were exam- nociceptive processes at the spinal cord level. ined for several days. Second, the ability of nitrous oxide (10– A challenging hypothesis is that the nitrous oxide an- 40%) to limit opioid-induced hyperalgesia induced by fentanyl algesic effect is not limited to its antinociceptive effect was evaluated in nonsuffering rats. Third, antihyperalgesic ef- fects of various nitrous oxide concentrations (20–50%) were via endogenous opioids systems but may also be due to assessed in both inflammatory and incisional pain models in the preventive blockade of pain hypersensitivity induced fentanyl-treated rats (4 ؋ 100 ␮g/kg subcutaneously). Finally, by nociceptive inputs. It is generally recognized that the analgesic effect of a single dose of morphine was evaluated tissue damage associated with surgical lesions or inflam- 24 h after fentanyl administration and nitrous oxide (D )to 0 mations often produces peripheral and central sensitiza- assess its preventive effect on acute morphine tolerance in both tion that may outlast the stimuli leading to sustained nonsuffering and hind paw–incised rats. hyperalgesia, allodynia, and persistent pain.4–7 At the Results: When applied on D0, nitrous oxide reduced delayed hyperalgesia induced by inflammation. Exposure to nitrous central level, many experimental studies have shown a oxide on D0 also reduced opioid-induced hyperalgesia in non- critical role for excitatory amino acids to injury-induced suffering rats in a dose-dependent manner. In fentanyl-treated pain sensitization via N-methyl-D-aspartate (NMDA) re- rats with inflammatory or incisional pain, nitrous oxide strongly ceptors. Because the aim of preemptive analgesia is to limited both magnitude and duration of hyperalgesia. Moreover, reduce central sensitization that arises from surgical nox- nitrous oxide exposure on D opposed development of acute tol- 0 ious inputs, many clinical studies have evaluated the erance to analgesic effects of morphine administered on D1 in both nonsuffering and incised fentanyl-treated rats. effectiveness of several NMDA receptor antagonists for Conclusions: Nitrous oxide, an N-methyl-D-aspartate receptor improving postoperative pain management. Clinical antagonist, prevented the enhancement of pain sensitivity in- studies using intravenous low doses of NMDA receptor duced by both nociceptive inputs and fentanyl and opposed antagonists have reported controversial results in hu- acute morphine tolerance. Results suggest that perioperative mans.8 However, ketamine and dextromethorphan have nitrous oxide use reduces exaggerated postoperative pain and morphine consumption. demonstrated promising antihyperalgesic effects in sev- eral clinical trials leading to a reduction in both postop- erative pain and morphine consumption.9–11 Because nitrous oxide was recently shown to be an * Staff Anesthesiologist, Department of Anesthesia and Intensive Care II, 12 Centre Hospitalier et Universitaire de Bordeaux, Bordeaux, France. Ph.D. Stu- NMDA receptor antagonist, one hypothesis is that ni- dent, Laboratoire Home´ostasie-Allostasie-Pathologie EA3666, Universite´ Victor trous oxide should have beneficial antihyperalgesic Se´galen-Bordeaux 2. † Staff Researcher, # Professor, Laboratoire Home´ostasie- Allostasie-Pathologie EA3666, Universite´ Victor Se´galen-Bordeaux 2. ‡ Assistant properties mimicking the ketamine ones, especially Professor Universite´ Bordeaux 1, Bordeaux, France. § Professor, Department of when large opioid doses are used during surgery. Exper- Anesthesia and Intensive Care III, Centre Hospitalier et Universitaire de Bor- 13–18 19–21 deaux, Bordeaux, France. ࿣ M.D., Medical Gases Domain Director, Research and imental and clinical studies have reported that Development, Claude-Delorme Research Center, Air Liquide, Jouy-en-Josas, opioids may paradoxically facilitate the activation of France. NMDA-dependent pronociceptive systems leading to ex- Received from the Laboratoire Home´ostasie-Allostasie-Pathologie EA3666, Uni- versite´ Victor Segalen-Bordeaux 2, Bordeaux, France. Submitted for publication aggerated postoperative pain. March 22, 2005. Accepted for publication July 2, 2005. Supported by the The purpose of the current study was to evaluate the Universite´ Victor Segalen Bordeaux 2, Bordeaux, France; the Ministe`re de l’Education nationale, de l’Enseignement supe´rieur et de la Recherche, Paris, nitrous oxide potency for preventing pain sensitization France; the Conseil Re´gional d’Aquitaine, Bordeaux, France; and Air Liquide induced by nociceptive inputs and high doses of fenta- Research and Development, Jouy-en-Josas, France. Presented at the Annual Meet- 16–18 ings of the Socie´te´ Franc¸aise d’Anesthe´sie et Re´animation, Paris, France, Septem- nyl. To assess such a hypothesis, we first evaluated ber 19, 2003, and the American Society of Anesthesiologists, San Francisco, the effect of a 50/50% N O–O mixture in rats with October 13, 2003. 2 2 inflammatory pain induced by a unilateral hind paw Address reprint requests to Dr. Simonnet: Laboratoire Home´ostasie-Allostasie- Pathologie EA3666, Universite´ Victor Segalen-Bordeaux 2, 146, rue Le´o Saignat, injection of the proinflammatory drug carrageenan. Sec- 33076 Bordeaux Cedex, France. Address electronic mail to: [email protected]. Individual article reprints may be purchased through the Journal Web site, www. ond, we tested the effects of different nitrous oxide anesthesiology.org. concentrations on the development of hyperalgesia in-

Anesthesiology, V 103, No 4, Oct 2005 845 846 RICHEBE´ ET AL. duced by high doses of fentanyl in nonsuffering rats. sliding door on one side to insert rats. Five rats were Third, we evaluated preventive effects of different ni- introduced in each chamber. Fresh gases were fed into trous oxide concentrations on the fentanyl enhancement the chamber through an inlet port (4 l/min) and purged of hyperalgesia induced by inflammatory or incisional by a vacuum set for sucking out the gas at the same rate nociceptive stimuli. Fourth, the 50/50% N2O–O2 mix- as the fresh gas inflow. Oxygen and nitrous oxide con- ture perioperative use was tested for evaluating its effec- centrations were continuously monitored to confirm the tiveness to prevent acute morphine tolerance observed gases’ concentrations. All gas exposures were initiated after such a hind paw surgery associated with perioper- 15 min before the beginning of each experiment and ative high doses of fentanyl.18 were followed for4hofexposure. The total gas expo- sure time was4h15min.

