トップ ごあいさつ ジャーナルオブアネステージア モニターワールド抄録集 ダウンロード
  1. Volatile anesthetic antagonism by long-chain free fatty acids
  2. Changes in body temperature during profound hypothermic cardiopulmonary bypass in adult patients undergoing
  3. Effects of withdrawal of phasic lung in?ation during normocapnia and hypercapnia on the swallowing re?ex in humans
  4. Antagonizing potencies of saturated and unsaturated long-chain free fatty acids to iso?urane in gold?sh
  5. Multi-Unit and Multi-Path system of the neural network can explain the steep dose-response of MAC
  6. Single sodium channels from human skeletal muscle in planar lipid bilayers: characterization and response to pentobarbital
  7. Toborinone and olprinone, phosphodiesterase III inhibitors, inhibit human platelet aggregation due to the inhibition of both calcium release from intracellular stores and calcium entry
  8. Spinal neurotoxicity and tolerance after repeated intrathecal administration of YM 872, an AMPA receptor antagonist, in rats
  9. Monitoring magnesium to guide magnesium therapy for heart surgery
  10. Severe Legionella pneumophila pneumonia associated with the public bath on a cruise ship in Japan
  11. Brachial plexus injury related to improper positioning during general anesthesia
  12. Successful management of a patient with neuroleptic malignant syndrome associated with marked elevation of serum creatine kinase
  13. Pneumothorax associated with epidural anesthesia
  14. Integration of suppression ratio in the bispectral index
  15. Letters to the editor

<JN>J Anesth (2004) 18:71-72
<PT>Editorial
<CT>Volatile anesthetic antagonism by long-chain free fatty acids
<CA>Tomohiro Yamakura
<ADD>Division of Anesthesiology, Niigata University Graduate School of Medical and Dental Sciences, 1-757 Asahimachi, Niigata 951-8510, Japan
<para1>If there were a speci?c anesthetic antagonist available without side effects, clinical management of general
anesthesia would bene?t greatly. However, such an
anesthetic antagonist is not available, partly because the mechanisms of general anesthesia in the central nervous system have yet to be fully elucidated. In this issue of the Journal of Anesthesia, Hanada et al. [1] expand their earlier interesting ?nding that myristate, a saturated free fatty acid with 14 carbons, antagonizes volatile anesthetics in gold?sh [2]; they show that the antagonizing effects of long-chain free fatty acids are determined not only by their hydrophobicity but also by the ability of their molecular con?guration to perturb lipid membrane structures [1]. Thus, they suggest that free fatty acids alter the function of membrane protein by both a direct action on membrane protein, and an indirect action through lipid bilayers.
  <para2>As a result of recent research into anesthetic mechanisms, membrane proteins, especially ligand-gated and other ion channels, have been considered as plausible target molecules of general anesthetics [3]. Although free fatty acids are shown to regulate the activity of ion channels, the reported effects do not seem very consistent. For example, in respect of the effects of unsaturated free fatty acids on the agonist binding or function of GABAA (A-aminobutyric acid type A) receptors, both potentiation [4,5] and inhibition/no effect [6,7] have been reported, although the experimental conditions were different. Thus, to understand the mechanism of volatile anesthetic antagonism by free fatty acids, more evidence concerning interactions between volatile anesthetics and free fatty acids may be required for each ion channel. Because the enhancement of the agonist binding of GABAA receptors by pentobarbital is decreased by unsaturated fatty acids, which themselves enhance the agonist binding [4], it may not be necessary for free fatty acids to have opposite effects on ion channels to those of anesthetics to antagonize the anesthetic effects.
  <para2>Recently, a target molecule for in vivo anesthesia by propofol has been identi?ed, using the elegant knock-in mouse technique. With this technique, in mice, a mutation was introduced into a GABAA receptor ?3 subunit that eliminated the propofol potentiation of GABAA receptors but did not alter other physiological functions [8]. The elimination of or profound reduction in anesthetic behaviors by propofol in these mice strongly suggests that the GABAA receptor is a major determinant of propofol anesthesia in vivo. Thus, the antagonism of propofol anesthesia in rats by the GABAA receptor antagonists picrotoxin and gabazine [9] may be an
example of anesthetic antagonism at a target molecule level. However, propofol anesthesia can also be reversed by an anticholinesterase agent, physostigmine [10,11]. Because physostigmine does not affect GABAA receptors [12], the antagonism of propofol anesthesia by physostigmine may be an example of anesthetic antagonism induced by an action on a molecule other than the anesthetic target. Thus, it may not be necessary for anesthetic antagonists to act on the anesthetic target itself. In this context, free fatty acids may not directly affect an anesthetic target protein to antagonize the action of volatile anesthetics.
  <para2>Although it is dif?cult to clarify whether anesthetics bind to a speci?c site on the membrane protein, evidence using the techniques of photoaf?nity labeling [13] and sulfhydryl-speci?c agents [14] suggests that anesthetic actions on ion channels are due to binding at a speci?c site. Thus, the ideal selective anesthetic antagonist would be a competitive antagonist at such an anesthetic binding site. However, as discussed by Hanada et al. [1], free fatty acids may not compete at a speci?c anesthetic binding site on the ion channel, because free fatty acids are negatively charged and iso?urane is uncharged. Nevertheless, free fatty acids may be attractive as antagonists, because no remarkable adverse effects from free fatty acids were observed in gold?sh [1], and free fatty acids are essential human nutrients. Further studies in mammals will be needed to more fully examine the abilities and safety of free fatty acids as volatile anesthetic antagonists.
<A>References
<REF> 1. Hanada R, Tatara T, Iwao Y (2004) Antagonizing potencies of saturated and unsaturated long-chain free fatty acids to iso?urane in gold?sh. J Anesth 18:89-93
<REF> 2. Tatara T, Kamaya H, Ueda I (2002) Myristate, a 14-carbon fatty acid, effectively reverses anesthesia. Anesthesiology 97:518-520
<REF> 3. Yamakura T, Bertaccini E, Trudell JR, Harris RA (2001) Anesthetics and ion channels: molecular models and sites of action. Annu Rev Pharmacol Toxicol 41:23-51
<REF> 4. Koenig JA, Martin IL (1992) Effect of free fatty acids on GABAA receptor ligand binding. Biochem Pharmacol 44:11-15
<REF> 5. Witt MR, Poulsen CF, Lukensmejer B, Westh-Hansen SE, Nabekura J, Akaike N, Nielsen M (1999) Structural requirements for the interaction of unsaturated free fatty acids with recombinant human GABAA receptor complexes. Ann NY Acad Sci 868:697-700
<REF> 6. Schwartz RD, Yu X (1992) Inhibition of GABA-gated chloride channel function by arachidonic acid. Brain Res 585:405-410
<REF> 7. Nabekura J, Noguchi K, Witt MR, Nielsen M, Akaike N (1998) Functional modulation of human recombinant A-aminobutyric acid type A receptor by docosahexaenoic acid. J Biol Chem 273:11056-11061
<REF> 8. Jurd R, Arras M, Lambert S, Drexler B, Siegwart R, Crestani F, Zaugg M, Vogt KE, Ledermann B, Antkowiak B, Rudolph U (2003) General anesthetic actions in vivo strongly attenuated by a point mutation in the GABAA receptor ?3 subunit. FASEB J 17:250-252
<REF> 9. Sonner JM, Zhang Y, Stabernack C, Abaigar W, Xing Y, Laster MJ (2003) GABAA receptor blockade antagonizes the immobilizing action of propofol but not ketamine or iso?urane in a dose-related manner. Anesth Analg 96:706-712
<REF>10. Fassoulaki A, Sarantopoulos C, Derveniotis C (1997) Physostigmine increases the dose of propofol required to induce anaesthesia. Can J Anaesth 44:1148-1151
<REF>11. Meuret P, Backman SB, Bonhomme V, Plourde G, Fiset P (2000) Physostigmine reverses propofol-induced unconsciousness and attenuation of the auditory steady state response and bispectral index in human volunteers. Anesthesiology 93:708-717
<REF>12. Li CY, Wang H, Xue H, Carlier PR, Hui KM, Pang YP, Li ZW, Han YF (1999) Bis(7)-tacrine, a novel dimeric AChE inhibitor, is a potent GABAA receptor antagonist. Neuroreport 10:795-800
<REF>13. Pratt MB, Husain SS, Miller KW, Cohen JB (2000) Identi?cation of sites of incorporation in the nicotinic acetylcholine receptor of a photoactivatible general anesthetic. J Biol Chem 275:29441-29451
<REF>14. Mascia MP, Trudell JR, Harris RA (2000) Speci?c binding sites for alcohols and anesthetics on ligand-gated ion channels. Proc Natl Acad Sci USA 97:9305-9310

ページのTOPへ


<JN>J Anesth (2004) 18:73-81
<PT>Original articles
<CT>Changes in body temperature during profound hypothermic cardiopulmonary bypass in adult patients undergoing
aortic arch reconstruction
<CA>Takashi Akata, Ken Yamaura, Tadashi Kandabashi, Shinya Sadamatsu, and Shosuke Takahashi
<ADD>Department of Anesthesiology and Critical Care Medicine, Faculty of Medicine, Kyushu University, 3-1-1 Maidashi,
Higashi-ku, Fukuoka 812-8582, Japan
<AB>Abstract
<AB>Purpose. Our aim was to characterize changes in body temperatures during profound hypothermic cardiopulmonary
bypass (CPB) conducted with the sternum opened.
<AB>Methods. In ten adult patients who underwent profound hypothermic (,20°C) CPB for aortic arch reconstruction, pulmonary arterial temperature (PAT), nasopharyngeal temperature (NPT), forehead deep-tissue temperature (FHT), and urinary bladder temperature (UBT) were recorded every 1 min throughout the surgery. In addition, the CPB venous line temperature (CPBT), a reasonable indicator of mixed venous blood temperature during CPB and believed to best re?ect core temperature during stabilized hypothermia on CPB, was recorded during the period of total CPB.
<AB>Results. PAT began to change immediately after the start of cooling or rewarming, closely matching the CPBT (r 5 0.98). During either situation, the other four temperatures lagged behind PAT (P , 0.05); however, NPT followed PAT more closely than the other three temperatures (P , 0.05). During stabilized hypothermia, PAT, NPT, and FHT, but not UBT, closely matched the CPBT, with gradients of less than 0.5°C.
<AB>Conclusion. During induction of profound hypothermia and its reversal on total CPB with the heart in situ, a PA catheter thermistor, presumably because of its placement immediately behind the superior vena cava, would provide a reliable measure of the mixed venous blood temperature. During stabilized profound hypothermia, PAT, NPT, and FHT, but not UBT, serve as a reliable index of core temperature.
<KW>Key words Induced hypothermia ・ Deep hypothermia ・ Core temperature ・ Cardiopulmonary bypass ・ Thoracic aortic
aneurysm
<A>Introduction
<para1>During profound hypothermic cardiopulmonary bypass (CPB) or circulatory arrest, it is essential to monitor body temperatures at several sites, to ensure that the organs vulnerable to decreased O2 delivery actually receive the bene?t of the desired degree of hypothermia, to assess evenness of cooling, and to diagnose hazardous hypothermia [1,2]. It is thus particularly important to use temperature monitoring sites most likely to
re?ect brain temperature. In this regard, pulmonary
arterial, nasopharyngeal, tympanic, and distal esophageal monitoring sites have been used to estimate brain temperature during CPB [1-3]. However, each of these sites has unique problems [3], and none of them may re?ect brain temperature reliably throughout the CPB [4]. In addition to core temperature monitoring, temperature monitoring at intermediate or peripheral zones (e.g., rectum, urinary bladder, muscle, skin) has been recommended to assess adequate whole body
rewarming during CPB [5-7].
  <para2>Stone et al. [4] previously demonstrated changes in temperatures measured at various standard monitoring sites (i.e., nasopharynx, esophagus, pulmonary artery, tympanic membrane, urinary bladder, rectum, axilla, sole of the foot) during profound hypothermic CPB conducted for repair of cerebral aneurysms without
the sternum opened. They found that, among those monitoring sites, measurements from the nasopharynx, esophagus, and pulmonary artery tended to match brain temperature measured during neurosurgical procedures with the brain exposed to the cool surroundings. If the sternum had been opened, temperatures measured at some of those standard monitoring sites (e.g., pulmonary artery, esophagus) might have changed differently during the profound hypothermic CPB.
  <para2>Less information is available regarding changes in body temperatures during profound hypothermic
CPB conducted for intrathoracic procedures with the sternum opened. In this study, we therefore investigated changes in pulmonary arterial, nasopharyngeal, forehead deep-tissue, urinary bladder, and ?ngertip skin-surface temperatures during profound hypothermic CPB in adult patients with thoracic aortic aneurysms who underwent aortic arch reconstruction. The CPB venous line temperature is a reliable indicator of the mixed venous blood temperature during CPB, and is believed to best re?ect core temperature (i.e., brain temperature) during CPB when no active core warming or cooling is occurring (i.e., during stabilized hypothermia) [1,2]. Thus, in order to help characterize changes in the above temperatures during hypothermic CPB, we compared them with the CPB venous line temperature.
<A>Patients and methods
<para1>With institutional approval and informed consent, we studied ten adult patients with thoracic aortic aneurysm who underwent profound deep hypothermic (,20°C) CPB and circulatory arrest for aortic arch reconstruction. The patient demographics are summarized in Table 1.
  <para2>The patients were premedicated with oral nitrazepam (2-7.5 mg) and roxatidine acetate hydrochloride (75 mg) 90 min before entering the operating room. Anesthesia was induced with intravenous midazolam (40-150 Ig kg21) and fentanyl (2-10 Ig kg21), and tracheal intubation was facilitated with vecuronium bromide
(5-10 mg i.v.). Anesthesia was subsequently maintained with midazolam, fentanyl, and sevo?urane (0.5%-3%) in oxygen. The lungs were ventilated mechanically to maintain PaCO2 at approximately 35 mmHg.
  <para2>In addition to the standard anesthetic safety monitors, radial, femoral, and pulmonary arterial catheters and transesophageal echocardiography were used to monitor cardiovascular functions. Pulmonary arterial temperature (PAT) was monitored with a thermistor at the tip of a thermodilution catheter placed in the right pulmonary artery (Swan-Ganz CCOmbo CCO/SvO2/VIP; Edwards Lifesciences LLC, Irvine, CA, USA), the placement of which was con?rmed by preoperative chest X-ray in every patient. Nasopharyngeal temperature (NPT) was monitored by placing a thermistor probe in the posterior nasopharynx (~5 cm from the external naris), and sealing the external naris with
cotton gauze. Forehead deep-tissue temperature (FHT) was monitored by placing a 4.5-cm-diameter sensor probe designed for the measurement of deep-tissue temperature (PD-11; Terumo, Tokyo, Japan) on the forehead. Urinary bladder temperature (UBT) was monitored using a thermistor-tipped urinary bladder catheter (Respiratory Support Products, Irvine, CA, USA). Fingertip skin-surface temperature (FSST) was monitored by placing a thermistor probe on the tip of the index ?nger opposite the nail bed and surrounding the probe with gauze folded in eight (~5 mm in thickness). Mixed venous blood temperature during total CPB was monitored using a thermistor probe (Avecor Cardiovascular, Plymouth, MN, USA) placed in the CPB venous line. All temperature sensors were interfaced with electronic thermometers (AA-900P thermometer [Nihon Kohden, Tokyo, Japan] for pulmonary arterial, nasopharyngeal, forehead deep-tissue, urinary bladder, and ?ngertip skin-surface temperatures; and
a YSI Precision 4000A thermometer [Nikkiso-YSI,
Tokyo, Japan] for CPB venous line temperature) whose synchronous digital output was continuously displayed. The data derived from the AA-900P thermometer were electronically sampled and stored at 1-min intervals, while the data displayed on the YSI Precision 4000A thermometer were manually recorded at several (3 to 5)-min intervals. The operating room temperature was thermostatically maintained at 20°C.
  <para2>Patients were anticoagulated with heparin
(300 U ・ kg21) and subsequent doses were titrated to keep the activated clotting times above 450 s. A 28-Fr cannula (DLP Single Stage Venous Cannula; Medtronics, Minneapolis, MN, USA) and a 32-Fr
cannula (Venous Return Catheter; Polystan A/S, Walgerholm, Denmark) were passed into the superior vena cava and inferior vena cava, respectively, via the right atrium. Roller pumps with a membrane oxygenator were used to achieve a 2.5 l・min21・m22 extracorporeal ?ow via a 10-mm Microvel vascular graft (Hemashield Gold; Boston Scienti?c, Natick, MA, USA) sutured to the femoral artery. The bypass circuit was primed using acetated Ringer solution (Veen-F; Nikken Kagaku, Tokyo, Japan) with or without adding packed red blood cells to maintain hematocrit between 20% and 25%.
  <para2>Profound hypothermia (NPT ,20°C) was rapidly (~20 min) induced with a separate water-bath heat exchanger which was initially set at 10°C (Table 2). After NPT became lower than 20°C, the temperature of the water bath was allowed to rise to 15°C-20°C, where it was maintained until rewarming began. After induction
of the profound hypothermia, cold (4°C) crystalloid cardioplegic solution, containing 20 mEq ・ l21 KCl, was infused into the coronary circulation to arrest the heart (Table 2). During aortic arch reconstruction, the circulation was arrested, except for the cerebral circulation, into which the hypothermic CPB perfusate was infused retrograde at a rate of 150-200 ml・min21 (with central venous pressure raised to 10-15 mmHg) via the 28-Fr cannula placed in the superior vena cava. The CPB was resumed after the aortic arch was reconstructed, and the hypothermia was gradually reversed on CPB (Table 2). During the rewarming period, the heat exchanger was initially set at 39°C and later adjusted downward. In addition, a water-?lled heating mattress was activated at 39°C. When UBT became higher than 36°C, the CPB was terminated and protamine given.
  <para2>Temperatures at the four standard core temperature monitoring sites (i.e., PAT, NPT, FHT, and UBT) were compared with the CPB venous line temperature, using correlation coef?cients and Bland and Altman analyses. In these analyses, the data obtained during the ?rst 20 and 60 min of the cooling and rewarming, respectively, were used. Because the data on UBT were relatively variable, we also investigated the relation between urine volume and the changes in UBT during either cooling or rewarming, using simple (either linear or non-linear) regression analyses. The temperature-time relationship during either cooling or rewarming was analyzed using analysis of variance (ANOVA), the Tukey-Kramer test (in case of homogeneous population variances), and the Games-Howell test (in case of heterogeneous population variances). Comparison of the correlation coef?cients was made using Fisher's
Z-transformation. Any other necessary comparisons between two groups were made by either two-tailed, unpaired Student's t-test (in case of homogeneous population variances) or Welch's t-test (in case of
heterogeneous population variances).
  <para2>All the above analyses were made on a computer, using GB-Stat v 6.5.6 PPC (Dynamic Microsystems, Silver Spring, MD, USA), PowerStats v 0.9 (Shinko Trading, Tokyo, Japan), or Excel (Microsoft, Redmond, WA, USA). Differences were considered signi?cant at P , 0.05. Values were expressed as either means 6 SD, means 6 2SD (in the Bland and Altman analyses) or means 6 SEM (for clarity in the temperature-time
relationship).
<A>Results
<B>Changes in body temperatures during cooling
<para1>Temperatures monitored at the four standard core
temperature monitoring sites (i.e., PAT, NPT, FHT, and UBT) were closely matched (P . 0.05) until the cooling was started (Fig. 1). However, only PAT began to decrease immediately after the start of the cooling (Fig. 1), and NPT, FHT, UBT, and FSST signi?cantly lagged (P , 0.05) behind PAT, with lag times of one to several min (Fig. 1). NPT followed PAT more closely (P , 0.05) than FHT, UBT, or FSST (Fig. 1). PAT and NPT were signi?cantly lower (P , 0.05) than either FHT or UBT at all time points after 3 min and 18 min, respectively, of the cooling (Fig. 1). However, no signi?cant differences (P . 0.05) were noted in the temperature-time relationship between FHT and UBT (Fig. 1).
  <para2>During the cooling, PAT, NPT, FHT, and UBT were all signi?cantly correlated with the CPB venous line temperature (Fig. 2). Among them, PAT was best correlated with the CPB venous line temperature (Fig. 2). The correlation coef?cient for the CPB venous line-pulmonary arterial relation (r 5 0.977) was signi?cantly larger (P , 0.05) than that for the relation of the CPB venous line temperature with either NPT (r 5 0.925), FHT (r 5 0.883), or UBT (r 5 0.857). However, no difference (P . 0.05) was observed in the correlation coef?cients among the CPB venous line temperature-NPT, CPB venous line temperature-FHT, and CPB venous line temperature-UBT relations. In the Bland-Altman analyses (Fig. 2), the difference from the CPB venous line temperature was signi?cantly lower for PAT (20.21°C) than for either NPT (22.6°C), FHT (25.1°C), or UBT (24.6°C) (P , 0.05). The SD of the difference between CPB venous line temperature and PAT (1.7°C) was smaller than that for the difference between CPB venous line temperature and either NPT (2.5°C), FHT (3.2°C), or UBT (3.4°C).
<B>Comparison of body temperatures during stabilized profound hypothermia
<para1>During the period of stabilized hypothermia (i.e., for 20 min before the start of rewarming), PAT (18.5 6 2.0°C), FHT (18.2 6 1.6°C), and NPT (18.6 6 1.8°C) closely matched the CPB venous line temperature (18.1 6 1.8°C vs PAT, P 5 0.715; vs FHT, P 5 0.893; vs NPT, P 5 0.622). However, UBT (20.0 6 1.9°C) was signi?cantly higher than the CPB venous line temperature (18.1 6 1.8°C vs UBT; P 5 0.043). In addition, as shown in Fig. 3, no signi?cant differences were observed among PAT, FHT, and NPT. However, UBT was slightly, although signi?cantly (P , 0.05), higher than these three temperatures (Fig. 3).
<B>Changes in body temperatures during rewarming
<para1>Only PAT began to increase immediately after the start of the rewarming, and the other four temperatures signi?cantly lagged (P , 0.05) behind PAT (Fig. 3). NPT followed PAT more closely (P , 0.05) than the other three temperatures (Fig. 3). As shown in Fig. 3, PAT and NPT were signi?cantly higher (P , 0.05) than either FHT, UBT, or FSST at many time points during the rewarming. However, no signi?cant differences (P . 0.05) were noted in the temperature-time relationship between FHT and UBT (Fig. 3).
  <para2>During the rewarming, temperatures at the four
standard core temperature monitoring sites were also signi?cantly correlated with the CPB venous line temperature (Fig. 4). Again, among them, PAT was best correlated with the CPB venous line temperature (Fig. 4). The correlation coef?cient for the CPB venous line temperature-PAT relation (r 5 0.981) was signi?cantly larger (P , 0.05) than that for the relation of the CPB venous line temperature with either NPT (r 5 0.946), FHT (r 5 0.939), or UBT (r 5 0.801). In addition, the correlation coef?cient for either the CPB venous line temperature-NPT relation or the CPB venous line
temperature-FHT relation was signi?cantly larger (P , 0.05) than that for the CPB venous line temperature-UBT relation. No signi?cant difference (P . 0.05) was observed in the correlation coef?cients between the CPB venous line temperature-NPT and CPB venous line temperature-FHT relations. In the Bland-Altman analyses, the difference from the CPB venous line temperature was signi?cantly lower for PAT (20.23°C) than for either NPT (20.78°C), FHT (2.00°C), or UBT (2.3°C) (P , 0.05). The SD of the difference between CPB venous line temperature and PAT (1.6°C) was smaller than that of the difference between CPB venous line temperature and either NPT (2.2°C), FHT (2.3°C), or UBT (3.9°C) (Fig. 4).
<B>Relation between changes in bladder temperature and urine ?ow rate
<para1>A signi?cant linear relation was found between changes in UBT (y) and urine volume (x) during the ?rst 30 min of cooling (y 5 10.7 1 0.0094x; r 5 0.69; P 5 0.03; n 5 10), but not during the ?rst 60 min of rewarming (y 5 7.2 1 0.013x; r 5 0.61; P 5 0.06; n 5 10). No signi?cant non-linear (power, exponential, logarithmic, reciprocal)
relation was found between changes in UBT and urine volume during either situation.
<A>Discussion
<para1>The immediate response of PAT to the start of active core cooling or rewarming on CPB (i.e., acute changes in blood temperature) implies that PAT closely follows the changes in blood temperature during either the cooling or rewarming. Indeed, during the cooling or rewarming, PAT closely (r 5 0.98) matched the CPB venous line temperature, a reliable estimate of the
perfusing blood temperature during CPB. From another viewpoint, the CPB venous line temperature is a reasonable indicator of mixed venous blood temperature during CPB, and thus would re?ect the average temperature within highly perfused organs during CPB. The highly perfused organs during CPB at lower hematocrit (~25%) would include brain, spinal cord, stomach, gut, liver, spleen, kidney, thyroid gland, skeletal muscles, and skin [8]. Because the speci?c heats of
these organs and blood are almost identical (3.56-3.85 kJ・kg21・°C21) [9], their temperatures would change at similar rates during the cooling or rewarming. In other words, the changes in the CPB venous line temperature would closely re?ect those in brain temperature. Indeed, it was previously shown in anesthetized sheep undergoing moderate hypothermia that, during either cooling or rewarming, the temperature of central venous blood in the right atrium closely matched brain temperature, measured with a thermometer inserted deep into the cerebral cortex via a small drill-hole in the skull [10]. Thus, PAT, which excellently correlated with the CPB venous line temperature during the cooling or rewarming, may possibly serve as a reliable index of brain temperature during active cooling or rewarming on CPB.
  <para2>During total CPB, in spite of cessation of pulmonary blood ?ow and the intrathoracic procedure with the sternum opened, PAT closely matched the CPB venous line temperature (i.e., mixed venous blood temperature during CPB). We speculate that the temperature measured with the pulmonary arterial catheter thermistor closely re?ected the temperature of venous blood ?owing at a high rate in the superior vena cava (SVC), because the thermistor was placed in the right pulmonary artery immediately behind the SVC (Fig. 5). The SVC blood temperature would represent the average temperature within highly perfused (i.e., vessel-rich) regions of the upper body (i.e., brain, thyroid gland, skeletal muscles, and skin). Because of the aforementioned identity of speci?c heat among the highly perfused organs [9], the SVC blood temperature would be identical to the mixed venous blood temperature during CPB. Thus, it is conceivable that, in spite of cessation of pulmonary blood ?ow, PAT closely matched the CPB venous temperature in this study. During our measurements, the right pulmonary arterial segment where the thermistor was placed had not been exposed to the air with the heart in situ (i.e., without being overturned), and the pleural cavity or the pericardium was not ?lled with cold irrigating solution. Thus, PAT was, presumably, little in?uenced by the cool surroundings, closely matching the CPB venous line temperature.
  <para2>Based on the assumption that PAT closely re?ected changes in brain temperature (as discussed above), none of NPT, FHT, and UBT could be considered as a reliable index of brain temperature during rapid induction of profound hypothermia and its reversal on CPB. However, during stabilized profound hypothermia on CPB, both NPT and FHT, but not UBT, could be considered as a reliable index of brain temperature.
  <para2>Posterior NPT has long been used to estimate core or brain temperature during hypothermic CPB in the clinical setting [5,11-13]. However, in earlier animal studies [10,14], as well as in a recent human study [4], NPT has been shown to modestly, although signi?cantly, either overestimate or underestimate brain temperature (measured with a thermometer placed in the cerebral cortex) during rapid (20-40 min) cooling or rewarming. However, in those studies [4,10], during stabilized profound hypothermia, the NPT closely matched the brain temperature with gradients of less than 1.0°C. All these ?ndings are not inconsistent with our ?ndings.
  <para2>Deep tissue temperature can be estimated using an insulated thermistor probe placed on the skin surface that creates an area of zero thermal ?ux between the skin surface and subcutaneous deep tissue [15-18]. In practice, the probe is insulated by electrically heating the upper surface of the probe and thereby eliminating thermal gradients between its upper and lower surfaces. This insulation would eventually lead to the creation of an area of zero thermal ?ux between the skin surface and subcutaneous deep tissue. Thus, utilizing this deep-tissue thermometry, brain temperature could be estimated noninvasively by placing the insulated thermistor probe at the forehead skin surface. However, the principle of deep-tissue thermometry suggests its slow responsiveness, and this thermometry may not be useful in detecting rapid changes in core temperature [19]. Indeed, this thermometry has been shown to be useful in estimating relatively slow changes in core temperature during general surgery [20], but not in estimating rapid changes in core temperature during the induction of moderate hypothermia (25°C-28°C) on CPB [19]. However, previous studies have yielded con?icting results regarding its usefulness in estimating changes in core temperature during reversal of moderate hypothermia on CPB [19,21,22].
  <para2>This study, for the ?rst time, investigated the possible usefulness of forehead deep-tissue thermometry in estimating core blood (or brain) temperature during profound hypothermic (#20°C) CPB, i.e., during cooling, stabilization, and rewarming. In our patients, FHT signi?cantly lagged behind PAT and NPT during either the cooling or rewarming, consistent with the previously reported discrepancy between NPT and FHT during the induction of moderate hypothermia (25°C-28°C) and its reversal on CPB [19]. As inferred from its principle, deep-tissue thermometry would fail to exteriorize the deep-tissue temperature if the ambient temperature is higher than the deep-tissue temperature (because of incomplete thermal insulation). Nevertheless, in our measurements made with the ambient temperature thermostatically controlled at 20°C, FHT closely matched both PAT and NPT during the profound hypothermia stabilized at ~18°C. Because the air movement is controlled using vertical ?ow in our operating rooms, the vertical air?ow may have decreased the temperature in the vicinity of the patient's forehead to lower than 18°C and thereby enabled FHT to re?ect the deep-tissue temperature. Forehead deep-tissue thermometry may be useful in estimating brain temperature during a period of retrograde cerebral circulation. However, we did not investigate this issue because of the lack of information on the reference temperature, i.e., directly measured brain temperature or average temperature of the blood returning from brain.
  <para2>UBT and rectal temperature, commonly used to estimate core temperature during general surgery, have both been shown to signi?cantly lag behind PAT, NPT, or esophageal temperature during induction of hypothermia and its reversal [23-26], consistent with our results. Because urine is a ?ltrate of blood, a thermistor-tipped urinary catheter would provide a reliable measure of core temperature if the urine ?ow rate were high. Thus, during a period of hypothermic CPB when the urine ?ow rate normally decreases, UBT would fail to accurately re?ect changes in core temperature.
Indeed, it was reported that, during rewarming on CPB, the difference between UBT and NPT increased with lower urine ?ow rates [25]. In this study, we con?rmed the dependence of UBT on the urine ?ow rate.
  <para2>Distal esophageal temperature has been suggested
to serve as a reliable index of brain or central blood temperature during the induction of hypothermia or hyperthermia, and its reversal [10,27-29]. However, in our recent clinical practice, it is becoming rare to insert a thermometer into the esophagus during cardiac surgery because of the routine intraoperative use of transesophageal echocardiography.
  <para2>Earlier studies [30,31] had proposed that tympanic membrane temperature could be considered as a gold reference for brain or hypothalamic temperature.
However, more recent studies have suggested that the tympanic temperature can be signi?cantly in?uenced by changes in ambient temperature (i.e., face or head skin temperature), and may not accurately re?ect brain
temperature [32,33]. Particularly in operating rooms in which air movement is controlled using vertical ?ow, the tympanic temperature would not re?ect brain
temperature precisely because cold air is blowing on the patient's head and face. In addition, perforation of
tympanic membrane was previously reported as a complication of tympanic thermometry during anesthesia [34]. Thus, tympanic thermometry is rarely used in our clinical practice.
  <para2>In conclusion, during a period of total CPB when pulmonary blood ?ow has nearly ceased and distal pulmonary arteries are not exposed to the air, a pulmonary arterial catheter thermistor located immediately behind the superior vena cava appears to provide a reliable estimate of the temperature of the blood returning from the upper body and, thus, possibly, the brain temperature. During the induction of profound hypothermia and its reversal on total CPB, the pulmonary arterial temperature, but neither the nasopharyngeal, forehead deep-tissue, nor urinary bladder temperature, would closely re?ect changes in the mixed venous blood temperature, indicative of the ef?ciency of active core cooling or rewarming. On the other hand, during stabilized profound hypothermia on CPB, the pulmonary arterial, nasopharyngeal and forehead deep-tissue temperatures, but not urinary bladder temperature, appear to provide a reliable measure of core temperature (i.e., brain temperature).
<ACK>Acknowledgments. The authors thank a number of anesthesia residents and staff members at the Department of Anesthesiology and Critical Care Medicine, Faculty of Medicine, Kyushu University (Fukuoka, Japan) for their kind cooperation in this work.
<A>References
<REF> 1. Kurusz M, Davis RF, Conti VR (2000) Conduct of cardiopulmonary bypass. In: Gravlee GP, Davis RF, Kurusz M, Utley JR (eds) Cardiopulmonary bypass. Lippincott Williams & Wilkins, Philadelphia, pp 549-577
<REF> 2. Skeehan TM, Jopling M (2003) Monitoring the cardiac surgical patient. In: Hensley FA Jr, Martin DE, Gravlee GP (eds) A practical approach to cardiac anesthesia. Lippincott Williams & Wilkins, Philadelphia, pp 98-140
<REF> 3. Davis RB, Kauffman JN, Cobbs TL, Mick SL (1995) Assembling and monitoring the extracorporeal circuit. In: Mora CT, Guyton RA, Finlayson DC, Rigatti RL (eds) Cardiopulmonary bypass. Springer, Berlin 1-1 Heidelberg Tokyo New York, pp 238-246
<REF> 4. Stone JG, Young WL, Smith CR, Solomon RA, Wald A, Ostapkovich N, Shrebnick DB (1995) Do standard monitoring sites re?ect true brain temperature when profound hypothermia is rapidly induced and reversed? Anesthesiology 82:344-351
<REF> 5. Muravchick S, Conrad DP, Vargas A (1980) Peripheral temperature monitoring during cardiopulmonary bypass operation. Ann Thorac Surg 29:36-41
<REF> 6. Azar I (1981) Rectal temperature is best indicator of adequate rewarming during cardiopulmonary bypass. Anesthesiology 55:189-190
<REF> 7. Ramsay JG, Ralley FE, Whalley DG, DelliColli P, Wynands JE (1985) Site of temperature monitoring and prediction of afterdrop after open heart surgery. Can J Anaesth 32:607-612
<REF> 8. Rudy LW Jr, Heymann MA, Edmunds LH Jr (1973) Distribution of systemic blood ?ow during cardiopulmonary bypass. J Appl Physiol 34:194-200
<REF> 9. Werner J, Buse M (1988) Temperature pro?les with respect to inhomogeneity and geometry of the human body. J Appl Physiol 65:1110-1118
<REF>10. Hercus V, Cohen D, Bowring AC (1959) Temperature gradients during hypothermia. BMJ 1:1439-1441
<REF>11. Noback CR, Tinker JH (1980) Hypothermia after cardiopulmonary bypass in man: amelioration by nitroprusside-induced vasodilation during rewarming. Anesthesiology 53:277-280
<REF>12. Rajek A, Lenhardt R, Sessler DI, Kurz A, Laufer G, Christensen R, Matsukawa T, Hiesmayr M (1998) Tissue heat content and distribution during and after cardiopulmonary bypass at 31°C and 27°C. Anesthesiology 88:1511-1518
<REF>13. Rajek A, Lenhardt R, Sessler DI, Brunner G, Haisjackl M, Kastner J, Laufer G (2000) Ef?cacy of two methods for reducing postbypass afterdrop. Anesthesiology 92:447-456
<REF>14. Stefaniszyn HJ, Novick RJ, Keith FM, Salerno TA (1983) Is the brain adequately cooled during deep hypothermic cardiopulmonary bypass? Current Surgery 40:294-297
<REF>15. Fox RH, Solman AJ (1971) A new technique for monitoring the deep body temperature in man from the intact skin surface. J Physiol (Lond) 212:8-10
<REF>16. Fox RH, Solman AJ, Isaacs R, Fry AJ (1973) A new method for monitoring deep body temperature from the skin surface. Clin Sci 44:81-86
<REF>17. Kobayashi T, Nemoto T, Kamiya A, Togawa T (1975) Improvement of deep body thermometer for man. Ann Biomed Eng 3:181-188
<REF>18. Togawa T, Nemoto T, Yamazaki T, Kobayashi T (1976) A modi?ed internal temperature measurement device. Med Biol Engineering 14:361-364
<REF>19. Muravchick S (1983) Deep body thermometry during general anesthesia. Anesthesiology 58:271-275
<REF>20. Matsukawa T, Sessler DI, Ozaki M, Hanagata K, Iwashita H, Kumazawa T (1997) Comparison of distal oesophageal temperature with "deep" and tracheal temperatures. Can J Anaesth 44:433-438
<REF>21. Sakuragi T, Mukai M, Dan K (1993) Deep body temperature during the warming phase of cardiopulmonary bypass. Br J Anaesth 71:583-585
<REF>22. Yamakage M, Iwasaki S, Namiki A (2002) Evaluation of a newly developed monitor of deep body temperature. J Anesth 16:354-357
<REF>23. Sellick BA (1957) A method of hypothermia for open heart surgery. Lancet 1:443-446
<REF>24. Molnar GW, Read GW (1974) Studies during open-heart surgery on the special characteristics of rectal temperature. J Appl Physiol 36:333-336
<REF>25. Horrow JC, Rosenberg H (1988) Does urinary catheter temperature re?ect core temperature during cardiac surgery? Anesthesiology 69:986-989
<REF>26. Bone ME, Feneck RO (1988) Bladder temperature as an estimate of body temperature during cardiopulmonary bypass. Anaesthesia 43:181-185
<REF>27. Cooper KE, Kenyon JR (1957) A comparison of temperatures measured in the rectum, oesophagus, and on the surface of the aorta during hypothermia in man. Br J Surg 44:616-619
<REF>28. Cohen D, Hercus V (1959) Controlled hypothermia in infants and children. BMJ 1:1435-1439
<REF>29. Shiraki K, Konda N, Sagawa S (1986) Esophageal and tympanic temperature responses to core blood temperature changes during hyperthermia. J Appl Physiol 61:98-102
<REF>30. Benzinger TH (1969) Tympanic thermometry in surgery and anesthesia. JAMA 209:1207-1211
<REF>31. Baker MA, Stocking RA, Meehan JP (1972) Thermal relationship between tympanic membrane and hypothalamus in conscious cat and monkey. J Appl Physiol 32:739-742
<REF>32. McCaffrey TV, McCook RD, Wurster RD (1975) Effect of head skin temperature on tympanic and oral temperature in man. J Appl Physiol 39:114-118
<REF>33. Shiraki K, Sagawa F, Tajima F, Yokota A, Hashimoto M, Brengelmann GL (1988) Independence of brain and tympanic temperatures in an unanesthetized human. J Appl Physiol 65:482-486
<REF>34. Wallace CT, Marks WE, Adkins WY, Mahafey JE (1974) Perforation of the tympanic membrane, a complication of tympanic thermometry during anesthesia. Anesthesiology 41:290-291

