- Email: [email protected]
Effect of domperidone on the ventilatory response to transient hyperoxia in patients anaesthetized with isoflurane
British Journal of Anaesthesia 1997; 79: 41–46
Effect of domperidone on the ventilatory response to transient hyperoxia in patients anaesthetized with isoflurane
I. T. FOO, P. M. WARREN AND G. B. DRUMMOND
Summary We have studied the ventilatory responses to transient hyperoxia in two groups of patients (n:10) anaesthetized with isoflurane (0.3 MAC); patients were allocated randomly to receive either domperidone or placebo orally before anaesthesia. In each patient, five two-breath oxygen tests were averaged and minute ventilation (VEinst) or mean inspiratory flow rate (VT/TI) for each post-test breath was compared with the mean values for these variables during baseline ventilation. A decrease to less than the 95% confidence limits of mean baseline values was considered a definite response. According to this definition, transient hyperoxia decreased VEinst in nine of 10 patients in the placebo group and in all patients in the domperidone group. Similar changes occurred in VT/TI, with eight of 10 definite responses in the placebo group and 10 of 10 in the domperidone group. Compared with placebo, in the domperidone group there were larger changes in VEinst (0.30 vs 0.55 litre min91 (P:0.05)) and VT/TI (8.5 vs 26.6 ml s91 (P:0.02)) from respective baselines. Peripheral chemoreceptors appeared to be active during isoflurane anaesthesia and domperidone pretreatment enhanced this activity by increasing respiratory drive. (Br. J. Anaesth. 1997; 79: 41–46).
remains unclear. Limited studies in humans suggest that the carotid bodies are the most likely site of action.9–11 However, the peripheral chemosensors appear to be active during light halothane anaesthesia. Clergue and co-workers12 used a method derived from Dejours’ test to assess the response to hypoxia from the transient decrease in ventilation after oxygen was substituted for room air, where the time course of response indicates a rapid (peripheral) reflex. We have assessed if peripheral chemoreceptors are active during isoflurane anaesthesia by observing ventilatory responses to transient hyperoxia during light isoflurane anaesthesia. This test, derived from Dejours’ test, is based on the concept that the initial response to hyperoxia (transient ventilatory depression) is mediated by peripheral chemoreceptors which respond more promptly, and that this effect occurs before the onset of secondary responses. The second aim of this study was to see if domperidone, a selective dopamine D2 receptor antagonist which does not cross the blood–brain barrier,13 could augment the response to this test in patients anaesthetized with isoflurane. We showed in a previous study14 that domperidone offset the suppressive effect of isoflurane on HVR, presumably by antagonism of dopamine D2 receptors in the carotid bodies.
Patients and methods Key words Anaesthetics volatile, isoflurane. Pharmacology, domperidone. Reflexes, chemoreceptors. Ventilation, spontaneous. Ventilation, hypoxic response.
Knill and co-workers have shown that the ventilatory response to hypoxia (HVR) in humans is easily depressed by volatile anaesthetics.1–3 Anaesthetic concentrations of halothane, enflurane and isoflurane virtually abolished HVR and even subanaesthetic concentrations of these agents reduced the response considerably. However, recent studies have shown that HVR is not as sensitive to isoflurane and ventilatory response to hypoxia remained in patients anaesthetized with 0.6–1.1 MAC of isoflurane.4–6 Animal studies also provide conflicting data as to whether or not HVR is depressed by volatile anaesthetics.7 8 The site of action of volatile agents on HVR
This double-blind, placebo-controlled study was approved by the local Research Ethics Committee. Written informed consent was obtained from 20 patients who were ASA I, not overweight and aged 18–65 yr. Patients were allocated randomly to receive either domperidone 20 mg or placebo tablets 1 h before arrival in the anaesthetic room. In the anaesthetic room, an i.v. infusion was commenced and non-invasive monitoring of arterial pressure, ECG and pulse oximetry started. General IRWIN T. FOO*, BSC(HONS), MB, BCHIR, MRCP(UK), FRCA, GORDON B. DRUMMOND, MA, MB, CHB, FRCA, Department of Anaesthetics, University of Edinburgh, Edinburgh. PATRICIA M. WARREN, BSC, PHD, Rayne Laboratory, Unit of Respiratory Medicine, Department of Medicine, Royal Infirmary, Edinburgh. Accepted for publication: March 18, 1997. *Address for correspondence: Department of Anaesthetics, Royal Infirmary of Edinburgh, Lauriston Place, Edinburgh EH3 9YW.
42 anaesthesia was induced with etomidate 0.3 mg kg91 i.v. and anaesthesia was deepened with isoflurane and nitrous oxide in oxygen, with the patient breathing spontaneously. When anaesthesia was judged to be of adequate depth for tracheal intubation, laryngoscopy was performed and the larynx sprayed with a metered dose of a lignocaine preparation (Xylocaine, Astra Pharmaceuticals) using no more than 40 mg. After a short delay, the trachea was intubated and the inspired oxygen concentration was reduced to and maintained at 20% for the duration of the study. At the same time, the inspired isoflurane concentration was adjusted to obtain a stable end-tidal concentration of 0.4%, as measured by a Brüel and Kjaer gas monitor 1304, sampling at 90 ml min91, with a response time of less than 360 ms.15 Headphones were placed on the patient’s ears and white noise was played to prevent noise affecting breathing, and the arterial pressure cuff was not automatically inflated routinely in case it altered breathing pattern. Measurements of ventilatory variables during anaesthesia were started after a stable end-tidal isoflurane concentration had been present for at least 20 min. The tracheal tube was connected via a screen pneumotachograph (F100L, Mercury Electronics) to an Ambu Hesse non-rebreathing valve. The pneumotachograph pressure signal was measured with a Furness FC044 differential pressure transducer (⫾10 mm H2O). The inspiratory side of the valve was connected to a large bore two-way tap which allowed rapid changes between two supply systems. Anaesthetic gases (20% oxygen, isoflurane and nitrous oxide) were provided from a 4-litre reservoir bag fitted with a spill valve to reduce positive pressure in the supply tubing. Oxygen was supplied (flow rate of 6 litre min91) from a T-piece system using a 1-m length of anaesthetic tubing as a reservoir. The deadspace of the breathing system was approximately 80 ml. Inspired and expired carbon dioxide and oxygen partial pressures and inspired and expired concentrations of isoflurane were measured continuously by the Brüel and Kjaer gas monitor. Gas monitor and flow signals were recorded on a lap-top computer using the commercial respiratory monitoring software Cardas (version 2.07 advanced, Oxcams, Oxford), running in MS-DOS (version 6.2) for later off-line analysis. The flow signal was integrated to obtain values for tidal volume (VT) and instantaneous minute ventilation (VEinst) calculated from tidal volume and respiratory cycle duration. Before each study, the pneumotachograph was calibrated against a 1-litre calibration syringe and the gas monitor was calibrated using a reference gas. When clinically stable anaesthesia was achieved, the tap was turned during expiration so that two breaths of 100% oxygen were given. Tests were repeated at 2-min intervals. Five satisfactory two-breath oxygen tests were recorded. At the end of the testing sequence, a venous blood sample was obtained for measurement of plasma concentrations of domperidone and surgery started.
British Journal of Anaesthesia
Figure 1 A: Mean inspiratory flow rate (VT/TI) from one patient showing the 10 breaths used as baseline and the post-test breaths analysed. The vertical lines indicate the two oxygen test breaths and the two breaths omitted from the analysis. B: Method used for analysis of transients. Baseline mean and 95% confidence limits (CL) used for testing against each post-test breath. Breaths below the horizontal dashed line are below the 95% CL of baseline and were considered definite responses (see text for further explanation).