Materials and Methods Measurement of Nociceptive Threshold Downloaded from http://pubs.asahq.org/anesthesiology/article-pdf/103/4/845/360312/0000542-200510000-00024.pdf by guest on 01 October 2021 Animals Nociceptive thresholds in handheld rats were deter- Experiments were performed on adult male Sprague- mined by a modification of the Randall-Selitto method,22 Dawley rats (Charles River Laboratories, l’Abresle, the paw-pressure vocalization test, in which a constantly France) weighing 300–350 g, housed in groups of five increasing pressure is applied to the hind paw until the per cage with a 12-h light–12-h dark cycle (lights on at rat squeaks. The Basile analgesimeter (Apelex, Massy, 7:00 AM) at a constant room temperature of 23° Ϯ 2°C. France; stylus tip diameter, 1 mm) was used. A 600-g The animals had access to food and water ad libitum. cutoff value was determined to prevent tissue damage. Pharmacologic tests and care of the animals were con- ducted in accordance with the Animals Care and Use Carrageenan Injection manual of the National Institutes of Health (Bethesda, On D0, the basal value of the nociceptive threshold MD, National Institutes of Health, 1999). This study, was evaluated, and rats were placed in a plastic cage and including care of the animals involved, was conducted then anesthetized with 3% halothane for 3 min. Carra- according to the official edict presented by the French geenan (0.2 ml of a 1% carrageenan solution in saline) was Ministry of Agriculture (Paris, France) and the recom- then injected into one rat plantar hind paw subcutane- mendations of the Helsinki Declaration. When the ex- ously. Injections were performed with a 25-gauge needle. periments were done, the rats were killed with carbon dioxide. These experiments were conducted in an au- Surgical Procedure thorized laboratory and under the supervision of an Just before the surgery, rats were anesthetized with authorized researcher (J.-P. L.). 1–3% halothane vaporized via a nose cone. The plantar aspect of the operated hind paw was prepared in a Drugs sterile manner with 5% povidone iodine solution, and Fentanyl citrate, morphine, naloxone, and carrageenan the foot was placed through a hole in a sterile drape. As ␭ (Sigma-Aldrich, Saint-Quentin Fallavier, France) were previously described,23 a 1-cm long incision, starting dissolved in physiologic saline (0.9%). Fentanyl (60 and 0.5 cm from the heel and extending toward the toes, was 100 ␮g/kg), morphine (2 and 3 mg/kg), and naloxone made with a No. 11 blade, through skin and fascia of the (1 mg/kg) were administered subcutaneously (1 ml/kg plantar aspect of the left hind paw including the under- body weight). Control animals received an equal volume lying muscle. The plantaris muscle was then raised and of saline injections. Carrageenan (0.2 ml of a 1% carra- incised longitudinally, leaving the muscle origin and in- geenan solution in saline) was prepared 24 h before each sertion intact. After hemostasis with gentle pressure, the experiment. With regard to the incisional pain model, an skin was apposed with two mattress sutures of 5-0 nylon on ointment with 2% Fucidine (Le´o, St. Quentin-en-Yve- a curved needle. The wound site was covered with an lines, France) and Primyxine (oxytetracycline hydrochlo- antibiotic mixture of polymyxin B, oxytetracycline, and ride and polymyxin B sulfate; Chemineau, Vouvray, fusidate. At the end of the surgery, halothane was stopped, France) was placed on the wound after the surgery. and rats were allowed to recover in the plastic box breath- Nitrous oxide (Air Liquide Sante´ France, Paris, France) ing the nitrous oxide–oxygen–nitrogen mixture according was delivered via bottles containing premixed nitrous to the group to which they were allocated. oxide, oxygen, and nitrogen. Different concentrations were used for nitrous oxide, oxygen, and nitrogen: ox- General Procedure ygen was set at 50% in all cases, nitrous oxide varied After arrival in the laboratory, animals were acclima- from 10 to 50%, and nitrogen varied from 0 to 40%. tized to the animal care unit for 4 days. To avoid stress resulting from the experimental conditions that might Exposures to Gas affect measurement of the nociceptive threshold, the All exposures to gas were performed in a plexiglas experiments were performed by the same experimenter chamber (42 cm long, 26 cm wide, 26 cm high) with a in quiet conditions in a testing room close to the animal

Anesthesiology, V 103, No 4, Oct 2005 NITROUS OXIDE IS AN ANTIHYPERALGESIC DRUG 847

care unit. For 2 weeks before the experiments, the during the 7 subsequent days (D1–D7). When rats had animals were weighted daily, handled during 5 min returned to the basal nociceptive threshold value, one gently, and placed in the test room for 2 h (from 9:00 AM naloxone injection (1 mg/kg subcutaneously) was per-

to 11:00 AM), where they were left to become acclima- formed on D7, and the nociceptive threshold was mea- tized. All experiments began at 9:00 AM and were per- sured 5 min later. formed on groups of 10 animals during the light part of Experiment 3: Administration of Various Con- the cycle. Rats were also acclimatized to the Plexiglas centrations of Nitrous Oxide–Oxygen–Nitrogen chamber for 2 weeks before the experiments with an air Treatment in Rats with a Carrageenan Inflamma-

inflow rate set at 4 l/min. Nociceptive threshold assess- tion and Treated by Fentanyl on D0. Fentanyl was ments were performed for the 2 days preceding the administered as described in the second set, resulting in ␮ experimental day (i.e.,onDϪ2, and DϪ1) and repeated on a total dose of 400 g/kg. Five min after the first fentanyl Downloaded from http://pubs.asahq.org/anesthesiology/article-pdf/103/4/845/360312/0000542-200510000-00024.pdf by guest on 01 October 2021 the experimental day (D0), before the exposure to gas and injection, rats received a carrageenan injection. Expo- the carrageenan injection or surgery. Next, the basal noci- sure to gas began 15 min before the first fentanyl admin-

ceptive threshold was determined several times on D0 istration. Rats breathed air or different concentrations of according to the various experimental protocols and once nitrous oxide–oxygen–nitrogen (20/50/30, 30/50/40, daily until the rats recovered the basal values. Experiments 40/50/10, and 50/50/0%, respectively) for 4 h after the were only initiated when no statistical change of the basal first fentanyl administration. Nociceptive threshold mea- nociceptive threshold was observed for 3 successive days surements were performed 2, 4, 6, and 8 h after fentanyl Ͼ (DϪ2,DϪ1, and D0; one-way analysis of variance, P 0.05). injection (D0) and once daily during the 12 subsequent The reference value of the nociceptive threshold was days (D1–D12). When rats had returned to the basal chosen as the basal value on D0. The experimenter was nociceptive threshold value, one naloxone injection unaware of the administered treatment. (1 mg/kg subcutaneously) was performed on D12, and the nociceptive threshold was measured 5 min later.