ページのTOPへ


<JN>J Anesth (2004) 18:82-88
<PT>
<CT>Effects of withdrawal of phasic lung in?ation during normocapnia and hypercapnia on the swallowing re?ex in humans
<CA>Toshihito Sai, Shiro Isono, and Takashi Nishino
<ADD>Department of Anesthesiology, Graduate School of Medicine, Chiba University, 1-8-1 Inohanacho, Chuo-ku, Chiba 260-8760, Japan
<AB>Abstract
<AB>Purpose. This study was done to test the hypothesis that hypercapnia has a direct, inhibitory effect on swallowing.
<AB>Methods. We investigated changes in the frequency and
timing of repeated swallows induced by continuous infusion
of water into the pharynx before, during, and after transient airway occlusion at normocapnia and hypercapnia in 12 healthy volunteers. Hypercapnia was induced by adding a dead space. Ventilation was monitored using a pneumotachograph, and swallowing was identi?ed by submental
electromyogram.
<AB>Results. We found that hypercapnia decreased the frequency of swallows (8.2 6 3.7 vs 11.4 6 5.3 swallows・min21 [mean 6 SD]: hypercapnia vs normocapnia; P , 0.05),
together with a loss of the preponderant coupling of swallows with expiratory phase observed at normocapnia. We also found that the withdrawal of phasic lung in?ation produced by airway occlusion at end-expiration suddenly increased the swallowing frequency, both at normocapnia (from 11.4 6 5.3 to 16.7 6 3.7 swallows・min21; P , 0.01) and at hypercapnia (from 8.2 6 3.7 to 22.0 6 6.7 swallows・min21; P , 0.01). Although the degree of increased swallowing frequency during airway occlusion was more prominent at hypercapnia than at normocapnia (P , 0.05), the distribution of the timing of swallows in relation to the phase of the respiratory cycle during airway occlusion at hypercapnia was similar to that during airway occlusion at normocapnia.
<AB>Conclusion. The results of our study strongly suggest that the attenuation of the swallowing re?ex during hypercapnia is not due to the direct, inhibitory effect of CO2 on the swallowing center, but, rather, is due to the increased inhibitory in?uence of a lung-volume-related re?ex.
<KW>Key words Swallowing re?ex ・ Hypercapnia ・ Lung-volume-related re?ex
<A>Introduction
<para1>Pulmonary aspiration is a major complication during perioperative periods, and the upper airway re?exes, including the swallowing re?ex, have obvious protective values against the aspiration of foreign material into the respiratory tract. There is some evidence to suggest that airway protective re?exes interact with background chemical ventilatory drive [1-3]. For example, it has been shown that, in anesthetized humans, an increase
in CO2 ventilatory drive decreases the degree and
duration of respiratory responses to airway irritation, whereas a decrease in CO2 ventilatory drive has the opposite effect [3]. Although the physiological signi?cance of this interaction is not entirely clear, it is conceivable that the automatic respiratory control system may prevail over the airway protective re?exes when the maintenance of ventilation is particularly important in a condition of hypercapnia. Considering that the swallowing re?ex functions as a protective re?ex, it is likely that the swallowing re?ex interacts with background chemical ventilatory drive. In fact, our previous study [4] showed that hypercapnia not only decreased the frequency of the repetitive swallowing induced by continuous infusion of water into the pharynx but also changed the timing of swallows in relation to the phase of the respiratory cycle.
  <para2>Although the mechanisms that cause these changes in the rate and timing of swallowing are not entirely clear, at least two possible mechanisms might be considered: (1) vagally mediated inhibitory re?exes for swallowing, and (2) direct effects of hypercapnia on the swallowing center in the medulla. Concerning the ?rst possibility, there are several reports [5-8] to support the idea that the vagally mediated re?exes such as lung/pulmonary re?exes and upper airway re?exes play an important role in the control of the swallowing re?ex. On the other hand, no substantial evidence to support possibility (2) above has been reported. The purpose of this study was to test the hypothesis that hypercapnia, per se, may have a direct effect on the integrative brainstem network underlying the swallowing re?ex. We investigated changes in the rate and timing of repetitive swallowing in response to a sudden withdrawal of phasic lung movements during normocapnia and hypercapnia. We reasoned that the sudden withdrawal of phasic vagal in?uence may disclose the swallowing-respiration relationship produced in the brainstem network.
<A>Subjects, materials, and methods
<B>Study subjects
<para1>Twelve healthy male volunteers, aged 21 to 54 years, were studied. None had histories of dysphagia, or of neuromuscular, cardiovascular, or pulmonary disease. Each subject provided informed consent, and the study protocol was approved by the Institutional Ethics Committee. All subjects were told about various procedures that would take place, but none was familiar with the hypothesis being tested.
<B>Study design
<para1>Each subject was seated during the experiment and breathed through a tightly ?tting face mask connected to a pneumotachograph and then a T-piece system. Details of the experimental setup are given elsewhere [7,8]. In brief, we measured ventilatory air?ow, tidal volume (VT), end-tidal CO2 tension (PETCO2), and mask pressure (Pmask). Swallowing was determined by a burst of the submental electromyogram (EMG) with interruption of air?ow and visual observation of the characteristic laryngeal movements. Re?ex swallows were induced by continuous infusion of water into the pharynx (2 ml・min21) through a thin nasopharyngeal catheter placed without the use of topical anesthesia. During the experiment, hyperoxia was maintained by passing 100% oxygen with a total ?ow of 10 l・min21 through the T-piece.
<B>Methods
<para1>Each subject breathed through the face mask with or without an additional dead space (1.3 l), and a period of 5-7 min was allowed for the establishment of stable breathing patterns at normocapnia and hypercapnia. Subsequently, continuous infusion of distilled water into the pharynx (2 ml・min21) was started to induce the repetitive swallowing re?ex. When the swallowing
response to continuous infusion of water was stable, breathing and swallowing were recorded for 3 min (baseline period). Then, the airway was occluded at end-expiration by in?ating a balloon placed near the entrance port of the face mask, and the occlusion was maintained for 15 s. During the occlusion of the airway (airway occlusion period) the subjects were asked to continue their breathing efforts as normally as possible. The changes in respiration and swallowing were analyzed using the data of the last 60 s during the baseline period, the 15 s of airway occlusion, and the ?rst 60 s of a 2-min recovery period following airway occlusion. For quantitative analysis of the effects of sudden changes in ventilation on the swallowing re?ex, changes in respiratory variables (air?ow, VT, PETCO2, Pmask) and swallowing rate during the baseline, airway occlusion, and recovery periods were analyzed. In addition, the timing of the swallows in relation to the phase of the respiratory cycle before and after airway occlusion was determined as described previously [4,7,8]. In brief, swallows preceded by and followed by inspiratory ?ow were marked as inspiratory (I) swallows, whereas swallows preceded by and followed by expiratory ?ow were designated as expiratory (E) swallows. Swallows occurring at the transition between inspiration and expiration were designated inspiratory-expiratory (I-E) swallows, and swallows occurring at the transition between expiration and the inspiratory phase of the next breath were designated expiratory-inspiratory (E-I) swallows (Fig. 1). The timing of swallows during airway occlusion was determined by analyzing the relationships between submental EMG signals and negative inspiratory airway pressure developed during airway occlusion (Fig. 2). The submental EMG signal during the swallowing
act was easily distinguishable from the EMG activity generated by inspiratory efforts, because the amplitude of EMG activity during the swallowing act was much greater than that generated by inspiratory efforts, even during airway occlusion.
<B>Analysis
<para1>Statistical analysis was performed by using repeated measures analysis of variance (ANOVA) (two-way) or Friedman repeated-measures ANOVA on ranks and paired t-test, where appropriate. The post-hoc test following ANOVA was performed by using Bonferroni's t-test. P , 0.05 was considered signi?cant.
<A>Results
<para1>All 12 subjects tolerated the continuous infusion of
water and transient airway occlusion procedures during normocapnia and hypercapnia, and completed the
experimental protocol. A total of 538 swallows were analyzed.
  <para2>Table 1 shows the mean values of respiratory variables before and during continuous infusion of water into the pharynx at normocapnia and hypercapnia.
Although repetitive swallows induced by continuous
infusion of water caused a slight decrease in respiratory frequency at normocapnia, there was no signi?cant
difference in the values of other respiratory variables between before and during continuous infusion of water into the pharynx.
  <para2>Figure 3 shows experimental records illustrating changes in respiration and swallowing in response to airway occlusion at normocapnia (Fig. 3A) and
hypercapnia (Fig. 3B) obtained in a single subject.
At normocapnia, a one-to-two rhythmic coupling of
swallowing and respiration was seen before airway
occlusion. In response to airway occlusion, the one-to-two coupling of swallowing and respiration changed
to a one-to-one rhythmic coupling, causing an increase in swallowing frequency. Immediately after the release of airway occlusion, there was a slight increase in
ventilation and the swallowing frequency decreased considerably. During hypercapnia, the responses of swallowing and respiration to airway occlusion were basically similar to those observed during normocapnia. However, the changes appeared to be more exaggerated, because the frequency of swallows was slower
before the airway occlusion and was higher during
airway occlusion, compared with normocapnia. Figure
4 shows changes in the frequency of swallowing in
response to airway occlusion obtained from all 12 subjects. At normocapnia, the values for swallowing frequency before, during, and after airway occlusion were 11.4 6 5.3, 16.7 6 3.7, and 7.6 6 3.6 swallows・min21, respectively (Fig. 4A). There were signi?cant differences among these values. At hypercapnia, the values for swallowing frequency before, during, and after airway occlusion were 8.2 6 3.7, 22.0 6 6.7, and 8.3 6 3.3 swallows・min21, respectively (Fig. 4B). The frequency
of swallows during airway occlusion was signi?cantly higher than those before and after the airway occlusion. In addition, compared with the values for swallowing frequency at normocapnia, the value before airway
occlusion was signi?cantly lower (P , 0.05) and the value during airway occlusion was signi?cantly higher (P , 0.05).
  <para2>Figure 5 shows changes in the timing of swallows in relation to the phase of the respiratory cycle in response to airway occlusion during normocapnia (Fig. 5A) and hypercapnia (Fig. 5B). Although there was a wide variation among subjects, the majority of swallows occurred at the expiratory phase at normocapnia before airway occlusion. During airway occlusion at normocapnia,
expiratory swallows decreased and E-I swallows
increased. Under hypercapnic conditions the preponderant occurrence of expiratory swallows was not
observed, and I-E as well as E-I transition swallows were more frequently observed. During airway occlusion under hypercapnia, I-E transition swallows decreased and inspiratory swallows increased. Thus, the distribution of timing of swallows in relation to the phase of the respiratory cycle during airway occlusion
at hypercapnia was quite similar to that during airway occlusion at normocapnia.
<A>Discussion
<para1>In this study we con?rmed our previous observa-
tion that the swallowing re?ex was attenuated during hyperpnea due to hypercapnia. We also demonstrated that airway occlusion caused considerable changes in the frequency and timing of swallows induced by continuous infusion of water into the pharynx. The major ?ndings in this study were that: (1) airway occlusion at end-expiration suddenly increased the frequency of swallows; (2) the degree of increased swallowing frequency during airway occlusion was more prominent at hypercapnia than at normocapnia; (3) the distribution of the timing of swallows in relation to the phase of
the respiratory cycle during airway occlusion at hypercapnia was similar to that during airway occlusion
at normocapnia, despite the marked difference in the distribution of the timing of swallows between normocapnia and hypercapnia before the airway occlusion. These results suggest that the attenuation of the swallowing re?ex during hypercapnia is not due to the direct, inhibitory effect of CO2 on the swallowing center, but, rather, is due to the increased inhibitory in?uence of a lung-volume-related re?ex.
  <para2>In the present study we also con?rmed our previous observation [4] that hypercapnia not only decreased the frequency of swallowing but also changed the timing of swallows in relation to the phase of the respiratory cycle. It is possible that hypercapnia, per se, may directly inhibit the swallowing center, thereby decreasing the frequency of swallows. However, this possibility is unlikely, because airway occlusion at both normocapnia and hypercapnia considerably increased the swallowing frequency immediately after the start of airway occlusion. Furthermore, the ?nding that the frequency of swallows during airway occlusion at hypercapnia
was higher than that during airway occlusion at normocapnia does not support the possibility that
hypercapnia exerts a direct, inhibitory effect on the swallowing center.
  <para2>The ?nding that the frequency of swallows increased during airway occlusion suggests that the withdrawal of phasic lung in?ation may exert an excitatory effect on the swallowing re?ex. Concerning this possibility, it has been shown that nasal continuous positive airway pressure (CPAP) or negative extrathoracic pressure applied in normal adult humans decreases the frequency of repeated swallows [6,7]. Furthermore, our recent study showed that hyperpnea decreased and breath-holding increased the frequency of repeated swallows [8]. All these ?ndings are compatible with the idea that a lung-volume related re?ex plays an important role in the control of the swallowing re?ex while exerting an inhibitory effect on re?ex swallowing. Thus, assuming that a greater phasic vagal activity can be produced by the increases in tidal volume and minute ventilation during hypercapnia [9], the observed decrease in swallowing frequency during hypercapnia can be explained exclusively by the augmented inhibitory effect of a lung-volume related re?ex.
  <para2>The possibility exists that the observed increase in swallowing frequency during airway occlusion may be associated with behavioral and/or emotional responses. For example, Fonagy and Calloway [10] showed that experimental tasks aimed at inducing emotional arousal increased the spontaneous swallowing rate in normal human subjects. Because airway occlusion may cause anxiety and a dyspneic sensation, particularly during hypercapnia, one cannot deny the possibility that the increase in swallowing frequency during airway occlusion may be due to emotional arousal. Another possibility comes from the consideration of a large negative pressure in the upper airway during inspiration that would lead to airway deformation and stimulation of many receptors in the upper airway. Upper airway receptors that initiate swallowing have not been identi?ed histologically, but slowly adapting receptors responding to water and tactile sensation may be responsible for initiating swallow from the upper airway [11].
  <para2>The marked increase in swallowing frequency during airway occlusion at hypercapnia may indicate that the swallowing center is activated rather than inhibited by hypercapnia in the absence of phasic vagal in?uence. Although the above-mentioned behavioral/emotional responses may, in part, contribute to the activation of the swallowing center, it is also possible that this activation of the swallowing center may be associated with the simultaneous activation of the respiratory neurons. In this context, it is perhaps no accident that the swallowing center is located at the same site of the brainstem in which the respiratory pattern generator is located. In fact, there is much evidence to indicate that, within the dorsal and ventral medulla, there exists a common pool of neurons that may have a multifunctional role [12-14]. It is possible that some neurons in the respiratory
pattern generator may participate in activities of swallowing. Thus, it is conceivable that an increase in
respiratory drive due to hypercapnia activates not only the respiratory pattern generator but also the swallowing center.
  <para2>Several studies [15-18] have shown that swallowing tends to occur preferentially during the expiratory phase of the respiratory cycle during eupneic conditions in normal adult human subjects. In a previous study [4], we have also shown that the preponderant coupling of swallows with expiratory phase is lost during hypercapnia. In the present study, we con?rmed these previous observations and showed a marked difference in the timing of swallows between normocapnia and hypercapnia before airway occlusion. Despite the considerably different distributions of timing of swallows at normocapnia and hypercapnia before airway occlusion, the distribution of the timing of swallows at these
two different conditions became quite similar during airway occlusion. These observations may suggest that CO2 cannot be a factor that determines the timing of swallows in relation to the phase of the respiratory cycle. On the other hand, the phasic vagal in?uence may be a crucial factor determining the timing of swallows. Thus, it is likely that the difference in the timing of swallows observed between normocapnia and hypercapnia may be due to the different degrees of phasic vagal inputs.
  <para2>A simple extrapolation of our results to clinical situations may not be entirely valid, because hypercapnia does not always accompany hyperpnea in clinical situations. For example, hypercapnia is frequently caused by hypoventilation due to depression of the central nervous system. However, in this situation, the depression of the central nervous system may decrease not only respiratory activity but also the activity of the swallowing center, and the interpretation of the effect of hypercapnia on the swallowing re?ex is very dif?cult. It is also possible that hypercapnia may be produced by changes in respiratory mechanics even in the absence of depression of the central nervous system. In this context, it has been reported that, in conscious subjects, the addition of an external resistive load caused a decrease in minute ventilation with a concomitant increase in PETco2, but that no change in the swallowing rate occurred in response to the continuous infusion of water [19]. Thus, it is clear that the swallowing re?ex is not attenuated during hypercapnea in the absence of hyperpnea.
  <para2>Obviously, respiration and swallowing cannot coexist in the upper airway, and the two activities must be coordinated to allow maximal operation of one function without compromising the other function. It has been shown that hypercapnia not only attenuates the swallowing re?ex but also changes the timing of swallows, and thereby enhances the chance of laryngeal irritation [4]. The results of the present study strongly suggest that the attenuation of the swallowing re?ex during hypercapnia is not due to the direct, inhibitory effect of CO2 on the swallowing center, but, rather, is due to the increased inhibitory in?uence of a lung-volume-related re?ex.
<ACK>Acknowledgments. This study was supported in part by Grant-in-Aid for Cancer Research (11-1) from the Ministry of Health, Labour and Welfare of Japan.
<A>References
<REF> 1. Nishino T, Yonezawa T, Honda Y (1981) Modi?cation of laryngospasm in response to changes in PaCO2 and PaO2 in the cat. Anesthesiology 55:286-291
<REF> 2. Lawson EE (1982) Recovery from central apnea: effect of stimulus duration and end-tidal CO2 partial pressure. J Appl Physiol 53:105-109
<REF> 3. Nishino T, Hiraga K, Honda Y (1989) Inhibitory effects of CO2 on airway defensive re?exes in en?urane-anesthetized humans. J Appl Physiol 66:2642-2646
<REF> 4. Nishino T, Hasegawa R, Ide T, Isono S (1998) Hypercapnia enhances the development of coughing during continuous infusion of water into the pharynx. Am J Respir Crit Care Med 157:815-821
<REF> 5. Wilson SL, Thach BT, Brouillette RT, Abu-Osba YK (1981) Coordination of breathing and swallowing in human infants. J Appl Physiol 50:851-858
<REF> 6. Nishino T, Sugimori K, Kohchi A, Hiraga K (1989) Nasal constant positive airway pressure inhibits the swallowing re?ex. Am Rev Respir Dis 140:1290-1293
<REF> 7. Kijima M, Isono S, Nishino T (2000) Modulation of swallowing re?ex by lung volume changes. Am J Respir Crit Care Med 162:1855-1858
<REF> 8. Yamamoto F, Nishino T (2002) Phasic vagal in?uence on the rate and timing of re?ex swallowing. Am J Respir Crit Care Med 165:1400-1403
<REF> 9. Sant'Ambrogio G (1982) Information arising from the tracheobronchial tree of mammals. Physiol Rev 62:531-569
<REF>10. Fonagy P, Calloway SP (1986) The effect of emotional arousal on spontaneous swallowing rates. J Psychosom Res 30:183-188
<REF>11. Storey AT (1968) Laryngeal initiation of swallowing. Exp Neurol 20:359-365
<REF>12. Sumi T (1963) The activity of brainstem respiratory neurons
and spinal respiratory motoneurons during swallowing. J Neurophysiol 26:466-477
<REF>13. Larson CR, Yajima Y, Ko P (1994) Modi?cation in activity of medullary respiratory-related neurons for vocalization and swallowing. J Neurophysiol 71:2294-2304
<REF>14. Shiba K, Satoh I, Kobayashi N, Hayashi F (1999) Multifunctional laryngeal motoneurons: an intracellular study in the cat. J Neurosci 19:2717-2727
<REF>15. Nishino T, Yonezawa T, Honda Y (1985) Effects of swallowing on the pattern of continuous respiration in human adults. Am Rev Respir Dis 132:1219-1222
<REF>16. Selley WG, Flack FC, Ellis RE, Brooks WA (1989) Respiratory patterns associated with swallowing: part 2. Neurologically impaired dysphagic patients. Age Ageing 18:173-176
<REF>17. Smith J, Wolkove N, Colacone A, Kreisman H (1989) Coordination of eating, drinking and breathing in adults. Chest 96:578-582
<REF>18. Preiksaitis HG, Mayrand S, Robins K, Diamant NE (1992) Coordination of respiration and swallowing: effect of bolus volume in normal adults. Am J Physiol 263:R624-R630
<REF>19. Kijima M, Isono S, Nishino T (1999) Coordination of swallowing and phase of respiration during added respiratory loads in awake subjects. Am J Respir Crit Care Med 159:1898-1902