ANALYSIS AND STATISTICS
We measured the 10 breaths before the two hyperoxic breaths as baseline. We ignored the two hyperoxic breaths (test breaths) and the following two breaths, as it required these additional breaths to elapse for exhaled oxygen to return to pre-test values. We then measured the eight subsequent breaths (fig. 1). Previous work has shown that maximum change in ventilation occurs after six breaths from the change in inspired oxygen.11 To reduce the influence of normal variation in breathing, the five hyperoxic tests for each patient were averaged on a breath-by-breath basis. We calculated means and 95% confidence limits (CL) for VEinst and mean inspiratory flow rate (VT/TI) for the 10 breaths used as baseline, and inspected the post-test breaths for a reduction to less than 95% CL of baseline. If any post-test breath was reduced below the 95% CL of baseline, this was considered a definite response provided that breaths increased again towards baseline after reaching a minimum value (fig. 1). From the VT/TI results, we recorded the post-test breath number with the smallest VT/TI. This breath was used subsequently for analysis of frequency, and inspiratory and expiratory time changes from baseline. The time from the hyperoxic stimulus to the breath with the smallest VT/TI was also noted. To assess the reliability and consistency of this method of analysis, a random selection of five sets of 22 breaths was obtained in each patient during baseline breathing and subjected to the same averaging and analysis to assess if similar results could be obtained without the hyperoxic test breaths. Definite responses for both the placebo and
Domperidone, isoflurane and transient hyperoxia
43
domperidone groups were compared for any significant difference using the chi-square test with Yates’ correction. To assess if domperidone augmented the response to the hyperoxic test breaths, the mean maximal reductions in VEinst and VT/TI from baseline was calculated for both the placebo and domperidone groups. These values, and TI and TE changes from baseline, based on the minimum VT/TI post-test breath, were compared between groups using the Mann–Whitney test. Statistical significance was set at the 5% level.
and in these cases, the first transient was used for analysis. Considering VEinst, nine of 10 patients who received placebo and 10 of 10 patients who received domperidone demonstrated definite responses to the test. When the analysis was repeated for VT/TI, there were eight definite responses in the placebo group and 10 in the domperidone group. There was no significant difference in response frequency between groups. For both placebo and domperidone groups, mean times to minimum VT/TI after two hyperoxic breaths were 13.5 (SD 2.0) s and 14.9 (2.6) s, respectively. This corresponded to the minimum VT/TI breath occurring between breaths 5 and 11. This appeared to depend on the ventilatory frequency of the individual patient when the testing was performed, with higher ventilatory frequencies associated with a larger number of breaths to the minimum. Randomly selected sequences of breaths in each patient subjected to the same analysis did not show any transient reduction in minute ventilation with recovery towards baseline values. Mean baseline VEinst and minimum minute ventilation after the test for each patient in each group are shown in table 1. The magnitude of the post-test reduction ranged from 0.04 to 0.53 litre min91 of mean baseline ventilation in the placebo group and from 0.21 to 1.74 litre min91 in the domperidone group. The overall mean reduction in ventilation was 0.30 litre min91 for the placebo group and 0.55 litre min91 for the domperidone group (table 2). These reductions differed significantly (P:0.05). Mean baseline VT/TI and minimum VT/TI after the test for each patient in each group are shown in table 1. Changes in VT/TI after the test varied from an increase of 1 ml s91 to a decrease of 27 ml s91 compared with mean baseline values (placebo) and
Results All patients completed the study without complications. The placebo and domperidone groups were comparable in age (mean 33 (range 26–53) and 34 (22–56) yr, respectively), weight (mean 73 (SD 15) and 72 (15) kg, respectively) and sex distribution (5M, 5F and 6M, 4F, respectively). Mean oxygen saturations were 92.7 (SD 2.3) % and 93.3 (1.9) %, respectively, and no patient had an oxygen saturation less than 90% when breathing an inspired oxygen concentration of 20%. Mean baseline end-tidal carbon dioxide partial pressures before the oxygen tests were 5.5 (range 4.5–6.0) kPa for the placebo group and 5.3 (4.2–6.3) kPa for the domperidone groups (ns). After oxygen, all patients showed a transient reduction in VEinst and VT/TI with recovery towards baseline values by the end of the analysis period. VEinst and VT/TI analysis showed that the post-test breath with the maximum reduction occurred consistently within the same breath or two for each patient. In three patients, there were two consecutive transient reductions in VEinst and VT/TI
Table 1 Mean (95% confidence limits) baseline VE, VT/TI, TI and TE, and post-test values after averaging of the five hyperoxic tests for each patient in the placebo and domperidone groups. Post-test VE and VT/TI, are minimum values and *indicates definite response, that is outside the 95% confidence limit of baseline VE, and VT/TI, respectively. Post-test TI and TE, values were derived from the minimum post-test VT/TI, breath number
Patient No.
Baseline VE (litre min91)
Placebo group 1 5.88 (5.81, 5.95) 2 7.76 (7.70, 7.82) 3 6.86 (6.80, 6.92) 4 6.76 (6.71, 6.81) 9 4.82 (4.78, 4.86) 11 5.01 (4.96, 5.06) 14 6.45 (6.36, 6.54) 16 5.37 (5.31, 5.43) 17 4.72 (4.67, 4.77) 18 5.88 (5.82, 5.94) Domperidone group 6 6.82 (6.72, 6.92) 7 5.43 (5.33, 5.53) 8 7.66 (7.50, 7.82) 10 6.11 (6.03, 6.19) 12 8.66 (8.54, 8.78) 13 7.17 (7.00, 7.34) 15 5.73 (5.66, 5.80) 19 5.76 (5.71, 5.81) 21 6.52 (6.44, 6.60) 22 6.81 (6.59, 7.03)
VE Minimum post-test (limit min91)
Baseline VT/TI, (ml s91)
VT/TI minimum post-test (ml s91)
Baseline TI (s)
Posttest TI (s)
Baseline TE (s)
Posttest TE (s)
5.84 7.34* 6.33* 6.25* 4.64* 4.74* 6.24* 4.99* 4.58* 5.59*
222 (217, 227) 295 (290, 300) 264 (260, 268) 253 (250, 256) 184 (181, 186) 183 (181, 185) 235 (232, 239) 207 (205, 209) 188 (186, 189) 203 (201, 206)
219 288* 237* 236* 176* 177* 237 199* 183* 196*
1.17 (1.14, 1.21) 1.00 (0.98, 1.02) 0.97 (0.95, 0.99) 0.95 (0.94, 0.97) 1.11 (1.10, 1.13) 0.89 (0.88, 0.90) 1.26 (1.24, 1.28) 1.12 (1.11, 1.14) 1.00 (0.98, 1.01) 0.89 (0.87, 0.91)