Experimental Protocols Experiment 4: Administration of 50/50% N2O–O2 In a preliminary experiment, the 50/50% O2–N2 treat- in Rats Scheduled for Left Foot Plantar Incision and ment administered on D0 was compared with air in naive Treated by Fentanyl on D0. Fentanyl was administered rats. This experiment was conducted to evaluate as described in the second set, resulting in a total dose of ␮ whether 50% O2 concentration used in the following five 400 g/kg. Exposure to nitrous oxide–oxygen began in sets of experiment had any effect per se on the nocicep- the Plexiglas chamber 15 min before the first fentanyl tive threshold. injection. All rats received a left foot plantar incision Experiment 1: Administration of the 50/50% under halothane anesthesia 15 min after the first fentanyl

N2O–O2 Treatment in Rats with a Carrageenan In- injection. Then, they recovered in the Plexiglas chamber flammation. Rats were allocated to one of the follow- breathing for 4 h either the equimolar mixture of nitrous

ing groups: (1) N2O–O2 50/50% for4hor(2)airforthe oxide–oxygen or air according to the original group to same duration. Carrageenan injection was performed 15 which they were allocated. Nociceptive thresholds were min after starting the exposure to gas. Nociceptive thresh- estimated according the design of the third set of experi-

old measurement were performed 2, 4, and 6 h after car- ments, except that the naloxone test was performed on D8. rageenan injection (D0) and once daily during the 7 subse- Experiment 5: Morphine Administration on D1 quent days (D1–D7). When rats had returned to the basal after Exposure to the 50/50% N2O–O2 Treatment in nociceptive threshold value, one naloxone injection the Incisional Pain Model. In a first phase (experi-

(1 mg/kg subcutaneously) was performed on D7, and the ment 5A), the analgesic efficiency of 2 mg/kg subcuta- nociceptive threshold was measured 5 min later. neous morphine was estimated 24 h after fentanyl injec- Experiment 2: Administration of Different Con- tion (4 ϫ 60 ␮g/kg) in rats that breathed air or the

centrations of Nitrous Oxide–Oxygen–Nitrogen in 50/50% N2O–O2 for 4 h during the analgesic effect of Fentanyl-treated Rats. Two hours after the basal noci- fentanyl (D0). Control of morphine effectiveness was esti- ceptive threshold measurement on D0, one fentanyl mated in rats receiving saline instead of fentanyl on D0. (100 ␮g/kg) injection (or saline) was performed four Nociceptive threshold was estimated every 30 min for 6 h

consecutive times every 15 min, resulting in a total dose after fentanyl injection on D0 and for 3 h after morphine ␮ of 400 g/kg. Exposure to gas began 15 min before the injection on D1. Nociceptive thresholds were also esti- first fentanyl administration. Rats breathed either air mated every day for the 7 subsequent days. A naloxone test

(control group) or different concentrations of nitrous was performed on D7 as described previously. oxide–oxygen–nitrogen (10/50/40, 20/50/30, 30/50/20, In a second phase (experiment 5B), analgesic effi-

and 40/50/10%) for the 4 h after the first fentanyl admin- ciency of morphine (3 mg/kg) was estimated on D1,24h istration (total gas exposure time:4h15min). Nocicep- after one plantar incision (D0) in fentanyl-treated rats ϫ ␮ tive threshold measurements were performed every 30 (4 100 g/kg) breathing air or the 50/50% N2O–O2 for min for 6 h after fentanyl injection (D0) and once daily 4 h during the analgesic effect of fentanyl (D0). Nocicep-

Anesthesiology, V 103, No 4, Oct 2005 848 RICHEBE´ ET AL.

Fig. 1. Effect of premixed 50/50% N2O–O2 treatment on carrageenan-induced hy- peralgesia in both the ipsilateral (A) and contralateral (B) hind paws. The carra-

geenan injection was performed on D0. Nitrous oxide or air was administered to rats 15 min before the carrageenan injec- tion and for 4 h. Nociceptive threshold

,was assessed on D؊2, D؊1, and D0; then 2 4, and 6 h after carrageenan injection; and subsequently once daily for 7 days. When the rats had returned to their basal

nociceptive threshold (D7), naloxone was injected (1 mg/kg subcutaneously),

and the nociceptive threshold was evalu- Downloaded from http://pubs.asahq.org/anesthesiology/article-pdf/103/4/845/360312/0000542-200510000-00024.pdf by guest on 01 October 2021 ated 5 min later. Nociceptive threshold is expressed as mean ؎ SD. * Dunnett test, P

< 0.05 compared with the D0 basal value. ؍ ؍ ؍ # Dunnett test, P < 0.05 for comparison between groups. Open circles 50/50% N2O–O2-treated rats (n 10); filled circles .(10 ؍ air-treated rats (n