ページのTOPへ


<JN>J Anesth (2004) 18:89-93
<PT>
<CT>Antagonizing potencies of saturated and unsaturated long-chain free fatty acids to iso?urane in gold?sh
<CA>Ryuzo Hanada, Tsuneo Tatara, and Yasuhide Iwao
<ADD>Department of Anesthesiology, Kyorin University School of Medicine, 6-20-2 Shinkawa, Mitaka, Tokyo 181-8611, Japan
<AB>Abstract
<AB>Purpose. We have previously reported that myristate, a saturated free fatty acid (FFA) with 14 carbons (C14), antagonizes volatile anesthetics in gold?sh. The hydrophobicity and molecular con?guration of FFAs may play an important role in the antagonizing effect. To examine their contribution, we investigated the antagonizing potencies of saturated and
unsaturated long-chain FFAs in gold?sh.
<AB>Methods. Saturated and monounsaturated FFAs of C14-18 were tested. We determined the anesthetic concentration
producing a 50% effect (EC50) of iso?urane in the absence or presence of FFA by observing the escape reaction of gold?sh against an electrical stimulus.
<AB>Results. All FFAs increased the EC50 of iso?urane dose-dependently compared with reactions in the absence of FFA (P , 0.05). For saturated FFAs, the relationship between chain lengths and antagonizing potencies was not linear. C18 was the most effective and C16 was the least effective antagonist (P , 0.05). Among unsaturated FFAs, C14 was the most effective antagonist (P , 0.05). In a comparison of saturated and unsaturated FFAs, saturated C14 and C18 were more effective antagonists than unsaturated FFAs of the same
carbon numbers (P , 0.05).
<AB>Conclusion. The hydrophobicity of FFAs increases as the chain length increases. Therefore, our ?ndings suggest that the antagonizing effect of long-chain FFAs in gold?sh, in terms of their capacity to perturb the lipid membrane structure, may be determined not solely by their hydrophobicity but also by their molecular con?guration.
<KW>Key words Free fatty acid ・ Volatile anesthetics ・ Antagonist
<A>Introduction
<para1>Free fatty acids (FFAs) play an important role in cellular signaling by acting as messengers [1]. Long-chain FFAs (chain length between 14 and 18 carbons) directly regulate the activity of speci?c ion channels and receptors, such as potassium channels, estradiol, and muscarinic receptors [1-3]. In recent years, oleamide, the amide derivative of oleic acid, has attracted attention as an endogenous sleep-inducing substance by potentiating A-aminobutyric acidA (GABAA) receptor function [4]. Furthermore, lipid emulsions, oil-in-water formulations clinically used as drug carriers, have been reported to affect N-methyl-d-aspartate (NMDA) receptor activity [5]. Because GABAA and NMDA receptors are considered to be possible acting sites of general anesthetics [6,7], these ?ndings suggest that FFAs may modulate anesthetic action. However, little work has been
performed to explore the effect of FFAs on anesthetic potency in vivo.
  <para2>Tatara et al. [8] reported that myristate (saturated free fatty acid with 14 carbons, saturated C14) increased the anesthetic concentration producing a 50% effect (EC50) of volatile anesthetics in gold?sh. That was the ?rst report to demonstrate the action of an FFA on anesthetic potency in vivo. A previous in vitro study, using ?re?y luciferase, showed that anesthetics decreased the thermal transition temperature of ?re?y luciferase by relaxing its molecular conformation, whereas saturated C14 increased the transition temperature by tightening the conformation [9]. This contrasting effect of anesthetics and saturated C14 may explain the antagonizing effect of saturated C14 in gold?sh. Furthermore, Matsuki et al. [10] reported that this tightening effect was observed in saturated FFAs with a carbon chain longer than 10. The effects were enhanced as chain length increased, suggesting that the hydrophobicity of FFAs may play an important role in the tightening of molecular conformation and thus anesthetic antagonism.
  <para2>Based on these earlier ?ndings, we investigated the antagonizing potencies of saturated and monounsaturated C14-18 to iso?urane in gold?sh in order to examine the contribution of the hydrophobicity and the molecular con?guration of FFAs to their antagonizing potency in vivo. The presence of a double bond in the carbon chain may affect antagonizing potencies because it changes the molecular con?guration. We used gold?sh to determine anesthetic potency because of (1) easy test administration, (2) inexpensiveness, and (3) commercial availability at any season [11].
<A>Materials and methods
<para1>This study was approved by the Animal Experimentation Committee of Kyorin University. Gold?sh, 4 to 6 cm in body length, were purchased from Edogawa Fish Farm (Tokyo, Japan) and acclimated at 28°C in aerated tap water for at least 1 week before the experiment. All chemicals were purchased from Sigma (St. Louis, MO, USA). Iso?urane was obtained from Abbott (North Chicago, IL, USA). The following FFAs were studied (structural formulas shown in Fig. 1); Na-myristate (saturated C14), myristoleic acid (unsaturated C14), Na-palmitate (saturated C16), palmitoleic acid (unsaturated C16), Na-stearate (saturated C18), and oleic acid (unsaturated C18). Because long-chain FFAs are sparingly soluble in water, we made stock solutions in methanol. The ?nal concentration of methanol was adjusted to 5 mM. Control studies without FFA were performed in water with the same concentration of methanol. Because the ?nal concentration of methanol in water was much lower than the EC50 of methanol (590 mM) [12], the effect of dissolved methanol on the EC50 of iso?urane was considered to be negligible.
  <para2>We used a glass tank 200 mm in diameter and 90 mm in depth for the experiment. A pair of circular stainless steel screens 180 mm in diameter were placed on the surface and the bottom to apply electrical stimuli
by a constant current generator (Tokushu Keisoku, Yokohama, Japan). The threshold of electrical current intensity at which the gold?sh showed escape motion was determined. Because all gold?sh responded to an electrical current of 8.0 mA and no adverse residual effect was observed after stimulation, we adopted 8.0 mA of square-wave current for the electric stimulation throughout the experiments.
  <para2>Before each experiment, gold?sh were placed in a bucket with distilled water containing 5, 10, or 20 IM of each FFA for 2 h. A suspension micelle was formed at a high concentration of FFA. Saturated C18, the most hydrophobic FFA, became muddy at over 20 IM, so the concentration of each FFA was maintained at less than 20 IM. The glass tank was placed on a hot stirrer (AS ONE, Osaka, Japan), and the temperature of the water was maintained at 21 6 1°C throughout the experiment. Iso?urane was bubbled for 30 min and stirred in the glass tank, which contained 3200 ml of distilled water with the same concentration of each FFA as in the bucket. Ten gold?sh were placed in the test tank and exposed to the anesthetic for 30 min. Then, they were electrically stimulated at 8.0 mA for 200 ms four times, with a 10-s interval between stimulations. Each gold?sh was evaluated as unanesthetized when it showed an escape motion at least once. The movements of each gold?sh were recorded with a digital video camera, and the number of gold?sh in an unanesthetized state was
rechecked. We used 20 gold?sh for each concentration of iso?urane.
  <para2>The concentration-response curves of iso?urane with or without FFA were obtained according to the following equation:
<EQ>y 5 1/(1 1 (p/EC50)n)
<para1>where y is the fraction of gold?sh responding to the stimulus, p is the concentration of iso?urane (%), and n is the slope of the curve [13]. The EC50 was calculated by logistic regression with GraphPad Prism 4 (GraphPad Software, San Diego, CA, USA). The EC50 values in the presence of FFA were compared with that in the control using the extra sum-of-squares F test [14]. P , 0.05 was considered statistically signi?cant. Data are presented as the EC50 6 SE.
<A>Results
<para1>The gold?sh showed no remarkable adverse effects, such as toxicity or agitation in response to the FFAs. The application of each FFA before the administration of iso?urane did not affect the escape reaction of gold?sh to an electrical stimulus. All gold?sh apparently recovered after the experiments.
<B>Concentration-response curves of iso?urane
<para1>The concentration-response curves of iso?urane with saturated or unsaturated FFAs are shown in Figs. 2 and 3. As the concentration of iso?urane increased, the fraction of gold?sh responding to the stimulation decreased. The concentration-response curves thus were sigmoid. All FFAs shifted the curve in the control to the right in a concentration-dependent manner. In the presence of saturated C18 at a concentration higher than 10 IM, one-third of the gold?sh showed an escape reaction, even at 5% iso?urane, whereas no gold?sh in the control exhibited an escape reaction at the same iso?urane concentration.
<B>Effects of FFA on the EC50 of iso?urane
<para1>The changes of EC50 values in the presence of 5, 10, or 20 IM FFAs are shown in Table 1. All FFAs signi?cantly increased the EC50 of iso?urane, except for 5 IM unsaturated C16. For the saturated FFAs, it was found that the relationship between the chain length of the FFA and the antagonizing effect was not linear. C18 increased the EC50 most effectively, to about 2.5 times that of the control, at concentrations higher than 10 IM. C16, the weakest in the saturated group, increased the EC50 about 1.5-fold. For the unsaturated FFAs, C14 increased the EC50 of iso?urane most effectively, to about twice that of the control, at 20 IM. Unsaturated C16 and C18, which were weaker than C14, increased the EC50 of iso?urane about 1.5-fold. In a comparison of the EC50 values between saturated and unsaturated FFAs, we found that saturated C14 and C18 were signi?cantly more effective than the unsaturated group of the same carbon numbers.
<A>Discussion
<para1>Both saturated and monounsaturated FFAs of C14-18 antagonized iso?urane in gold?sh. Our results suggest that the antagonizing effect to volatile anesthetics in gold?sh may be a nonspeci?c characteristic of long-chain FFAs. It is unlikely that iso?urane in water is depleted by partitioning into FFAs, because iso?urane was constantly bubbled into water during the experiments. Therefore, FFAs may affect the active site or mode of action of iso?urane in gold?sh and thus decrease its potency.
  <para2>FFAs bind some kinds of proteins in living bodies, such as albumin [15] and fatty acid binding proteins [16]. There are many hydrophobic cavities at the ・-helices in the subdomain of serum albumin, and myristate binds these pockets [17]. NMDA regions have regions similar to fatty acid binding proteins in their known amino acid sequences, and these regions may be involved in the binding of FFAs and the modulation of the receptor activity by FFAs [18]. These studies suggest that membrane proteins may be the targets of FFAs in their modulation of the action of volatile anesthetics in gold?sh, although we could not identify the exact binding site of FFAs in the present study.
  <para2>On the assumption that the targets of FFAs are
membrane proteins, two possible mechanisms can
explain the antagonizing potency of FFAs in gold?sh. First, FFAs may compete for a speci?c binding site of anesthetics on channel proteins [19]. However, it is unlikely that negatively charged FFAs bind the same site as uncharged iso?urane. Second, FFAs may change the conformation of membrane proteins and modulate the action of iso?urane. This scenario may be supported by the contrasting actions on the structural change of ?re?y luciferase between volatile anesthetics and FFAs in vitro [9,10].
  <para2>Saturated C18 was the most effective antagonist among the saturated FFAs tested in the present study.
If hydrophobicity alone determines the antagonizing potencies of FFAs, the order of tightening effects on molecular conformation should be C14 , C16 , C18, as observed in the luciferase study, because FFA becomes more hydrophobic as the carbon chain length increases [10]. However, saturated C16 showed less effectiveness than saturated C14 in gold?sh. This discrepancy may be explained by the difference between luciferase and membrane proteins. Luciferase is a water-soluble protein, and membrane proteins are bound with hydrophobic lipids. Modi?cation of fatty acid levels in
the membrane changes the function of membrane-
associated proteins, including transporters, receptors, enzymes, and signaling molecules [20]. The activities of membrane proteins are regulated by the carbon chain length of the lipid bilayer [21], and there may exist a speci?c lipid thickness for optimum activity. Assuming that not only a direct action on membrane proteins but also partitioning into the lipid bilayer participate in the antagonism exerted by FFAs, it is not unreasonable that the order of antagonism may not rigidly correspond to the hydrophobicity level of the FFAs.
  <para2>Saturated C14 and C18 were more effective antagonists than were the monounsaturated types. Reports on the anesthetic actions on GABAA receptors have shown different effects between saturated and unsaturated FFAs. The binding of diazepam or muscimol to GABAA receptors was more greatly enhanced by unsaturated FFAs than by saturated types [22]. These effects depend on the length of the carbon chain, and FFAs with a medium chain length were the most effective [23]. In the present study, all the unsaturated FFAs were cis type. The cis double bond placed centrally in the carbon chain bends the molecular con?guration of FFAs. The double-bonded FFAs may perturb the packing of lipid bilayer structures [24] and decrease the
activity of membrane proteins. This indirect action of unsaturated FFAs via the lipid bilayer may weaken the tightening effect of FFAs and may be the cause of the weaker antagonism than that of the saturated FFAs.
  <para2>At present, we have no clinical antagonists of volatile anesthetics. FFAs are contained in our food in the form of glyceryl esters and essential nutrients. Our result suggests that FFAs are a safe, new type of antagonist of volatile anesthetics. To investigate their antagonizing potency and the relevant concentration of FFAs, further studies are needed, using mammalian subjects.
  <para2>In conclusion, we demonstrated that saturated and monounsaturated long-chain FFAs of C14-18 antagonized iso?urane in gold?sh. For saturated FFAs, the relationship between chain lengths and antagonizing potencies was not linear. C18 was the most effective and C16 was the least so. Among unsaturated FFAs, C14 was the most effective antagonist. Saturated C14 and C18 were more effective antagonists than unsaturated FFAs of the same carbon numbers. These ?ndings suggest that the antagonizing effect of long-chain FFAs, in terms of their ability to perturb the lipid membrane structure, was determined not solely by their hydrophobicity but also by their molecular con?guration.
<ACK>Acknowledgments. We are indebted to Professor Issaku Ueda (University of Utah) and Dr. Kuniaki Nakanishi (National Defense Medical College) for their kind advice regarding this study. We also wish to thank Chikako Okada for gold?sh care.
<A>References
<REF> 1. Ordway RW, Singer JJ, Walsh JV Jr (1991) Direct regulation of ion channels by fatty acids. Trends Neurosci 14:96-100
<REF> 2. Hanley MR (1978) Crotoxin effects on Torpedo californica cholinergic excitable vesicles and the role of its phospholipase A activity. Biochem Biophys Res Commun 82:392-401
<REF> 3. Okun IM, Merezhinskaya NV, Rakovich AA, Volkovets TM, Aksentsev SL, Konev SV (1986) Inactivation of muscarinic acetylcholine receptors in brain synaptic membranes by free fatty acids. Evaluation of the role of lipid phase. Gen Physiol Biophys 5:243-258
<REF> 4. Nicholson RA, Zheng J, Ganellin CR, Verdon B, Lees G (2001) Anesthetic-like interaction of the sleep-inducing lipid oleamide with voltage-gated sodium channels in mammalian brain. Anesthesiology 94:120-128
<REF> 5. Weigt HU, Georgieff M, Beyer C, Fohr KJ (2002) Activation of neuronal N-methyl-d-aspartate receptor channels by lipid emulsions. Anesth Analg 94:331-337
<REF> 6. Tanelian DL, Kosek P, Mody I, Maclver MB (1993) The role of the GABAA receptor/chloride channel complex in anesthesia. Anesthesiology 78:757-776
<REF> 7. Yang J, Zorumski CF (1991) Effect of iso?urane on N-methyl-d-aspartate gated ion channels in cultured rat hippocampal neurons. Ann NY Acad Sci 625:287-289
<REF> 8. Tatara T, Kamaya H, Ueda I (2002) Myristate, a 14-carbon fatty acid, effectively reverses anesthesia. Anesthesiology 97:518-520
<REF> 9. Ueda I, Suzuki A (1998) Irreversible phase transition of ?re?y luciferase: contrasting effects of volatile anesthetics and myristic acid. Biochim Biophys Acta 1380:313-319
<REF>10. Matsuki H, Suzuki A, Ueda I (1999) Speci?c and non-speci?c binding of long-chain fatty acids to ?re?y luciferase: cutoff at octanoate. Biochim Biophys Acta 1426:143-150
<REF>11. Selye H (1943) The ?sh assay for the anesthetic effect of the steroids. Anesthesiology 4:36-47
<REF>12. Ali?moff JK, Firestone LL, Miller KW (1989) Anaesthetic potencies of primary alkanols: implications for the molecular dimensions of the anaesthetic site. Br J Pharmacol 96:9-16
<REF>13. Waud DR (1972) On biological assays involving quantal responses. J Pharmacol Exp Ther 183:577-607
<REF>14. Motulsky H, Christopoulos A (2003) Fitting models to biological data using linear and nonlinear regression. GraphPad Software, San Diego, pp 138-142
<REF>15. Ashbrook JD, Spector AA, Santos EC, Fletcher JE (1975) Long chain fatty acid binding to human plasma albumin. J Biol Chem 250:2333-2338
<REF>16. Bakowies D, van Gunsteren WF (2002) Simulations of apo and holo-fatty acid binding protein: structure and dynamics of protein, ligand and internal water. J Mol Biol 315:713-736
<REF>17. Curry S, Mandelkow H, Brick P, Franks N (1998) Crystal structure of human serum albumin complexed with fatty acid reveals an asymmetric distribution of binding sites. Nat Struct Biol 5:827-835
<REF>18. Petrou S, Ordway RW, Singer JJ, Walsh JV Jr (1993) A putative fatty acid-binding domain of the NMDA receptor. Trends Biochem Sci 18:41-42
<REF>19. Franks NP, Lieb WR (1993) Selective action of volatile general anesthetics at molecular and cellular levels. Br J Anaesth 71:65-76
<REF>20. Johnson RA, Hamilton JA, Worgall TS, Deckelbaum RJ (2003) Free fatty acids modulate intermembrane traf?cking
of cholesterol by increasing lipid mobilities: novel 13C NMR analyses of free cholesterol partitioning. Biochemistry 42:1637-1645
<REF>21. Mouritsen OG, Bloom M (1984) Mattress model of lipid-protein interactions in membranes. Biophys J 46:141-153
<REF>22. Koenig JA, Martin IL (1992) Effect of free fatty acids on GABAA receptor ligand binding. Biochem Pharmacol 44:11-15
<REF>23. Witt MR, Nielsen M (1994) Characterization of the in?uence of unsaturated free fatty acids on brain GABA/benzodiazepine receptor binding in vitro. J Neurochem 62:1432-1439
<REF>24. Rehberg B, Urban BW, Duch DS (1995) The membrane lipid cholesterol modulates anesthetic actions on a human brain ion channel. Anesthesiology 82:749-758
<FIG>

ページのTOPへ


<JN>J Anesth (2004) 18:94-99
<PT>
<CT>Multi-Unit and Multi-Path system of the neural network can explain the steep dose-response of MAC
<CA>Yoshiroh Kaminoh1, Hiroshi Kamaya2, Chikara Tashiro1, and Issaku Ueda2
<ADD>1 Department of Anesthesiology, Hyogo College of Medicine, 1-1 Mukogawa-cho, Nishinomiya 663-8501, Japan
<ADD>2 Department of Anesthesiology, University of Utah School of Medicine, Salt Lake City, UT, USA
<AB>Abstract
<AB>Purpose. The slope of the dose-response curves of inhalation anesthetics is steep around the minimum alveolar concentration of inhalation anesthetics (MAC) value. Contrastingly, the anesthetic dose-response curves of ion channels and
enzymes are gradual. This discrepancy in the steepness may
be a key to solve the mechanisms of anesthesia. To explain the steepness we propose a mathematical model of the neural network related to MAC.
<AB>Methods. We assumed that, in order to show movement in response to a noxious stimulus, a signal needed to be transmitted from A to B. There are m conduction pathways (Multi-Path) in the nerve network between A and B, and there are n conduction units (Multi-Unit) in each conduction pathway. Anesthetics bind to each conduction unit and block signal transmission. Anesthetics prevent movement in response to a stimulus, when at least one conduction unit among all conduction pathways has been blocked. We derived the equation for the probability of the signal being blocked by anesthetics.
<AB>Results. The steep dose-response curve of in vivo anesthesia requires a very large number of conduction units (n . 100) and conduction pathways (m . 106). The EC50 for each conduction unit was at least 3.8-fold larger than the apparent EC50 for the whole system under the experimental condition of simulation.
<AB>Conclusions. We constructed a model for the neural networks that relates to MAC as a Multi-Unit and Multi-Path system (MUMPS). To obtain highly cooperative dose-
response curves comparable to those of in vivo anesthesia, at least 106 conduction pathways and more than 100 conduction units are required for each pathway. In these systems, the apparent anesthetic potency on the whole system (MAC) is much stronger than the anesthetic action on each unit. Because of this discrepancy, it is important to set anesthetic concentrations appropriately for experiments with in vitro systems.
<KW>Key words Dose-response curve ・ Anesthesia mechanism ・ MAC
<A>Introduction
<para1>The MAC (minimum alveolar concentration of inhalation anesthetics) is recognized as the standard mea-
sure of anesthetic potency, because immobilization by
inhalation anesthetics to noxious stimuli is the most desirable and obvious endpoint of anesthesia [1,2].
Anesthesia mechanisms have been extensively studied using physical chemistry, molecular modeling, physiology, pharmacology, and whole-animal experiments [2].
  <para2>The slope of the dose-response curves of inhalation anesthetics is steep around the MAC value. These steep dose-response curves are analyzed by the logistic plot, and the slope is calculated by the Waud equation [3]. The slope values of clinical anesthesia are often expressed by the Hill coef?cient (nH), and are in two-digit values ranging from about 6 to 20 [4]. The Hill equation is a convenient method to determine the steep dose-response curve, but it is often misused. The steep dose-response curves guarantee that, in clinical anesthesia practice, almost all patients become immobilized. Contrastingly, the anesthetic dose-response curves of ion channels and enzyes are gradual, and the nH ~ value is usually less than 3 [4-6].
  <para2>The discrepancy in the steepness between whole-
animal and in vivo study systems may be due to the lack of a method for functional analysis of the neural
network, and such a method will be a key to solve the missing information on anesthesia mechanisms. The present communication deals with the derivation of a mathematical model of the neural network, and examines the effect of an anesthetic on this model.
<A>Methods
<para1>We propose the following model for the neural network between points A and B. It is assumed that there are m conduction pathways (Multi-Path) in the network, and n conduction units (Multi-Unit) in each conduction pathway. Patients do not respond to noxious stimuli when the conduction between A and B is lost.
  <para2>Figure 1 shows a schematic presentation of the proposed mathematical model. Anesthetics affect each conduction unit. When an anesthetic molecule binds to that conduction unit, conduction by the unit is interrupted (Fig. 1A). When one of the n conduction units in each conduction pathway is interrupted, the conduction by that pathway is blocked (Fig. 1B). Additionally, we assume that the patient responds to noxious stimuli when the conduction is maintained at least in one among m conduction pathways. Immobilization ensues when the conduction is totally shutdown (Fig. 1C).
  <para2>An equilibrium constant K determines the possibility of inducing anesthesia.
<EQ> (1)
<para1>where A is the anesthetic, E is the awake state, and R is the resting state (R state, unable to conduct) of the conduction unit.
  <para2>The probability p of each conduction unit being in the R state is:
<EQ> (2)
<para1>where EC50unit is the anesthetic concentration that inhibits 50% of the conduction unit.
  <para2>The probability that the i-th pathway fails to conduct, Pi, is:
<EQ> (3)
  <para2>Therefore, the probability, Y, that all of the conduction roots are incapable of signal conduction (immobilized) is:
<EQ> (4)
  <para2>In this highly complicated system with many conduction units and pathways, there is a possibility that the anesthetic potency on each conduction unit (EC50unit) may not match the apparent anesthetic potency on the total system (EC50system). Equation 5 is obtained by setting
Y 5 1/2 and [A] 5 EC50system:
<EQ> (5)
  <para2>To evaluate the steepness of the dose-response curve, we also calculated the ratio of EC50system and EC95system (effective concentration of 95%):
<EQ> (6)
<A>Results
<B>Conformational changes in the dose-response curve
<para1>Figure 2 shows the effect of the number of conduction units (n) when there is one conduction pathway (m 5 1). An increase in the number of conduction units
decreases the conformation of the dose-response curve. When n exceeds 100 (n . 100), the change becomes negligible. This indicates that the variation in n has
little effect on the conduction pathway. The intense cooperativity (Hill number, [nH] above 20) of clinical anesthesia cannot be achieved by an increase of n
values.
  <para2>Figure 3 shows the effect of the number of conduction pathways (m) when there is one conduction unit in each conduction pathway (n 5 1). Here again, the conformation of the dose-response curve stayed unchanged
when the number exceeded 100 (m . 100). A high cooperativity, comparable to in vivo dose-response curves, could not be obtained by an increase in m values either.
  <para2>Figure 4 shows the dose-response curves when both the number of conduction units and the number of pathways are varied. The conformation of the dose-response curves became steeper with the increase in both parameters. When both numerical values exceeded 104, the dose-response curves became comparable to those for in vivo anesthesia.
<B>ED95: the 95% effective concentration
<para1>When the Hill number is 20, the 95% effective concentration is calculated to be 1.16-fold of MAC. The MAC95 of sevo?urane was also reported to be 1.16 of MAC [7]. Table 1 shows the ratio of EC50system and EC95system for the combination of conduction units and pathways (EC95system/EC50system; these values were calculated from Eq. 6). The shadowed area in Table 1 signi?es the EC95system/EC50system values that match the in vivo values (1.11-1.21, where we allowed 5% error). The shadowed area shows the number of units above 100 and the number of pathways above 106. The results indicate that in vivo anesthesia requires very large numbers of conduction units and pathways.
<B>Anesthetic potency on the conduction unit and
the total system
<para1>Figure 5 shows the ratio between the conduction unit and the whole system (EC50unit/EC50system) when their numbers are altered. The vertical axis is plotted in logarithmic scale. In the positive region (EC50unit . EC50system) the anesthetic potency on the conduction unit is smaller than the apparent anesthetic potency on the whole
system. This means that the action of anesthetics on the system is strong despite the anesthetic effect on each unit being weak.
  <para2>When the number of conduction units (n) increases, anesthetic action on the whole system becomes more potent. This is because, when one unit fails, each pathway fails to conduct (Fig. 5; arrow A). In contrast, an increase in the number of conduction pathways (m) makes the system less vulnerable to interruption (Fig. 5; arrow B).
  <para2>Area C in Fig. 5 is calculated from the combination of conduction units and conduction pathways in which the EC95system/EC50system values match the in vivo values (1.11-1.21), and shows that, under these conditions, the log (EC50unit/EC50system) is always positive; also the value for EC50unit/EC50system becomes 3.8 3 100 (m 5 1010 and n 5 103) to 7.1 3 108 (m 5 1010 and n 5 1010). Even though the anesthetic effect on each conduction unit is small, the overall effect on the system is large.
<A>Discussion
<para1>Eckenhoff and Johansson [8] assumed that mobilization was maintained by multiple elements. They constructed a model composed of multiple elements, where anesthetic effects on each element may be small, but the composite effect may become large. By considering
that the actual Hill number of in vivo studies is in the range of 20, they concluded that in vitro EC50 values could be larger than the actual MAC. The EC50 values of in vitro systems do not necessarily match the EC50 values of in vivo systems. With in vitro studies, the
EC50 values can be higher than MAC. The curves
shown in Fig. 1 in the article by Eckenhoff and Johansson [8] were generated by the Hill equation,
with changes of the EC50 and the Hill number, without
a description of the relationship between Kd and EC50.
It is a mystery why EC50 is 50 IM and the Hill number
is 2, when Kd is 1 mM and ten different sites exist with equal additive effects (curve b in the same ?gure). We notice that the Hill equation is often used for a steep dose response curve, but it is often misused, being
used just as a curve-?tting procedure in most cases. The conclusion that there are possible dissociations be-tween MAC and potency at the action sites deserves attention [8], despite the ?nding that the Hill equation failed to serve an apparent model of anesthesia
mechanism [9].
  <para2>Yamakura et al. [4] and Eger's group [10] assumed that the in vitro actions of anesthetics on each element were a non-quantal (analog) response. It is possible to convert a non-quantal response to quantal by setting
a certain borderline and converting it to population distribution (hit or no-hit). Figure 1 in the article by Yamakura et al. [4] shows that the Hill number of the dose-response curves for six oocytes was 1.25. When the data were converted to a categorical response the number increased to 8.4. They stated that the slope of categorical responses for each individual does not depend on the Hill value of the underlying dose-response relationship, but depends on the population distribution. Thus, the steep slope of the MAC response may not necessarily be related to mechanisms of anesthetic
actions on targets [4].
  <para2>Yamakura et al. [4] assumed that the threshold was 0.5, but there is no reason for this assumption. Eger's group [10] discussed the details of the threshold value, and concluded that the threshold should lie within values of 0.1 to 0.9. But this conclusion was not based on necessity, either. They stated that the Hill coef?cient of the receptor probably exceeded 2, when the threshold was less than 0.1 or greater than 0.9 [10]. There seemed to be no clear reason for this. Furthermore, steep dose-response curve of anesthetics in a non-quantal response have been reported [9].
  <para2>The observed response of oocytes is a result of the opening and closing of ion channels. The open or closed state of channels is quantal, and the response rates
represent the mean value of the probability of an open (or closed) state of every channel on each oocyte. The obtained dose-response curves were already converted from a quantal parameter to a non-quantal parameter. Yamakura et al. [4] and Eger's group [10] converted this non-quantal parameter to quantal one by categorization. This categorization should be done for each oocyte ?rst, using a binomial distribution, and then the probability should be obtained (see Appendix). However, we know neither how much is the threshold value nor how many channels are in each oocyte.
  <para2>We analyzed the in vivo effects of anesthetics by constructing a model composed of two variables: a conduction unit and a conduction path. We found that the minimum requirements to obtain high cooperativity, which is speci?c to in vivo anesthesia, were: the number of conduction units had to be above 100 and the number of conduction pathways had to be above 106. This indicates that a total of 108 units (100 3 106) are required to obtain steep dose-response curves. The appropriateness is considered below.
  <para2>First, we stress that the unit number (n) does not represent the number of neurons. By considering that anesthesia becomes effective locally when part of the neuron is affected, it can be seen that there are many conduction units in a neuron.
  <para2>The number of conduction paths (m) in our model represents linear pathways (see Fig. 1). In reality however, neurons ramify into branches and form networks. In these complex systems, it is possible to form many conduction pathways in a single neuron. The assumption of 106 conduction paths is not unrealistic. To construct a model with branched neurons means the addition of another variable and only makes the analyses more dif?cult, without practical bene?t for the analyses.
  <para2>When the strength of the anesthetic effect on the conduction unit (EC50unit) and the whole system (EC50system 5 MAC) is compared, the contribution of EC50unit is
much stronger than that of EC50system under the high cooperativity condition. The numerical value was estimated to be 3.8 3 100 to 7.1 3 108. At clinical anesthetic concentrations, the anesthetic effect on the conduction unit may be very small. The concentrations of aqueous volatile anesthetics in equilibrium with MAC in the gas phase are in the range of 0.19-0.64 mM [11]. Therefore, the EC50 for conduction units requires high anesthetic concentrations, of 1 mM or more. The eight-digit difference between the EC50unit and EC50system values was the mathematical value calculated from the model, and may be almost meaningless when we discuss the actual anesthetic potency on the conduction unit. However, these ?ndings suggested that the anesthetic concentrations for studies on channels, enzymes, membrane proteins, etc. require reconsideration. The effects of these high anesthetic concentrations suggest that the anesthetic effects are directed to the physical properties of the membranes as well as to high-af?nity binding sites.
<ACK>Acknowledgments. This study was supported by a Grant-in-Aid for Scienti?c Research of the JSPS and Veterans Affairs Medical Research.
<A>Appendix
<B>Categorization of the non-quantal response of oocytes
<para1>The response rate of the i-th oocyte at the anesthetic concentration [A], Poi([A]) is the mean value of the probability of the open (or closed) state of each
channel.
<EQ> (A-1)
<para1>Here, pci,j([A]) is the probability of the open (or closed) state of the j-th channel of the i-th oocyte, and nci is the number of channels on the i-th oocyte. This equation represents the dose-response curve of each oocyte (for example, Fig. 1 of Yamakura et al. [4] and Fig. 5 of Eger et al. [10]), and the probability of the quantal phenomenon was converted to the non-quantal parameter, the response rate.
  <para2>Yamakura et al. [4] and Eger's group [10] categorized the non-quantal response of each oocyte by threshold, and the probability of the response of oocytes (or immobility), PIM([A]), is
<EQ> (A-2)
<para1>Here, Th is the threshold value of the response, and
N is the number of oocytes which have been examined. The if (BOOLEAN, VALUE 1, VALUE 2) function becomes VALUE 1, if the BOOLEAN expression is true. If the BOOLEAN expression is false, the function becomes VALUE 2.
<B>Probability of the quantal response of each oocyte, using binomial distribution
<para1>The response rate of each oocyte, Poi([A]), is
<EQ> (A-3)
<para1>Here, nic,Th is the number of channels required to classify the response of the oocytes, and is
<EQ> (A-4)
<para1>In this case, the response of the oocytes is quantal (on-off, or mobile-immobile). Thus, PIM([A]) is the average of Poi([A]).
<EQ> (A-5)
<A>References
<REF> 1. Sonner JM, Antognini JF, Dutton RC, Flood P, Gray AT, Harris RA, Homanics GE, Kendig J, Orser B, Raines DE, Trudell J, Vissel B, Eger EI (2003) Inhaled anesthetics and immobility: mechanisms, mysteries, and minimum alveolar anesthetic concentration. Anesth Analg 97:718-740
<REF> 2. Urban BW, Bleckwenn M (2002) Concepts and correlations relevant to general anaesthesia. Br J Anaesthesia 89:2-16
<REF> 3. Waud DR (1972) On biological assays involving quantal responses. J Pharmacol Exp Ther 183:577-607
<REF> 4. Yamakura T, Batrtaccini D, Trudell JR, Harris RA (2000) Anesthetics and ion channels: molecular model and sites of action. Annu Rev Pharmacol Toxicol 41:23-51
<REF> 5. Kumamoto E, Murata Y (1996) Enhancement by lanthanide of general anesthetic-induced GABAA-receptor current in rat septal cholinergic neurons in culture. J Neurophysiol 75:2294-2299
<REF> 6. Jenkins A, Franks NP, Lieb WR (1999) Effects of temperature and volatile anesthetics on GABA(A) receptors. Anesthesiology 90:484-491
<REF> 7. Katoh T, Ikeda K (1992) Minimum alveolar concentration of sevo?urane in children. Br J Anaesthesia 68:139-141
<REF> 8. Eckenhoff RG, Johansson JS (1999) On the relevance of
"clinically relevant concentrations" of inhaled anesthetics in in vitro experiments. Anesthesiology 91:856-860
<REF> 9. Ueda I (2002) The steep dose-response curves of anesthesia. Anesthesiology 96:252
<REF>10. Eger EI, Fisher DM, Dilger JP, Sonner JM, Evers A, Franks NP, Harris RD, Kendig JJ, Lieb WR, Yamakura T (2001) Relevant concentrations of inhaled anesthetics for in vitro studies of anesthetic mechanisms. Anesthesiology 94:915-921
<REF>11. Franks NP, Lieb WR (1996) Temperature dependence of the potency of volatile general aensthetics: implications for in vitro experiments. Anesthesiology 84:716-720