1.15 1.05 1.05 1.08 1.17 0.91 1.23 1.14 1.02 0.94
1.48 (1.46, 1.50) 1.27 (1.25, 1.29) 1.26 (1.23, 1.29) 1.18 (1.16, 1.20) 1.43 (1.41, 1.45) 1.06 (1.04, 1.08) 1.49 (1.46, 1.52) 1.42 (1.40, 1.44) 1.39 (1.37, 1.41) 0.94 (0.92, 0.96)
1.37 1.27 1.30 1.36 1.48 1.14 1.57 1.59 1.39 1.05
6.28* 5.12* 7.17* 5.65* 6.92* 6.56* 5.45* 5.55* 6.10* 6.39*
241 (233, 248) 210 (206, 214) 272 (263, 280) 226 (220, 231) 372 (365, 378) 266 (255, 277) 188 (187, 190) 206 (204, 209) 290 (281, 300) 240 (238, 242)
219* 197* 252* 191* 296* 229* 183* 200* 251* 228*
1.30 (1.25, 1.35) 1.39 (1.37, 1.42) 1.26 (1.21, 1.31) 1.11 (1.08, 1.14) 1.07 (1.05, 1.09) 1.10 (1.05, 1.15) 0.93 (0.92, 0.94) 0.90 (0.89, 0.91) 1.12 (1.07, 1.17) 0.64 (0.63, 0.65)
1.47 1.47 1.35 1.34 1.24 1.25 0.99 0.92 1.29 0.65
1.44 (1.38, 1.50) 1.84 (1.81, 1.87) 1.40 (1.37, 1.43) 1.32 (1.28, 1.36) 1.67 (1.64, 1.69) 1.31 (1.26, 1.36) 0.90 (0.89, 0.91) 1.03 (1.01, 1.05) 1.42 (1.39, 1.45) 0.72 (0.71, 0.73)
1.56 1.91 1.47 1.41 1.75 1.33 0.97 1.07 1.40 0.74
44
British Journal of Anaesthesia Table 2 Mean (range) baseline VE, VT/TI, TI and TE, and post-test values for the placebo and domperidone groups. Post-test VE and VT/TI are minimum values and post-test TI and TE values were derived from the minimum post-test VT/TI breath number. Magnitude of change between baseline and post-test values are shown and *indicates significant difference (P-0.05) from placebo Baseline VE (litre min91) Placebo Domperidone VT/TI (ml s91) Placebo Domperidone TI (s) Placebo Domperidone TE (s) Placebo Domperidone
Minimum post-test
5.96 (4.72, 7.76) 6.67 (5.43, 8.66) 223 (183, 295) 251 (188, 372) Baseline
5.65 (4.58, 7.34) 6.12 (5.12, 6.92) 215 (176, 288) 224 (183, 296) Post-test
⌬ Baseline—minimum post-test 0.30 (0.04, 0.53) 0.55 (0.21, 1.74)* 9 (91, 27) 27 (6, 76)* ⌬ Baseline—post-test
1.04 (0.89, 1.26) 1.08 (0.64, 1.39)
1.07 (0.91, 1.23) 1.20 (0.65, 1.47)
90.04 (90.13, 0.03) 90.12 (90.23, 90.01)*
1.29 (0.94, 1.49) 1.31 (0.72, 1.84)
1.35 (1.05, 1.59) 1.36 (0.74, 1.91)
90.06 (90.18, 0.11) 90.06 (90.12, 0.02)
decreases from 6 to 76 ml s91 (domperidone). The mean change was 99 ml s91 for the placebo group and 927 ml s91 for the domperidone group (table 2). These changes were statistically significantly (P:0.02). Timing variables (TI and TE) for the minimum post-test VT/TI breath were compared with respective mean baseline values. With the exception of data from two patients in the placebo group, inspiratory times were lengthened after the test in both groups compared with baseline values (table 1). The mean increase in inspiration time for the placebo group was 0.04 s and 0.12 s for the domperidone group (table 2). There was a significant difference between groups (P:0.02). Expiration times were more variable but the trend was towards increasing expiration time in both groups (table 1). The mean increase for both groups was 0.06 s (table 2). Mean plasma concentration of domperidone in the domperidone group was 7.9 (range 2.0–19.3) ng ml91. There was no correlation between plasma concentrations of domperidone and the size of the transient reductions in VEinst or VT/TI.
Discussion We have shown that peripheral chemoreceptor activity was present in a majority of patients during isoflurane anaesthesia. Pre-existing stimulus to breathing was shown by transient depression of both VEinst and VT/TI after two breaths of 100% oxygen. With domperidone pretreatment, patients showed greater reductions in both VEinst and VT/TI after two breaths of 100% oxygen. We reduced the respiratory depresent effects of other anaesthetic agents in this study by using etomidate for induction of anaesthesia, which has a short duration of action, and by avoiding the use of opioids. We averaged the repeated tests in each individual to reduce breath-to-breath variability so that small changes in ventilatory variables (as produced by this transient hyperoxia test) were more likely to be detected. Our method of testing for a definite response to hyperoxia is similar to that used in human studies involving single-breath carbon dioxide tests of peripheral chemosensitivity.16
Randomly selected breaths subjected to the same analysis did not demonstrate the clear pattern of reduction followed by recovery towards baseline and we believe that this method of analysis reliably detected transient changes in responses. In our analysis, we ignored the two hyperoxic breaths (test breaths) and the following two breaths as there was a delay for the expired oxygen to return to baseline and we did not correct the pneumotachograph for changes in gas composition. As previous work showed that the maximum change occurred six breaths after the change in inspired oxygen,11 we were unlikely to have missed the maximum change. We used a variation of Dejours’ test to assess the activity of the peripheral chemoreceptors during isoflurane anaesthesia. With this technique in resting awake humans, the magnitude of ventilatory change is approximately 8–10% of pre-stimulus baseline ventilation and occurs approximately 10 s after the stimulus.17 In our study, during isoflurane anaesthesia, mean ventilatory change was 4.9% in the placebo group and 7.7% in the domperidone group and mean time to minimum VT/TI after stimulus was 13.5 s in the placebo group and 14.9 s in the domperidone group. This is longer than in awake subjects. Washin of oxygen to the alveoli and hence the change in arterial oxygen content may take longer during anaesthesia because ventilation is less in the anaesthetized subject. Alternatively, control of ventilation may be different from the awake state. There are possible drawbacks of this test to assess chemoreceptor activity during anaesthesia. Hyperoxia can increase blood PaCO2 via the Haldane effect and thereby cause an increase in ventilation. Furthermore, hyperoxia can reduce cerebral blood flow which would tend to increase brain PCO2 and thus also ventilation.18 These effects, which are probably less rapid than effects at the peripheral chemoreceptors, would tend to oppose the decrease in ventilation caused by hyperoxia and therefore underestimate the magnitude of the ventilatory output of the peripheral chemoreceptors. Duffin, Triscott and Whitwam11 studied patients after opioid premedication and found that 0.7–0.8% halothane in air substantially decreased the ventilatory response to two breaths of oxygen compared with the response during thiopentone anaesthesia.