tive threshold was estimated every 2 h for 8 h after threshold on the ipsilateral and contralateral paws on D0, fentanyl injection on D0 and every 30 min for 1.5 h after which lasted 4 and 2 days, respectively (figs. 1A and B; morphine injection on D1. Nociceptive thresholds were Dunnett test, P Ͻ 0.05). In rats exposed to 50/50% also estimated daily for the 8 subsequent days. A nalox- N2O–O2, a smaller decrease of the nociceptive threshold one test was performed on D as described previously. 8 was observed on D0 and on D1 for both hind paws (figs. Ͻ 1A and B; Dunnett test, P 0.05). When injected on D7, Calculation and Statistical Analysis naloxone induced a marked decrease of the nociceptive To evaluate the time course effects of treatments on threshold, which was smaller for both hind paws in nociceptive threshold (basal reference value: precarrag- nitrous oxide–oxygen-breathing rats as compared with eenan or presurgery value on D0 for all experiments and the reduction observed in air-breathing rats (Dunnett initial reference value for the premorphine value on D1 test, P Ͻ 0.05). for morphine analgesia), an analysis of variance followed by post hoc analysis using the Dunnett test was per- formed on D0, and another one was performed on the days after the treatments in each group. The Mann– Effect of Different Concentrations of Nitrous Oxide– Whitney test was used to compare the morphine anal- Oxygen–Nitrogen on Delayed Fentanyl-induced gesic indexes. Analgesic indexes for morphine-induced Hyperalgesia in Rats (Experiment 2) analgesia represented by the area under the curve were As described previously, fentanyl administration in- calculated for each rat by the trapezoidal method and duced analgesia followed by both immediate (hours) and expressed as a mean percentage (Ϯ SD) of the reference delayed hyperalgesia for several days (figs. 2A–D). Expo- index (100%: analgesic index associated with analgesia sure to nitrous oxide (10, 20, 30, or 40%) completely observed in the control group). The paired Student t test reduced the immediate hyperalgesia observed after an- Ͼ was used for comparing the hyperalgesic effect induced algesia on D0 (Dunnett test, P 0.05). Exposure to by naloxone on D8. The statistical significance criterion nitrous oxide on D0 also induced a dose-dependent re- was P Ͻ 0.05. duction in the delayed nociceptive threshold decrease observed for several days in air-treated rats. When ni- trous oxide was used at only 10% concentration, the Results nociceptive threshold decrease was still significant for 2 days (Dunnett test, P Ͻ 0.05; fig. 2A) and completely Effect of 50% O2 Concentration on the Nociceptive Threshold (Preliminary Experiment) suppressed for exposures to the highest nitrous oxide No effect of oxygen was observed on the nociceptive concentrations (Dunnett test, P Ͼ 0.05). Comparison threshold in rats breathing 50% O2 as compared with rats between air and nitrous oxide–treated rats indicated a breathing air (data not shown) (P Ͼ 0.05). significant difference for 3 days with 10 and 20% con- centrations and for 5 days with 30 and 40% concentra- tions (Dunnett test, P Ͻ 0.05). Naloxone injected on D Effect of 50/50% N2O–O2 on Nociceptive Threshold 7 in Rats with a Hind Paw Inflammation Induced by induced a significant decrease of the nociceptive thresh- One Plantar Carrageenan Injection (Experiment 1) old in rats preexposed to air or 10% N2O (Student t test, In rats exposed to air, the plantar carrageenan injec- P Ͻ 0.05) but not in rats preexposed to 20, 30, and 40% Ͼ tion in one foot induced a decrease of the nociceptive N2O (Student t test, P 0.05).

Anesthesiology, V 103, No 4, Oct 2005 NITROUS OXIDE IS AN ANTIHYPERALGESIC DRUG 849 Downloaded from http://pubs.asahq.org/anesthesiology/article-pdf/103/4/845/360312/0000542-200510000-00024.pdf by guest on 01 October 2021

Fig. 2. Effect of different concentrations of premixed nitrous oxide–oxygen–nitrogen (N2O–O2–N2) treatment on fentanyl-induced ؋ ␮ hyperalgesia. The fentanyl injections were performed on D0 (total dose of 4 100 g/kg). Gas mixtures were administered to rats ؍ ؍ ؍ starting 15 min before the fentanyl injection and for 4 h. (A)N2O–O2–N2 10/50/40%; (B)N2O–O2–N2 20/50/30%; (C)N2O–O2–N2 ؍ D)N2O–O2–N2 40/50/10%. Nociceptive threshold at the hind paw level was assessed on D؊2, D؊1, and D0; then every 30 min) ;30/50/20% after the first fentanyl injection until 420 min on D0; and subsequently once daily for 7 days. Once the rats had returned to the basal nociceptive threshold (D7), naloxone was injected (1 mg/kg subcutaneously), and the nociceptive threshold was evaluated 5 min later. ؎ Nociceptive threshold is expressed as mean SD. * Dunnett test, P < 0.05 compared with the D0 basal value. # Dunnett test, P < 0.05 for ؍ ؍ ؍ ؍ comparison between groups. Open circles N2O–O2–N2-treated rats (n 10); filled circles air-treated rats (n 10).

Effect of Various Concentrations of Nitrous Oxide– rats breathing air (Dunnett test, P Ͼ 0.05; fig. 3A).

Oxygen–Nitrogen on Both Fentanyl-induced Preexposure of rats to 30% N2O induced a significant Analgesia and Long-lasting Hyperalgesia in Rats difference for 1 day as compared with rats breathing air with Hind Paw Inflammation (Experiment 3) (Dunnett test, P Ͻ 0.05; fig. 3B). When used at 40 and As shown in figure 3, exposure to nitrous oxide en- 50% concentrations, nitrous oxide reduced the nocicep- hanced fentanyl analgesic effect for the highest gas con- tive threshold decrease for several days as compared centrations (Dunnett test, P Ͻ 0.05). Fentanyl analgesic with rats breathing air (Dunnett test, P Ͻ 0.05; figs. 3C effect was followed by a large and sustained decrease of and D). When injected on D12, naloxone induced a the nociceptive threshold for several days in air-treated significant decrease in the nociceptive threshold in rats Ͻ Ͻ rats (Dunnett test, P 0.05). When nitrous oxide was preexposed to air or 20–40% N2O (Student t test, P administered on D0 at 20% concentration, no change in 0.05) but not in rats preexposed to 50% N2O (Student t the nociceptive threshold decrease was noticed with test, P Ͼ 0.05).