ページのTOPへ


<JN>J Anesth (2004) 18:100-106
<PT>
<CT>Single sodium channels from human skeletal muscle in planar lipid bilayers: characterization and response to pentobarbital
<CA>Hans C. Wartenberg and Bernd W. Urban
<ADD>Department of Anesthesiology, University of Bonn, Sigmund-Freud-Strase 25, 53105 Bonn, Germany
<AB>Abstract
<AB>Purpose. To investigate the response to general anesthetics of different sodium-channel subtypes, we examined the effects of pentobarbital, a close thiopental analogue, on single sodium channels from human skeletal muscle and compared them to existing data from human brain and human ventricular muscle channels.
<AB>Methods. Sodium channels from a preparation of human skeletal muscle were incorporated into planar lipid bilayers, and the steady-state behavior of single sodium channels and their response to pentobarbital was examined in the presence of batrachotoxin, a sodium-channel activator. Single-channel currents were recorded before and after the addition of pentobarbital (0.34-1.34 mM).
<AB>Results. In symmetrical 500 mM NaCl, human skeletal muscle sodium channels had an averaged single-channel conductance of 21.0 6 0.6 pS, and the channel fractional open time was 0.96 6 0.04. The activation midpoint potential was 296.2 6 1.6 mV. Extracellular tetrodotoxin blocked the
channel with a half-maximal concentration (k1/2) of 60 nM at
0 mV. Pentobarbital reduced the time-averaged conductance of single skeletal muscle sodium channels in a concentration-dependent manner (inhibitory concentration 50% [IC50] 5 0.66 mM). The steady-state activation was shifted to
more hyperpolarized potentials (216.7 mV at 0.67 mM
pentobarbital).
<AB>Conclusion. In the planar lipid bilayer system, skeletal muscle sodium channels have some electrophysiological properties that are signi?cantly different compared with those of sodium channels from cardiac or from central nervous tissue. In contrast to the control data, these different human sodium channel subtypes showed the same qualitative and quantitative response to the general anesthetic pentobarbital. The implication of these effects for overall anesthesia will depend on the role the individual channels play within their neuronal networks, but suppression of both central nervous system and peripheral sodium channels may add to general anesthetic effects.
<KW>Key words Conduction (block) ・ Membrane potential ・ Na channel ・ Single-channel currents ・ Skeletal muscle function
<A>Introduction
<para1>General anesthetics have a wide spectrum of actions on sodium channels of the central nervous system. A recent paper [1] showed that isoforms of the general anesthetic ketamine had a voltage-dependent effect on rat neuronal and human skeletal muscle sodium channels, indicating a local anesthetic effect of the ketamines.
In the present investigation, we wished to examine whether pentobarbital, another general anesthetic agent, which we showed previously to have no voltage-dependent effects on central nervous system (CNS)
sodium channels, in contrast to ketamine, nevertheless has an effect on skeletal muscle sodium channels.
  <para2>Dysfunctions of skeletal muscle sodium channels have serious clinical implications, as life-threatening complications resulting from severe muscle rigidity during induction of anesthesia have been observed in patients with hereditary sodium channel myopathies [2,3]. The functional expression of a speci?c subunit of the skeletal muscle sodium channel (SkM2) very likely plays an important role in patients susceptible to malignant hyperthermia [4]. Finally, succinylcholine produced masseter muscle rigidity and activated myotonic discharges, followed by a neuromuscular block, suggesting that either succinylcholine or its metabolites may interfere directly with voltage-operated sodium channels of the sarcolemma [5]. These examples show that modi?cations of human skeletal sodium channels may have important clinical consequences. Therefore, anesthetic actions on sodium channels from healthy skeletal muscle should be studied, because not only sodium channels from the CNS but also skeletal muscle sodium channels may be involved in important clinical effects (immobility) and side effects caused by general
anesthetics.
  <para2>This study continues the investigation of the role sodium channels may play in overall anesthetic sensitivity and in the speci?c clinical pro?les of anesthetics. Sodium channels from three different human tissues have been examined so far. Sodium channels from two
human heart tissues (atrium and ventricle) and sodium channels from human CNS differed in their electrophysiological characteristics, while their anesthetic
responses to a thiopental analogue (pentobarbital) showed little difference [6]. Here, we examine a sodium channel from a fourth human tissue in the lipid bilayer system: the sodium channel from human skeletal muscle.
<A>Patients, materials, and methods
<B>Preparation
<para1>With the approval of the local Committees on Human Rights in Research (the investigation conformed to the principles outlined in the Declaration of Helsinki),
human skeletal muscle samples were acquired. After written consent was provided, these samples were
obtained from ?ve patients undergoing lower-limb
amputation. The muscle tissue was derived from the gastrocnemius muscle of the calf. Samples were taken immediately after amputation and frozen at 280°C. Membrane preparation was as described for canine heart [7].
<B>Bilayer procedures
<para1>Most materials and experimental methods are described elsewhere [8,9]; a brief description is given
below.
  <para2>Experiments were conducted in symmetrical Te?on bilayer chambers with 5-ml compartments separated by a Te?on partition bearing a hole of approximately
300-Im diameter in its center. All experiments were conducted at room temperature (22°C-24°C) in symmetrical 500-mM NaCl buffered at pH 7.4 with
10 mM HEPES (4-(2-hydroxyethyl)-1-piperazine ethanesulfonic acid, United States Biochemical, Lleveland, OH, USA); no corrections were made for temperature differences between experiments. Planar bilayers were formed over the aperture from neutral phospholipid solutions containing (4 : 1) 1-palmitoyl-2-oleoyl-phosphatidylethanolamine and 1-palmitoyl-2-oleoyl-phosphatidylcholine (Avanti Polar Lipids, Birmingham, AL, USA) in decane (5% wt/vol, 99.9% pure; Wiley Organics, Columbus, OH, USA). Tetrodotoxin (TTX) was purchased from Sigma Chemical (St. Louis, MO, USA). Batrachotoxin (BTX) was a gift from Dr. J. Daly, NIH (Bethesda, MD, USA). The incorporation of sodium channels into the bilayer was achieved by adding, with a glass pipette, small amounts of preparation close to the preformed bilayer membrane in the presence of 0.5 IM BTX. In case of fusion of individual channels with the lipid bilayer channel, orientation was determined by channel-gating characteristics. Only experiments with equally oriented
channels were included in this study. The electrophysiological sign convention was used for the presentation of the results (i.e., the side to which TTX binds is the reference for the potentials).
  <para2>Channel currents were recorded under voltage-clamp conditions at holding potentials (as indicated) between 2100 mV and 1100 mV, using a standard current-to-voltage ampli?er (Axopatch 200 and Axon TL-1 DMA Interface; Axon Instruments, Foster City, CA, USA) and ?ltered at 50 Hz (model 902; Frequency Devices, Haverhill, MA, USA). Current was sampled, and time-averaged conductances were calculated by computer software (pClamp 5.51; Axon Instruments). After the incorporation of a sodium channel into the bilayer, control currents were measured for at least 60 min,
using a series of standard protocols with changing membrane potentials (discrete steps of 10 mV or 15 mV,
between 2100 mV and 1100 mV). In one experiment, increasing concentrations of TTX were achieved by adding small amounts from a stock solution to the extracellular side of the channels.
  <para2>In some experiments, steady-state activation properties were determined as described [9]. In the range
of channel activation, channel fractional open time
(fo) could be described by a two-level distribution (Boltzmann distribution) with one open and one closed state:
<EQ>fo 5 fmax/{1 1 exp(2zaF[V 2 Va])/RT}
<para1>containing as parameter the maximum channel fractional open time (fmax), the steady-state midpoint
potential (Va 5 the potential at which the channel is open fmax/2), and the valence of the effective gating charge (za); (V 5 membrane potential, F 5 Faraday constant, R 5 gas constant, T 5 absolute temperature).
  <para2>In a number of experiments, pentobarbital was added from an ethanol stock solution to the aqueous compartment facing either the extracellular or the intracellular side of the channel. Pentobarbital-block (Block[PTB]) was calculated according to:
<EQ>Block[PTB] 5 1 2 g[PTB]/g[control]
<para1>where g[PTB] is the time-averaged conductance in the presence of pentobarbital and g[control] is the time-
averaged conductance in the absence of pentobarbital. The protocol of the control measurements was repeated with increasing concentrations of pentobarbital.
  <para2>Standard deviations (SD) were used when expressing averages; to examine signi?cance we analyzed our data using a two-step procedure-a Kruskal-Wallis test (non-parametric counterpart of an analysis of variance [ANOVA]), followed by a multiple comparison test (Dunnett, multiple groups vs control). In all tests the data were found to be signi?cantly different under a probability level of 0.05.
<A>Results
<para1>In the experimental system of the planar lipid bilayer the incorporation of a sodium channel is a rare event. Table 1 shows the number of overall experiments and the-by comparison-small numbers of successful experiments with single-channel incorporations suitable for pharmacological investigation. As data obtained from individual patients showed no signi?cant differences, they were subsequently pooled.
  <para2>The typical recordings of a single skeletal muscle sodium channel under our control conditions are shown in Fig. 1a. The skeletal muscle channel stayed open most of the time (96%) and showed only rare and brief closures, of the order of 100 ms or less [10], because BTX impairs channel fast inactivation [11].
  <para2>Analysis of the transitions between open and closed states revealed a single-channel current-voltage relationship that was symmetrical and independent of
membrane potential under symmetrical electrolyte
conditions.
  <para2>Extracellular TTX (but not intracellular TTX, which had no effect at all) blocked sodium channels in an all-or-none manner, manifesting itself as closures of long duration. This block was almost complete at 145 mV when 0.1 IM TTX was added to the extracellular side of the experimental setup. TTX sensitivity was assessed
by the addition of increasing concentrations of the toxin to the extracellular side of the incorporated channels. The block could be shown to be voltage-dependent, with block decreasing as the potential across the
channel was increasingly depolarized. Subconductance states were observed in none of the experiments with sodium channels from human skeletal muscle.
  <para2>In 7 out of 12 experiments, single channels were successfully exposed to the general anesthetic pentobarbital. When pentobarbital was added to either side of the sodium channels, these underwent frequent transitions between a fully open and a fully closed state (Fig. 1b, c). As this action became too rapid for full resolution, it was quanti?ed by measuring the current averaged over time (Fig. 2a); from these data, the fractional open times were calculated. Pentobarbital induced
a concentration-dependent block of the channels (Fig. 2b).
  <para2>At membrane potentials more negative than 280 mV, the fractional open time (but not the single-channel conductance) usually became dependent on membrane potential. Below this voltage, the fractional open time decreased successively with increasing negative potentials, until channels were totally closed at a potential below 2120 mV. This effect re?ects channel deactivation, which-due to BTX modi?cation-is shifted to more hyperpolarized potentials. Only single-channel experiments with more than three successive activation curves were included in this analysis. With increasing concentrations of pentobarbital, the midpoints of activation were shifted to increasingly more negative potentials (Fig. 3). At all examined pentobarbital concentrations, these shifts were signi?cant (340 IM, 27.4 mV; 670 IM, 216.7 mV; 1340 IM, 222.7 mV).
<A>Discussion
<para1>The lipid bilayer technique provides stable experimental conditions needed for pharmacological studies. Individual sodium channels can be observed for several hours, making it possible to distinguish steady-state properties (e.g., an intrinsic variability) from a natural run-down of the preparation [12], a consideration that is particularly important in the study of drug actions [10,11]. In addition, data are not only sampled in a true steady-state but also different pharmacological situations (such as increasing concentrations of a drug) can be explored suf?ciently and repeatedly with the same channel. Because pharmacological agents may act by causing changes in the physiochemical properties of the lipid membrane, the option to modify and control the lipid environment of the channel is important.
  <para2>When pharmacological effects on sodium channels from different tissues are compared, one must allow for the possibility that differences in sensitivity may not only arise from changes in the amino-acid sequence of the ion channel proteins [7,13,14] but also from post-translational modi?cation [15]. Thus, differences in potency may not become apparent when sodium channels are expressed heterologously in foreign cells using
molecular-biological methods. In contrast to the patch-clamp method, native human sodium channels are
reconstituted into planar lipid bilayers. The main
advantages of the bilayer system are that the protein is kept in a well-characterized lipid environment and that the complete channel with all its subunits and associated structures can be investigated.
  <para2>The number of successful experiments in this study was small but typical, and not different from other experimental series with human sodium channels. The sodium channel of human skeletal muscle examined in this study had the qualitative biophysical characteristics expected for BTX-modi?ed sodium channels in planar lipid bilayers (Table 2). Comparing the sodium channels of three human tissues with other channels revealed different electrophysiological properties, although they shared many similarities.
  <para2>Compared with all other (BTX-modi?ed) human
sodium channels in planar lipid bilayers, the single-channel conductance of the skeletal muscle sodium channel was smaller (Table 2; channel conductance). Other authors reported similar differences between
sodium channels from skeletal muscle and cardiac muscle in lipid bilayers [7,16].
  <para2>Only brief closures, of the order of 100 ms or less, could be seen. This resembles sodium channels from human brain and it is in contrast to cardiac channels from human atrium and ventricle, which show longer closures (up to seconds) [17]. This observation was
supported by the analysis of the fractional open time (Table 2), which was lower in cardiac sodium channels than in channels from skeletal muscle or brain.
  <para2>Subconductance states were not detected in the 12 experiments with skeletal muscle sodium channels. This is different from ?ndings in human heart ventricular sodium channels [18] and human brain sodium channels [9] in lipid bilayers where, apart from the predominant single-channel conductance level, we also observed smaller current transitions.
  <para2>BTX caused sodium channels from human skeletal muscle to stay open 96% of the time by practically abolishing fast inactivation [11]. A similar fractional open time (0.94) has been reported previously for human brain sodium channels [19]. In contrast, the fractional open time of human heart sodium channels was clearly less (0.85) [18], resulting from a population of longer-lasting closing events (longer than 100 ms).
  <para2>High sensitivity to the speci?c sodium-channel blocker TTX is one of the main characteristics of
neuronal and muscle sodium channels, while cardiac sodium channels demonstrated less sensitivity [20]. We found the same pattern in lipid bilayers, with a tenfold difference in sensitivity (Fig. 4) for brain and skeletal muscle sodium channels compared to human cardiac sodium channels of ventricle [18] and of atrium [17]. The voltage-dependence of the TTX block, another characteristic of BTX-modi?ed sodium channels in
bilayers [19,21,22], was similar in all human sodium channels (Fig. 4).
  <para2>The steady-state activation midpoint of human skeletal muscle sodium channels was not different from that of the two human cardiac channels (Table 2). However, the activation midpoints of these three muscle-type
sodium channels were some 210 mV more negative compared to human brain sodium channels (Table 2). Activation midpoint potentials that were more negative for skeletal muscle and cardiac sodium channels than for neuronal sodium channels have been consistently reported by other groups working with expression
systems and with native cells [23-25].
  <para2>Pentobarbital has been used as a reference substance to characterize the response of sodium channels to general anesthetic agents in lipid bilayers [6,10,11]. At a free clinical concentration of 88 mM pentobarbital, our conductance block was 8.4%, which is a signi?cant
effect. This block in fact seems small, but it is not known if a suppression of 50 % is needed to induce clinical anesthesia. The clinical concentration for the suppression of a status epilepticus, on the other hand, is much closer to our inhibitory concentration 50% (IC50) (~400 mM) [26]. Considering the electrophysiological differences between sodium channels from different tissue sources, it is surprising that almost no signi?cant difference in anesthetic actions could be found (Table 2). The IC50 values for conductance suppression were almost identical, while the hyperpolarizing shifts in the steady-state activation caused by pentobarbital were slightly different.
  <para2>Among other targets, sodium channels are considered to be very likely targets involved in anesthetic immobility, which is considered to be one of the essential components of anesthesia [27]. In this article, we show that pentobarbital at a clinical concentration also affects a peripheral human sodium channel, the skeletal muscle channel, in addition to its previously reported effects on sodium channels from the human CNS [10]. Considering human sodium channels, immobility may not only be a response of the CNS but even peripheral sodium channels may play a part in the full anesthetic effects.
  <para2>Still, considering today's sketchy knowledge of the underlying neuronal networks, it is not possible to make conclusive statements about the speci?c role that sodium channels play in overall anesthesia. Too little is known about the effects that changes in the function of a single type of ion channel have within a complex neuronal network, as these networks have not yet been identi?ed. The location of a molecular structure inside this network may be of greater importance to the overall anesthetic effects than the absolute anesthetic suppression of this molecular structure. Besides, CNS and skeletal muscle sodium channels are only two of many molecular players affected at clinical concentrations of anesthetics. Knowledge of all other molecular participants affected by anesthesia, as well as the network that connects them, is required before a more complete picture of the mechanisms of anesthesia can be obtained.
  <para2>In conclusion, the general anesthetic pentobarbital has been shown to affect sodium channels of human skeletal muscle at clinical concentrations, and in a
concentration-dependent manner. Despite their electrophysiological differences, human sodium channel subtypes show only small differences in their qualitative and quantitative anesthetic responses. The importance of these effects for overall anesthesia can only be fully evaluated when the neuronal networks in which these channels function have been identi?ed. However, as sodium channels are considered likely targets for anesthetic actions causing immobility, this study shows that even peripheral sodium channels may contribute to overall anesthetic effects.
<ACK>Acknowledgments. We thank Z. Dorner and A. Brambeer for technical assistance and Dr. B. Rehberg and Dr. C. Frenkel for many helpful discussions.
<A>References
<REF> 1. Haeseler G, Tetzlaff D, Bu?er J, Dengler R, Munte S, Hecker H, Leuwer M (2003) Blockade of voltage-operated neuronal and skeletal muscle sodium channels by S(1)- and R(2)-ketamine. Anesth Analg 96:1019-1026
<REF> 2. Haeseler G, Stormer M, Mohammadi B, Bu?er J, Dengler R, Piepenbrock S, Leuwer M (2001) The anesthetic propofol modulates gating in paramyotonia congenita mutant muscle sodium channels. Muscle Nerve 24:736-743
<REF> 3. Vita GM, Olckers A, Jedlicka AE, George AL, Heiman-Patterson T, Rosenberg H, Fletcher JE, Levitt RC (1995)
Masseter muscle rigidity associated with glycine1306-to-alanine mutation in the adult muscle sodium channel alpha-subunit gene. Anesthesiology 82:1097-1103
<REF> 4. Fletcher JE, Wieland SJ, Karan SM, Beech J, Rosenberg H (1997) Sodium channel in human malignant hyperthermia. Anesthesiology 86:1023-1032
<REF> 5. Haeseler G, Petzold J, Hecker H, Wurz A, Dengler R, Piepenbrock S, Leuwer M (2000) Succinylcholine metabolite succinic acid alters steady state activation in muscle sodium channels. Anesthesiology 92:1385-1391
<REF> 6. Wartenberg HC, Wartenberg JP, Urban BW (2001) Human
cardiac sodium channels are affected by pentobarbital. Eur J Anaesthesiol 18:306-314
<REF> 7. Guo XT, Uehara A, Ravindran A, Bryant SH, Hall S, Moczydlowski E (1987) Kinetic basis for insensitivity to tetrodotoxin and saxitoxin in sodium channels of canine heart and denervated rat skeletal muscle. Biochemistry 26:7546-7556
<REF> 8. Recio-Pinto E, Duch DS, Levinson SR, Urban BW (1987) Puri?ed and unpuri?ed sodium channels from eel electroplax in planar lipid bilayers. J Gen Physiol 90:375-395
<REF> 9. Duch DS, Recio-Pinto E, Frenkel C, Urban BW (1988) Human brain sodium channels in bilayers. Brain Res 464:171-177
<REF>10. Frenkel C, Duch DS, Urban BW (1990) Molecular actions of pentobarbital isomers on sodium channels from human brain cortex. Anesthesiology 72:640-649
<REF>11. Wartenberg HC, Urban BW, Duch DS (1999) Distinct molecular sites of anaesthetic action: pentobarbital block of human brain sodium channels is alleviated by removal of fast inactivation. Br J Anaesth 82:76-80
<REF>12. Rehberg B, Duch DS, Urban BW (1994) The voltage dependent action of pentobarbital on batrachotoxin-modi?ed human brain sodium channels. Biochim Biophys Acta 1194:215-222
<REF>13. Wang DW, George AL, Bennett PB (1996) Comparison of heterologously expressed human cardiac and skeletal muscle sodium channels. Biophys J 70:238-245
<REF>14. Mandel G (1992) Tissue-speci?c expression of the voltage-sensitive sodium channel. J Membr Biol 125:193-205
<REF>15. Recio-Pinto E, Thornhill WB, Duch DS, Levinson SR, Urban BW (1990) Neurominidase treatment modi?es the function of electroplax sodium channels in planar lipid bilayers. Neuron 5:675-684
<REF>16. Zamponi GW, Doyle DD, French RJ (1993) Fast lidocaine block of cardiac and skeletal muscle sodium channels: one site with two routes of access. Biophys J 65:80-90
<REF>17. Wartenberg HC, Wartenberg JP, Urban BW (2003) Pharmacological modi?cation of sodium channels from human heart atrium in planar lipid bilayers: electrophysiological characterization of responses to batrachotoxin and pentobarbital. Eur J Anaesthesiol 20:354-362
<REF>18. Wartenberg HC, Wartenberg JP, Urban BW (2001) Single sodium channels from human ventricular muscle in planar lipid bilayers. Basic Res Cardiol 18:306-314
<REF>19. Frenkel C, Wartenberg HC, Duch DS, Urban BW (1998) Steady-state properties of sodium channels from healthy and tumorous human brain. Brain Res 59:22-34
<REF>20. Catterall WA (1995) Structure and function of voltage-gated ion channels. Annu Rev Biochem 64:493-531
<REF>21. French RJ, Worley JF, Krueger BK (1984) Voltage-dependent block by saxitoxin of sodium channels incorporated into planar lipid bilayers. Biophys J 45:301-310
<REF>22. Green WN, Weiss LB, Andersen OS (1987) Batrachotoxin-modi?ed sodium channels in planar lipid bilayers. Characterization of saxitoxin- and tetrodotoxin-induced channel closures. J Gen Physiol 89:873-903
<REF>23. O'Leary ME (1998) Characterization of the isoform-speci?c differences in the gating of neuronal and muscle sodium channels. Can J Physiol Pharmacol 76:1041-1050
<REF>24. Yang XC, Labarca C, Nargeot J, Ho BY, Elroy-Stein O, Moss B, Davidson N, Lester HA (1992) Cell-speci?c posttranslational events affect functional expression at the plasma membrane but not tetrodotoxin sensitivity of the rat brain IIA sodium channel alpha-subunit expressed in mammalian cells. J Neurosci 12:268-277
<REF>25. Arreola J, Spires S, Begenisich T (1993) Na1 channels in cardiac and neuronal cells derived from a mouse embryonal carcinoma cell line. J Physiol (Lond) 472:289-303
<REF>26. Claassen J, Hirsch LJ, Emerson RG, Mayer SA (2003) Treatment of refractory status epilepticus with pentobarbital, propofol, or midazolam. A systematic review. Epilepsia 43:146-153
<REF>27. Sonner JM, Antognini JF, Dutton RC, Flood P, Gray AT, Harris RA, Homanics GE, Kendig J, Orser B, Raines DE, Trudell J, Vissel B, Eger EI (2003) Inhaled anesthetics and immobility: mechanisms, mysteries, and minimum alveolar anesthetic concentration. Anesth Analg 97:718-740
<REF>28. Wartenberg HC, Wang J, Rehberg B, Urban BW, Duch DS (1994) Molecular actions of pentobarbitone on sodium channels in lipid bilayers: role of channel structure. Br J Anaesth 72:668-673