Domperidone, isoflurane and transient hyperoxia They concluded that halothane reduced the activity and responsiveness of peripheral chemoreceptors. As we did not have a group of patients in this study who did not receive isoflurane, we were unable to determine the effect of isoflurane anaesthesia per se. Our results only indicate that peripheral chemoreceptors appeared to be active during isoflurane anaesthesia. Different anaesthetic techniques such as the use of opioids and nitrous oxide and different ways of analysing the breaths may account for some of the differences observed. Furthermore, the anaesthetic concentrations used were different in the two studies; Duffin, Triscott and Whitwam11 used approximately 1 MAC of halothane while we used approximately 0.3 MAC of isoflurane and 80% nitrous oxide. This difference could have influenced the degree of suppression of the peripheral chemoreceptors by these volatile agents. Nevertheless, it is likely that in our patients with an end-tidal isoflurane concentration of 0.4% (approximately 0.3 MAC), peripheral chemoreceptor activity was still present and responsive to changes in oxygen tensions, but perhaps not as vigorously as in awake subjects. Others have studied patients anaesthetized with 1.5% halothane, using a method derived from Dejours’ test, and found that peripheral chemoreceptor activity was depressed but not abolished.12 Knill and coworkers3 found that even at 0.1 MAC of isoflurane, there was a 50% reduction in the ventilatory response to isocapnic hypoxia. This was confirmed recently by van den Elsen and colleagues10 who extended the concentration of isoflurane to 0.2 MAC and found a 65% reduction in isocapnic HVR. At anaesthetic concentrations of 1.1 MAC, HVR was virtually abolished.3 This suggested that HVR and therefore the peripheral chemoreflex pathway was highly sensitive to the effects of isoflurane. However, our results and those of Clergue and co-workers12 suggest that peripheral chemoreceptors are not “silent” at anaesthetic concentrations of isoflurane and halothane, respectively, suggesting that although peripheral chemoreceptor responses are depressed, this is perhaps not to the degree as suggested by previous investigators.3 Our findings also support recent studies by Lindahl’s group4–6 who found persistent ventilatory responses to hypoxia under poikilocapnic conditions in patients anaesthetized with 0.6–0.85 MAC of isoflurane and a reduction of 60–70% in HVR under isocapnic conditions with 0.85–1.1 MAC of isoflurane. Therefore, anaesthetic concentrations of isoflurane do not appear to suppress HVR to the same extent as that found by previous investigators.3 10 Differences in experimental conditions may account for some of these differences.10 Washin studies of subanaesthetic halothane and isoflurane9 10 suggest that the most likely site of action of volatile agents is at the peripheral chemoreceptors. Depression of hypoxaemic-driven ventilation by halothane and isoflurane occurred extremely rapidly, within the first minute of inhalation. For isoflurane, it was estimated that after 30 s of isoflurane inhalation, the carotid body isoflurane tension would be approximately 90% of end-tidal, whereas the brain isoflurane tension would be
45 approximately 8% of end-tidal.10 Therefore, ventilatory depression occurring within 30 s of isoflurane inhalation is best explained by an action of isoflurane at the peripheral chemoreceptors. However, Whitwam and co-workers19 demonstrated in anaesthetized patients that there was a mean delay of only 10 s in central chemoreceptor ventilatory response to i.v. injection of sodium bicarbonate (causing an increase in PaCO2 ) when peripheral chemoreceptor activity was depressed by hyperoxia. This suggests that the time lag between the effect of isoflurane at central and peripheral chemoreceptor sites during washin could also be in the region of 10 s. Therefore, a central effect of subanaesthetic isoflurane cannot be excluded in the first 30 s of isoflurane washin and overall suppression of HVR at anaesthetic doses could be caused by a combination of peripheral and central effects. The second aim of this study was to determine if domperidone, a selective dopamine D2 receptor antagonist which does not readily cross the blood–brain barrier,13 could augment the response to transient hyperoxia in patients anaesthetized with isoflurane. Domperidone augments HVR in humans20 and can offset the suppression caused by isoflurane on the ventilatory response to hypoxia.14 We found that with domperidone pretreatment, all patients showed positive results to transient hyperoxia suggesting that with increased ventilatory drive, it was easier to demonstrate the transient reductions in both VEinst and VT/TI after two breaths of 100% oxygen. Analysis using VT/TI instead of breath-by-breath minute ventilation demonstrated a greater difference between placebo and domperidone. VT/TI is a measure of respiratory drive allowing minute ventilation to be separated into drive and timing components: VEinst is the product of VT/TI and TI/TTOT (timing component where TTOT is the sum of TI and TE).21 When inspiratory and expiratory times were compared between groups, only TI was significantly different, with domperidone causing an increase in inspiration time. These results support the observations of Clergue and co-workers who found a prompt change in TI after oxygen breathing commenced.12 In summary, our data support the view that peripheral chemoreceptors remain active during isoflurane anaesthesia. Their function may be depressed by isoflurane, but perhaps not as much as suggested previously. Domperidone enhanced peripheral chemosensor activity by increasing respiratory drive during isoflurane anaesthesia.
Acknowledgements We thank Sanofi Winthrop for supplies of domperidone and Dr R. Woestenborghs, Janssen Research Foundation, for measurement of plasma concentrations of domperidone. I. T. F. was a British Journal of Anaesthesia Research Fellow.
References 1. Knill RL, Gelb AW. Ventilatory responses to hypoxia and hypercapnia during halothane sedation and anaesthesia in man. Anesthesiology 1978; 49: 244–251.
46 2. Knill RL, Manninen PH, Clement JL. Ventilation and chemoreflexes during enflurane sedation and anaesthesia in man. Canadian Anaesthetists Society Journal 1979; 29: 353–360. 3. Knill RL, Kieraszewicz HT, Dodgson BG, Clement JL. Chemical regulation of ventilation during isoflurane sedation and anaesthesia in humans. Canadian Anaesthetists Society Journal 1983; 30: 607–614. 4. Sjögren D, Sollevi A, Ebberyd A, Lindahl SGE. Poikilocapnic hypoxic ventilatory response in humans during 0.85 MAC isoflurane anaesthesia. Acta Anaesthesiologica Scandinavica 1994; 38: 149–155. 5. Sjögren D, Sollevi A, Ebberyd A, Lindahl SGE. Isoflurane anaesthesia (0.6 MAC) and hypoxic ventilatory responses in humans. Acta Anaesthesiologica Scandinavica 1995; 39: 17–22. 6. Sollevi A, Lindahl SGE. Hypoxic and hypercapnic ventilatory responses during isoflurane sedation and anaesthesia in women. Acta Anaesthesiologica Scandinavica 1995; 39: 931–938. 7. Davies RO, Edwards MW, Lahiri S. Halothane depresses the response of carotid body chemoreceptors to hypoxia and hypercapnia in the cat. Anesthesiology 1982; 57: 153–159. 8. Koh SO, Severinghaus JW. Effect of halothane on hypoxic and hypercapnic ventilatory responses of goats. British Journal of Anaesthesia 1990; 65: 713–717. 9. Knill RL, Clement JL. Site of selective action of halothane on the peripheral chemoreflex pathway in humans. Anesthesiology 1984; 61: 121–126. 10. van den Elsen M, Dahan A, DeGoede J, Berkenbosch A, van Kleef J. Influences of subanesthetic isoflurane on ventilatory control in humans. Anesthesiology 1995; 83: 478–490. 11. Duffin J, Triscott A, Whitwam JG. The effect of halothane and thiopentone on ventilatory responses mediated by the peripheral chemoreceptors in man. British Journal of
British Journal of Anaesthesia Anaesthesia 1976; 48: 975–980. 12. Clergue F, Ecoffey C, Derenne JP, Viars P. Oxygen drive to breathing during halothane anesthesia: effects of almitrine bismesilate. Anesthesiology 1984; 60: 125–131. 13. Champion MC, Hartnett M, Yen M. Domperidone, a new dopamine antagonist. Canadian Medical Association Journal 1986; 135: 457–461. 14. Foo IT, Martin SE, Lee RJ, Drummond GB, Warren PM. Interaction of the effects of domperidone and subanaesthetic concentrations of isoflurane on the immediate and sustained hypoxic ventilatory response in humans. British Journal of Anaesthesia 1995; 74: 134–140. 15. McPeak HB, Palayiwa E, Robinson GC, Sykes MK. An evaluation of the Brüel and Kjaer monitor 1304. Anaesthesia 1992; 47: 41–47. 16. Fernando SSD, Saunders KB. Ventilatory sensitivity to single breaths of CO2 around the control point in man. Respiration Physiology 1995; 99: 205–214. 17. Dejours P. Chemoreflexes in breathing. Physiological Reviews 1962; 42: 335–358. 18. Cunningham DJC, Robbins PA, Wolff CB. Integration of respiratory responses to changes in alveolar partial pressures of CO2 and O2 and in arterial pH. In: Cherniack NS, Widdicombe JG, eds. Control of Breathing, vol. II. Bethesda: American Physiological Society, 1986; 475–528. 19. Whitwam JG, Duffin J, Triscott A, Lewin K. Stimulation of the peripheral chemoreceptors with sodium bicarbonate. British Journal of Anaesthesia 1976; 48: 853–857. 20. Delpierre S, Fornaris M, Guillot C, Grimaud C. Increased ventilatory chemosensitivity induced by domperidone, a dopamine antagonist, in healthy humans. Bulletin of European Physiopathology Respiration 1987; 23: 31–35. 21. Milic-Emili J, Grunstein MM. Drive and timing components of ventilation. Chest 1976; 70: 131–136.