Anesthesiology, V 103, No 4, Oct 2005 850 RICHEBE´ ET AL. Downloaded from http://pubs.asahq.org/anesthesiology/article-pdf/103/4/845/360312/0000542-200510000-00024.pdf by guest on 01 October 2021

Fig. 3. Effect of different concentrations of premixed nitrous oxide–oxygen–nitrogen (N2O–O2–N2) treatment on fentanyl–carra- geenan-induced hyperalgesia. The first fentanyl injection (D0) was performed 5 min before the carrageenan injection and then every min for a total dose of 4 ؋ 100 ␮g/kg. Nitrous oxide or air was administered to rats 15 min before the first fentanyl injection and 15 ؍ ؍ ؍ ؍ for4h.(A)N2O–O2–N2 20/50/30%; (B)N2O–O2–N2 30/50/20%; (C)N2O–O2–N2 40/50/10%; (D)N2O–O2 50/50%. Nociceptive threshold was assessed on D؊2,D؊1, and D0; then every 2 h after the first fentanyl injection until8honD0; and subsequently once daily for 12 days. When the rats had returned to the basal nociceptive threshold (D12), naloxone was injected (1 mg/kg subcutane- ,ously), and the nociceptive threshold was evaluated 5 min later. Nociceptive threshold is expressed as mean ؎ SD. * Dunnett test ؍ P < 0.05 compared with the D0 basal value. # Dunnett test, P < 0.05 for comparison between groups. Open circles N2O–O2–N2- .(10 ؍ air-treated rats (n ؍ filled circles ;(10 ؍ treated rats (n

Effect of 50/50% N2O–O2 on Both Fentanyl-induced When injected on D8 in rats that had returned to the basal Analgesia and Delayed Fentanyl-induced nociceptive threshold, naloxone induced a smaller de- Hyperalgesia in Rats with Plantar Incision crease in the nociceptive threshold as compared with rats (Experiment 4) preexposed to air (Dunnett test, P Ͻ 0.05).

As shown in figure 4, exposure to 50/50% N2O–O2 during the analgesic effect of fentanyl induced an en- Effect of 50/50% N2O–O2 on Acute Tolerance to hancement of analgesia (Dunnett test, P Ͻ 0.05). In Morphine Analgesic Effect (Experiment 5) air-treated rats, fentanyl analgesic effect was followed by In a first experiment (experiment 5A), tolerance to a large and sustained decrease of the nociceptive thresh- morphine analgesic effect was assessed in rats on D1 old for 4 days (Dunnett test, P Ͻ 0.05). In rats treated during hyperalgesia induced by fentanyl administration with nitrous oxide on D0, decrease of nociceptive thresh- performed on the day before (D0). As compared with the old was limited to 3 days (Dunnett test, P Ͻ 0.05; fig. 4). analgesic effect observed in non–fentanyl-treated rats,

Anesthesiology, V 103, No 4, Oct 2005 NITROUS OXIDE IS AN ANTIHYPERALGESIC DRUG 851 Downloaded from http://pubs.asahq.org/anesthesiology/article-pdf/103/4/845/360312/0000542-200510000-00024.pdf by guest on 01 October 2021

Fig. 4. Effect of premixed 50/50% N2O–O2 treatment on fenta- Fig. 5. Effect of premixed 50/50% N2O–O2 treatment on fenta- nyl–incision-induced hyperalgesia. The fentanyl injection was nyl-induced hyperalgesia and acute morphine tolerance. The performed on D0. The first fentanyl injection was performed 15 fentanyl or saline injections were performed on D0 (total dose: min before the plantar incision and every 15 min for a total 4 ؋ 60 ␮g/kg or same volume of saline). The nitrous oxide or air dose of 4 ؋ 100 ␮g/kg. The nitrous oxide or air was adminis- was administered to rats from 15 min before the first fentanyl tered to rats starting from 15 min before the first fentanyl injection and for 4 h. Morphine (2 mg/kg subcutaneously) was injection and for 4 h. Nociceptive threshold was assessed on injected on D1 24 h later. Nociceptive threshold was assessed at D؊2,D؊1, and D0; then every 2 h after the first fentanyl injection the hind paw level on D؊2,D؊1, and D0; every 30 min after the until8honD0; and subsequently once daily for 8 days. When first fentanyl injection and to6honD0; every 30 min after the rats had returned to the basal nociceptive threshold (D8), morphine injection on D1 for 3 h; and subsequently once daily naloxone was injected (1 mg/kg subcutaneously), and the no- for 7 days. On D7, all rats were injected with naloxone (1 mg/kg ciceptive threshold was evaluated 5 min later. Nociceptive subcutaneously), and the nociceptive threshold was evaluated 5 .threshold is expressed as mean ؎ SD. * Dunnett test, P < 0.05 min later. Nociceptive threshold is expressed as mean ؎ SD compared with the D0 basal value. # Dunnett test, P < 0.05 for * Dunnett test, P < 0.05 compared with the D1 basal value. ؍ comparison between groups. Open circles N2O–O2-treated # Dunnett test, P < 0.05 for comparison between groups. Open ؍ ؍ ؍ ؍ ؍ ؍ rats (n 10); filled circles air-treated rats (n 10). circles N2O–O2–fentanyl-treated rats (n 10); filled circles -air–saline ؍ open triangles ;(10 ؍ air–fentanyl-treated rats (n –air ؍ Analgesic indexes: black square .(10 ؍ treated rats (n ;air–saline-treated rats ؍ the 2-mg/kg morphine injection in pretreated fentanyl fentanyl-treated rats; white square ؍ rats induced a similar time course and no difference in dashed square N2O–O2–fentanyl-treated rats. area under the curve (Mann–Whitney test, P Ͼ 0.05) notwithstanding a large shift in the basal nociceptive duced by nociceptive inputs and its enhancement by threshold (Dunnett test, P Ͻ 0.05). By administering fentanyl, coadministration of 50/50% N2O–O2 with fen- nitrous oxide only on D0, the analgesic effect of 2 mg/kg tanyl reduced acute tolerance to the analgesic effect of subcutaneous morphine was totally restored on D1 (Man- postoperative morphine. n–Whitney test, P Ͼ 0.05; fig. 5). In a second experiment As expected, when applied for4hinrats with unilat- (experiment 5B), tolerance to morphine analgesic effect eral inflammation, the 50/50% N2O–O2 mixture induced was assessed in rats 24 h after plantar incision in fenta- an antinociceptive effect as indicated by the reduction of nyl-treated rats. Figure 6 shows that rats that breathed nociceptive threshold decrease at the inflamed paw 50/50% N2O–O2 on D0 had a smaller decrease of the level. Interestingly, reduction of nociception largely out- nociceptive threshold on D1 as compared with air- lasted the4h15minexposure time to nitrous oxide breathing rats (Dunnett test, P Ͻ 0.05), leading to an because reduction of nociceptive threshold decrease enhancement of morphine maximum effect notwith- was still observed at the inflamed hind paw level 24 h standing both unchanged time course and area under the after stopping nitrous oxide treatment. Noteworthy is curve (Mann–Whitney test, P Ͼ 0.05). our observation that the 50/50% N2O–O2 treatment also strongly reduced the nociceptive threshold decrease ob- Discussion served for 2 days after injury at the non–carrageenan- injected hind paw that had not received any nociceptive This experimental investigation on animals shows that input. This indicates that a time-limited exposure to nitrous oxide, an NMDA receptor antagonist, is able to nitrous oxide may oppose development of secondary reduce fentanyl-induced hyperalgesia observed after an- hyperalgesia or allodynia, which have been previously algesia in a dose-dependent manner. Moreover, by pre- described as mainly resulting from a central pain sensi- venting the development of pain hypersensitivity in- tization process.17,24,25 We have previously reported that