ページのTOPへ


<JN>J Anesth (2004) 18:107-112
<PT>
<CT>Toborinone and olprinone, phosphodiesterase III inhibitors, inhibit human platelet aggregation due to the inhibition of both calcium release from intracellular stores and calcium entry
<CA>Kyoko Kageyama, Toshiki Mizobe, Shinji Nozuchi, Noriko Hiramatsu, Yasufumi Nakajima, and Hiroshi Aoki
<ADD>Department of Anesthesiology, Kyoto Prefectual University of Medicine, Kawaramachi-Hirokoji, Kamigyo-ku, Kyoto 602-8566, Japan
<AB>Abstract
<AB>Purpose. We investigated the inhibitory effects of toborinone and olprinone on human platelet aggregation and calcium mobilization.
<AB>Methods. Washed human platelets were preincubated with toborinone or olprinone, then exposed to 0.015 U・ml21 of thrombin. Aggregation curves were measured using an aggregometer. Effects of toborinone or olprinone on changes in intracellular calcium concentration ([Ca21]i) were measured ?uorometrically using fura-2 acetoxymethyl ester (fura-2). Levels of intracellular cyclic 39,59-adenosine monophosphate concentration ([cAMP]i) were also measured, using enzyme-linked immunosorbent assay (ELISA) techniques.
<AB>Results. The concentrations required to cause 50% inhibition of aggregation (IC50) induced by thrombin were
9.7 6 0.9 IM for toborinone and 3.6 6 0.2 IM for olprinone. Both drugs at IC50 signi?cantly elevated [cAMP]i levels and signi?cantly inhibited Ca21 release from intracellular stores. Release of [Ca21]i induced by thrombin was 272.9 6 87.1 nM, 153.3 6 28.7 nM, and 138.9 6 58.2 nM in the control, toborinone, and olprinone groups, respectively (P , 0.02). Calcium in?ux through calcium channels in the plasma membrane was also suppressed by toborinone and olprinone.
<AB>Conclusion. Toborinone (9.7 IM) and olprinone (3.6 IM) inhibit human platelet aggregation, though these concentrations are higher than their therapeutic plasma concentrations. The inhibitory effects of both drugs are related to the inhibition of both Ca21 release and Ca21 entry through [cAMP]i elevation.
<KW>Key words Toborinone ・ Olprinone ・ Phosphodiesterase III inhibitor ・ Platelet
<A>Introduction
<para1>Toborinone and olprinone are phosphodiesterase III (PDE III) inhibitors that are used to treat heart failure [1-3]. As noncatecholamine, nonglycosidic inotropic agents, they increase myocardial intracellular cyclic 39,59-adenosine monophosphate concentration ([cAMP]i) by selective inhibition of cardiac PDE III enzymes, and they increase intracellular calcium delivery, thereby increasing myocardial contractility. These agents also increase cAMP in vascular smooth muscle, resulting in direct vasodilator activity. These combined positive inotropic and vasodilatory effects are ideal for treating severe congestive heart failure. Conversely, however, by increasing cAMP, these agents may affect intracellular calcium concentration ([Ca21]i) and inhibit platelet aggregation. These agents evidently have the potential to provide bene?cial antithrombotic properties [4], particularly in unstable angina and myocardial infarction. We have recently reported that milrinone, at clinical concentrations (0.9 IM), inhibits thrombin-
induced platelet aggregation, an effect that is predominantly mediated by the suppression of calcium release from the dense tubular system [5]. Other studies have reported the inhibition of platelet aggregation and
intracellular calcium mobilization by the PDE III
inhibitors, milrinone and amrinone [6-8]. However, information regarding the inhibition of platelet aggregation and calcium mobilization in human platelet-rich plasma (PRP) is currently unavailable for toborinone and olprinone. Calcium signaling has been considered to be important in numerous platelet processes. The present study examined the effect of olprinone and toborinone on platelet aggregation and on intracellular calcium mobilization.
<A>Subjects, materials, and methods
<para1>The Ethics Committee on Human Research of Kyoto Prefectural University of Medicine approved this study. Written, fully informed consent was obtained from all participants.
<B>Platelet preparation
<para1>Blood samples from the antecubital veins of eight healthy volunteers who had not taken any drugs for at least 2 weeks were collected into plastic tubes containing acid-citrate-dextrose solution, comprising 85 mM sodium citrate, 70 mM citric acid, and 110 mM glucose. The ratio of blood to acid-citrate-dextrose solution was 4 : 1. Samples were centrifuged at 110 g for 15 min and the upper layer of PRP was removed. PRP samples were incubated with 3 mM fura-2 acetoxymethyl ester (fura-2) for 15 min at 37°C, then centrifuged at 250 g for 15 min in the presence of 0.3 U・ml21 apyrase. A washed platelet suspension (WPS) loaded with fura-2 was obtained by discarding the supernatant and resuspending the pellet in Tyrode-Hepes buffer at pH 7.4 [9]. To examine aggregation curves and intracellular calcium mobilization, and to measure cAMP concentrations, numbers of platelets in the WPS were adjusted to 105 cells・Il21, using a Coulter Counter Model II (Coulter Electronics, Hialeah, FL, USA).
<B>Determination of concentrations of toborinone and olprinone required to cause 50% inhibition of aggregation (IC50) induced by thrombin
<para1>Platelet aggregation curves were investigated, using a turbidimetric method, in samples. Aliquots of WPS (245 Il) were placed into a silicon-coated glass tube, maintained at 37°C, and stirred at 1000 rpm throughout the experiments. Extracellular free calcium concentrations were adjusted to 1 mM using CaCl2, and mixtures were incubated for 1 min with 5 Il of toborinone (seven concentrations, 0.2-200 IM) or olprinone (seven concentrations, 0.0656-65.6 IM). Platelet aggregation induced by thrombin (?nal concentration, 0.015 U・ml21) was measured for 5 min, using an aggregometer (Hema Tracer 601; Nikoh Bioscience, Tokyo, Japan). The baseline optical density point was de?ned as 0% for WPS and 100% for distilled water. To samples in the control group, 5 Il of distilled water was added. The IC50 values of toborinone and olprinone were calculated
using an equation of the ?rst degree.
<B>Measurement of intracellular cyclic 39,59-adenosine monophosphate concentration ([cAMP]i)
<para1>After the incubation of PRP with distilled water, or with toborinone (?nal concentration, 9.7 IM) or olprinone (?nal concentration, 3.6 IM) for 1 min at 37°C in the presence of 1 mM CaCl2, samples were stimulated for 5 min with 0.015 U・ml21 of thrombin. Reactions were stopped by the addition of ice-cold 99.8% ethanol. Samples were centrifuged at 1000 g for 20 min at 4°C, then [cAMP]i was measured in the supernatant, using an enzyme immunoassay kit, according to the non-acetylation protocol described by the manufacturer (Amersham International, Amersham, Buckinghamshire, UK). Values for Results were expressed as pmol・1028 platelets.
<B>Measurement of intracellular free Ca21 mobilization
<para1>Aliquots of WPS (490 Il) loaded with fura-2 were added to a ?uorometric cuvette and stirred at 1000 rpm in a ?uorometer (CAF-110; JASCO, Tokyo Japan). After the incubation of samples with 10 Il of distilled water (control) or with toborinone (?nal concentration, 9.7 IM) or olprinone (?nal concentration, 3.6 IM) for 1 min at 37°C in the presence of 1 mM CaCl2 to record basal ?uorescence intensity, stimulation was performed for 5 min, using 0.015 U・ml21 thrombin. Excitation wavelengths were 340 nm and 380 nm, and the emission wavelength was 505 nm [10]. Changes in [Ca21]i were monitored using the fura-2 340/380-?uorescence ratio, calculated according to the equation described by Tsien et al. [11], using a dissociation constant for fura-2 and [Ca21]i of 224 nM.
  <para2>As nickel ions (Ni21) block the in?ux of extracellular calcium, NiCl2 (?nal concentration, 1 mM) was added to fura-2-loaded human platelets instead of CaCl2. Under these conditions, the effects of toborinone or olprinone on thrombin-induced calcium release from intracellular stores such as the dense tubular system were examined [12,13].
  <para2>The entry of manganese ions (Mn21) was also evaluated, using a quenching technique, to investigate the in?ux of calcium into the cytosol. MnCl2 (?nal concentration, 1 mM) was added to fura-2-loaded human platelets, and fura-2 leakage was assessed as a decrease in
the ?uorescence signal after stimulation by thrombin
(calcium-insensitive excitation wavelength of 360 nm). The effects of toborinone or olprinone on thrombin-induced calcium in?ux through calcium channels in the plasma membrane were also examined [14,15]. Data values were expressed as percentages of the initial ?uorescence level.
<B>Materials
<para1>We purchased toborinone (Ohtsuka Pharmaceuticals, Tokyo, Japan), olprinone hydrochloride (Eizai,
Tokyo, Japan), fura-2 acetoxymethyl ester (Dojindo Laboratories, Kumamoto, Japan), and apyrase (Sigma Chemical, St. Louis, MO, USA). Thrombin was donated by Mochida Pharmaceuticals (Tokyo, Japan). All other compounds were purchased from commercial sources.
<B>Statistical analyses
<para1>Values were expressed as means 6 SDs. Aggregation ratios were analyzed using unpaired t-tests. Increases in [Ca21]i and the release of calcium from intracellular stores were analyzed using general linear regression models for one-way analysis of variance (ANOVA;
one between factor), followed by Scheffe multiple-
comparison tests. Differences in Ca21 in?ux measured using Mn21 were analyzed using general linear regression models for two-way ANOVA (one within one between factors). Values of P , 0.05 were regarded as signi?cant for all comparisons.
<A>Results
<para1>Toborinone and olprinone potently inhibited thrombin-induced platelet aggregation in a concentration-
dependent manner (Fig. 1). Both toborinone (20.0 IM) and olprinone (6.5 IM) completely blocked thrombin-induced platelet aggregation. The IC50 values were 9.7 6 0.9 IM for toborinone and 3.6 6 0.2 IM for olprinone.
  <para2>The effects of toborinone and olprinone on [cAMP]i for platelets stimulated with thrombin (0.015 U・ml21) are shown in Fig. 2. In the resting state without thrombin, [cAMP]i was 4.5 6 0.8 pmol・1028 platelets. After stimulation with 0.015 U・ml21 thrombin, [cAMP]i did not signi?cantly change (4.5 6 0.6 pmol・1028 platelets). Toborinone at 9.7 IM (IC50 value) signi?cantly increased [cAMP]i to 6.4 6 0.9 pmol・1028 platelets, compared with the control values (P , 0.05). Olprinone at 3.6 IM (IC50 value) signi?cantly increased [cAMP]i to 7.1 6 1.7 pmol・1028 platelets, compared with the control (P , 0.02). However, no signi?cant difference was noted between the toborinone and olprinone groups.
  <para2>Resting levels of [Ca21]i did not differ signi?cantly between groups (Table 1). Neither of the two agents signi?cantly changed [Ca21]i before thrombin
stimulation.
  <para2>Increases in [Ca21]i at the peak induced by thrombin in the toborinone (9.7 IM) and olprinone (3.6 IM) groups were signi?cantly reduced compared to the control value (P , 0.05 and P , 0.02, respectively). When Ni21 was added to samples instead of CaCl2, increases in [Ca21]i induced by thrombin in the toborinone (9.7 IM) and olprinone (3.6 IM) groups were signi?cantly lower than the control value (P , 0.02 for both; Table 1);
Fig. 3 shows traces representing eight independent
experiments.
  <para2>Figure 4 shows Mn21 entry, expressed as a percentage decrease in ?uorescence at 360 nm. After stimulation by thrombin, ?uorescence decreased rapidly within 1 min (early phase), then decreased gradually (1-5 min; delayed phase). The percentage decreases in ?uorescence in the early phase with toborinone (9.7 IM) and olprinone (3.6 IM) were signi?cantly lower (P , 0.02) than that in the control group (74 6 4%, 85 6 4%, and 88 6 3% in control, toborinone, and olprinone groups, respectively). The percentage decrease in ?uorescence in the delayed phase (5 min after stimulation by thrombin) with toborinone (9.7 IM) and olprinone (3.6 IM) groups was signi?cantly lower (P , 0.02) than that of the control group (63 6 5%, 78 6 3%, and 82 6 4% in the control, toborinone, and olprinone groups, respectively). However, differences in the percentage decrease in ?uorescence between the toborinone and olprinone groups were not signi?cant.
<A>Discussion
<para1>We investigated the effects of the PDE III inhibitors, toborinone and olprinone, on platelet aggregation and on changes in the intracellular second messenger, [Ca21]i, using human platelets. Toborinone and olprinone inhibited human platelet aggregation. The inhibitory effects of both drugs were related to the inhibition of both Ca21 release from intracellular stores and Ca21 entry through calcium channels in the plasma membrane.
  <para2>The therapeutic plasma concentration of toborinone obtained from a pharmacokinetic study in patients with congestive heart failure was 400-800 ng・ml21 (1.0-2.0 IM) [16]. The therapeutic plasma concentration of olprinone is 20 ng・ml21 (0.066 IM, unpublished data, Eizai Pharmaceuticals, Tokyo, Japan). Given that 90% of toborinone and 81% of olprinone is bound to plasma protein, 9.7 IM of toborinone and 3.6 IM of olprinone would be equivalent to plasma concentrations of 97.0 IM of toborinone and 18.9 IM of olprinone, respectively. Concentrations of either drug required to inhibit platelet aggregation were higher than the clinically
effective doses. We recently reported that 0.9 IM of milrinone inhibited platelet aggregation and increased intracellular Ca21 concentration, and was equivalent
to a plasma concentration of 3 IM [5]. As 1 IM of milrinone was de?ned as the concentration close to the therapeutic plasma level, milrinone exerts stronger inhibitory effects on platelet aggregation and calcium mobilization than toborinone or olprinone.
  <para2>We then studied whether the inhibitory effects of toborinone and olprinone in?uenced cytosolic Ca21 mobilization. When agonists bind to receptors on the platelet membrane, intracellular second messengers such as inositol 1,4,5-trisphosphate (IP3) and diacylglycerol (DG) are activated, resulting in increased free calcium concentrations in the platelet cytosol. Platelet release reactions (inside-out signaling) and tyrosine kinase activation (outside-in signaling) occur and change the conformation of ・IIb?3, making it competent to bind dimeric ?brinogen [16]. Stirring allows contact between ?brinogen bound on adjacent platelets, leading to platelet aggregation. Toborinone and olprinone signi?cantly inhibited the elevation of cytosolic Ca21, via the release of Ca21 from intracellular stores and via Ca21 entry across the plasma membrane. Intracellular Ca21 mobilization is antagonized by increasing levels of cAMP and the increasing levels of cAMP block many signal transduction pathways, resulting in the inhibition of platelet aggregation. The process of platelet activation is regulated by levels of the second messenger, cAMP. The [cAMP]i stimulated by thrombin (0.015 iu・ml21) was signi?cantly increased by both toborinone and olprinone. Increased cAMP concentrations in the toborinone and olprinone groups in which platelet aggregation was inhibited were only 1.4- to 1.5-fold higher than those of control groups. However, cAMP present in untreated platelets may fully occupy the high-af?nity cAMP-binding sites of the regulatory subunits of cyclic nucleotide-dependent protein kinase (PKA) at concentrations of 2.31 6 0.53 pmol・1028 platelets [18]. In our results, resting levels of platelet cAMP concentration were 4.5 6 0.8 pmol・1028 platelets, suf?cient to occupy the high-af?nity cAMP-binding sites of the regulatory subunits of PKA. The inhibitory effects of platelet aggregation occurring with toborinone and olprinone are related to the inhibition of both Ca21 release and Ca21 entry, through small increases in cAMP concentration.
  <para2>The major target molecules of cAMP in platelets are PKAs, whose effects are mediated through the phosphorylation of speci?c substrates. This phosphorylation directly affects receptor/G-protein activation and interferes with various signal transduction pathways [19-22], blocking several steps of cytosolic Ca21 elevation and leading to the inhibition of agonist-induced platelet
aggregation. Toborinone or olprinone may have had direct or indirect actions in their inhibitory effects on platelet aggregation or Ca21 mobilization.
  <para2>The present study demonstrated that toborinone and olprinone inhibited human platelet aggregation, though this occurred at concentrations higher than their therapeutic plasma concentrations. These inhibitory effects were related to the inhibition of both Ca21 release and Ca21 entry through [cAMP]i elevation. Cyclic nucleotides and their regulatory pathways are of particular interest for developing new approaches to treating thrombotic and cardiac disorders. Further study is required to fully understand the effects on platelet function of PDE III inhibitors, such as toborinone and olprinone.
<A>References
<REF> 1. Hirata T, Akamatsu T, Horinokuchi N, Uchida O, Kuro M (1999) Study on the effectiveness of toborinone (OPC-18790) in the treatment of heart failure in patients following cardiac surgery. Arzneimittelforschung 9:499-503
<REF> 2. Semeniuk LM, Belenkie I, Tyberg JV (1998) Acute effects of toborinone on vascular capacitance and conductance in experimental heart failure. Circulation 98:58-63
<REF> 3. Murakami R, Sano K, Murakami Y, Shimada T, Morioka S (1995) Effects of intracoronary infusion of an inotropic agent, E-1020 (loprinone hydrochloride), on cardiac function: evaluation of
left ventricular contractile performance using the end-systolic pressure-volume relationship. Int J Cardiol 51:57-63
<REF> 4. Patelunas DM, Carmint WJ, Willis JZ, Colatsky TJ, Fenichel RL (1991) Comparative antithrombotic activities of the phosphodiesterase inhibitors pelrinone (AY-26,768), AY-31,390 and milrinone. Thromb Res Suppl 62:389-400
<REF> 5. Hiramatsu N, Kageyama K (2003) Anti-thrombotic effect of milrinone is caused by inhibition of calcium release from the dense tubular system in human platelets. Acta Anesthesiol Scand 47:53-57
<REF> 6. Kikura M, Kazama T, Ikeda T, Sato S (2000) Disaggregatory effects of prostaglandin E1, amrinone and milrinone on platelet aggregation in human whole blood. Platelets 11:446-458
<REF> 7. Kikura M, Lee MK, Safon RA, Bailey JM, Levy JH (1995) The effects of milrinone on platelets in patients undergoing cardiac surgery. Anesth Analg 81:44-48
<REF> 8. Pattison A, Astley N, Eason CT, Bonner FW (1990) A comparison of the effects of three positive inotropic agents (amrinone, milrinone and medorinone) on platelet aggregation in human whole blood. Thromb Res Suppl 57:909-918
<REF> 9. Lagarde M, Bryon PA, Guichardant M, Dechavanne M (1980) A simple and ef?cient method for platelet isolation from their plasma. Thromb Res 17:581-588
<REF>10. Grynkiewicz G, Poenie M, Tsien RY (1985) A new generation of Ca21 indicators with greatly improved ?uorescence properties. J Biol Chem 260:3440-3450
<REF>11. Tsien RY, Rink TJ, Poenie M (1985) Measurement of cytosolic free Ca21 in individual small cells using ?uorescence micro-scopy with dual excitation wavelengths. Cell Calcium 6:145-157
<REF>12. Molino M, Di Lallo M, de Gaetano G, Cerletti C (1992) Intracellular Ca21 rise in human platelets induced by polymorphonuclear-leucocyte-derived cathepsin G. Biochem J 288:741-745
<REF>13. Sage SO, Rink TJ (1987) The kinetics of changes in intracellular calcium concentration in fura-2-loaded human platelets. J Biol Chem 262:16364-16369
<REF>14. Wang JP, Chang LC, Huang LJ, Kuo SC (2001) Inhibition of extracellular Ca21 entry by YC-1, an activator of soluble guanylyl cyclase, through a cyclic GMP-independent pathway in rat neutrophils. Biochem Pharmacol 62:679-684
<REF>15. Sugatani J, Iwai T, Watanabe M, Machida K, Tanaka T, Maeda T, Miwa M (2000) Inhibition of rabbit platelet aggregation by nucleoside 59-alkylphosphates: correlation with inhibition of agonist-induced calcium in?ux. Biochem Pharmacol 60:197-205
<REF>16. Feldman MD, Pak PH, Wu CC, Haber HL, Heesch CM, Bergin JD, Powers ER, Cowart TD, Johnson W, Feldman AM, Kass DA (1996) Acute cardiovascular effects of OPC-18790 in patients with congestive heart failure. Time- and dose-dependence analysis based on pressure-volume relations. Circulation 93:474-483
<REF>17. Clark EA, Shattil SJ, Brugge JS (1994) Regulation of protein tyrosine kinases in platelets. Trends Biochem Sci 19:464-469.
<REF>18. Eigenthaler M, Nolte C, Halbrugge M, Walter U (1992) Concentration and regulation of cyclic nucleotides, cyclic-nucleotide-dependent protein kinases and one of their major substrates in human platelets. Estimating the rate of cAMP-regulated and cGMP-regulated protein phosphorylation in intact cells. Eur J Biochem 205:471-481
<REF>19. Ryningen A, Olav Jensen B, Holmsen H (1998) Elevation of cyclic AMP decreases phosphoinositide turnover and inhibits thrombin-induced secretion in human platelets. Biochim Biophys Acta 1394:235-248
<REF>20. Cavallini L, Coassin M, Borean A, Alexandre A (1996) Prostacyclin and sodium nitroprusside inhibit the activity of the platelet inositol 1,4,5-trisphosphate receptor and promote its phosphorylation. J Biol Chem 271:5545-5551
<REF>21. El-Daher SS, Patel Y, Siddiqua A, Hassock S, Edmunds S, Maddison B, Patel G, Goulding D, Lupu F, Wojcikiewicz RJ, Authi KS (2000) Distinct localization and function of (1,4,5)IP(3) receptor subtypes and the (1,3,4,5)IP(4) receptor GAP1(IP4BP) in highly puri?ed human platelet membranes. Blood 95:3412-3422
<REF>22. Schwarz UR, Kobsar AL, Koksch M, Walter U, Eigenthaler M (2000) Inhibition of agonist-induced p42 and p38 mitogen-activated protein kinase phosphorylation and CD40 ligand/P-selectin expression by cyclic nucleotide-regulated pathways in human platelets. Biochem Pharmacol 60:1399-1407

ページのTOPへ


<JN>J Anesth (2004) 18:113-117
<PT>
<CT>Spinal neurotoxicity and tolerance after repeated intrathecal administration of YM 872, an AMPA receptor antagonist, in rats
<CA>Tomoki Nishiyama1, Sachiko Kawasaki-Yatsugi2, Tokio Yamaguchi2, and Kazuo Hanaoka1
<ADD>1 Department of Anesthesiology, The University of Tokyo, Tokyo, Japan
<ADD>2 Insitute for Drug Discovery Research, Yamanouchi Pharmaceutical Co. Ltd., Tsukuba, Japan
<AB>Abstract
<AB>Purpose. Although the ・-amino-3-hydroxy-5-methylisoxazole-4-propionic acid (AMPA) receptor antagonist, YM 872, has been considered to be useful in analgesia for both acute and chronic pain, there are no studies of its neurotoxicity and tolerance. We examined the spinal neurotoxicity and tolerance of YM 872 analgesia by repeated intrathecal administration in rats.
<AB>Methods. Male Sprague-Dawley rats with lumbar intrathecal catheters received YM 872 at 1 Ig ・ 10 Il21 (eight rats; YM group) or normal saline 10 Il (eight rats; C group) intrathecally once a day for 30 days. We evaluated the analgesic effects every 3 days, by tail-?ick test and behavioral side effects. On the 31st day, the lumbar spinal cord was removed from four randomly selected rats in each group for histological examination.
<AB>Results. The YM group showed signi?cantly longer tail-?ick latency when subjected to a high-intensity light beam than the C group at each measurement time point, although no signi?cant changes in the latency according to the time course of the study were observed for the entire study period of 30 days in either group. No rats showed any side effects. Histologically, only slight lymphocytic cell in?ltration and degeneration of myelinated ?bers occurred, similarly in both groups. No changes were observed in the spinal cord in either group.
<AB>Conclusion. Administration of YM 872 (1 Ig) once a day for 30 days did not induce any tolerance and caused no histological changes in the spinal cord.
<KW>Key words Analgesia ・ ・-Amino-3-hydroxy-5-methylisoxazole-4-propionic acid (AMPA) receptor antagonist ・ Tolerance ・ Toxicity ・ Spinal cord
<A>Introduction
<para1>Glutamate receptors, mainly N-methyl-d-aspartate (NMDA) and ・-amino-3-hydroxy-5-methylisoxazole-4-propionic acid (AMPA) receptors, have an important role in mediating pain in the spinal cord. NMDA receptors are involved in the hyperalgesic state underlying neuropathic pain but are not involved in acute pain [1]. NMDA receptor antagonists except for ketamine lack clinical value because of their side effects, such as a psychotomimetic action [2], learning impairment [3], and neurotoxicity [4]. On the other hand, it has been suggested that AMPA receptors have a role in both acute and persistent in?ammatory pain in the spinal cord [5,6]. In our previous studies, a new AMPA receptor antagonist, YM 872 {[2,3-dioxo-7-(1H-imidazol-
1-yl)-6-nitro-1,2,3,4-tetrahydro-1-quinoxalinyl] acetic acid; Yamanouchi Pharmaceutical, Tsukuba, Japan} showed analgesic effects on both acute thermal pain (50% effective dose; ED50 5 1 Ig) and formalin-induced in?ammatory pain (ED50 5 0.24 Ig in phase 1 and 0.21 Ig in phase 2) in rats [6]. However, YM 872 induced transient motor disturbance and ?accidity at doses of more than 10 Ig [6]. Although YM 872 had no neurotoxicity in cat brain [7], its toxicity in other organs should be elucidated before its clinical application. Thus, the ?rst purpose of this study was to investigate any histological changes of the spinal cord caused by repeated intrathecal administration of YM 872.
  <para2>For chronic pain, long-term administration of an
analgesic is often required. Although morphine and clonidine are used for the treatment of chronic pain, continuous exposure of I opioid receptors or ・2 adrenoceptors to an agonist will produce tolerance, caused by an unknown mechanism [8]. However, there are no studies on the tolerance of AMPA receptor antagonists. The second purpose of this study was to investigate whether tolerance would occur to the analgesic effects of YM 872 given by intrathecal administration.
<A>Materials and methods
<para1>The protocol was approved by the Research Committee of the University. Male Sprague-Dawley rats (280-300 g; Nippon Bio-Supply, Tokyo, Japan) were implanted with chronic lumbar intrathecal catheters, under halothane (2%) anesthesia. An 8.5-cm polyethylene catheter (PE-10; Clay Adams, Parsippany, NJ, USA) was advanced caudally through an incision in the atlanto-occipital membrane to the thoracolumbar level of the spinal cord. The external part of the catheter was tunneled subcutaneously to exit on the top of the skull and plugged with a 28-G stainless-steel wire. Sixteen rats with normal motor function and behavior 7 days after surgery were used. The position of the catheter was checked by the aspiration of cerebrospinal ?uid at implantation and was veri?ed directly after the rat was killed.
  <para2>YM 872 (Yamanouchi Pharmaceutical), 10 mg, was dissolved in 0.97 ml distilled water with 30 Il 1 N NaOH to adjust the pH to 7.3-7.5. Solutions of 1 Ig (2.86 nmol; ED50 for the tail-?ick test, as used in our previous study [6]) per 10 Il were made, using normal saline, before each injection. Normal saline 10 Il was used as the control. After each intrathecal drug injection, the catheter was ?ushed with a subsequent injection of 10 Il of
normal saline to clear the dead space of the catheter
(8 6 0.9 Il; mean 6 SD). Microinjector syringes were used for all injections.
  <para2>Starting on the seventh day after the catheter insertion (day 0), YM 872, 1 Ig ・ 10 Il21 (eight rats; YM group) or saline 10 Il (eight rats; C group) was administered intrathecally at 7 a.m. every day for 30 days. Every 3 days, the analgesic effect and behavioral side effects were evaluated 15 min after the intrathecal drug injection. The analgesic effect was tested by the tail-?ick test. We
measured tail-?ick latency and checked the side effects 15 min after the drug injection, because both the analgesic and side effects were greatest at that time in our previous study [6]. On the 31st day, the rats were
killed with an overdose of halothane. The location
of the catheter was veri?ed in every rat. Four rats in
each group were selected at random for the histological study.
  <para2>For the experiment, the rats were placed in a clear plastic cylindrical cage with their tails extending through a slot provided in the rear of the tube. Noxious stimulation was provided by a beam of high-intensity light (Tail-?ick Analgesia Meter MK-330A; Muromachi Kikai, Tokyo, Japan) focusing on the tail, 2 to 3 cm proximal to the end. The focus was in almost, but not exactly, the same place at every measurement. The response time was measured, and de?ned as the interval between the onset of the thermal stimulation and an abrupt ?ick of the tail. From our experience, the cutoff time in the absence of a response was set at 14 s to prevent tissue-burn injury.
  <para2>The behavior (including agitation and allodynia), motor function, ?accidity, pinna re?ex, and corneal
re?ex were examined by a blinded investigator after
the tail-?ick measurement. The behaviors were judged as present or absent. Agitation was judged as spontaneous irritable movement and/or vocalization. The presence of allodynia was examined by observation for agitation (escape and/or vocalization) evoked by lightly stroking the ?ank of the rat with a small probe. Motor function was evaluated by the placing/stepping re?ex and the righting re?ex. The former re?ex was evoked by drawing the dorsum of either hind paw across the edge of a table. Normally rats try to place the paw ahead
in a position to walk. The latter re?ex was assessed by placing the rat horizontally with its back on the table, a placement which normally gives rise to an immediate, coordinated twisting of the body to an upright position. Disturbance of the righting re?ex also shows impairment of the function of the central nervous system. Flaccidity was judged as a muscle weakness in raising the forepaw to a place 3-5 cm higher than the hind paw. When a 3- to 5-cm higher place is placed in front of a rat, normally, the rat will walk up to the higher place. Lack of a walking-up movement was judged as ?accidity. Pinna and corneal re?exes were examined with a paper string. When a paper string is inserted into the ear canal or touches the cornea, rats normally shake their head or blink, respectively.
  <para2>Just after being killed with halothane, four randomly selected rats in each group were perfused with 10% formalin through the ascending aorta. The lumbar spinal cord was removed with the ventral and dorsal roots at the lumbar enlargement where the tip of the catheter was located, ?xed in 15% formalin for 24 h, decalci?ed for 48 h, and then embedded in paraf?n. Four slices selected randomly by an animal pathologist were examined using light microscopy after hematoxylin-eosin (H&E) staining and luxor fast blue-H&E staining.
The animal pathologist, at Hatano Research Institute (Kanagawa, Japan) was blinded to the treatment.
Histology was indicated as negative, very slight, slight, moderate, or severe changes according to the usual judgment by the pathologist.
  <para2>Tail-?ick response latency was converted to percent maximum possible effect (%MPE), according to the fol-lowing formula: %MPE 5 [(postdrug latency 2 baseline latency) / (cutoff time 2 baseline latency)] 3100.
  <para2>Differences in the %MPE for tail-?ick latency were analyzed with repeated measures analysis of variance (ANOVA), followed by the Student Newman-Keuls test. Histological ?ndings were compared using the  ̄2 test. A P value of less than 0.05 was considered statistically signi?cant.
<A>Results
<para1>The tail-?ick latencies before the administration of YM 872 or saline were not different between the groups and did not change during the study period (Table 1). The YM group showed signi?cantly greater tail-?ick latency than the C group at each time point measured during the study, but no signi?cant changes were observed in tail-?ick latency according to the time course of the study in either group (Fig. 1). No rats showed any observable motor disturbances or behavioral abnormalities with the dose used.
  <para2>In the histological study, there were no signi?cant differences between the two groups. No histological changes were observed in the spinal cord in either group (Table 2).
<A>Discussion
<para1>Intrathecal administration of YM 872, 1 Ig (ED50 dose), once a day for 30 days, provided analgesia for thermally induced pain without inducing any tolerance, and showed no apparent histological changes of the spinal cord on examination by light microscopy.
  <para2>It is well known that the chronic administration of various receptor agonists may lead to the development of tolerance to the effects, in conjunction with a decrease in the number of membrane receptor molecules which speci?cally bind the agonist [9]. No tolerance to motor disturbance has been observed with long-term treatment with the 2,3-benzodiazepines, which are AMPA receptor antagonists [10]. In the present study, we did not detect any side effects, including motor disturbance, with YM 872.
  <para2>The speed at which tolerance develops depends greatly on the pharmacokinetic characteristics of the drug [11]. Speci?cally, shorter duration of action is associated with faster development of tolerance. The rate of onset of tolerance is also linked to the time during which the receptor is exposed to the ligand. It is possible that slower development of tolerance may be obtained by continuous administration of the drug rather than by intermittent exposure of the receptors to high concentrations from intermittent doses. Therefore, we chose intermittent administration rather than continuous infusion. If glutamate has some roles in the development
of tolerance, as shown in previous studies [9, 12], glutamate antagonists such as YM 872 should have a bene?t to inhibit tolerance.
  <para2>Before the clinical application of a drug, its safety and side effects should be tested. It has been reported that muscle tone was decreased by the intraperitoneal administration of an AMPA receptor antagonist, CNQX [13], while muscle tone was not affected by intraperitoneal GYKI-52466 [14]. Ataxia was induced by NBQX
in rats [15]. A potent AMPA receptor antagonist, YM 90K, has no psychotomimetic action and shows no
histological changes in the rat brain [16]. However, the clinical application of AMPA receptor antagonists has been limited because of poor water solubility and nephrotoxicity caused by precipitation of the drug in the kidney [17]. Thus, analogues with better water solubility have been desired. YM 872 used in our study is much more water-soluble than other formulations of AMPA receptor antagonists [18], and precipitation in the
kidney was not observed in rats [7]. With most AMPA receptor antagonists, the therapeutic dose is close to that causing severe motor impairment [19]. At the doses used to reduce nociception, ?accidity was found with the competitive AMPA receptor antagonist ACEA 2085, and motor disturbance was found at higher concentrations [5]. However, YM 872 did not induce any motor disturbances with the ED50 dose in the present study. Also, YM 872 had no neurotoxicity in cat brains in a cerebral ischemia model [7]. The present study suggests that no histological changes were produced in the spinal cord by the intermittent intrathecal administration of YM 872, although the number of rats tested was too small to reach a de?nite conclusion.
  <para2>Clinically, LY-293558, an AMPA receptor antagonist, was intravenously administered to volunteers, without serious adverse events, except for hazy vision that was resolved in 60 min [20]. A 3-h intravenous
infusion of the AMPA receptor antagonist, YM 90K, at 36 mg caused little change in blood or urine chemistry, electrocardiogram, vital signs, or in levels of insulin, glucagon, or blood glucose in healthy volunteers [21]. YM 872 should be tested, as these compounds have been, before clinical application.
  <para2>In conclusion, although a limited number of rats was examined, administration of the ED50 dose of YM 872 (1 Ig) once a day for 30 days did not induce any tolerance or any histological changes of the spinal cord on examination by light microscopy. Further studies are necessary, using different doses and different animals, to con?rm its safety and lack of induction of tolerance.
<A>References
<REF> 1. Ault B, Hildebrand LM (1993) Effects of excitatory amino acid receptor antagonists on a capsaicin-evoked nociceptive re?ex: a comparison with morphine, clonidine and baclofen. Pain 52:341-349
<REF> 2. Koek W, Woods JH, Winger GD (1988) MK-801, a proposed noncompetitive antagonist of excitatory amino acid neurotransmission, produces phencyclidine-like behavioral effects in pigeons, rats and rhesus monkeys. J Pharmacol Exp Ther 245:969-974
<REF> 3. Morris RGM, Anderson E, Lynch GS, Baudry M (1986) Selective impairment of learning and blockade of long-term potentiation by an N-methyl-D-aspartate receptor antagonist, AP5. Nature 319:774-776
<REF> 4. Olney JW, Labruyere J, Wang G, Wozniak DF, Price MT, Sesma MA (1991) NMDA antagonist neurotoxicity: mechanism and prevention. Science 254:1515-1518
<REF> 5. Nishiyama T, Yaksh TL, Weber E (1998) Effects of intrathecal NMDA and non-NMDA antagonists on acute thermal nociception and their interaction with morphine. Anesthesiology 89:715-722
<REF> 6. Nishiyama T, Gyermek L, Lee C, Kawasaki-Yatsugi S, Yamaguchi T (1999) The spinal antinociceptive effects of a
novel competitive AMPA receptor antagonist, YM872, on thermal or formalin-induced pain in rats. Anesth Analg 89:143-147
<REF> 7. Takahashi M, Ni JW, Yatsugi-Kawasaki S, Toya T, Yatsugi S, Shimizu-Sasamata M, Koshiya K, Shishikura J, Sakamoto S, Yamaguchi T (1998) YM 872, a novel selective ・-amino-3-hydroxy-5-methylisoxazole-4-propinoic acid receptor antagonist, reduces brain damage after permanent focal cerebral ischemia in cats. J Pharmacol Exp Ther 284:467-473
<REF> 8. Stevens CW, Monasky MS, Yaksh TL (1988) Spinal infusion of opiate and alpha-2 agonists in rats: tolerance and cross-tolerance studies. J Pharmacol Exp Ther 244:63-70
<REF> 9. Overstreet DH, Yamamura HI (1979) Receptor alterations and drug tolerance. Life Sci 25:1865-1878
<REF>10. De Sarro G, Di Paola ED, Gareri P, Gallelli L, Scotto G, De Sarro A (1999) Effects of some AMPA receptor antagonists on the development of tolerance in epilepsy-prone rats and in pentylenetetrazole kindled rats. Eur J Pharmacol 368:149-159
<REF>11. Kissin I, Lee SS, Arthur GR, Bradley EL (1996) Time course characteristics of acute tolerance development to continuously infused alfentanil in rats. Anesth Analg 83:600-605
<REF>12. McLemore GL, Kest B, Inturrisi CE (1997) The effects of LY293558, an AMPA receptor antagonist, on acute and chronic morphine dependence. Brain Res 778:120-126
<REF>13. Maj J, Rogoz Z, Skuza G, Jaros T (1995) Some behavioral effects of CNQX and NBQX, AMPA receptor antagonists. Pol J Pharmacol 47:269-277
<REF>14. Maj J, Rogoz Z, Skuza G, Kolodziejczyk K (1995) Some central effects of GYKI 52466, a non-competitive AMPA receptor antagonist. Pol J Pharmacol 47:501-507
<REF>15. Filliat P, Pernot-Marino I, Baubichon D, Lallement G (1998) Behavioral effects of NBQX, a competitive antagonist of AMPA receptors. Pharmacol Biochem Behav 59:1087-1092
<REF>16. Izumisawa N, Kawakami A, Ohata T, Hanada T, Okeda R (1995) YM 90K, an AMPA antagonist, has no neurotoxic effects on cerebrocortical neurons in rats. Exp Neurol 134:199-204
<REF>17. Xue D, Huang ZG, Barnes K, Lesiuk HJ, Smith KE, Buchan AM (1994) Delayed treatment with AMPA, but not NMDA, antagonists reduces neocortical infarction. J Cereb Blood Flow Metab 14:251-261
<REF>18. Kohara A, Okada M, Tsutsumi R, Ohno K, Takahashi M, Shimizu-Sasamata M, Shishikura J, Inami H, Sakamoto S, Yamaguchi T (1998) In-vitro characterization of YM872, a selective, potent and highly water-soluble ・-amino-3-hydroxy-5-methylisoxazole-4-propionate receptor antagonist. J Pharm Pharmacol 50:795-801
<REF>19. Kubova H, Vilagi I, Mikulecka A, Mares P (1997) Non-NMDA receptor antagonist GYKI 52466 suppresses cortical afterdischarges in immature rats. Eur J Pharmacol 333:17-26
<REF>20. Gilron I (2001) LY-293558 Eli Lilly & Co. Curr Opin Investig Drugs 2:1273-1278
<REF>21. Umemura K, Kondo K, Ikeda Y, Teraya Y, Yoshida H, Homma M, Uematsu T, Nakashima M (1997) Pharmacokinetics and safety of the novel amino-3-hydroxy-5-methylisoxazole-4-propionate receptor antagonist YM90K in healthy men. J Clin Pharmacol 37:719-727
<FN>