Effect of domperidone on the ventilatory response to transient hyperoxia in patients anaesthetized with isoflurane
I. T. FOO, P. M. WARREN AND G. B. DRUMMOND
Summary We have studied the ventilatory responses to transient hyperoxia in two groups of patients (n:10) anaesthetized with isoflurane (0.3 MAC); patients were allocated randomly to receive either domperidone or placebo orally before anaesthesia. In each patient, five two-breath oxygen tests were averaged and minute ventilation (VEinst) or mean inspiratory flow rate (VT/TI) for each post-test breath was compared with the mean values for these variables during baseline ventilation. A decrease to less than the 95% confidence limits of mean baseline values was considered a definite response. According to this definition, transient hyperoxia decreased VEinst in nine of 10 patients in the placebo group and in all patients in the domperidone group. Similar changes occurred in VT/TI, with eight of 10 definite responses in the placebo group and 10 of 10 in the domperidone group. Compared with placebo, in the domperidone group there were larger changes in VEinst (0.30 vs 0.55 litre min91 (P:0.05)) and VT/TI (8.5 vs 26.6 ml s91 (P:0.02)) from respective baselines. Peripheral chemoreceptors appeared to be active during isoflurane anaesthesia and domperidone pretreatment enhanced this activity by increasing respiratory drive. (Br. J. Anaesth. 1997; 79: 41–46).
remains unclear. Limited studies in humans suggest that the carotid bodies are the most likely site of action.9–11 However, the peripheral chemosensors appear to be active during light halothane anaesthesia. Clergue and co-workers12 used a method derived from Dejours’ test to assess the response to hypoxia from the transient decrease in ventilation after oxygen was substituted for room air, where the time course of response indicates a rapid (peripheral) reflex. We have assessed if peripheral chemoreceptors are active during isoflurane anaesthesia by observing ventilatory responses to transient hyperoxia during light isoflurane anaesthesia. This test, derived from Dejours’ test, is based on the concept that the initial response to hyperoxia (transient ventilatory depression) is mediated by peripheral chemoreceptors which respond more promptly, and that this effect occurs before the onset of secondary responses. The second aim of this study was to see if domperidone, a selective dopamine D2 receptor antagonist which does not cross the blood–brain barrier,13 could augment the response to this test in patients anaesthetized with isoflurane. We showed in a previous study14 that domperidone offset the suppressive effect of isoflurane on HVR, presumably by antagonism of dopamine D2 receptors in the carotid bodies.
Patients and methods Key words Anaesthetics volatile, isoflurane. Pharmacology, domperidone. Reflexes, chemoreceptors. Ventilation, spontaneous. Ventilation, hypoxic response.
Knill and co-workers have shown that the ventilatory response to hypoxia (HVR) in humans is easily depressed by volatile anaesthetics.1–3 Anaesthetic concentrations of halothane, enflurane and isoflurane virtually abolished HVR and even subanaesthetic concentrations of these agents reduced the response considerably. However, recent studies have shown that HVR is not as sensitive to isoflurane and ventilatory response to hypoxia remained in patients anaesthetized with 0.6–1.1 MAC of isoflurane.4–6 Animal studies also provide conflicting data as to whether or not HVR is depressed by volatile anaesthetics.7 8 The site of action of volatile agents on HVR
This double-blind, placebo-controlled study was approved by the local Research Ethics Committee. Written informed consent was obtained from 20 patients who were ASA I, not overweight and aged 18–65 yr. Patients were allocated randomly to receive either domperidone 20 mg or placebo tablets 1 h before arrival in the anaesthetic room. In the anaesthetic room, an i.v. infusion was commenced and non-invasive monitoring of arterial pressure, ECG and pulse oximetry started. General IRWIN T. FOO*, BSC(HONS), MB, BCHIR, MRCP(UK), FRCA, GORDON B. DRUMMOND, MA, MB, CHB, FRCA, Department of Anaesthetics, University of Edinburgh, Edinburgh. PATRICIA M. WARREN, BSC, PHD, Rayne Laboratory, Unit of Respiratory Medicine, Department of Medicine, Royal Infirmary, Edinburgh. Accepted for publication: March 18, 1997. *Address for correspondence: Department of Anaesthetics, Royal Infirmary of Edinburgh, Lauriston Place, Edinburgh EH3 9YW.
42 anaesthesia was induced with etomidate 0.3 mg kg91 i.v. and anaesthesia was deepened with isoflurane and nitrous oxide in oxygen, with the patient breathing spontaneously. When anaesthesia was judged to be of adequate depth for tracheal intubation, laryngoscopy was performed and the larynx sprayed with a metered dose of a lignocaine preparation (Xylocaine, Astra Pharmaceuticals) using no more than 40 mg. After a short delay, the trachea was intubated and the inspired oxygen concentration was reduced to and maintained at 20% for the duration of the study. At the same time, the inspired isoflurane concentration was adjusted to obtain a stable end-tidal concentration of 0.4%, as measured by a Brüel and Kjaer gas monitor 1304, sampling at 90 ml min91, with a response time of less than 360 ms.15 Headphones were placed on the patient’s ears and white noise was played to prevent noise affecting breathing, and the arterial pressure cuff was not automatically inflated routinely in case it altered breathing pattern. Measurements of ventilatory variables during anaesthesia were started after a stable end-tidal isoflurane concentration had been present for at least 20 min. The tracheal tube was connected via a screen pneumotachograph (F100L, Mercury Electronics) to an Ambu Hesse non-rebreathing valve. The pneumotachograph pressure signal was measured with a Furness FC044 differential pressure transducer (⫾10 mm H2O). The inspiratory side of the valve was connected to a large bore two-way tap which allowed rapid changes between two supply systems. Anaesthetic gases (20% oxygen, isoflurane and nitrous oxide) were provided from a 4-litre reservoir bag fitted with a spill valve to reduce positive pressure in the supply tubing. Oxygen was supplied (flow rate of 6 litre min91) from a T-piece system using a 1-m length of anaesthetic tubing as a reservoir. The deadspace of the breathing system was approximately 80 ml. Inspired and expired carbon dioxide and oxygen partial pressures and inspired and expired concentrations of isoflurane were measured continuously by the Brüel and Kjaer gas monitor. Gas monitor and flow signals were recorded on a lap-top computer using the commercial respiratory monitoring software Cardas (version 2.07 advanced, Oxcams, Oxford), running in MS-DOS (version 6.2) for later off-line analysis. The flow signal was integrated to obtain values for tidal volume (VT) and instantaneous minute ventilation (VEinst) calculated from tidal volume and respiratory cycle duration. Before each study, the pneumotachograph was calibrated against a 1-litre calibration syringe and the gas monitor was calibrated using a reference gas. When clinically stable anaesthesia was achieved, the tap was turned during expiration so that two breaths of 100% oxygen were given. Tests were repeated at 2-min intervals. Five satisfactory two-breath oxygen tests were recorded. At the end of the testing sequence, a venous blood sample was obtained for measurement of plasma concentrations of domperidone and surgery started.
British Journal of Anaesthesia
Figure 1 A: Mean inspiratory flow rate (VT/TI) from one patient showing the 10 breaths used as baseline and the post-test breaths analysed. The vertical lines indicate the two oxygen test breaths and the two breaths omitted from the analysis. B: Method used for analysis of transients. Baseline mean and 95% confidence limits (CL) used for testing against each post-test breath. Breaths below the horizontal dashed line are below the 95% CL of baseline and were considered definite responses (see text for further explanation).