Anesthesiology, V 103, No 4, Oct 2005 852 RICHEBE´ ET AL.

4 h, as observed with NMDA receptor antagonists, pre- vented, in a dose-dependent manner, development of both immediate and delayed fentanyl-induced hyperalge- sia for several days. This shows that nitrous oxide anti- hyperalgesic properties are not limited to pain hypersen- sitivity induced by nociceptive inputs but might oppose NMDA-dependent central pain sensitization processes induced by opioids. These results led us to evaluate the nitrous oxide capability of preventing the fentanyl enhancement of long-lasting hyperalgesia induced by inflammatory or in- cisional nociceptive inputs. In animal experimental stud- Downloaded from http://pubs.asahq.org/anesthesiology/article-pdf/103/4/845/360312/0000542-200510000-00024.pdf by guest on 01 October 2021 ies, it has been reported that an opioid such as fentanyl enhances the long-lasting hyperalgesia observed after inflammation or surgical lesion.17,18 The clinically avail- able NMDA receptor antagonist ketamine prevented this pain enhancement when administered just before fenta- nyl injections and tissue injury.17,18 In accord with these Fig. 6. Effect of premixed 50/50% N2O–O2 treatment on fenta- nyl–incision-induced hyperalgesia and acute morphine toler- experimental data, some clinical studies have reported ance at D1. The fentanyl or saline injections were performed on ؋ ␮ that major surgeries with opioid-based anesthesia were D0 (total dose: 4 100 g/kg or same saline volume). The first injection was done 15 min before incision, then every 15 min. associated with a high incidence of exaggerated postop- Nitrous oxide or air was administered to rats from 15 min erative pain and morphine requirement.19,20,30,31 More- before the first fentanyl injection and for 4 h. Morphine over, it has been reported that patients receiving peri- (3 mg/kg subcutaneously) was injected on D . Nociceptive 1 operative ketamine administration showed significantly threshold was assessed on D؊ ,D؊ , and D ; every 30 min after 2 1 0 32 the first fentanyl injection and to8honD0; every 30 min after less residual pain until the sixth postoperative month. morphine injection on D1 for1h30min; and subsequently In humans, it seems that the larger the intraoperative once daily for 8 days. On D8, all rats were injected with nalox- one (1 mg/kg subcutaneously), and the nociceptive threshold fentanyl or remifentanil dose is, the greater the postop- was evaluated 5 min later. Nociceptive threshold is expressed as erative opioid requirement is.19 Therefore, although an ؎ mean SD. * Dunnett test, P < 0.05 compared with the D1 basal excess of nociceptive inputs generally explains exagger- value. # Dunnett test, P < 0.05 for comparison between groups. ated postoperative pain, another explanation is that it ؍ ؍ Open circles N2O–O2–incision–fentanyl-treated rats (n 10); -An- also results from an enhanced activation of NMDA-de .(10 ؍ air–incision–fentanyl-treated rats (n ؍ filled circles -air–incision–fentanyl-treated pendent pronociceptive systems by opioids them ؍ algesic indexes: black square .N O–O –incision–fentanyl-treated rats ؍ rats; white square 2 2 selves.26 Our study shows that a single nitrous oxide– an NMDA receptor antagonist such as ketamine prevents oxygen treatment for 4 h reduced, in a dose-dependent such secondary hyperalgesia.17 Because the 50/50% manner, the fentanyl enhancement of long-lasting hyper-

N2O–O2 treatment did not induce any reduction of car- algesia observed after inflammation or surgical incision rageenan-induced hind paw inflammation, this effect pain. Although nitrous oxide has a number of receptor suggests that the pharmacologic effect of nitrous oxide is interactions,1 this suggests that NMDA receptor antago- not limited to its antinociceptive effect during exposure nist properties of nitrous oxide play a critical role in its but might also partially oppose mechanisms of pain antihyperalgesic effect. However, comparison of the re- sensitization initiated by tissue damage.7 sults showed that the preventive antihyperalgesic effect For a better evaluation of this new effect of nitrous was stronger on the inflammatory pain model than on oxide, we studied the effect of nitrous oxide on an the incisional pain model. The meaning of such a differ- experimental model of hyperalgesia developed in the ence has to be explained. absence of tissue damage, i.e., the opioid-induced hyper- Although the current results have been gathered from algesia model in the rat.13,26 We demonstrated previ- animal preclinical studies, this potent NMDA-like antihy- ously that a single administration of an opioid such as peralgesic effect of nitrous oxide could partly explain heroin or fentanyl in rats induced, in a dose-dependent the controversial results observed in clinical studies manner, two kinds of NMDA-dependent hyperalgesia: an about the preemptive potency of NMDA receptor antag- early, short-duration hyperalgesia after analgesia and a onists because the published meta-analysis did not take delayed, sustained hyperalgesia for several days.27,28 In- into account whether nitrous oxide was used during the terestingly, it was also previously reported that NMDA anesthetic procedure.9–11 This critical point should be receptor antagonists, especially ketamine, prevented studied in the future for better assessing in humans the opioid-induced hyperalgesia in experimental animal therapeutical interest of NMDA receptor blockade in a models13,27,28 but also in human volunteers.21,29 Our preemptive strategy for postoperative pain management. study showed that a nitrous oxide–oxygen treatment for During the past decade, acute tolerance has been re-