ページのTOPへ


<JN>J Anesth (2004) 18:118-128
<PT>Review
<CT>Monitoring magnesium to guide magnesium therapy for heart surgery
<CA>Terry L. Shirey
<ADD>Nova Biomedical, 200 Prospect Street, Waltham, MA 02454-9141, USA
<KW>Key words Magnesium ・ Ionized ・ Heart ・ Surgery
<A>Introduction
<para1>For many years it has been recognized that magnesium levels play an important role in morbidity associated with heart surgery. However, considerable confusion exists in the literature concerning whether Mg should be administered to these patients, and, if so, how much and when. The goals of this paper are to review (1) methods used to evaluate Mg status in patients, (2) causes and consequences of abnormal Mg levels perioperatively, (3) outcome improvements and risks with Mg supplementation, and (4) guidelines for administering Mg therapy.
<A>Methods used to evaluate Mg status in patients
<B>Physiological indicators
<para1>Pulse, mean arterial pressure, deep tendon re?exes, hourly diuresis, respiratory recordings, and hypotension are used to monitor Mg status in patients [1]. Signi?cant prolongation of intraatrial and atrioventricular (AV) nodal conduction times, as seen by ECG, may also re?ect Mg activity [2].
<B>Mg measurement
<para1>Total magnesium (TMg) represents the concentration of Mg present in blood plasma or serum. TMg, a measure of all of the Mg in the plasma or serum sample, equals protein-bound Mg plus ligand-bound Mg plus ionized Mg (iMg). Measurement of TMg requires the isolation of plasma (centrifugation of blood sample) or serum (clotting and centrifugation of blood sample). Measurement is made by atomic absorption spectrophotometry or colorimetry. Reference values for plasma and serum TMg typically range from 0.66 to 1.07 mmol・l21. There is virtually no correlation between plasma/serum TMg and intracellular TMg.
  <para2>Ionized magnesium (iMg 5 Mg21) represents the activity of unbound Mg in whole blood plasma, plasma, and/or serum. It is the physiologically active Mg fraction, i.e., the fraction to which tissues respond. Reference values typically range from 0.45 to 0.62 mmol・l21. The fact that it can be measured in a whole blood sample by electrode produces a rapid result (,150 s) on a small sample (,200 Il). A rapid result can be very helpful for patients (1) with arrhythmia, (2) with changes in cardiac output, (3) receiving cardiovascular drugs, (4) sustaining hypoxic damage, and (5) receiving Mg therapy. Blood plasma iMg correlates with intracellular iMg and therefore represents a better indicator of Mg status than TMg. It is typically 70% of the TMg value, but varies with the protein and small ligand
concentrations in the blood. The iMg value may be substantially less than 70% of the TMg value in critically ill patients where binding ligand concentrations (heparin, citrate, lactate, phosphate, bicarbonate, etc.) have increased.
  <para2>Comparisons of iMg to TMg were illustrated in nine clinical settings (hypertension, acute myocardial infarction, head trauma, noninsulin-dependent diabetes, stroke, pregnancy, ischemic heart disease, cyclosporin recipients, and asthma) [3]. Ionized magnesium was found to be a better indicator of disease than TMg. In summary, iMg (1) represents the physiologically important Mg measurement, (2) is a better indicator of disease than TMg, and (3) is a more rapid measurement.
  <para2>The literature should be interpreted with caution. Frequently, if not generally, the Mg values reported in the literature are TMg, even though they may be reported as Mg21. Measurement of iMg has recently been made possible (early 1990s), but this does not change the fact that many recent papers reporting Mg21 really mean TMg.
<A>Causes and consequences of abnormal Mg
levels perioperatively
<B>Presurgery
<para1>Although TMg de?ciency is rare in healthy subjects, 16% of 98 heart surgery patients were hypomagnesemic [4]. Kidney disease, reduced glomerular ?ltration rate (,30 ml・min21) [2], reduced tubular reabsorption (frequently caused by the use of diuretics), reduced oral intake, and intravenous ?uids with inadequate Mg led to abnormal Mg levels in these patients. Patients receiving digoxin for heart failure had preoperative
total hypomagnesemia more frequently than patients not on digoxin (36% vs 10%) [4]. A history of high-grade ventricular dysrhythmia was associated with signi?cantly lower preoperative mean TMg [4]. Patients who may have asymptomatic hypomagnesemia presurgery may then undergo a surgical procedure that, in the perioperative period, can cause them to progress
to symptomatic hypomagnesemia [5]. Hyperaldosteronism [6] and noradrenaline [7] also promote
hypomagnesemia.
<B>During surgery
<para1>The frequency of hypomagnesemia increased to 71% (71/100 patients) following cardiopulmonary bypass (CPB) surgery [8]. Several factors reduce Mg concentration during surgery. These include hemodilution from prime volume [9], chelation of iMg by heparin and acid-citrate-dextrose when donated blood is used to prime the CPB circuit [9], and intramyocyte hypoxia [10-12]. In recently introduced off-pump coronary artery bypass grafting (CABG) surgery, many patients are not receiving Mg supplementation, and, consequently, are hypomagnesemic postsurgery (TMg averaging 0.61 mM after surgery) [13]. Magnesium administration (via cardioplegia, bolus doses, and ?uid supplements) can minimize Mg depletion, if not lead to hypermagnesemia, perioperatively.
<A>Outcome improvements and risks with
magnesium supplementation
<para1>Adequate Mg levels are required for normal cardiovascular activity (conduction and contraction), tissue protection from oxygen free radicals and the in?ammatory response, and blood ?ow. Table 1 identi?es improvements in outcomes following cardiac surgery when Mg supplements were used.
  <para2>Even though Mg supplementation provides strong bene?ts for the cardiac patient, the risks and complications of hypermagnesemia need to be considered. Hypermagnesemia may develop from inadequate renal function or high dosage levels of supplemented Mg. CPB surgery, itself, may cause renal failure suf?cient to give hypermagnesemia with Mg therapy [35].
  <para2>Hypermagnesemia can cause excessive ?ushing, sweating, warm sensations, hypotension, and vasodilation [5]. Magnesium may act as a laxative [1]. As an anticoagulant, Mg inhibits platelet aggregation, P-selectin expression, and ?brinogen binding to platelet GP IIb/IIIa receptor in vitro [36]. The number and energy of direct-current shocks to initialize and to sustain de?brillation was greatest in patients who received Mg therapy before bypass and in those whose plasma TMg was greater than 0.93 mmol・l21 [37].
  <para2>Table 2 presents a rough guideline correlating Mg values and clinical observations. Mg therapy in the immediate postoperative period following cardiac surgery can result in hypotension and bradycardia [5]. In one study, St. Thomas II cardioplegia solution, containing 16 mM MgCl2, was infused (10 ml・kg21 every 30 min) during aortic cross-clamping [44]. Ionized magnesium levels of 1.5-1.58 mmol・l21 were noted after unclamping. After discontinuation of extracorporeal circulation, vascular resistance decreased by 40%, while 78% (14/18) of patients required atrial or ventricular pacing in order to maintain a physiological level of heart rate for obtaining better hemodynamics. The hypotension and bradycardia extended the time for extracorporeal circulation. Hypotension following cardiac surgery may be unresponsive to pressors [5].
  <para2>Magnesium therapy also poses risk when used with other medications. In CPB patients receiving nitrate (another vasodilator) or an angiotensin-converting enzyme inhibitor, Mg therapy is not currently justi?ed [36]. Considering that Mg is a natural calcium channel blocker, individuals with low ionized calcium levels may experience compromised contraction. Magnesium also affects the properties attributed to quinolones, tetracyclines, aminoglycosides, and vancomycin [1]; neuromuscular blocking agents; and volatile anesthetics [19].
<A>Guidelines for administering Mg therapy
<para1>Magnesium supplementation is generally considered to be safe and has resulted in few side effects [2]. While there is documentation for people/equipment-related overdoses, inadequately functioning kidneys pose the most serious risk. Consequently, it is recommended that kidney function be assessed prior to the use of Mg therapy [2]. Outcomes may differ depending on the variables that follow, making it very dif?cult to tie outcomes to Mg therapy based on dosage and timing
considerations.
<B>Mg salt
<para1>Many reports in the literature quantify the magnesium salt used in grams. Considering that multiple hydrates of Mg salts are available, the molar concentrations of Mg administered should always be stated. MgSO4 is the traditional formulation used in cardioplegic solutions and i.v. bolus infusions [1]. MgCl2 provides higher bioavailability than other commercial Mg preparations [45]. It is more advantageous pharmacologically and toxicologically than MgSO4 [1], and it avoids the reduction of Ca21 seen with the use of MgSO4 [46]. Mg-ATP improves organ blood ?ow, microcirculation, energy balance, and immune competence, leading to survival. It is a fast-acting vasodilator used for the management of acute pulmonary hypertensive crises, for maintenance of blood pressure during aortic cross-clamping, and as a therapeutic adjunct in patients with multiple organ failure [47]. It has been shown to improve organ function and survival time in a variety of animal models of oligemic shock, ischemia, and sepsis, and in human volunteers and patients with shock [48]. Mg oxide is better absorbed from the intestine than other commonly available compounds for oral Mg repletion [49].
<B>Routes of Mg administration
<para1>Oral Mg therapy may be used to maintain Mg levels under conditions associated with chronic Mg loss, e.g., use of diuretics [2]. It resulted in arterial pressure decreasing from 91 to 87 mm Hg, and a signi?cant reduction in all classes of arrhythmia (VPBs, couplets, and nonsustained ventricular tachycardia) [50,51]). Taking 16 mmol MgCl2・day21 (~400 mg elemental Mg) orally as tablets raised TMg from 0.87 mM to 0.92 mM after 6 weeks, and was accompanied by a large increase in urine excretion [50].
  <para2>Intravenous Mg therapy provides greater bioavailability than oral Mg therapy [52], acutely increasing intracellular Mg levels [53]. While a 30-mmol oral dose of Mg gave a 4.4% increase in TMg (0.91 to 0.95 mM) and a 6.3 % increase in iMg (0.48 to 0.51 mM) 2 h after ingestion, 15 mmol of Mg by i.v. gave a 50% increase in TMg (0.88 to 1.32 mM) and a 44% increase in iMg (0.48 to 0.69 mM) at 4 h [52].
  <para2>Cardioplegia solution with 15 mM Mg offered maximum recovery of myocardial function after an ischemia-reperfusion sequence. This is the concentration used in the St. Thomas II cardioplegia solution [53]. Using three cardioplegia solutions (3-4 mM Mg, 8-10 mM Mg, and 16-18 mM Mg), it was found that the 8- to 10-mM Mg cardioplegia solution yielded higher auto-resuscitation following surgery, shorter mechanical ventilation, shorter intensive care unit (ICU) stays, and the lowest creatine kinase MB isoenzyme (CK-MB) and troponin I levels [54]. The addition of Mg to warm blood cardioplegia resulted in a lower incidence of intraoperative and postoperative arrhythmias in patients undergoing urgent CABG for unstable angina [55].
  <para2>Intramuscular Mg therapy, by attaining a higher and more sustained serum Mg concentration, converted multifocal atrial tachycardia to normal sinus rhythm more rapidly (1-2 h) than i.v. Mg therapy (4-8 h) [56].
<B>Speed of Mg infusion
<para1>The pharmacologic effect of supplemented Mg is quite rapid, while the replenishment of body stores is much more gradual and slow [57]. Slower infusions (32-48 mmol Mg・24 h21) are appropriate unless cardiac arrhythmias or seizures are present [2]. Usually, at least 50% of an infused dose is wasted in the urine, even in patients who are profoundly Mg-depleted [5]. Replacement over 12-24 h will cause the plasma level of Mg to exceed Tmax (plasma TMg ~1.2 mM, the plasma level at which the kidney tubule resorbs less Mg from the glomerular ?ltrate) less often and will lead to less wasting [5]. Eight millimole Mg by i.v. over 1-2 min, followed by an additional 40 mmol over the next 5 h, is considered safe and probably effective for arrhythmias [2].
  <para2>In patients with normal renal function it is very
dif?cult to induce hypermagnesemia with commonly used regimens. Treatment of mild hypomagnesemia
with 50 ml of 4 mmol MgSO4 over 1 h or 25 mmol
MgSO4 over 6 h was acceptable [5]. For severe hypomagnesemia (life-threatening emergencies with known hypomagnesemia complicated by seizures, unremitting cardiac dysrhythmias, etc.), 8-12 mmol Mg over 5-10 min was found acceptable [5]. Rapid administration of Mg preparations can cause excessive ?ushing, sweating, warm sensations, hypotension, and vasodilation [5].
<B>Timing of Mg infusion perioperatively
<para1>Preoperative Mg supplementation protects the heart from anoxic and toxic challenges [14]. Preoperative treatment of patients undergoing mitral valve replacement with a slow-releasing oral MgCl2 showed a decrease in dysrhythmias postoperatively [5]. Extreme caution should be employed when Mg is given in combination with drugs that act synergistically, such as nitrates.
  <para2>Intraoperative Mg infusion is as effective as preoperative infusion in decreasing the rate of new-onset atrial ?brillation [7]. Cardioplegia containing Mg yielded CABG patients with higher Mg levels, fewer ischemic ECG changes, and fewer ventricular arrhythmias postoperatively [19].
  <para2>Postoperative Mg supplementation for several days may be used to maintain Mg levels. This may be important, considering that atrial arrhythmias frequently occur 2-3 days following surgery. One recommendation for postoperative Mg supplementation is 48 mmol for the ?rst 24 h postsurgery, followed by 12 mmol of Mg・day21 for the next 3 days (provided that kidney function is adequate; i.e., creatinine ,2 mg・dl21). Monitoring TMg daily is recommended [19].
<B>Dynamics of serum/plasma Mg perioperatively
<para1>Plasma Mg levels decrease during heart surgery. With no Mg supplementation, 18% (18/99) of patients were hypomagnesemic (TMg ,0.80 mM) preinduction, going to 71% (71/100) following CPB [8]. Table 3 describes several studies in which TMg was measured at various times over the perioperative course when no Mg was administered. Five studies giving preoperative TMg
values and those measured 1 day after surgery suggest that the median TMg decrease is 27%, ranging from 13% to 34% [4,15,24,60,61]. The lowest serum/plasma TMg values are generally seen after surgery and on day 1 postsurgery, gradually returning to their normal ranges over the next several days. Ionized magnesium activity also drops during surgery, but recovers more quickly than TMg after surgery, returning to its normal range within 48 h [31,62].
  <para2>Perioperative plasma Mg levels depend on the dosage and timing of Mg supplements. Table 3 tracks TMg in patients receiving Mg supplements. Predictably, TMg values obtained on the day of surgery will depend on the amount of Mg administered, when it was administered, and when samples were obtained (e.g., after CPB, ICU entry). Dose regimens in Table 3 include best estimates from the articles cited, including Mg present in cardioplegia solutions (volumes delivered) and molar estimates when solutions were reported in terms of grams of salt used (neglecting to mention water of hydration). Values obtained on the days following surgery will depend on the Mg status preoperatively, Mg administered during surgery, and on subsequent additions. Without the addition of Mg postsurgery, TMg levels were lower the day after surgery [4].
  <para2>Urine Mg levels rose in all patients following CPB no matter when or if they received supplementary Mg [61]. Subjects with normal Mg balance and renal function excrete most of a parenterally administered Mg load within 24 h [2].
  <para2>Plasma TMg does not correlate with myocardial TMg. Hypomagnesemia predated decreases in myocardial TMg by 2 to 6 weeks in patients with heart failure who commonly have persistent hypomagnesemia. Repletion of myocardial TMg occurred some weeks later than normalization of serum TMg levels [63]. Low myocardial TMg may contribute to the high incidence of fatal arrhythmic events [11]. Myocardial TMg is lower in patients with postoperative arrhythmia compared to those without arrhythmia [7]. Myocardial calcium levels are particularly high in Mg-depleted subjects [63].
<B>Dose-time approach to guiding Mg therapy
<para1>Table 4 illustrates observations of arrhythmias, made by several authors who related their ?ndings to dose and time of Mg administration. The studies included in Table 4 were chosen because each made mention of a TMg measurement either on the day of surgery or the day following, giving more information than simply dose and time. Generally, the incidence and intensity of the arrhythmias decreased with increasing Mg levels. A general impression one gets from the literature is that ventricular arrhythmias are of greater concern for the ?rst 24 h after surgery, while atrial arrhythmias tend to develop during the second and third postoperative days [6,7,19,20,66]. While higher levels of Mg supplementation favorably reduce the danger of arrhythmias, bradycardia may become a concern [35].
  <para2>In some dose-time studies, Mg supplementation was made via cardioplegia, only. Cardioplegia solution that was 15 mM in MgSO4 reduced the ischemic electrocardiographic changes, the frequency of ventricular ectopia, and the number of patients with ventricular arrhythmia relative to the control group. Atrial ?brillation occurred in 5/25 patients receiving Mg, compared to 8/25 not receiving Mg [16]. The addition of Mg to warm blood cardioplegia resulted in a lower incidence of intraoperative and postoperative arrhythmias [55].
  <para2>An interoperative bolus of 2 g MgCl2 (value in millimoles not given) in 100 ml normal saline, given over 30 min after termination of CPB, halved the frequency of postoperative ventricular dysrhythmias. It also increased the stroke volume, and thereby the cardiac
index, in the early postoperative period. Magnesium therapy also reduced the need for drugs to combat arrhythmias, and shortened the time patients spent on respirators [4]. Magnesium, given before cardioversion in patients suffering atrial arrhythmias following mitral and/or aortic valve surgery, diminished repolarization abnormalities and ventricular arrhythmias [67].
  <para2>An 8 mmol bolus of Mg after surgery reduced ventricular ectopia and lowered its grade, as categorized by a modi?ed Lown classi?cation [65].
  <para2>Intravenous Mg delivery during surgery and for the ?rst 24 h after surgery [23] reduced ventricular arrhythmias and helped control hypotension. Fifty patients, receiving 16 mmol MgSO4 by i.v. from induction to aortic cross-clamping and a second dose (32 mmol) starting at release of cross-clamp until 24 h later, were compared to 48 control patients who received no Mg. The patients receiving Mg had a lower incidence of ventricular arrhythmias (2% [1/50] vs 29% [14/48]) and were hypotension-free more frequently (4% [2/50] vs 33% [16/48]) [23].
  <para2>Continuous Mg infusion after surgery (no Mg in cardioplegia or prime) reduces atrial ?brillation. It was shown that 48 mmol Mg over the ?rst 24 h followed by 12 mmol・day21 for the next 3 days reduced the number and severity (requiring less cardioversion and multiple drug intervention) of episodes of atrial ?brillation, with no recognized adverse effects [15].
  <para2>Mg supplementation along with other drugs may alter clinical effects. Magnesium infusion with low doses of lidocaine is more effective for ventricular arrhythmia than lidocaine alone [1]. Magnesium and a low-dose beta-blocker were key to an aggressive atrial ?brillation prophylaxis regimen, cutting the atrial ?brillations in half (20% to 10%) [68]. Administration of MgSO4, resulting in iMg levels of 1.3 mM, prolonged neuromuscular blockade by 30-35 min (from 42 to 74 min) using cisatracurium [31].
<B>Risks associated with the dose-time approach to guiding Mg therapy
<para1>Perioperative Mg levels depend on the Mg status of the patient preoperatively, the amount of Mg administered, the rate of Mg delivery, kidney function, hemodilution, binding ligands, pH, and intraoperative hypoxia. Outcomes appear to depend on Mg levels at speci?c times over the postoperative period, making it dif?cult to predict outcomes from Mg therapy solely on dosage and timing.
  <para2>Contradictory dose-time outcomes have been documented. For example, one study reports that, of 100 patients receiving 12 mmol MgSO4 over 2 h preoperatively, perioperatively, and on postoperative days 0, 1, 2, and 3, 15% developed atrial ?brillation. Of the 100 patients not receiving the MgSO4, 16% also developed atrial ?brillation. This led to the conclusion that Mg infusion alone was not suf?cient for the prophylaxis of atrial ?brillation [69]. Another study reports that 100 patients who received 6 mmol MgSO4 in 100 ml of 0.9% NaCl solution over 4 h the day before surgery, just before CPB, and once daily for 4 days after surgery, were compared to 100 control patients who received the saline solution only. Intermittent 16-mM Mg cardioplegia was given to both the Mg and the control patients. Of the patients in the Mg group, 2% (2/100) had postoperative atrial ?brillation; 21% (21/100) had atrial ?brillation in the control group. It was concluded that the use of Mg in the preoperative and early postoperative periods is highly effective in reducing the incidence of atrial ?brillation after CABG [7].
  <para2>Other evidence demonstrates that the dose-time approach to Mg therapy may be inadequate. The ef?cacy of Mg therapy for ventricular arrhythmias has been debated [1,15]. In addition, the prevention of atrial ?brillation could not be attained with only routine (dose-time) administration of MgSO4 [69].
<B>Monitoring TMg as an approach to
guiding Mg therapy
<para1>Giving Mg supplements before or during CPB surgery, or not giving it at all, did not affect the incidence of arrhythmias and ventricular ?brillation after aortic unclamping; however, the plasma TMg levels did [19]. Even then, the correlation between serum TMg levels and clinical outcomes is poor [7,70], in part because a normal serum Mg level may coexist with tissue Mg de?ciency [7]. Clinical signs of hypomagnesemia lag considerably behind falling total serum levels. Therefore, early treatment of serum Mg changes may be important in preventing or reducing clinical dysfunction [71].
  <para2>Table 5 suggests that once TMg drops near the
lower limit of its reference range (~0.75 mmol・l21), the frequency of atrial and supraventricular arrhythmias
increases dramatically. The combination of signal-
averaged P wave duration before surgery, along with low serum TMg on the ?rst postoperative day, identi?ed the majority of patients with atrial ?brillation after coronary artery bypass surgery [6]. Of the patients with lower TMg (,0.8 mM) persisting into day 1 (66% of patients), 35% (22/63) required prolonged ventilatory support (mechanical ventilation .24 h after surgery),
in contrast to 12% (4/33) of the normomagnesemic
patients [4,8]. Maintaining TMg between 1.5 and 3.0 mmol・l21 is generally considered to be the level required to prevent and treat arrhythmia [31]. Only 4% (2/50) of patients with high Mg administration (TMg . 2 mM) had ventricular arrhythmias [23].
<B>Monitoring iMg as an approach to guiding therapy
<para1>Magnesium therapy during CPB should be based on iMg rather than TMg values [36,72]. Postoperatively, Mg supplementation should be continued despite normal serum TMg levels, because iMg hypomagnesemia may occur [19]. A large drop in iMg (but not in TMg) 24 h after CPB was seen in patients receiving 16 mM MgCl2 at bypass. The iMg drop correlated highly with preoperative iMg (r 5 0.96) [73]. In addition, the risk of hypermagnesemia from giving Mg to heart patients with kidney disease (e.g., creatinine .2 mg・dl21) might be reduced with careful monitoring of iMg.
  <para2>Samples (n 5 237), collected from 31 patients before, during, and after cardiac surgery, during which Mg-containing cardioplegia was used, were monitored for both TMg and iMg [74]. Scatter about the regression line between iMg and TMg (Sy・x 5 0.0914) suggested that the iMg value could vary between 0.54 and 0.92 mM in patients having a TMg value of 1.0 mM. In some instances, even greater differences were noted (e.g., in one sample, iMg was 0.3 mM and TMg was 1.1 mM).
It was suggested that the scatter about the regression line re?ects the differences in Mg bound to protein
and small ligands from sample to sample. Most of the samples with elevated iMg had values ranging from 0.7 to 1.1 mM; 3% (7/237) of samples were in the 1.2- to 1.3-mM range. The slope of the least-squares line comparing iMg to TMg was 0.71; suggesting that, on average, 71% of the Mg in the samples was unbound (ionized). Interestingly 2 of the 31 patients maintained a high iMg/TMg ratio for all samples following the addition of the cardioplegia, while 1 of the 31 maintained a low iMg/TMg ratio. As time progressed, both sets of Mg values (iMg and TMg) dropped, but, surprisingly, maintained either continuously high or low iMg/TMg ratios for each of the 3 patients. These 3 patients illustrate that iMg and TMg cannot be used interchangeably. It is not clear whether or not the iMg/TMg ratio may suggest the
underlying Mg status in the patients.
  <para2>Of 186 pediatric patients undergoing congenital cardiac surgery augmented with Mg in cardioplegia, 34% had low iMg (2 SD below the mean of the age-adjusted reference range) before surgery, 40% during CPB (particularly during cooling), 30% post-CPB, and 21% on admission to the ICU. Intraoperative ionized hypomagnesemia was more prevalent in patients who took furosemide before surgery. Patients with low iMg had higher mean lactate levels and required 1.65 3 the mechanical ventilation time of normomagnesemic patients. Low-iMg patients had longer cardiac ICU (CICU) stays that correlated most strongly with presurgical iMg values. Among the 14 patients who died during hospitalization, 12 (86%) had ionized hypomagnesemia during CPB [9,75].
  <para2>In one study, cardioplegia (St. Thomas II cardioplegia solution) containing 16.0 mmol・l21 of MgCl2 was infused at 10 ml・kg21 every 30 min during aortic cross-clamping into 18 patients. This amounted to an average of 38 6 9 mmol of Mg administered per patient. For the ?rst 20 min after unclamping, iMg values averaged around 1.55 mM (ranging from 1.2 to 2.0 mM). Vascular resistance dropped by 40%, and 14/18 patients required atrial or ventricular pacing. Hypotension and bradycardia after unclamping delayed weaning from extracorporeal circulation. It was concluded that monitoring iMg was needed to avoid side effects of a high Mg concentration when Mg-rich cardioplegia solutions were used [44].
<B>Re?ections and recommendations
<para1>The dynamics of Mg metabolism suggest that myocytes store Mg, 90% to 991% of which is bound to ATP, protein, and other ligands in the cytosol. Intracellular iMg, the remaining small percentage of the intracellular TMg, may pass through the cell membrane, and appears to establish equilibrium with extracellular iMg over a time span of minutes to hours, depending on the intra- to extracellular iMg gradient [3]. Chronic Mg wasting presurgery would result in low levels of intracellular TMg. An acute decrease in blood plasma iMg during cardiac surgery would reduce further the intracellular iMg. An acute bolus of Mg from supplementation would start to reverse the intracellular loss, but the kidneys rapidly void much of the added Mg. Unless the intracellular shortage of Mg is made up over the time following surgery (days), intracellular Mg may remain low. This could explain why extracellular Mg might drop, even after bolus Mg supplementation. Reduced myocardial iMg likely plays a major role in the clinical consequences following cardiac surgery.
  <para2>A target range for plasma iMg, which should pro-
vide therapeutic value without signi?cant risk during the time surrounding heart surgery, might be 0.60-1.0 mmol ・ l21. This range is based on the following considerations: (1) the clinical observations detailed in Tables 2, 4, and 5 are reported as a function of TMg, (2) on average, iMg in 70% of the TMg value, (3) iMg is a more clinically relevant value than TMg, (4) there is
a delay in intracellular iMg equilibration with plasma iMg, and (5) plasma iMg changes with time. The suggested range needs to be con?rmed
<A>Summary
<para1>Many factors affect the activity of the physiologically important ionized Mg fraction, iMg, in patients who undergo heart surgery. These factors include the use of diuretics before surgery; hemodilution, binding ligands, and hypoxia introduced during surgery; and the administration of Mg via pump prime, cardioplegia, or injection perioperatively. The consequences of abnormal Mg in these patients range from arrhythmia and respiratory complications when the patients are hypomagnesemic, to hypotension and bradycardia when they are hypermagnesemic.
  <para2>While Mg supplements are helpful for these patients, their clinical effects are best predicted by measurements of blood plasma iMg rather than by dosing schedules.
<A>References
<REF> 1. Pages N, Guiet-Bara A, Bac P, Bara M, Dupont C, Durlach J (2001) Pharmacological properties and indications of magnesium. In: Rayssiguier Y, Mazur A, Durlach J (eds) Advances in magnesium research: nutrition and health. John Libbey, Eastleigh, England, pp 453-458
<REF> 2. Noronha JL, Matuschak GM (2002) Magnesium in critical illness: metabolism, assessment, and treatment. Intensive Care Med 28:667-679
<REF> 3. Shirey TL (2000) Ionized magnesium (iMg): clinical importance and measurement [1]. Life Support and Anesthesia 7:265-270
<REF> 4. England MR, Gordon G, Salem M, Chernow B (1992) Magnesium administration and dysrhythmias after cardiac surgery. JAMA 268:2395-2402
<REF> 5. Brusco L (1994) Magnesium homeostasis in the operating room and intensive care unit. In: Eisenkraft JB (ed) Progress in anesthesiology VIII, chapter 3. WB Saunders, Philadelphia, pp 47-56
<REF> 6. Zaman AG, Alamgir F, Richens T, Williams R, Rothman MT, Mills PG (1997) The role of signal averaged P wave duration and serum magnesium as a combined predictor of atrial ?brillation after elective coronary bypass surgery. Heart 77:527-531
<REF> 7. Toraman F, Karabulut EH, Alhan HC, Dagdelen S, Tarcan S (2001) Magnesium infusion dramatically decreases the incidence of atrial ?brillation after coronary artery bypass grafting. Ann Thorac Surg 72:1256-1262
<REF> 8. Aglio LS, Stanford GG, Maddi R, Boyd JL, Nussbaum S, Chernow B (1991) Hypomagnesemia is common following cardiac surgery. J Cardiothorac Vasc Anesth 5:201-208
<REF> 9. Munoz R, Laussen PC, Palacio G, Zienko L, Piercy G, Wessel D (2000) Whole blood ionized magnesium: age-related differences in normal values and clinical implications of ionized hypomagnesemia in patients undergoing surgery for congenital cardiac disease. J Thorac Cardiovasc Surg 119:891-898
<REF>10. Kirkels JH, van Echteld CJA, Ruigrok TJC (1989) Intracellular magnesium during myocardial ischemia and reperfusion: possible consequences for postischemic recovery. J Mol Cell Cardiol 21:1209-1218
<REF>11. Millane TA, Ward DE, Camm AJ (1992) Electrophysiology, pacing, and arrhythmia: is hypomagnesemia arrhythmogenic? Clin Cardiol 15:103-108
<REF>12. Satur CMR, Anderson JR, Jennings A, Newton K, Martin PG, Unikrishnan N, Walker DR (1994) Magnesium ?ux caused by coronary artery bypass operation: three patterns of de?ciency. Ann Thorac Surg 58:1674-1678
<REF>13. Huss MG, Wasnick JD (1999) Magnesium and off-pump coronary artery bypass. J Cardiothorac Vasc Anesth 13:374-375
<REF>14. Birch RFH, Lake CL (1991) Magnesium therapy in cardiac surgery: Pro: magnesium is a valuable therapy in the cardiac surgical patient. J Cardiothorac Vasc Anesth 5:518-521
<REF>15. Fanning WJ, Thomas CS, Roach A, Tomichek R, Alford WC, Stoney WS (1991) Prophylaxis of atrial ?brillation with magnesium sulfate after coronary artery bypass grafting. Ann Thorac Surg 52:529-533
<REF>16. Shakerinia T, Ali IM, Sullivan JA (1996) Magnesium in cardioplegia: is it necessary? Can J Surg 39:397-400
<REF>17. Dorman BH, Sade RM, Burnette JS, Wiles HB, Pinosky ML, Reeves ST, Bond BR, Spinale FG (2000) Magnesium supplementation in the prevention of arrhythmias in pediatric patients undergoing surgery for congenital heart defects. Am Heart J 139:522-528
<REF>18. Roscoe A, Ahmed AB (2003) A survey of peri-operative use of magnesium sulphate in adult cardiac surgery in the UK. Anaesthesia 58:363-365
<REF>19. Gomez MN (1998) Magnesium and cardiovascular disease. Anesthesiology 89:222-240
<REF>20. Soucier RJ, Mirza S, Abordo MG, Berns E, Dalamagas HC, Hanna A, Silverman DI (2001) Predictors of conversion of atrial ?brillation after cardiac operation in the absence of Class I or III antiarrhythmic medications. Ann Thorac Surg 72:694-698
<REF>21. Creswell LL, Schuessler RB, Rosenbloom M, Cox JL (1993) Hazards of postoperative atrial arrhythmias. Ann Thorac Surg 56:539-549
<REF>22. Lowe JE, Hendry PJ, Hendrickson SC, Wells R (1991) Intraoperative identi?cation of cardiac patients at risk to develop postoperative atrial ?brillation. Ann Surg 213:388-392
<REF>23. Caspi J, Rudis E, Bar I, Safadi T, Saute M (1995) Effects of magnesium on myocardial function after coronary artery bypass grafting. Ann Thorac Surg 59:942-947
<REF>24. Karmy-Jones R, Hamilton A, Dzavik V, Allegreto M, Finegan BA, Koshal A (1995) Magnesium sulfate prophylaxis after cardiac operations. Ann Thorac Surg 59:502-507
<REF>25. Altura BT, Altura BM (1991) Measurement of ionized magnesium in whole blood, plasma and serum with a new ion-selective electrode in healthy and diseased human subjects. Magnes Trace Elem 10:90-98
<REF>26. Reynolds TR, Gef?n GA, Titus JS, O'Keefe DD, Daggett WM (1989) Myocardial preservation related to Mg content of hyperkalemic cardioplegic solutions at 8°C. Ann Thorac Surg 47:907-913
<REF>27. Storm W, Zimmerman JJ (1997) Magnesium de?ciency and cardiogenic shock after cardiopulmonary bypass. Ann Thorac Surg 64:572-577
<REF>28. Sjogren A, Edvinsson L, Fallgren B (1989) Magnesium de?ciency in coronary artery disease and cardiac arrhythmias. J Intern Med 226:213-222
<REF>29. Finlayson DC (1991) Mg: its time has come. J Cardiothorac Vasc Anesth 5:199-200
<REF>30. Altura BM, Altura BT (1995) Magnesium and cardiovascular biology: an important link between cardiovascular risk factors and atherogenesis. Cell Mol Biol Res 41:347-359
<REF>31. Pinard AM, Donati F, Martineau R, Denault AY, Taillefer J, Carrier M (2003) Magnesium potentiates neuromuscular blockade with cisatracurium during cardiac surgery. Can J Anaesth 50:172-178
<REF>32. Rayssiguier Y, Malpuech C, Nowacki W, Rock E, Gueux E, Mazur A (1997) In?ammatory response in magnesium de?ciency. In: Smetana R (ed) Advances in magnesium research. John Libbey, London, pp 415-421
<REF>33. Benzer W (1997) Signi?cance of supraphysiologic administration of magnesium after myocardial infarct. Wien Klin Wochenschr Suppl 2:38-41
<REF>34. Brown PS, Holland FW, Parenteau GL, Clark RE (1991) Magnesium ion is bene?cial in hypothermic crystalloid cardioplegia. Ann Thorac Surg 51:359-366
<REF>35. Wistbacka J-OM, Koistinen J, Karlqvist KEV, Lepojarvi MVK, Hanhela R, Laurila J, Nissinen J, Pokela R, Salmela E, Ruokonen A, Nuutinen LS (1995) Magnesium substitution in elective coronary artery surgery: a double-blind clinical study. J Cardiothorac Vasc Anesth 9:140-146
<REF>36. Grigore AM, Mathew JP (2000) Con: magnesium should not be administered to all coronary artery bypass graft surgery patients undergoing cardiopulmonary bypass. J Cardiothorac Vasc Anesth 14:344-346
<REF>37. Hecker BR, Lake CL, Kron IL, Mentzer RM, Crosby IK, Nolan SP, Crampton RS (1985) In?uence of magnesium ion on human ventricular de?brillation after aortocoronary bypass surgery. Am J Cardiol 55:61-64
<REF>38. Polancic JE (1991) Magnesium: metabolism, clinical importance, and analysis. Clin Lab Sci 4:105-108
<REF>39. Kasaoka S, Tsuruta R, Inoue T, Yamamoto M, Ueda S, Sadamitsu D, Tateishi A, Maekawa T (1994) Evaluation of a new ion selective electrode for ionized magnesium. J Jpn Soc Inten Care Med 2:119-121
<REF>40. Greenway DC, Hindmarsh JT, Wang J, Khodadeen JA, Hebert PC (1996) Reference interval for whole blood ionized magnesium in a healthy population and the stability of ionized magnesium under varied laboratory conditions. Clin Biochem 29:515-520
<REF>41. Springhouse (1997) Fluids and electrolytes made incredibly easy. Springhouse Corp., Springhouse, PA, p 123
<REF>42. Burtis CA, Ashwood ER (1999) Tietz textbook of clinical chemistry (3rd edn). WB Saunders, Philadelphia, p 1825
<REF>43. Hoshino K, Ogawa K, Hishitani T, Kitazawa R (2001) Characteristics of improved NOVA magnesium ion-selective electrode: changes of ionized magnesium values and reference interval in healthy children. Magnes Res 14:203-210
<REF>44. Ohashi Y, Uchida O, Kuro M (1996) The effect of magnesium rich cardioplegic solution upon blood ionized magnesium level during open heart surgery. Anesthesiology 85(3A):A138
<REF>45. Rodrigues-Moran M, Guerrero-Romero F (2003) Oral magnesium supplementation improves insulin sensitivity and metabolic control in type 2 diabetic subjects. Diabetes Care 26:1147-1152
<REF>46. Wilson RF, Binkley L, Sabo FM, Wilson JA, Munkarah MM, Dulchavsky SA, Diebel LN (1992) Electrolyte and acid-base changes with massive blood transfusions. Am Surg 58:535-545
<REF>47. Nalos M, Asfar P, Ichai C, Rademacher P, Leverve XM, Froba G (2003) Adenosine triphosphate-magnesium chloride: relevance for intensive care. Intensive Care Med 29:10-18
<REF>48. Soliman HM, Mercan D, Lobo SSM, Melot C, Vincent J-L (2003) Development of ionized hypomagnesemia is associated with higher mortality rates. Crit Care Med 31:1082-1087
<REF>49. Salem M, Munoz R, Chernow B (1991) Hypomagnesemia in critical illness: a common and clinically important problem. Crit Care Clin 7:225-252
<REF>50. Bashir Y, Sneddon JF, Staunton HA, Haywood GA, Simpson IA, McKenna WJ, Camm AJ (1993) Effects of long-term oral magnesium chloride replacement in congestive heart failure secondary to coronary artery disease. Am J Cardiol 72:1156-1162
<REF>51. Feyertag J, Laimer H, Herglotz P, Douglas T, Boettcher E, Ekmekcioglu C, Marktl W (1995) Effect of an oral magnesium supplementation on magnesium status and on ventricular arrhythmias. Magnes Res 8[Suppl 1]:27-28
<REF>52. Meludu SC, Castillo MJ, Scheen AJ, Lefebvre PJ (1995) Magnesium retention tests: marked differences in intravenous and oral magnesium administration. Magnes Res 8[Suppl 1]:52
<REF>53. Boyd WC, Thomas SJ (2000) Pro: magnesium should be administered to all coronary artery bypass graft surgery patients undergoing cardiopulmonary bypass. J Cardiothorac Vasc Anesth 14:339-343
<REF>54. Ji B, Feng Z, Liu J, Long C (2002) Myocardial protection related to magnesium content of cold blood hyperkalemic cardioplegic solutions in CABG. J Extra Corpor Technol 34:107-110
<REF>55. Yeatman M, Caputo M, Narayan P, Lotto AA, Ascione R,
Bryan AJ, Angelini GD (2002) Magnesium-supplemented warm blood cardioplegia in patients undergoing coronary artery revascularization. Ann Thorac Surg 73:112-118
<REF>56. Cohen L, Kitzes R, Shnaider H (1995) Mg in cardiac arrhythmias-digoxin-induced non-paroxysmal atrioventricular junctional tachycardia and multifocal atrial tachycardia responsive to parenteral magnesium sulphate. Magnes Res 8[Suppl 1]:18
<REF>57. Foley C, Zaritsky A (1998) Should we measure ionized magnesium? Crit Care Med 26:1949-1950
<REF>58. Harris MNE, Crowther A, Jupp RA, Aps C (1988) Magnesium and coronary revascularization. Br J Anaesth 60:779-783
<REF>59. Togashi H, Kasuda H, Inoue S, Hotta Y, Fukuda H (1997) Serum and urinary magnesium during and after cardiac surgery. Masui 46:1336-1341
<REF>60. Van den Bergh WM, Albrecht KW, Berkelbach Van Der Sprenkel JW, Rinkel GJ (2003) Magnesium therapy after aneurysmal subarachnoid haemorrhage: a dose-?nding study for long term treatment. Acta Neurochir 145:195-199
<REF>61. Skerrett PJ (1992) Magnesium touted for preventing arrhythmia complications following open heart surgery. Anesthesiology News 5:22
<REF>62. Hoshino K, Ogawa K, Hishitani T, Kitazawa R (2003) In?uence of heart surgery on magnesium concentrations in pediatric patients. Pediatr Int 45:39-44
<REF>63. Millane T (1993) Studies of magnesium in heart transplantation: further insights into magnesium and cardiac physiology. Magnes Res 6:361-368
<REF>64. Bert AA, Reinert SE, Singh AK (2001) A beta-blocker, not magnesium, is effective prophylaxis for atrial tachyarrhythmias after coronary artery bypass graft surgery. J Cardiothorac Vasc Anesth 15:204-209
<REF>65. Yurvati AH, Sanders SP, Dullye LJ, Carney MP, Archer RL, Koro PP (1992) Antiarrhythmic response to intravenously administered magnesium after cardiac surgery. South Med J 85:714-717
<REF>66. Rubin DA, Nieminski KE, Reed GE, Herman MV (1987) Predictors, prevention, and long-term prognosis of atrial ?brillation after coronary artery bypass graft operations. J Thorac Cardiovasc Surg 94:331-335
<REF>67. Szyszka A, Baszynska-Wachowiak H, Paluszkiewicz L, Tomczyk J (1995) Magnesium sulphate diminishes abnormalities of repolarization and ventricular arrhythmias after electrical cardioversion. Magnes Res 8[Suppl 1]:70
<REF>68. Gott JP, Cooper WA, Schmidt FE, Brown WM, Wright CE, Merlino JD, Fortenberry JD, Clark WS, Guyton RA (1998) Modifying risk for extracorporeal circulation: trial of four antiin?ammatory strategies. Ann Thorac Surg 66:747-754
<REF>69. Kaplan M, Kut MS, Icer UA, Demirtas MM (2003) Intravenous magnesium sulfate prophylaxis for atrial ?brillation after coronary artery bypass surgery. J Thorac Cardiovasc Surg 125:344-352
<REF>70. Zaloga GP (1989) Interpretation of the serum magnesium level. Chest 95:257-258
<REF>71. Marcus JC, Valencia G, Altura BM, Altura BT, Jean-Baptiste D, Sinha K (1998) Serum ionized magnesium (iMg) levels in premature and full-term infants. Pediatr Neurol 18:311-314
<REF>72. Salem M, Stacey J, Chernow B (1993) Ionized magnesium values in critically ill patients-a novel ion selective electrode for determining free extracellular magnesium concentrations. Crit Care Med 21 [Suppl]:S256
<REF>73. Brookes CI, Fry CH (1993) Ionised magnesium and calcium in plasma from healthy volunteers and patients undergoing cardiopulmonary bypass. Br Heart J 69:404-408
<REF>74. Altura BT, Shirey, TL, Young CC, Dell'Orfano K, Hiti J, Welsh R, Yeh Q, Barbour RL, Altura BM (1994) Characterization of a new ion selective electrode for ionized magnesium in whole blood, plasma, serum, and aqueous samples. Scand J Clin Lab Invest Suppl 217:21-36
<REF>75. Munoz R, Laussen PC, Zienko L, Palacio G, Piercy G, Wessel D (1999) Ionized hypomagnesemia in patients undergoing congenital cardiac surgery: frequency and clinical implications. J Am Coll Cardiol 33:Abstract 530A