ANALYSIS AND STATISTICS
We measured the 10 breaths before the two hyperoxic breaths as baseline. We ignored the two hyperoxic breaths (test breaths) and the following two breaths, as it required these additional breaths to elapse for exhaled oxygen to return to pre-test values. We then measured the eight subsequent breaths (fig. 1). Previous work has shown that maximum change in ventilation occurs after six breaths from the change in inspired oxygen.11 To reduce the influence of normal variation in breathing, the five hyperoxic tests for each patient were averaged on a breath-by-breath basis. We calculated means and 95% confidence limits (CL) for VEinst and mean inspiratory flow rate (VT/TI) for the 10 breaths used as baseline, and inspected the post-test breaths for a reduction to less than 95% CL of baseline. If any post-test breath was reduced below the 95% CL of baseline, this was considered a definite response provided that breaths increased again towards baseline after reaching a minimum value (fig. 1). From the VT/TI results, we recorded the post-test breath number with the smallest VT/TI. This breath was used subsequently for analysis of frequency, and inspiratory and expiratory time changes from baseline. The time from the hyperoxic stimulus to the breath with the smallest VT/TI was also noted. To assess the reliability and consistency of this method of analysis, a random selection of five sets of 22 breaths was obtained in each patient during baseline breathing and subjected to the same averaging and analysis to assess if similar results could be obtained without the hyperoxic test breaths. Definite responses for both the placebo and
Domperidone, isoflurane and transient hyperoxia
43
domperidone groups were compared for any significant difference using the chi-square test with Yates’ correction. To assess if domperidone augmented the response to the hyperoxic test breaths, the mean maximal reductions in VEinst and VT/TI from baseline was calculated for both the placebo and domperidone groups. These values, and TI and TE changes from baseline, based on the minimum VT/TI post-test breath, were compared between groups using the Mann–Whitney test. Statistical significance was set at the 5% level.
and in these cases, the first transient was used for analysis. Considering VEinst, nine of 10 patients who received placebo and 10 of 10 patients who received domperidone demonstrated definite responses to the test. When the analysis was repeated for VT/TI, there were eight definite responses in the placebo group and 10 in the domperidone group. There was no significant difference in response frequency between groups. For both placebo and domperidone groups, mean times to minimum VT/TI after two hyperoxic breaths were 13.5 (SD 2.0) s and 14.9 (2.6) s, respectively. This corresponded to the minimum VT/TI breath occurring between breaths 5 and 11. This appeared to depend on the ventilatory frequency of the individual patient when the testing was performed, with higher ventilatory frequencies associated with a larger number of breaths to the minimum. Randomly selected sequences of breaths in each patient subjected to the same analysis did not show any transient reduction in minute ventilation with recovery towards baseline values. Mean baseline VEinst and minimum minute ventilation after the test for each patient in each group are shown in table 1. The magnitude of the post-test reduction ranged from 0.04 to 0.53 litre min91 of mean baseline ventilation in the placebo group and from 0.21 to 1.74 litre min91 in the domperidone group. The overall mean reduction in ventilation was 0.30 litre min91 for the placebo group and 0.55 litre min91 for the domperidone group (table 2). These reductions differed significantly (P:0.05). Mean baseline VT/TI and minimum VT/TI after the test for each patient in each group are shown in table 1. Changes in VT/TI after the test varied from an increase of 1 ml s91 to a decrease of 27 ml s91 compared with mean baseline values (placebo) and
Results All patients completed the study without complications. The placebo and domperidone groups were comparable in age (mean 33 (range 26–53) and 34 (22–56) yr, respectively), weight (mean 73 (SD 15) and 72 (15) kg, respectively) and sex distribution (5M, 5F and 6M, 4F, respectively). Mean oxygen saturations were 92.7 (SD 2.3) % and 93.3 (1.9) %, respectively, and no patient had an oxygen saturation less than 90% when breathing an inspired oxygen concentration of 20%. Mean baseline end-tidal carbon dioxide partial pressures before the oxygen tests were 5.5 (range 4.5–6.0) kPa for the placebo group and 5.3 (4.2–6.3) kPa for the domperidone groups (ns). After oxygen, all patients showed a transient reduction in VEinst and VT/TI with recovery towards baseline values by the end of the analysis period. VEinst and VT/TI analysis showed that the post-test breath with the maximum reduction occurred consistently within the same breath or two for each patient. In three patients, there were two consecutive transient reductions in VEinst and VT/TI
Table 1 Mean (95% confidence limits) baseline VE, VT/TI, TI and TE, and post-test values after averaging of the five hyperoxic tests for each patient in the placebo and domperidone groups. Post-test VE and VT/TI, are minimum values and *indicates definite response, that is outside the 95% confidence limit of baseline VE, and VT/TI, respectively. Post-test TI and TE, values were derived from the minimum post-test VT/TI, breath number
Patient No.
Baseline VE (litre min91)
Placebo group 1 5.88 (5.81, 5.95) 2 7.76 (7.70, 7.82) 3 6.86 (6.80, 6.92) 4 6.76 (6.71, 6.81) 9 4.82 (4.78, 4.86) 11 5.01 (4.96, 5.06) 14 6.45 (6.36, 6.54) 16 5.37 (5.31, 5.43) 17 4.72 (4.67, 4.77) 18 5.88 (5.82, 5.94) Domperidone group 6 6.82 (6.72, 6.92) 7 5.43 (5.33, 5.53) 8 7.66 (7.50, 7.82) 10 6.11 (6.03, 6.19) 12 8.66 (8.54, 8.78) 13 7.17 (7.00, 7.34) 15 5.73 (5.66, 5.80) 19 5.76 (5.71, 5.81) 21 6.52 (6.44, 6.60) 22 6.81 (6.59, 7.03)
VE Minimum post-test (limit min91)
Baseline VT/TI, (ml s91)
VT/TI minimum post-test (ml s91)
Baseline TI (s)
Posttest TI (s)
Baseline TE (s)
Posttest TE (s)
5.84 7.34* 6.33* 6.25* 4.64* 4.74* 6.24* 4.99* 4.58* 5.59*
222 (217, 227) 295 (290, 300) 264 (260, 268) 253 (250, 256) 184 (181, 186) 183 (181, 185) 235 (232, 239) 207 (205, 209) 188 (186, 189) 203 (201, 206)
219 288* 237* 236* 176* 177* 237 199* 183* 196*
1.17 (1.14, 1.21) 1.00 (0.98, 1.02) 0.97 (0.95, 0.99) 0.95 (0.94, 0.97) 1.11 (1.10, 1.13) 0.89 (0.88, 0.90) 1.26 (1.24, 1.28) 1.12 (1.11, 1.14) 1.00 (0.98, 1.01) 0.89 (0.87, 0.91)