Anesthesiology, V 103, No 4, Oct 2005 NITROUS OXIDE IS AN ANTIHYPERALGESIC DRUG 853 ported as a new adverse effect related to short-term geries.35 As reported previously, our results showed that opioid use for surgery.19,31 Clinical studies have shown naloxone precipitated hyperalgesia when this opioid re- that fentanyl or remifentanil administration for abdomi- ceptor antagonist was administered, after return to basal nal19,20,33 or orthopedic surgeries31 increased morphine nociceptive threshold, in rats treated by a previous her- requirement, suggesting short-term tolerance. As re- oin16 or fentanyl administration.18 The fact that admin- ported for postoperative hyperalgesia, it seems that the istration of an opioid-receptor antagonist induced no larger the intraoperative fentanyl or remifentanil dose is, effect in naive rats but induced a pharmacologic effect the greater the postoperative morphine requirement such as hyperalgesia in rats without apparent pain has is.19 Interestingly, short-term tolerance has also been led us to suggest that rats with previous incisional pain observed in human volunteers 1–2 h after the beginning and opioid histories did not return to their initial equi- of low-dose remifentanil infusion.34 In the rat incisional librium (homeostasis) between opioid-dependent an- pain model, we recently demonstrated that the postop- tinociceptive systems and NMDA-dependent pronoci- Downloaded from http://pubs.asahq.org/anesthesiology/article-pdf/103/4/845/360312/0000542-200510000-00024.pdf by guest on 01 October 2021 erative decrease of morphine effectiveness is closely ceptive systems. We previously proposed they were in a related to the hyperalgesia level observed 24 h after new equilibrium (allostasis) with a high level balance incision and fentanyl administration.18 By reducing the between these two opposite pain-controlling systems hyperalgesia level, ketamine, when administered before that mask one another.16,18 By sharply blocking opioid- both surgery and fentanyl administration, improved the dependent antinociceptive systems, naloxone-precipi- postoperative effectiveness of morphine. Because ni- tated hyperalgesia would allow the level of pronocicep- trous oxide is effective in preventing postinjury hyper- tive system functioning in animal experimental models algesia in fentanyl-treated rats, we finally evaluated the to be unmasked. This might explain the pain vulnerabil- 17 effectiveness of 50/50% N2O–O2 treatment in preventing ity observed in rats with pain and opioid histories. acute morphine tolerance in both nonsuffering and pain- Noteworthy is our observation that preliminary treat- ful fentanyl-treated rats. As previously shown for hero- ment with nitrous oxide, as with ketamine,18 prevented in,28 we reported that both time course and area under naloxone-precipitated hyperalgesia when the opioid re- the curve related to the morphine analgesic effect were ceptor antagonist was administered after return to nor- unchanged when morphine was injected 24 h after fen- mal nociceptive threshold. Interestingly, all these bene- tanyl administration during the hyperalgesic period. In ficial effects of nitrous oxide on experimental models fact, during this period, the impression of less analgesia, were observed for low concentrations of nitrous oxide i.e., apparent tolerance, as seen by the decrease in mor- substantially below to the minimal alveolar concentra- phine maximum analgesic effect, was a consequence of tion in rats.36 the nociceptive threshold shift to lower values. This Because opioids are widely used for surgery, the re- confirms our original hypothesis26 that short-term toler- sults of this study suggest that nitrous oxide, an NMDA ance observed during the postoperative period is not receptor antagonist, could reduce the occurrence of mainly due to an actual decrease in the analgesic mor- exaggerated postoperative pain and tolerance observed phine potency per se as described classically but is re- in major surgeries with opioid-based anesthesia. Conse- lated to sustained pain hypersensitivity induced by an quently, this may facilitate postoperative rehabilitation initial opioid exposure. By totally reducing hyperalgesia, and perhaps limit the development of pain chroniciza- the 50/50% N2O–O2 pretreatment completely restored tion. morphine effectiveness in nonsuffering fentanyl-treated rats. Although it was not possible to demonstrate such a type of result in painful rats because no analgesic refer- References ence effect may be evaluated as in nonsuffering rats, our 1. Fujinaga M, Maze M: Neurobiology of nitrous oxide-induced antinociceptive study showed that the apparent enhancement of mor- effects. Mol Neurobiol 2002; 25:167–89 phine effectiveness by nitrous oxide pretreatment is 2. Ohashi Y, Guo T, Orii R, Maze M, Fujinaga M: Brain stem opioidergic and mainly due to the reduction of the nociceptive threshold GABAergic neurons mediate the antinociceptive effect of nitrous oxide in Fischer rats. ANESTHESIOLOGY 2003; 99:947–54 decrease i.e., pain hypersensitivity induced by the fent- 3. Orii R, Ohashi Y, Halder S, Giombini M, Maze M, Fujinaga M: GABAergic anyl exposure 24 h previously. Considering our previous interneurons at supraspinal and spinal levels differentially modulate the antino- ciceptive effect of nitrous oxide in Fischer rats. ANESTHESIOLOGY 2003; 98:1223–30 data regarding the NMDA receptor antagonist ketamine 4. Mao J, Price DD, Mayer DJ: Mechanisms of hyperalgesia and morphine 18 tolerance: A current view of their possible interactions. Pain 1995; 62:259–74 on the same incisional pain model, this also suggests 5. Coderre TJ: The role of excitatory amino acid receptors and intracellular that NMDA receptor antagonist properties of nitrous messengers in persistent nociception after tissue injury in rats. Mol Neurobiol 1993; 7:229–46 oxide play a critical role for preventing acute tolerance 6. Millan MJ: The induction of pain: An integrative review. Prog Neurobiol to analgesic effects of opioid agonists. 1999; 57:1–164 7. Woolf CJ, Salter MW: Neuronal plasticity: Increasing the gain in pain. These beneficial and prolonged effects of nitrous oxide Science 2000; 288:1765–9 on postoperative pain management led us to determine 8. Schmid RL, Sandler AN, Katz J: Use and efficacy of low-dose ketamine in the management of acute postoperative pain: A review of current techniques and whether such a treatment may protect against long-term outcomes. Pain 1999; 82:111–25 pain vulnerability as described after some types of sur- 9. McCartney CJ, Sinha A, Katz J: A qualitative systematic review of the role of

Anesthesiology, V 103, No 4, Oct 2005 854 RICHEBE´ ET AL.