ページのTOPへ


<JN>J Anesth (2004) 18:129-131
<PT>Clinical reports
<CT>Severe Legionella pneumophila pneumonia associated with the public bath on a cruise ship in Japan
<CA>Atsuko Kobayashi1, Yoshifumi Yamamoto2, Sumito Chou2, and Satoru Hashimoto3
<ADD>1 Department of Anesthesiology and Intensive Care, Saiseikai Suita Hospital, 1-2 Kawazono-cho, Suita 564-0013, Japan
<ADD>2 Department of Respiratory Medicine, Saiseikai Suita Hospital, Suita, Japan
<ADD>3 Department of Anesthesiology and Intensive Care, Kyoto Prefectural University of Medicine, Kyoto, Japan
<KW>Key words Legionella pneumophila ・ Public bath ・ Cruise ship
<A>Introduction
<para1>The relatively large number of patients with Legionella pneumophila pneumonia infected through attendance at bathing facilities is a Japanese characteristic [1]. Contamination of 24-h home bath water with legionella species has been reported, and eradication of the organism has been tried [2]. The Infectious Agents Surveillance Report (IASR) (http://idsc.nih.go.jp/iasr/iasr-ggl.html) reported that samples taken from 808 bathing facilities in Japan were contaminated by legionella species in 2003. However, the public baths on cruise ships traveling to and from Japan are poorly investigated, and there is no legal inspection by the appropriate public organization. We present a case of severe L. pneumophila pneumonia acquired from the publich bath-system on a cruise ship traveling from Osaka to Taiwan; this case raises the necessity for investigation of both systems on cruise ships.
<A>Case report
<para1>A 70-year-old Japanese man was admitted to the emergency room with the symptom of severe dyspnea. He had a 1-week history of diminished appetite, general fatigue, and a 2-week history of dry cough and mild fever. He had traveled to Taiwan on a cruise ship for 10 days. He had returned home 7 days prior to admission. He had used the public bath at least seven times on the cruise ship. As he showed dif?culty in breathing, and his SpO2 was 80% (ambient air) on admission, he was transferred to the intensive care unit (ICU). Because of his progressive respiratory failure, tracheal intubation and mechanical ventilation was needed, with continuous
intravenous administration of propofol. Arterial
gas analysis with ventilatory support (FIO2 1.0; PEEP, 5 cmH2O; SIMV 5 18 min21) showed pH 7.48, PaO2 66 mmHg, and PaCO2 55 mmHg. A chest radiograph
revealed bilateral diffuse pneumonia with existing
emphysema. A computed tomogram (CT) of the chest revealed multiple areas of in?ltration. Laboratory
examination showed a leukocyte count of 16 500/mm3; C-reactive protein (CRP) level was 36.4 mg/dl and total bilirubin level was 2.6 mg/dl. The administration of panapenem (1 g/day i.v.) was started. On hospital day 1, we obtained a specimen for determination of the causative microorganism by bronchoendscopy, but culture of the aspirated ?uid showed no signi?cant growth of organisms. Based on this ?nding, with the characteristics of the chest radiograph, we suspected interstitial pneumonia, and methylrednislone (1 g/day) was given for 3 days. In spite of the treatment, the in?ltrative shadow grew, and the patient's oxygenation remained unimproved. His condition continued to deteriorate, with ongoing multiple organ failure (MOF) and refractory hypotension.
  <para2>On hospital day 4, Legionella pneumophila was isolated from culture of the sputum on buffered charcoal yeast extract with alpha-ketoglutarate (BCYE-・) agar plates. Culture of bronchial lavage was positive for L. pneumophila serogroup 5. After con?rmation of these colonies, the antibiotic administration was switched
to erythromycin (2000 mg/day i.v.), cipro?oxacin (400 mg/day i.v.), and rifampicin (450 mg/day p.o.) in combination. Speci?c urinary antigen detection of L. pneumophila was later reported to be positive (Binax, Portland, ME, USA). His condition was ameliorated dramatically after the changing of antibiotics. Arterial gas analysis with ventilatory support (FIO2, 0.4; CPAP with PEEP, 5 cmH2O) showed pH 7.48, PaO2 78 mmHg, and PaCO2 41 mmHg on hospital day 9. His trachea was extubated but his oxygenation deteriorated in spite of treatment by nasal positive pressure ventilation. His trachea was reintubated on hospital day 10. He rapidly developed refractory hypoxemia and the PaO2/FIO2 (P/F)
ratio was under 70 on hospital day 11. CT of the chest revealed diffuse ?brotic change, especially in the lower lobe, and atelectasis of the dorsal side. Methylprednisolone (1 g/day) was given for 2 days for the alleviation of ?brotic change after the diffuse alveolar damage caused by L. pneumophila. His oxygenation gradually improved. A tracheotomy was performed on hospital day 13. The atelectasis of the dorsal side was ameliorated by respiratory therapy done with the patient in the prone position for the purpose of prompting excretion of the sputum. The patient was discharged from the ICU on hospital day 22.
  <para2>To identify the source of L. pneumophila, environmental sampling of the ship's water system was performed. Cultures taken from the water system in the men's bathroom yielded an isolate of L. pneumophila serogroup 5. Pulsed-?eld gel electrophoresis (PFGE) of DNA showed that the patient's pathogen and the strain obtained from the bath water were identical.
<A>Discussion
<para1>Legionella pneumophila is a common cause of sporadic community-acquired pneumonia. Cigarette smoking, chronic lung disease, advanced age, and immunosuppression have been consistently implicated as risk factors [3-6]. Our patient was 70 years old and a heavy smoker, with previously existing mild emphysema. His severe pneumonia required admission to the ICU and he rapidly showed MOF accompanied by septicemia. The pneumonia was progressive and characterized by acute multifocal alveolitis. Chest CT showed interstitial or intraalveolar edema with subsequent pulmonary ?brosis. We administered methylprednisolone after we diagnosed legionnaires' disease, because the patient's oxygenation deteriorated consistently in spite of the proper selection of antibiotics for L. pneumophila. On hospital day 9, his trachea was extubated, but his oxygenation deteriorated, and his trachea was reintubated on hospital day 10 in spite of the administration of erythromycin, cipro?oxacin, and rifampicin. Careful observation is required for L. pneumophila patients after the administration of methylprednisolone, because a withdrawal syndrome may occur, as happened in our case. Another reason for the treatment dif?culty was the delay of hospitalization in our patient. Besides, the patient's predisposing pulmonary condition affected the severity of the pneumonia.
  <para2>L. pneumophila is a facultative intracellular pathogen that infects human macrophages, monocytes, and epithelial cells [7-9]. In the aquatic environment, it can survive and multiply within ameba, which act as natural hosts. Bacteria growing within ameba are changed
phenotypically and exhibit an increased resistance to antibiotics and biocides when compared with cells grown in conventional media [10-12]. L. pneumophila is chlorine-tolerant; the organism survives the water
treatment process and passes into the water distribution system. The natural habitat of L. pneumophila appears to be aquatic bodies, including rivers, lakes, streams, and thermally polluted waters. In general, warm water seems to promote the growth of many legionellae, and very high numbers have been isolated from bio?lms associated with hot springs and from hot water plumbing systems. Jermigan et al. [13] reported an outbreak of legionnaires' disease associated with a cruise ship in 1996. This outbreak represented the ?rst documented instance of legionnaires' disease aboad a cruise ship docking in United States ports. Surveillance of cruise ships was reinforced after the outbreak. In Japan, the Ministry of Health, Labor, and Welfare reported that legionella species were detected in 2946 baths (16.7%) among 17 614 public baths that were investigated in 2003. The IASR reported, in 2003, that legionella species were detected in the samples taken from 808 bathing facilities in Japan. L. pneumophila serogroup 5 was detected in 155 (19.2%) of 808 samples. However, the public baths on cruise ships have been poorly investigated by the appropriate public organization. Soon after we reported this case, the public health center examined the bath system in this cruise ship. The men's and women's public baths proved to be contaminated by
L. pneumophila serogroups 5 and 1, respectively. L. pneumophila serograoup 5 was identi?ed as the strain in the water obtained from the men's bath, and PFGE
of DNA showed that the patient's pathogen and the strain obtained from the water of the men's bath mere identical. The samples obtained from the water of
the women's bath were positive for L. pneumophila serogroup 1. Another person (a woman) on the same ship suffered from legionnaires' disease and required admission to hospital in Tokyo at nearly the same time. Although culture of her sputum was negative, speci?c urinary antigen detection of L. pneumophila was positive. Our case suggests that surveillance for the presence of L. pneumophila and periodic check-ups of water supply systems on cruise ships are necessary to prevent legionnaires' disease outbreaks.
  <para2>In summary, a 70-year-old Japanese man presented with severe L. pneumophila pneumonia and MOF. The pathogen proved to have been acquired from the public bath on a cruise ship traveling from Osaka to Taiwan. The patient's strain of L. pneumophila and the strain of L. pneumophila obtained from the water in the public bath on the cruise ship were both identi?ed as serogroup 5. Although the patient showed MOF caused by septicemia, he was successfully treated with erythromycin, cipro?oxacin, rifampicin, and methylprednisolone. The public baths on cruise ships have been poorly investigated by the appropriate public organization in Japan. This case raises an important public health issue regarding the prevention of legionellosis in Japan.
<A>References
<REF> 1. Kuroki T, Yagita K, Yabuuchi E, Agata K, Ishima T, Katsube Y, Endo T (1998) Isolation of Legionella and free living amoebae at hot spring spas in Kanagawa, Japan. Kansenshougaku Zasshi (JJA Inf D) 72:1050-1055
<REF> 2. Li N, Aoyama T, Hori H, Ezaki T (1997) Isolation of Legionella pneumophila from 24-h home bath water and an eradication trial of the bacteria from the bath. Kansenshougaku Zasshi (JJA Inf D) 71:763-769
<REF> 3. Yu VL, Kroboth FJ, Shonnard J, Brown A, McDearman S, Magnussen M (1982) Legionnaires' disease: new clinical perspective from a prospective pneumonia study. Am J Med 73:357-361
<REF> 4. Roig J, Aguilar X, Ruiz J, Domingo C, Mesalles E, Manterola J, Morera J (1991) Comparative study of Legionella pneumophila and other nosocominal-acquired pneumonias. Chest 99:344-350
<REF> 5. Kool JL, Fiore AE, Kioski CM, Brown EW, Benson RF, Pruckler JM, Glasby C, Butler JC, Cage GD, Carpenter JC, Mandel RM, England B, Breiman RF (1998) More than 10 years unrecognized nosocominal transmission of legionnaires' disease among transplant patients. Infect Control Hosp Epidemiol 19:898-904
<REF> 6. Lepine L, Jernigan DB, Butler JC, Pruckler JM, Benson RF, Kim G, Hadler JL, Cartter ML, Fields BS (1998) A recurrent outbreak of nosocominal Legionnaires' disease detected by urinary antigen testing: evidence for long-term colonization of a hospital pumping system. Infect Control Hosp Epidemiol 19:905-910
<REF> 7. Pearlman E, Jiwa AH, Engleberg NC, Einstein BI (1988) Growth of Legionella pneumophila in human macrophage-like (u937) cell line. Microb Pathog 5:87-95
<REF> 8. Mody CH, Paine III R, Sharabadi MS, Simon RH, Pearlman E, Eisenstein BI, Toews GB (1993) Legionella pneumophila replicates with rat alveolar epithelial cells. J Infect Dis 167:1138-1145
<REF> 9. Horwitz MA (1983) Formation of a novel phagosome by the legionnaires' disease bacterium (Legionella pneumophila) in human monocytes. J Exp Med 158:1319-1331
<REF>10. Barker J, Brown MRW, Collier PJ, Farrell I, Gilbert P (1992) Relationship between Legionella pneumophila and Acanthamoeba polyohaga: physiological status and susceptibility to chemical inactivation. Appl Environ Microbiol 58:2420-2425
<REF>11. Kilvington S, Price J (1990) Survival of Legionella pneumophila within cysts of Acanthamoeba polyohaga following chlorine exposure. J Appl Bacteriol 68:519-525
<REF>12. Kuchta JM, Navratil JS, Sheoherd ME, Wodowsky RM, Dowling JN, States SJ, Yee RB (1993) Impact of chlorine and heat on the survival of Hartmannella verniformis and subsequent growth of Legionella pneumophila. Appl Environ Microbiol 56:4096-4100
<REF>13. Jermigan DB, Hofmann J, Cetron MS, Genese CA, Nuorti JP, Fields BS, Benson RF, Carter RJ, Edelstein PH, Guerrero IC, Paul SM, Lipman HB, Breiman RF (1996) Outbreak of Legionnaires' disease among cruise ship passengers exposed to a contaminated whirlpool spa. Lancet 347:494-499

ページのTOPへ


<JN>J Anesth (2004) 18:132-134
<PT>
<CT>Brachial plexus injury related to improper positioning during
general anesthesia
<CA>Pornswan Ngamprasertwong1, Vorapong Phupong2, and Ketchada Uerpairojkit1
<ADD>1 Department of Anesthesiology, Faculty of Medicine, Chulalongkorn University, Rama IV Road, Pathumwan, Bangkok 10330, Thailand
<ADD>2 Department of Obstetrics and Gynecology, Faculty of Medicine, Chulalongkorn University, Bangkok, Thailand
<KW>Key words Brachial plexus injury ・ Perioperative positioning
<A>Introduction
<para1>Brachial plexus injury is one of the most common nerve injuries associated with the improper positioning of the patient during general anesthesia [1,2]. Careful attention by the surgeon and anesthesiologist is essential to prevent such neurological injuries [3]. There have been a number of reports concerning brachial plexus injury following general anesthesia [2-4]. However, few of the reports are related to urological operations. We report an unusual case of brachial plexus injury related to the improper perioperative positioning of a patient who underwent laparoscopic radical nephrectomy.
<A>Case report
<para1>A 42-year-old man had a hand-assisted laparoscopic radical nephrectomy for renal cell carcinoma of the left kidney. His past medical history was remarkable for mild hypertension and renal insuf?ciency (creatinine clearance, 80 ml・min21). He had no preexisting neurological disorder. His height was 170 cm and weight, 71 kg (body mass index [BMI], 24.6). Anesthesia was induced in the usual manner using fentanyl, thiopental, and atracurium. The patient was placed in the right lateral decubitus position and leaned backward about 60° with a supported roll under the left scapula. His head was in the neutral position. His left arm was in hyperabduction of 120° and was suspended from an L-shaped bar
(Fig. 1). The procedure lasted for 7 h. The left radial arterial pulse was palpably strong and the pulse oximetry wave from the left hand was normal throughout the operation. There were neither prolonged hypotensive episodes nor other intraoperative complications. The
core temperature of the patient was kept above 35°C.
Anesthesia was maintained with N2O, O2, iso?urane, atracurium, and morphine.
  <para2>Postoperatively, the patient complained of weakness in his left upper extremity. The motor power was graded as follows: deltoid, 4; supraspinatus, 5; biceps, 4; brachioradialis, 4; triceps, 4; and wrist and ?nger ?exors/extensors, 3. Pinprick sensation was also decreased along his C7-T1 dermatome. Deep tendon re?exes of the left biceps and brachioradialis muscles were absent. A diagnosis of incomplete, total arm-type, brachial plexus injury was suspected. Four weeks after the
injury, an electrodiagnostic study was performed to
con?rm the diagnosis. The patient was treated conservatively with physiotherapy and a rehabilitation program. Over the next few days, sensation and motor power completely returned. At follow-up 1 month after discharge, the patient had hyperesthesia in the affected limb.
<A>Discussion
<para1>Brachial plexus injury is one of the most common nerve injuries related to malpositioning of the patient during general anesthesia. It is also an important cause of malpractice claims [1].
  <para2>The distribution of motor weakness and decreased sensation de?ned the lesion at the level of the brachial plexus in our patient. The absence of deep tendon re?exes effectively ruled out an upper motor neuron lesion. In addition, the pattern of motor weakness and sensory loss observed, combined with the absence of radicular pain, made a nerve root lesion quite unlikely as well. The extent of involved muscle was too great to be attributed purely to a peripheral nerve injury [5].
  <para2>Direct trauma, excessive stretching, external pressure, or some combination of these effects, was reported to have caused the acute onset of brachial plexus injury in such cases [3,4,6]. Our patient had no prior history of trauma or bleeding tendency. The surgical procedure was noninvasive, and there was no hematoma formation noted subsequently in his limb. These facts made stretching the most likely cause of his brachial plexus injury.
  <para2>The brachial plexus is susceptible to injury during anesthesia and surgery because it is long and ?rmly attached to the vertebrae and prevertebral fascia proximally, and the axillary fascia distally. Its lack of mobility and close proximity to bony structures such as the ?rst rib, clavicle, coracoid process, and head of the humerus cause it to be easily compressed when there is improper positioning of the patient [2]. This is especially true during general anesthesia, when muscle tone is already reduced by muscle relaxants and anesthetic agents [6]. Moreover, the anesthetized patient is unable to perceive pain or numbness caused by ischemia of the vasa nervorum arising from stretching and compression of the nerve bundle.
  <para2>In our patient, the injury to the brachial plexus was distal to the trunk level because of the sparing of the supraspinatus muscle. This muscle is supplied by the suprascapular nerve, the ?rst branch from the upper trunk of the brachial plexus. The most likely cause of the injury was thought to be a combination of the downward tilting of the head and the hyperabduction of the arm, which probably stretched the brachial plexus, especially at the cord level where it passes beneath the coracoid process of the scapula.
  <para2>The spectrum of injuries can vary from neurapraxia to axonotmesis [7]. Treatment of brachial plexus injury varies depending on the mechanism and the time the injury is discovered in relation to the inciting trauma
[6]. Current treatments include protection of the hypesthesic skin from further injuries, physical therapy to avoid muscle wasting and joint change, daily intermittent galvanic stimulation to the affected muscle, or surgery if recovery does not occur [3]. The recovery times may vary anywhere from hours to months. Sensation always returned ?rst, followed by motor function of the lower roots, then the upper roots [6].
  <para2>Not only the lateral decubitus position in urological procedures but also other types of operation and other positions cause brachial plexus injury [8,9]. In order to prevent or minimize the severity of brachial plexus injuries that may be related to the perioperative positioning of patients, the following recommendations have been made [7,10]:
<UL>? Avoid extension and external rotation in the supine position by limiting arm abduction to no more than 90° in the neutral position, using padded arm boards.
<UL>? Avoid extreme abduction in the prone position by tucking in and padding the arms by the patient's side, rather than abducting the arm above the head by more than 90°, which will unduly stretch the brachial plexus.
<UL>? In the steep head-down position, tuck the arms in at the patient's side with draw sheets. Suspension of the patient by the wrists or the use of shoulder braces to prevent the patient from sliding cephalad may increase the risk of brachial plexus injury.
<UL>? In the lateral decubitus position, always use a chest roll and avoid suspension of the arm from a L-shaped bar.
<UL>? In any position, always keep the head in neutral position. Rotation and lateral ?exion of the neck increase tension in the brachial plexus on the opposite side.
  <para2>In summary, although brachial plexus injury is one of the most common neuropathies occurring during anesthesia, awareness of the risk factors and positions which are likely to cause brachial plexus injury can limit their extent and occurrence. Constant attention to the positioning of the patient on the operating table can help to prevent this type of injury or the potential disability of patients.
<ACK>Acknowledgments. The authors express appreciation to Chairoj Uerpairojkit, M.D., orthopedic surgeon at the Institute of Orthopedics, Leardsin General Hospital, Bangkok, Thailand, for his critical review of this manuscript and his many insights and suggestions regarding this work.
<A>References
<REF> 1. Kroll DA, Caplan RA, Posner K, Ward RJ, Cheney FW (1990) Nerve injury associated with anesthesia. Anesthesiology 73:202-207
<REF> 2. Parks BJ (1973) Postoperative peripheral neuropathies. Surgery 74:348-357
<REF> 3. Po BT, Hansen HR (1969) Iatrogenic brachial plexus injury: a survey of the literature and of pertinent cases. Anesth Analg 48:915-922
<REF> 4. Bhardwaj D, Peng P (1999) An uncommon mechanism of brachial plexus injury: a case report. Can J Anesth 46:173-175
<REF> 5. Maurice V, Allan HR (2001) Diseases of the peripheral nerves. In: Maurice V, Allan HR (eds) Principles of neurology. Mc Graw-Hill, New York, pp 1370-1443
<REF> 6. Desai DC, Uribe A, Lachman T (1997) Brachial plexus injury due to compression: an alternate mechanism of injury: Case report and review of the literature. Am Surg 63:487-489
<REF> 7. Cooper DE, Jenkins RS, Bready L, Rockwood CA (1988) The prevention of injuries of the brachial plexus secondary to malposition of the patient during surgery. Clin Orthop 228:33-41
<REF> 8. Alan GP, Answorth AA, Ohannes AN (1997) Neurological injury in the upper extremity after total hip arthroplasty. Clin Orthop 345:181-186
<REF> 9. Souza EP, Durand PG, Sassolas F, Vial C, Lehot JJ (1998) Brachial plexus injury during cardiac catheterisation in children. Acta Anaesthesiol Scand 42:876-879
<REF>10. Anonymous (2000) Practice advisory for the prevention of perioperative peripheral neuropathies: a report by the American Society of Anesthesiologists Task Force on Prevention of Perioperative Peripheral Neuropathies. Anesthesiology 92:1168-1182
<FN>