1.15 1.05 1.05 1.08 1.17 0.91 1.23 1.14 1.02 0.94
1.48 (1.46, 1.50) 1.27 (1.25, 1.29) 1.26 (1.23, 1.29) 1.18 (1.16, 1.20) 1.43 (1.41, 1.45) 1.06 (1.04, 1.08) 1.49 (1.46, 1.52) 1.42 (1.40, 1.44) 1.39 (1.37, 1.41) 0.94 (0.92, 0.96)
1.37 1.27 1.30 1.36 1.48 1.14 1.57 1.59 1.39 1.05
6.28* 5.12* 7.17* 5.65* 6.92* 6.56* 5.45* 5.55* 6.10* 6.39*
241 (233, 248) 210 (206, 214) 272 (263, 280) 226 (220, 231) 372 (365, 378) 266 (255, 277) 188 (187, 190) 206 (204, 209) 290 (281, 300) 240 (238, 242)
219* 197* 252* 191* 296* 229* 183* 200* 251* 228*
1.30 (1.25, 1.35) 1.39 (1.37, 1.42) 1.26 (1.21, 1.31) 1.11 (1.08, 1.14) 1.07 (1.05, 1.09) 1.10 (1.05, 1.15) 0.93 (0.92, 0.94) 0.90 (0.89, 0.91) 1.12 (1.07, 1.17) 0.64 (0.63, 0.65)
1.47 1.47 1.35 1.34 1.24 1.25 0.99 0.92 1.29 0.65
1.44 (1.38, 1.50) 1.84 (1.81, 1.87) 1.40 (1.37, 1.43) 1.32 (1.28, 1.36) 1.67 (1.64, 1.69) 1.31 (1.26, 1.36) 0.90 (0.89, 0.91) 1.03 (1.01, 1.05) 1.42 (1.39, 1.45) 0.72 (0.71, 0.73)
1.56 1.91 1.47 1.41 1.75 1.33 0.97 1.07 1.40 0.74
44
British Journal of Anaesthesia Table 2 Mean (range) baseline VE, VT/TI, TI and TE, and post-test values for the placebo and domperidone groups. Post-test VE and VT/TI are minimum values and post-test TI and TE values were derived from the minimum post-test VT/TI breath number. Magnitude of change between baseline and post-test values are shown and *indicates significant difference (P-0.05) from placebo Baseline VE (litre min91) Placebo Domperidone VT/TI (ml s91) Placebo Domperidone TI (s) Placebo Domperidone TE (s) Placebo Domperidone
Minimum post-test
5.96 (4.72, 7.76) 6.67 (5.43, 8.66) 223 (183, 295) 251 (188, 372) Baseline
5.65 (4.58, 7.34) 6.12 (5.12, 6.92) 215 (176, 288) 224 (183, 296) Post-test
⌬ Baseline—minimum post-test 0.30 (0.04, 0.53) 0.55 (0.21, 1.74)* 9 (91, 27) 27 (6, 76)* ⌬ Baseline—post-test
1.04 (0.89, 1.26) 1.08 (0.64, 1.39)
1.07 (0.91, 1.23) 1.20 (0.65, 1.47)
90.04 (90.13, 0.03) 90.12 (90.23, 90.01)*
1.29 (0.94, 1.49) 1.31 (0.72, 1.84)
1.35 (1.05, 1.59) 1.36 (0.74, 1.91)
90.06 (90.18, 0.11) 90.06 (90.12, 0.02)
decreases from 6 to 76 ml s91 (domperidone). The mean change was 99 ml s91 for the placebo group and 927 ml s91 for the domperidone group (table 2). These changes were statistically significantly (P:0.02). Timing variables (TI and TE) for the minimum post-test VT/TI breath were compared with respective mean baseline values. With the exception of data from two patients in the placebo group, inspiratory times were lengthened after the test in both groups compared with baseline values (table 1). The mean increase in inspiration time for the placebo group was 0.04 s and 0.12 s for the domperidone group (table 2). There was a significant difference between groups (P:0.02). Expiration times were more variable but the trend was towards increasing expiration time in both groups (table 1). The mean increase for both groups was 0.06 s (table 2). Mean plasma concentration of domperidone in the domperidone group was 7.9 (range 2.0–19.3) ng ml91. There was no correlation between plasma concentrations of domperidone and the size of the transient reductions in VEinst or VT/TI.
Discussion We have shown that peripheral chemoreceptor activity was present in a majority of patients during isoflurane anaesthesia. Pre-existing stimulus to breathing was shown by transient depression of both VEinst and VT/TI after two breaths of 100% oxygen. With domperidone pretreatment, patients showed greater reductions in both VEinst and VT/TI after two breaths of 100% oxygen. We reduced the respiratory depresent effects of other anaesthetic agents in this study by using etomidate for induction of anaesthesia, which has a short duration of action, and by avoiding the use of opioids. We averaged the repeated tests in each individual to reduce breath-to-breath variability so that small changes in ventilatory variables (as produced by this transient hyperoxia test) were more likely to be detected. Our method of testing for a definite response to hyperoxia is similar to that used in human studies involving single-breath carbon dioxide tests of peripheral chemosensitivity.16
Randomly selected breaths subjected to the same analysis did not demonstrate the clear pattern of reduction followed by recovery towards baseline and we believe that this method of analysis reliably detected transient changes in responses. In our analysis, we ignored the two hyperoxic breaths (test breaths) and the following two breaths as there was a delay for the expired oxygen to return to baseline and we did not correct the pneumotachograph for changes in gas composition. As previous work showed that the maximum change occurred six breaths after the change in inspired oxygen,11 we were unlikely to have missed the maximum change. We used a variation of Dejours’ test to assess the activity of the peripheral chemoreceptors during isoflurane anaesthesia. With this technique in resting awake humans, the magnitude of ventilatory change is approximately 8–10% of pre-stimulus baseline ventilation and occurs approximately 10 s after the stimulus.17 In our study, during isoflurane anaesthesia, mean ventilatory change was 4.9% in the placebo group and 7.7% in the domperidone group and mean time to minimum VT/TI after stimulus was 13.5 s in the placebo group and 14.9 s in the domperidone group. This is longer than in awake subjects. Washin of oxygen to the alveoli and hence the change in arterial oxygen content may take longer during anaesthesia because ventilation is less in the anaesthetized subject. Alternatively, control of ventilation may be different from the awake state. There are possible drawbacks of this test to assess chemoreceptor activity during anaesthesia. Hyperoxia can increase blood PaCO2 via the Haldane effect and thereby cause an increase in ventilation. Furthermore, hyperoxia can reduce cerebral blood flow which would tend to increase brain PCO2 and thus also ventilation.18 These effects, which are probably less rapid than effects at the peripheral chemoreceptors, would tend to oppose the decrease in ventilation caused by hyperoxia and therefore underestimate the magnitude of the ventilatory output of the peripheral chemoreceptors. Duffin, Triscott and Whitwam11 studied patients after opioid premedication and found that 0.7–0.8% halothane in air substantially decreased the ventilatory response to two breaths of oxygen compared with the response during thiopentone anaesthesia.