N-methyl-D-aspartate receptor antagonists in preventive analgesia. Anesth Analg 23. Brennan TJ, Vandermeulen EP, Gebhart GF: Characterization of a rat 2004; 98:1385–400 model of incisional pain. Pain 1996; 64:493–501 10. Subramaniam K, Subramaniam B, Steinbrook RA: Ketamine as adjuvant 24. Kayser V, Guilbaud G: Local and remote modifications of nociceptive analgesic to opioids: A quantitative and qualitative systematic review. Anesth sensitivity during carrageenin-induced inflammation in the rat. Pain 1987; 28:99– Analg 2004; 99:482–95 107 11. Himmelseher S, Durieux ME: Ketamine for perioperative pain manage- 25. Fletcher D, Kayser V, Guilbaud G: The influence of the timing of bupiva- ment. ANESTHESIOLOGY 2005; 102:211–20 caine infiltration on the time course of inflammation induced by two carrageenin 12. Jevtovic-Todorovic V, Todorovic SM, Mennerick S, Powell S, Dikranian K, injections seven days apart. Pain 1997; 69:303–9 Benshoff N, Zorumski CF, Olney JW: Nitrous oxide (laughing gas) is an NMDA 26. Simonnet G, Rivat C: Opioid-induced hyperalgesia: Abnormal or normal antagonist, neuroprotectant and neurotoxin. Nat Med 1998; 4:460–3 pain? Neuroreport 2003; 14:1–7 13. Mao J, Price DD, Mayer DJ: Thermal hyperalgesia in association with the 27. Laulin JP, Larcher A, Ce´le`rier E, Le Moal M, Simonnet G: Long-lasting development of morphine tolerance in rats: roles of excitatory amino acid receptors and protein kinase C. J Neurosci 1994; 14:2301–12 increased pain sensitivity in rat following exposure to heroin for the first time. 14. Laulin JP, Ce´le`rier E, Larcher A, Le Moal M, Simonnet G: Opiate tolerance Eur J Neurosci 1998; 10:782–5 to daily heroin administration: an apparent phenomenon associated with en- 28. Ce´le`rier E, Laulin JP, Corcuff JB, Le Moal M, Simonnet G: Progressive hanced pain sensitivity. Neuroscience 1999; 89:631–6 enhancement of delayed hyperalgesia induced by repeated heroin administra-

15. Vanderah TW, Ossipov MH, Lai J, Malan Jr, TP Porreca F: Mechanisms of tion: A sensitization process. J Neurosci 2001; 21:4074–80 Downloaded from http://pubs.asahq.org/anesthesiology/article-pdf/103/4/845/360312/0000542-200510000-00024.pdf by guest on 01 October 2021 opioid-induced pain and antinociceptive tolerance: Descending facilitation and 29. Angst MS, Koppert W, Pahl I, Clark DJ, Schmelz M: Short-term infusion of spinal dynorphin. Pain 2001; 92:5–9 the mu-opioid agonist remifentanil in humans causes hyperalgesia during with- 16. Ce´le`rier E, Rivat C, Jun Y, Laulin JP, Larcher A, Reynier P, Simonnet G: drawal. Pain 2003; 106:49–57 Long-lasting hyperalgesia induced by fentanyl in rats: Preventive effect of ket- 30. Stubhaug A, Breivik H, Eide PK, Kreunen M, Foss A: Mapping of punctuate amine. ANESTHESIOLOGY 2000; 92:465–72 hyperalgesia around a surgical incision demonstrates that ketamine is a powerful 17. Rivat C, Laulin JP, Corcuff JB, Celerier E, Pain L, Simonnet G: Fentanyl suppressor of central sensitization to pain following surgery. Acta Anaesthesiol enhancement of carrageenan-induced long-lasting hyperalgesia in rats: Preven- Scand 1997; 41:1124–32 tion by the N-methyl-D-aspartate receptor antagonist ketamine. ANESTHESIOLOGY 31. Guignard B, Coste C, Costes H, Sessler DI, Lebrault C, Morris W, Simonnet 2002; 96:381–91 G, Chauvin M: Supplementing desflurane-remifentanil anesthesia with small-dose 18. Richebe´ P, Rivat C, Laulin JP, Maurette P, Simonnet G: Ketamine improves ketamine reduces perioperative opioid analgesic requirements. Anesth Analg the management of exaggerated postoperative pain observed in perioperative 2002; 95:103–8 fentanyl-treated rats. ANESTHESIOLOGY 2005; 102:421–8 32. De Kock M, Lavand’homme P, Waterloos H: “Balanced analgesia” in the 19. Guignard B, Bossard AE, Coste C, Sessler DI, Lebrault C, Alfonsi P, Fletcher perioperative period: Is there a place for ketamine? Pain 2001; 92:373–80 D, Chauvin M: Acute opioid tolerance: Intraoperative remifentanil increases 33. Cooper DW, Lindsay SL, Ryall DM, Kokri MS, Eldabe SS, Lear GA: Does postoperative pain and morphine requirement. ANESTHESIOLOGY 2000; 93:409–17 20. Chia YT, Liu K, Wang JJ, Kuo MC, Ho ST: Intraoperative high dose fentanyl intrathecal fentanyl produce acute cross-tolerance to i.v. morphine? Br J Anaesth induces postoperative fentanyl tolerance. Can J Anaesth 1999; 46:872–7 1997; 78:311–3 21. Koppert W, Sittl R, Scheuber K, Alsheimer M, Schmelz M, Schuttler J: 34. Vinik HR, Kissin I: Rapid development of tolerance to analgesia during Differential modulation of remifentanil-induced analgesia and postinfusion hyper- remifentanil infusion in humans. Anesth Analg 1998; 86:1307–11 algesia by S-ketamine and clonidine in humans. ANESTHESIOLOGY 2003; 99:152–9 35. Macrae WA: Chronic pain after surgery. Br J Anaesthesiol 2001; 87:88–98 22. Kayser V, Basbaum AI, Guilbaud G: Deafferentation in the rat increases 36. Gonsowski CT, Eger IIEI: Nitrous oxide minimum alveolar anesthetic mechanical nociceptive threshold in the innervated limbs. Brain Res 1990; concentration in rats is greater than previously reported. Anesth Analg 1994; 79: 508:329–32 710–2

Anesthesiology, V 103, No 4, Oct 2005