ページのTOPへ


<JN>J Anesth (2004) 18:135-137
<PT>
<CT>Successful management of a patient with neuroleptic malignant syndrome associated with marked elevation of serum creatine kinase
<CA>Daisuke Takizawa1,2, Koichi Nishikawa1, Haruhiko Hiraoka1, Hiroshi Hinohara2, Shigeru Saito1,
Fumio Goto1, and Fumio Kunimoto2
<ADD>1 Department of Anesthesiology, Gunma University Graduate School of Medicine, 3-39-22 Showa-machi, Maebashi 371-8511, Japan
<ADD>2 Intensive Care Unit, Gunma University Hospital, Maebashi, Japan
<KW>Key words Dantrolene ・ Bromocriptine ・ Propofol ・ Dopamine receptor ・ Antipsychotic medication
<A>Introduction
<para1>Neuroleptic malignant syndrome (NMS) is a severe complication that occurs during the administration of neuroleptics, antidepressants, and medications for the treatment of Parkinson's disease. The syndrome is characterized by the sudden appearance of fever, motor rigidity, autonomic dysfunction, and increases in liver enzymes and creatinine kinase (CK) [1-3]. We report the intact survival of a patient with NMS associated with markedly elevated CK levels, in whom puri?cation therapy led to successful management.
<A>Case report
<para1>A 60-year-old man, 76.3 kg in weight and 170 cm in height, had been diagnosed with schizophrenia at age 24 years. Since then, he had been in and out of mental hospitals repeatedly. He had been treated with bromperidol (9 mg/day) on an outpatient basis for the past 7 years. Although he had residual delusions and hallucinations, his mental status was relatively stable. His past history included diabetes mellitus and hypertension, which had been controlled by oral medications. He was found by neighbors in an unconscious state at his home, and was brought to our hospital, where he was immediately transferred to the intensive care unit (ICU). Arterial blood pressure was 220/120 mmHg, with a heart rate of 120 beats/min, and body temperature was over 38°C. His level of consciousness according to the Glasgow Coma Scale was E4V3M5. NMS was diagnosed, as blood examination revealed typical ?ndings of NMS, i.e., elevated levels of GOT (2120 IU・l21), GPT (540 IU・l21), lactic dehydrogenase (LDH; 15 500 IU・l21), blood urea nitrogen (BUN; 69.8 mg・dl21), creatinine (Cr; 3.8 mg・dl21), WBC (15 400 ・mm3-1), and markedly elevated serum CK, reaching 161 440 IU・l21 (normal, ,230 IU・l21) (Table 1). Blood gas analysis indicated metabolic acidosis (pH, 7.37; PaO2, 73.6 mmHg; PaCO2, 26.2 mmHg; HCO3, 14.7 mEq/l; and base excess (BE), 28.7 mEq/l). The patient was intubated and sedated using propofol (300 mg・h21) for respiratory management. The patient was hydrated to prevent renal failure, and urine was alkalinized with bicarbonate infusion to inhibit the production of nephrotoxic metabolites of myoglobin.
  <para2>Treatments and changes in laboratory ?ndings
after admission to the ICU are summarized in Fig. 1. Administration of dantrolene at 40 mg・day21 and bromocriptine at 7.5 mg・day21 was immediately initiated to treat NMS. Plasma exchange (PE) was performed to facilitate the excretion of myoglobin. Plasauto EZ (Asahi Medical, Tokyo, Japan) was used to maintain blood ?ow at 4.8 l・h21. A Plasma?o ?lter (Asahi Medical) made of a 0.5 m2 membrane area was used, allowing replacement of 30% of plasma with fresh frozen plasma (40 units, 3200 ml). Plasma exchange was performed three times. In addition, continuous hemodia?ltration (CHDF) was performed for 5 days. A JUN-500 (UBE Medical, Tokyo, Japan) was used to maintain blood ?ow at 4.8 l・h21. A Rena?o II hemo?lter (Minnteck, Minneapolis, MN, USA), made of a 0.71 m2 polysulfone membrane area was used. Fortunately, the patient responded well to these treatments, with his elevated myoglobin concentration (14 120 ng・ml21) on admission to the ICU normalizing within 1 week (see Fig. 1). CK and transaminase levels also gradually decreased within a week. CHDF was terminated on day 6, and hemodialysis (HD) was performed instead, every 2 days. The patient was extubated on day 10 and his charged from the ICU 2 weeks after his admission. HD was required three times a week to treat renal dysfunction after his discharge from the ICU. Diuresis was noted on day 20, so HD was discontinued on day 45. Oral administration of quietiapine (50 mg・day21), a low-potency atypical antipsychotic agent, was started on day 45, with the dosage gradually increased to 400 mg・day21. His mental condition improved, and the patient was able to begin the rehabilitation process.
<A>Discussion
<para1>Most neuroleptics are thought to be capable of inducing NMS, including the newer atypical antipsychotics [3]. Previous studies have reported NMS as a severe complication, with mortality rates of 15%-25% or higher [3-5]. In NMS patients with elevated CK, toxic intracellular substances such as myoglobin are thought to be released into the systemic circulation, leading to multiple organ failure (MOF) [6]. Treatments that prevent acute renal failure (ARF) and MOF are therefore
particularly important to improve prognosis.
  <para2>The clinical symptoms of NMS patients can be
divided into two categories; those of central nervous system (CNS) origin, and those of muscular origin. Symptoms such as mental disorder, disturbed thermoregulation, and autonomic dysfunction result from the neuroleptic drug binding to dopamine receptors in the CNS. Dopamine agonists should thus be considered as ?rst-choice agents to reverse the receptor blockade by neuroleptics. Intracellular Ca21 metabolism is also disturbed in severe cases; these patients require dantrolene sodium to inhibit Ca21 release from the sarcoplasmic reticulum [2,7]. In general, serum CK levels indicate the extent of muscle injury. Because the CK level was extremely elevated in our patient, reaching 161 440 IU・l21 (normal, ,230 IU・l21), dantrolene was administered in addition to bromocriptine mesilate, a central dopamine agonist. These two drugs were successfully used to treat this syndrome.
  <para2>Myoglobin, a substance that can cause ARF, should be removed as soon as possible after it enters the circulation, as myoglobin-induced renal failure is one of the most serious complications in NMS. However, myoglobin cannot be eliminated by HD, due to its relatively high molecular weight (MW, 17 800). We therefore
used PE prior to HD for our patient with markedly elevated CK. CHDF was then initiated instead of HD, because myoglobin was shown to be eliminated better by convection (ultra?ltration) than by dialysis (diffusion) [8-10]. The effectiveness of continuous renal replacement therapy (CRRT) remains controversial in terms of preventing renal failure [11]. To use CRRT more effectively, myoglobin should be removed prior to CHDF in patients with myoglobinemia. We could not show that myoglobin clearance was accelerated by the PE, compared to CHDF alone, because the PE was continued intermittently during CHDF, but we believe that the use of PE prior to CHDF was useful in this patient.
<ACK>Acknowledgments. The authors thank Forte Science Communication Inc. (Tokyo, Japan) and Mr. S. Hardy-Yamada for English editing.
<A>References
<REF> 1. Levenson JL (1985) Neuroleptic malignant syndrome. Am J Psychiatry 142:1137-1145
<REF> 2. Caroff SN, Mann SC (1993) Neuroleptic malignant syndrome. Med Clin North Am 77:185-202
<REF> 3. Pelonero AL, Levenson JL, Pandurangi AK (1998) Neuroleptic malignant syndrome: a review. Psychiatr Serv 49:1163-1172
<REF> 4. Levenson JL, Fisher JG (1988) Long-term outcome after neuroleptic malignant syndrome. J Clin Psychiatry 49:154-156
<REF> 5. Hermesh H, Aizenberg D, Weizman A, Lapidot M, Mayor C, Munitz H (1992) Risk for de?nite neuroleptic malignant syndrome. A prospective study in 223 consecutive in-patients. Br J Psychiatry 161:254-257
<REF> 6. Weinstein R (2000) Therapeutic apheresis in neurological disorders. J Clin Apheresis 15:74-128
<REF> 7. Caroff S, Rosenberg H, Gerber JC (1983) Neuroleptic malignant syndrome and malignant hyperthermia. Lancet 1:244
<REF> 8. Bellomo R, Daskalakis M, Parkin G, Boyce N (1991) Myoglobin clearance during acute continuous hemodia?ltration. Intensive Care Med 17:509
<REF> 9. Berns JS, Cohen RM, Rudnick MR (1991) Removal of myoglobin by CAVH-D in traumatic rhabdomyolysis. Am J Nephrol 11:73
<REF>10. Amyot SL, Leblanc M, Thibeault Y, Geadah D, Cardinal J (1999) Myoglobin clearance and removal during continuous venovenous hemo?ltration. Intensive Care Med 25:1169-1172
<REF>11. Zager RA (1996) Rhabdomyolysis and myohemoglobinuric acute renal failure. Kidney Int 49:314-326

ページのTOPへ


<JN>J Anesth (2004) 18:138-140
<PT>
<CT>Pneumothorax associated with epidural anesthesia
<CA>Kosuke Miura, Shiro Tomiyasu, Sungsam Cho, Tetsuya Sakai, and Koji Sumikawa
<ADD>Division of Anesthesiology, Nagasaki University Graduate School of Biomedical Sciences, 1-7-1 Sakamoto, Nagasaki 852-8588, Japan
<KW>Key words Epidural anesthesia ・ Technique-related complication ・ Pneumothorax
<A>Introduction
<para1>Technique-related complications associated with epidural anesthesia include intravascular injection, subarachnoid injection, and neurologic injury [1]. Pneumothorax after thoracic epidural anesthesia is a rare but possible complication [2]. We report a case of pneumothorax probably associated with epidural needle insertion.
<A>Case report
<para1>A 17-year-old boy (height, 176 cm; weight, 56 kg) was scheduled for right bullectomy with video-assisted mini-thoracotomy, because his right pneumothorax had not improved with conservative therapy. Preoperative computed tomography revealed a multilobular bulla in the apex of the right lung, while neither obvious bulla nor bleb was observed in the left lung. He had no history of spinal anomaly. Epidural catheter insertion for postoperative analgesia was scheduled before the induction of general anesthesia.
  <para2>The ?rst insertion was attempted by a skilled resident supervised by an experienced anesthesiologist. The
patient was placed in the ?exed left lateral decubitus position. After his back was sterilized, the skin and subcutaneous tissue was in?ltrated with 1% mepivacaine deeply enough to contact the ligamentous
structure and lamina. A 17-gauge Tuohy needle was introduced 1.5 cm laterally from the midline via the left paramedian approach at the level of the Th7-Th8 interspace. After contacting the vertebral lamina, the angle of the needle was adjusted to identify the epidural space. The Tuohy needle was directed at an angle of 45° to cephalad, and, presumably, less than 15° toward the midline. A loss-of-resistance technique, using a saline-?lled glass syringe, detected a likely space at a depth of 6 cm from the skin. A test aspiration was done with the syringe to con?rm negative blood or cerebrospinal ?uid, when air was aspirated unexpectedly. At that moment, the patient leaned slightly toward the resident. Although no symptom of pneumothorax, such as respiratory distress or decreased breath sounds, was observed, the needle was withdrawn. The other, experienced,
anesthesiologist tried to insert the epidural needle via the left paramedian approach at the level of the Th8-Th9 interspace after adjusting the patient's position, but without changing it to another position. The Tuohy needle was introduced 1.0 cm laterally from the midline. After contacting the lamina, "walking" on the lamina facilitated the loss-of-resistance feeling obtained at a depth of 6 cm from the skin. The angle of the needle was then about 45° to cephalad, and 15° toward the midline. A test dose of 3 ml of 2% lidocaine was injected
from the epidural catheter. Bilateral Th7-Th9 thermal hypesthesia was con?rmed 5 min later with cold test. Subsequently, general anesthesia was induced with intravenous fentanyl, 100 Ig; propofol, 90 mg; and vecuronium, 9 mg. To achieve left-sided one-lung ventilation during the right bullectomy, a single-lumen endotracheal tube with a bronchial blocker was intubated. Anesthesia during one-lung ventilation was maintained with sevo?urane, 70% oxygen in nitrogen and intravenous fentanyl, and appropriate oxygenation and carbon dioxide elimination were maintained. The anesthesia and operative procedure were then uneventful.
  <para2>A routine postoperative chest radiograph, to verify proper pulmonary expansion, was taken about 4 h after the ?rst attempt at inserting the epidural catheter, and it revealed a left-sided pneumothorax (Fig. 1). Although left respiratory sound was slightly weak, the patient had normal oxygenation and tidal volume without symptoms of dyspnea. The endotracheal tube was extubated, and chest drainage was not placed. The pneumothorax disappeared spontaneously within 5 days (Fig. 2), and the patient was discharged on the tenth postoperative day.
<A>Discussion
<para1>Pleural puncture with subsequent pneumothorax is a rare but serious complication of epidural anesthesia, and several cases have been reported. Koch and Nielsen [3], and Furuya et al. [4] have reported the intrapleural misplacement of an epidural catheter found by surgeons during open thoracotomy. Iida et al. [2] reported a patient who required chest drainage after cholecystectomy because of hemothorax and the patient's complaint
of dyspnea. In these patients and ours, a "loss-of-
resistance feeling" was obtained. Thus, direct entrance of the Tuohy needle into the pleural cavity through relatively tight paravertebral tissue was, possibly, the course of the pleural puncture.
  <para2>The conventional insertion point of a thoracic epidural needle is 1 to 2 cm lateral to the superior margin of the spinous process, and the needle should be advanced at angles of 45° to 55° to cephalad, and 15° to 30° toward the midline to encounter the epidural space [5]. An alternative method is to insert the needle perpendicular to the skin in all planes until the lamina is encountered. The needle is then "walked" off the superior edge of the lamina and into the epidural space [6]. As Oku and Nishimoto [7] reported, conforming to the small angle toward the midline did not prevent the advancement of the Tuohy needle toward the opposite intravertebral foramen during puncture of the ligamentum ?avum. In these previous reports, misplacement of the epidural catheter occurred at the opposite side of the thoracic space [2,4,7]. In contrast, ipsilateral pleural puncture occurred in the present patient. The difference in the side of pleural puncture may be attributed to the different thickness of the subcutaneous fat tissue. In fact, the patients in the previous reports were obese, while our patient was relatively thin.
  <para2>Other causes of this adverse event could include a smaller angle to compensate for the lateral insertion point, a larger distance to the midline, and inadequate positioning of the patient. The small angle toward the midline itself was not inadequate [6]. The distance to the midline during the ?rst insertion was larger than that of the second insertion. The patient's position was, possibly, inadequate, i.e., he leaned slightly toward the resident. Thus, the combination of the long distance to the midline and seemingly adequate angle stemming from the inadequate patient's position would have contributed to the adverse event. As for patient positioning, Brown [1] suggested that inadequate positioning of the patient negates meticulous technique.
  <para2>To reduce the risk of misplacement of an epidural catheter, the insertion point should be made just beside the lateral margin of the spinous process, and the needle should be advanced with a smaller angle toward the midline than is generally accepted. Ramamurthy [5] noted that the epidural needle should be inserted from just next to the lateral margin of the cephalad edge of the spine. Minimizing the angle required to make the puncture in the ligamentum ?avum will provide a safer technique without the in?uence of variable subcutaneous fat density, because the epidural space can be
entered by "walking" off the superior margin of the lamina. The "walking" could signi?cantly increase the safety of the technique [5].
  <para2>Brismar et al. [8] reported that 4 of 21 patients who received pleural puncture, carried out to insert an intrapleural catheter for postoperative analgesia, developed minor pneumothorax. In their study, the need for great care to prevent this complication was emphasized. They recommended that patients should be instructed to hold their breath with an open airway after submaximal
inspiration at the time of catheter insertion. Although there is not much necessity to pay attention to the patient's respiration during epidural catheter insertion, the relatively massive air aspiration in our patient may have been in?uenced by his respiration.
  <para2>In summary, we encountered a case of ipsilateral pneumothorax associated with the conventional technique of thoracic epidural anesthesia. To reduce the risk of this complication, the thoracic epidural approach should be conducted with the insertion point of the needle near the midline and a small angle toward the midline, while meticulous attention is given to maintain an adequate patient position.
<A>References
<REF>1. Brown DL (2000) Spinal, epidural, and caudal anesthesia. In: Miller RD (ed) Anesthesia, 5th edn. Churchill Livingstone, New York, pp 1491-1519
<REF>2. Iida Y, Kashimoto S, Matsukawa T, Kumazawa T (1994) A hemothorax after thoracic epidural anesthesia. J Clin Anesth 6:505-507
<REF>3. Koch J, Nielsen JU (1986) Rare misplacement of epidural catheters. Anesthesiology 65:556-557
<REF>4. Furuya A, Matsukawa T, Ozaki M, Kumazawa T (1998) Interpleural misplacement of an epidural catheter. J Clin Anesth 10:425-426
<REF>5. Ramamurthy S (2001) Thoracic epidural nerve block. In: Waldman SD (ed) Interventional pain management, 2nd edn. WB Saunders, Philadelphia, pp 390-395
<REF>6. Bernards CM (2001) Epidural and spinal anesthesia. In: Barash PG, Cullen BF, Stoelting RK (eds) Clinical anesthesia, 4th edn. Lippincott Williams & Wilkins, Philadelphia, pp 689-713
<REF>7. Oku S, Nishimoto N (1977) A case of epidural catheter accidentally inserted along the left 11th intercostal nerve (in Japanese with English abstract). Masui (Jpn J Anesthesiol) 26:190-193
<REF>8. Brismar B, Pettersson N, Tokics L, Strandberg A, Hedenstierne
G (1987) Postoperative analgesia with intrapleural administration of bupivacaine-adrenaline. Acta Anaesthesiol Scand 31:515-520
<FIG>

ページのTOPへ


<JN>J Anesth (2004) 18:141-143
<PT>Short communication
<CT>Integration of suppression ratio in the bispectral index
<CA>Toshiya Koitabashi
<ADD>Department of Anesthesiology, Ichikawa General Hospital, Tokyo Dental College, 5-11-13 Sugano, Ichikawa 272-8513, Japan
<KW>Key words Algorithm ・ Bispectral index ・ Suppression ratio
<para1>The bispectral index (BIS) is derived from the electroencephalogram (EEG) and is a speci?c and sensitive monitor for assessing the hypnotic component of general anesthesia and sedation. The BIS has been described in the literature as a composite index consisting of a weighted combination of four components; however, the details of the BIS algorithm have not been described in their entirety by the manufacturer of the BIS monitors. It has been reported that the four component subparameters are derived from time-domain, bispectral, and power spectral analyses of the EEG [1]. The time domain subparameter includes Quazi suppression detection and the suppression ratio (SR) [1], and these two subparameters represent the deepest anesthesia conditions. The bispectral domain subparameter is called SyncFastSlow [1]. It represents the low-
frequency feature and is associated with moderate
anesthetic effect. The frequency domain subparameter is called the relative beta ratio [1]. It represents the high-frequency feature and is associated with light anesthetic effect and beta activation. Therefore, the BIS is a combination of the four subparameters described above, and the individual subparameters make the BIS a precise, near-linear function across the continuum of clinical states from awake to isoelectric EEG. However, of the four subparameters, only the SR is available for recording via the processed EEG port. In this article, the author presents the relationship between the BIS and the SR at deeper anesthesia and identi?es how the SR is incorporated into the BIS.
  <para2>After gaining approval by the Human Investigations Committee (Ichikawa General Hospital, Tokyo Dental College, Chiba, Japan) and obtaining informed written consent, the author enrolled 40 patients (30 men and 10 women) who underwent elective surgery under general anesthesia monitored using the BIS. Patients were excluded if any of the following conditions were met: seizure disorder; or longterm opioid, sedative, or alcohol use. Premedication consisted of 0.5 mg・kg21 of atropine, with or without midazolam 1-3 mg, intramuscularly 30 min prior to the anesthetic induction. After arrival in the operating room, patients were connected to standard physiological monitors. EEG (At1-Fpzt or At2-Fpzt) was measured using a self-preparing sensor (BIS Sensor) and BIS monitor (A-1050; BIS version 3.4;
Aspect Medical Systems, Newton, MA, USA). The BIS and SR were recorded continuously by a computer. Anesthesia was maintained with either propofol or sevo?urane, with or without regional anesthesia. The selection of the agents and their concentrations were at the discretion of the attending anesthesiologists. After the surgery, the author evaluated the relationship between the BIS and the SR from the obtained data.
  <para2>Statistical analysis was performed using the Pearson correlation coef?cient to evaluate linear regression.
  <para2>The patients' average age was 42 6 14 years. The average weight and height were 65 6 16 kg and 160 6 11 cm, respectively. After IV induction, all patients
were endotracheally intubated with vecuronium and ventilated mechanically. End-expiratory CO2 was maintained between 30 and 35 mmHg.
  <para2>The BIS decreased in an anesthetic dose-dependent fashion in all patients and reached a plateau of 30 (Fig. 1). In all patients with BIS in the 30s, the SR value was between 0 and 40%. In seven of the 40 patients, the SR exceeded 40%. In these situations, the author observed an inverse proportional relationship between the BIS and the SR. For an SR greater than 40%, the BIS was completely determined from the SR component. The author found a 1-point decrease of the BIS for each 2% increase of the SR for an SR greater than 40%.
  <para2>This study evaluated the relationship between the BIS and the SR in various anesthetic states, especially in deep anesthesia (strong hypnotic effect). As the concentrations of the anesthetics increased, the BIS decreased in response to the hypnotic effect of the anesthetics. The BIS decreased to 40 without suppression occurring in the EEG. With an additional increase in concentration, the BIS decreased to 30 with or without suppression
in the EEG. When suppression occurred more than 40% of the time (i.e., SR . 40%), the BIS decreased to below 30 with increasing anesthetic concentration, in direct inverse proportion to the increase of detected suppression.
  <para2>Anesthesiologists have sought a direct and reliable method of measuring anesthetic drug effects on the brain. The EEG is an obvious brain-monitoring modality, because it is a continuous, noninvasive measure of brain activity. However, until the middle 1990s, there were signi?cant problems in using traditional EEG monitors in the operating theater. The most important problem was the dif?culty of using only power spectral analysis to interpret the patients' precise hypnotic state. Anesthetic agents typically alter the low-amplitude, high-frequency EEG of the awake state to produce a high-amplitude, low-frequency signal. If this change were the sole effect of anesthetics on the EEG, the depth of the anesthesia would be easy to understand. However, there are some other effects of anesthetic agents that are more dif?cult to quantify, including beta activation, near suppression, burst suppression, and synchronization. Sedative-hypnotic agents, such as benzodiazepines and barbiturates, produce biphasic effects on the EEG. Low doses increase the high-frequency activity, while larger doses decrease the high-frequency activity. Very large doses of these hypnotics or inhalational agents produce near suppression and burst suppression, the latter consisting of periods of isoelectric (?at) EEG, interspersed with bursts of high-amplitude activity [2]. The BIS monitor is different from previous monitors, as it is the ?rst one to adopt other methods of analyzing raw EEG data [3]. As described above, the combination of the four subparameters produces a single number at which each of the subparameters is chosen to have a speci?c range of anesthetic effect where that subparameter performs best. The BIS weights the relative beta ratio most heavily when the EEG has the characteristics of light sedation [1]. The SyncFastSlow subparameter is well correlated with behavioral responses during moderate sedation or light anesthesia [1]. The burst-suppression ratio detects deep anesthesia [1]. With regard to suppression in the current version the BIS monitor recognizes ?at periods of the EEG based on a linear combination of the log of the total power between 2 and 30 Hz and the log of the total power between 31 and 40 Hz [4]. A linear combination means that there is a weighted ratio between the two frequency ranges, but the details of this ratio are not available because of proprietary concerns.
  <para2>However, this study revealed that suppression was not weighted heavily until it exceeded 40%. Once suppression occurred more than 40% of the time, the BIS was completely determined by the SR. Therefore, when the BIS is in the 30s, the SR information must be
ignored, and either Quazi suppression or SyncFastSlow may be the dominant subparameter expressed in the BIS. If the algorithm does not work like this; for example, if the SR information is re?ected to the BIS directly in a continual fashion, sudden changes of the BIS value will occur when low levels of suppression are detected. Because this situation can often be seen at many different concentrations of anesthetics, the manufacturer of the BIS monitors established the algorithm in this way in order to reduce the possibility that there would be ?uctuations in the BIS due to intermittent incorrect detection of these short episodes of suppression (personal communication from Paul J. Manberg, Ph. D., Vice President of Aspect Medical Systems). Actually, it is very dif?cult to correctly identify brief periods of true suppression versus low-amplitude signals or false suppression, and the transition into a burst-suppression pattern is quite variable between patients, so the author believes that the manufacturer has adopted the correct approach in not relying on the SR until it becomes signi?cant (.40%). However, from this study, a BIS between 30 and 40 has the possibility of indicating a wide variety of EEG states (i.e., with or without suppression), suggesting that a modi?cation of the algorithm will be necessary in order to make the BIS have a more linear function, especially for BIS values in the 30s.
  <para2>In conclusion, there is a zone of anesthesia in which the detection of some levels of EEG suppression does not result in a decrease in the BIS, although it was intended by the manufacturer to avoid such ?uctuations in the BIS. The BIS is completely determined from the SR information when the SR exceeds 40%.
<A>References
<REF>1. Rampil IJ (1998) A primer for EEG signal processing in anesthesia. Anesthesiology 89:980-1002
<REF>2. Bowdle TA (1999) The bispectral index (BIS): routine measurement of depth of hypnosis during anesthesia. Curr Rev Clin Anesth 19:169-180
<REF>3. Johansen JW, Sebel PS (2000) Development and clinical application of electroencephalographic bispectrum monitoring. Anesthesiology 93:1336-1344
<REF>4. Sigl JC, Manberg PJ, Chamoun NG, Chiang H-H, Devlin PH, Rampil IJ, Greenwald SD (1995) Quanti?cation of EEG suppression during anesthesia: correlation with iso?urane dose and patient responsiveness. Anesth Analg 80:S447
<FN>
<TB>

ページのTOPへ


<JN>J Anesth (2004) 18:144-145
<PT>Letters to the editor
<LCT>Modi?ed long Trachlight wand for a double-lumen endobronchial tube
<LCA>Ryuiku Watanabe
<LA>Division of Anesthesia, Hakuaikai Hospital, 28-25 Sasaoka 1-chome, Chuo-ku, Fukuoka 810-0034, Japan
<para1>To the editor: Tracheal intubation using a lightwand device is a useful technique. However, a lightwand is not long enough to intubate a double-lumen endobronchial tube (DLT). We, therefore, modi?ed a Trachlight wand and stylet (Laerdal, Wappingers Falls, NY, USA) and were easily able to intubate a DLT in a patient with protruding teeth. This simple modi?cation has made it possible to develop a wand of any length, even for extraordinarily long endotracheal tubes, and to use a Trachlight for the intubation of these tubes, especially in patients in whom laryngoscopy is dif?cult.
  <para2>The stem of the original Trachlight wand is a simple vinyl tube that contains two electric wires. These wires pass through two small channels on either side of the main channel of a stylet. Because the bore of these wire channels is relatively large, two wires can pass through each channel. Although the stem of the wand for adults is elliptical and measures 4.5 mm along the major axis and 35 cm in length, the diameter of the light bulb used is 5.5 mm.
  <para2>Using two wands for adults, we developed a single wand, 45 cm in length. The base of one of the wands was cut off, and its vinyl tube was cut, 33 cm from the tip, while preserving the two electric wires. The light bulb of another wand was cut off, and its vinyl tube was cut, leaving 12 cm in a similar manner (Fig. 1A). After the removal of rust from the electric wires, these two wands were joined by placing four electric wires into the two wire channels (two wires in each channel) to maintain electrical contact until the cut surfaces were put together exactly (Fig. 1B). A thin tape, 5 cm in width (Kitchen Tape W; Cemedine, Tokyo, Japan), was wound around the joint. The surfaces of the tape were nonconductive, and its adhesiveness was so strong that the joint supported a weight of 5 kg in our test until the vinyl of the wand was stretched to 1.3 times its original length.
  <para2>A stylet, with a diameter of 2 mm and a length of 49 cm, was made from stainless steel wire. Both its ends were rounded using sandpaper. It was bent at a right angle, leaving 44.5 cm at one end, to form a trigger. Because these four wires, the stylet, and the tape worked as internal and external stents, the structural stability and isolation of the wand could be secured.
  <para2>For the intubation of the DLT, the wand, along with the stylet, was attached to the handle. An optimal length of the wand was inserted into the DLT and ?xed to the handle with adhesive plaster at a site close to the clamp lever, instead of locking a tube connector (Fig. 1C). As per the usual procedure, the wand and the DLT were inserted into the trachea, with the patient's head in a neutral position. Because the joint in the wand was most likely to come apart during its removal, we ensured that the distal part was secured as soon as the joint came out of the DLT. After the removal of the wand, the DLT was inserted into the desired bronchus, with the patient's head in an extended position.
  <para2>The internal diameter of a DLT is narrowest at the Y-bifurcation, and it varies according to the manufacturer. The wand we constructed passed through BlueLine DLTs (SIMS Portex, Keene, NH, USA) of 35 Fr and 37 Fr relatively easily, and it narrowly passed through a Broncho-Cath DLT (Mallinckrodt Medical, Athlone, Ireland) of 37 Fr, but could not pass through one of 35 Fr.
<LFN>Address correspondence to: R. Watanabe
<LFN>Received: October 7, 2003 / Accepted: December 3, 2003
<JN>J Anesth (2004) 18:146
<LCT>Pulmonary embolism after minor surgery in a patient with low-risk thrombocythemia
<LCA>Satoshi Hosoi, Takehiko Adachi, Tomoko Hara,
Tatsushi Nakagawa, Yukiko Sasaki, Yasushi Hara,
and Nobukata Urabe
<LA>Department of Anesthesiology, Kitano Hospital, 2-4-20 Ohgimachi, Kita-ku, Osaka 530-8480, Japan
<para1>To the editor: Essential thrombocythemia (ET) is a myeloproliferative disorder with chronic peripheral thrombocytosis. The risk of thrombosis in ET patients aged less than 60 years, with no thrombotic history, and a platelet count of less than 1 500 000/mm3 is not increased, and clinical observation alone is recommended, based on the results of a prospective study [1]. However, pulmonary embolism after abdominal surgery in a patient with low-risk thrombocythemia has recently been reported [2]. We report here the occurrence of pulmonary embolism shortly after hand and shoulder surgery in a patient with low-risk ET.
  <para2>A 54-year-old woman (height, 160 cm; weight, 57 kg) presented to our hospital for the removal of subcutaneous tumors in the right wrist and shoulder. Two years previously, she had been diagnosed with hypertension, and medication for it
was started. Thrombocytosis was also detected at that time, but was left untreated. Baseline laboratory studies on
admission revealed hemoglobin, 13.1 g ・ dl21; platelet count, 787 000 ・ mm321, and white blood cell count, 11.6 3 109 ・ l21. The patient underwent removal of the subcutaneous tumors in the right wrist and shoulder in the prone position. Anesthesia was induced with 200 mg of thiopental, 100 Ig of fentanyl, 2.5 mg of droperidol, and 8 mg of vecuronium, and maintained with sevo?urane and nitrous oxide. The trachea was intubated and an elastic bandage was applied to both legs to prevent deep venous thrombosis (DVT). During the removal of the right-wrist tumor, a tourniquet was used on the right arm for 1 h (250 mmHg). The time of surgery was 1 h 5 min, and that of anesthesia, 3 h 10 min. Blood loss was indeterminable; total infusion was 1000 ml, and urine output was 200 ml during the operation. No intraoperative hypotension was observed. Recovery from anesthesia was uneventful, and the patient was transferred to the surgical ward.
  <para2>Six hours after the surgery, she walked, uneventfully. Twelve hours after the surgery, she walked a second time
and developed nausea, cold sweats, and hypotension (systolic blood pressure, 60 mmHg). Infusion of Ringer's lactate
solution was started, but her heart rate increased to 140-150 ・ min21. Blood gas analysis revealed severe hypoxia
with mild hypercapnia (PaO2, 37 mmHg; PaCO2, 47.4 mmHg).
As pulmonary embolism was suspected, oxygen inhalation (5 l ・ min21) was begun, and 5000 units of heparin was given. Thrombus was identi?ed in the pulmonary artery by computed tomography (CT) with contrast. Echocardiography revealed thrombus in the right atrium and right ventricle.
  <para2>Fifteen thousand units of heparin and 120 000 units of urokinase were given per day. Warfarin was started 1 week after the onset day. The thrombus was not identi?ed in the pulmonary artery on CT with contrast after 7 days of urokinase therapy. She was discharged from the hospital on warfarin and cilostazol 22 days after the surgery, without sequelae.
  <para2>The present patient had no previous history of thrombus, was 54 years old, and had a platelet count of 787 000 ・ mm321. Although a bone-marrow biopsy was not performed, she was clinically considered to have low-risk ET, and clinical observation alone was performed. The duration of surgery and the amount of blood loss were both minimal. Accordingly, we employed only elastic bandages on the lower extremities during anesthesia, and used no other means of prevention of DVT. However, pulmonary embolism occurred, probably at the time of the second walk by the patient 12 h after the surgery. The ?ndings in this patient and a recent case report [2] suggest that the prevention of DVT by intermittent pneumatic compression and/or anticoagulant prophylaxis, such as low-dose heparin administration, should be considered in patients with low-risk ET, even for minor surgery.
<A>References
<REF>1. Ruggeri M, Finazzi G, Tosetto A, Riva S, Rodeghiero F, Barbui T (1998) No treatment for low-risk thrombocythemia: results from a prospective study. Br J Haematol 103:772-777
<REF>2. Akcevin A, Mert M, Turkoglu H, Paker T (2003) Right ventricular mass and pulmonary embolism in a patient with essential thrombocythemia. Blood Coagul Fibrinolysis 14: 79-81
<LFN>Address correspondence to: T. Adachi
<LFN>Received: July 23, 2003 / Accepted: December 13, 2003

ページのTOPへ

ジャーナルオブアネステージアTOP