Domperidone, isoflurane and transient hyperoxia They concluded that halothane reduced the activity and responsiveness of peripheral chemoreceptors. As we did not have a group of patients in this study who did not receive isoflurane, we were unable to determine the effect of isoflurane anaesthesia per se. Our results only indicate that peripheral chemoreceptors appeared to be active during isoflurane anaesthesia. Different anaesthetic techniques such as the use of opioids and nitrous oxide and different ways of analysing the breaths may account for some of the differences observed. Furthermore, the anaesthetic concentrations used were different in the two studies; Duffin, Triscott and Whitwam11 used approximately 1 MAC of halothane while we used approximately 0.3 MAC of isoflurane and 80% nitrous oxide. This difference could have influenced the degree of suppression of the peripheral chemoreceptors by these volatile agents. Nevertheless, it is likely that in our patients with an end-tidal isoflurane concentration of 0.4% (approximately 0.3 MAC), peripheral chemoreceptor activity was still present and responsive to changes in oxygen tensions, but perhaps not as vigorously as in awake subjects. Others have studied patients anaesthetized with 1.5% halothane, using a method derived from Dejours’ test, and found that peripheral chemoreceptor activity was depressed but not abolished.12 Knill and coworkers3 found that even at 0.1 MAC of isoflurane, there was a 50% reduction in the ventilatory response to isocapnic hypoxia. This was confirmed recently by van den Elsen and colleagues10 who extended the concentration of isoflurane to 0.2 MAC and found a 65% reduction in isocapnic HVR. At anaesthetic concentrations of 1.1 MAC, HVR was virtually abolished.3 This suggested that HVR and therefore the peripheral chemoreflex pathway was highly sensitive to the effects of isoflurane. However, our results and those of Clergue and co-workers12 suggest that peripheral chemoreceptors are not “silent” at anaesthetic concentrations of isoflurane and halothane, respectively, suggesting that although peripheral chemoreceptor responses are depressed, this is perhaps not to the degree as suggested by previous investigators.3 Our findings also support recent studies by Lindahl’s group4–6 who found persistent ventilatory responses to hypoxia under poikilocapnic conditions in patients anaesthetized with 0.6–0.85 MAC of isoflurane and a reduction of 60–70% in HVR under isocapnic conditions with 0.85–1.1 MAC of isoflurane. Therefore, anaesthetic concentrations of isoflurane do not appear to suppress HVR to the same extent as that found by previous investigators.3 10 Differences in experimental conditions may account for some of these differences.10 Washin studies of subanaesthetic halothane and isoflurane9 10 suggest that the most likely site of action of volatile agents is at the peripheral chemoreceptors. Depression of hypoxaemic-driven ventilation by halothane and isoflurane occurred extremely rapidly, within the first minute of inhalation. For isoflurane, it was estimated that after 30 s of isoflurane inhalation, the carotid body isoflurane tension would be approximately 90% of end-tidal, whereas the brain isoflurane tension would be
45 approximately 8% of end-tidal.10 Therefore, ventilatory depression occurring within 30 s of isoflurane inhalation is best explained by an action of isoflurane at the peripheral chemoreceptors. However, Whitwam and co-workers19 demonstrated in anaesthetized patients that there was a mean delay of only 10 s in central chemoreceptor ventilatory response to i.v. injection of sodium bicarbonate (causing an increase in PaCO2 ) when peripheral chemoreceptor activity was depressed by hyperoxia. This suggests that the time lag between the effect of isoflurane at central and peripheral chemoreceptor sites during washin could also be in the region of 10 s. Therefore, a central effect of subanaesthetic isoflurane cannot be excluded in the first 30 s of isoflurane washin and overall suppression of HVR at anaesthetic doses could be caused by a combination of peripheral and central effects. The second aim of this study was to determine if domperidone, a selective dopamine D2 receptor antagonist which does not readily cross the blood–brain barrier,13 could augment the response to transient hyperoxia in patients anaesthetized with isoflurane. Domperidone augments HVR in humans20 and can offset the suppression caused by isoflurane on the ventilatory response to hypoxia.14 We found that with domperidone pretreatment, all patients showed positive results to transient hyperoxia suggesting that with increased ventilatory drive, it was easier to demonstrate the transient reductions in both VEinst and VT/TI after two breaths of 100% oxygen. Analysis using VT/TI instead of breath-by-breath minute ventilation demonstrated a greater difference between placebo and domperidone. VT/TI is a measure of respiratory drive allowing minute ventilation to be separated into drive and timing components: VEinst is the product of VT/TI and TI/TTOT (timing component where TTOT is the sum of TI and TE).21 When inspiratory and expiratory times were compared between groups, only TI was significantly different, with domperidone causing an increase in inspiration time. These results support the observations of Clergue and co-workers who found a prompt change in TI after oxygen breathing commenced.12 In summary, our data support the view that peripheral chemoreceptors remain active during isoflurane anaesthesia. Their function may be depressed by isoflurane, but perhaps not as much as suggested previously. Domperidone enhanced peripheral chemosensor activity by increasing respiratory drive during isoflurane anaesthesia.
Acknowledgements We thank Sanofi Winthrop for supplies of domperidone and Dr R. Woestenborghs, Janssen Research Foundation, for measurement of plasma concentrations of domperidone. I. T. F. was a British Journal of Anaesthesia Research Fellow.
References 1. Knill RL, Gelb AW. Ventilatory responses to hypoxia and hypercapnia during halothane sedation and anaesthesia in man. Anesthesiology 1978; 49: 244–251.
46 2. Knill RL, Manninen PH, Clement JL. Ventilation and chemoreflexes during enflurane sedation and anaesthesia in man. Canadian Anaesthetists Society Journal 1979; 29: 353–360. 3. Knill RL, Kieraszewicz HT, Dodgson BG, Clement JL. Chemical regulation of ventilation during isoflurane sedation and anaesthesia in humans. Canadian Anaesthetists Society Journal 1983; 30: 607–614. 4. Sjögren D, Sollevi A, Ebberyd A, Lindahl SGE. Poikilocapnic hypoxic ventilatory response in humans during 0.85 MAC isoflurane anaesthesia. Acta Anaesthesiologica Scandinavica 1994; 38: 149–155. 5. Sjögren D, Sollevi A, Ebberyd A, Lindahl SGE. Isoflurane anaesthesia (0.6 MAC) and hypoxic ventilatory responses in humans. Acta Anaesthesiologica Scandinavica 1995; 39: 17–22. 6. Sollevi A, Lindahl SGE. Hypoxic and hypercapnic ventilatory responses during isoflurane sedation and anaesthesia in women. Acta Anaesthesiologica Scandinavica 1995; 39: 931–938. 7. Davies RO, Edwards MW, Lahiri S. Halothane depresses the response of carotid body chemoreceptors to hypoxia and hypercapnia in the cat. Anesthesiology 1982; 57: 153–159. 8. Koh SO, Severinghaus JW. Effect of halothane on hypoxic and hypercapnic ventilatory responses of goats. British Journal of Anaesthesia 1990; 65: 713–717. 9. Knill RL, Clement JL. Site of selective action of halothane on the peripheral chemoreflex pathway in humans. Anesthesiology 1984; 61: 121–126. 10. van den Elsen M, Dahan A, DeGoede J, Berkenbosch A, van Kleef J. Influences of subanesthetic isoflurane on ventilatory control in humans. Anesthesiology 1995; 83: 478–490. 11. Duffin J, Triscott A, Whitwam JG. The effect of halothane and thiopentone on ventilatory responses mediated by the peripheral chemoreceptors in man. British Journal of
British Journal of Anaesthesia Anaesthesia 1976; 48: 975–980. 12. Clergue F, Ecoffey C, Derenne JP, Viars P. Oxygen drive to breathing during halothane anesthesia: effects of almitrine bismesilate. Anesthesiology 1984; 60: 125–131. 13. Champion MC, Hartnett M, Yen M. Domperidone, a new dopamine antagonist. Canadian Medical Association Journal 1986; 135: 457–461. 14. Foo IT, Martin SE, Lee RJ, Drummond GB, Warren PM. Interaction of the effects of domperidone and subanaesthetic concentrations of isoflurane on the immediate and sustained hypoxic ventilatory response in humans. British Journal of Anaesthesia 1995; 74: 134–140. 15. McPeak HB, Palayiwa E, Robinson GC, Sykes MK. An evaluation of the Brüel and Kjaer monitor 1304. Anaesthesia 1992; 47: 41–47. 16. Fernando SSD, Saunders KB. Ventilatory sensitivity to single breaths of CO2 around the control point in man. Respiration Physiology 1995; 99: 205–214. 17. Dejours P. Chemoreflexes in breathing. Physiological Reviews 1962; 42: 335–358. 18. Cunningham DJC, Robbins PA, Wolff CB. Integration of respiratory responses to changes in alveolar partial pressures of CO2 and O2 and in arterial pH. In: Cherniack NS, Widdicombe JG, eds. Control of Breathing, vol. II. Bethesda: American Physiological Society, 1986; 475–528. 19. Whitwam JG, Duffin J, Triscott A, Lewin K. Stimulation of the peripheral chemoreceptors with sodium bicarbonate. British Journal of Anaesthesia 1976; 48: 853–857. 20. Delpierre S, Fornaris M, Guillot C, Grimaud C. Increased ventilatory chemosensitivity induced by domperidone, a dopamine antagonist, in healthy humans. Bulletin of European Physiopathology Respiration 1987; 23: 31–35. 21. Milic-Emili J, Grunstein MM. Drive and timing components of ventilation. Chest 1976; 70: 131–136.