Effect of low-dose enflurane on the ventilatory response to hypoxia in humans

Transcription

1 British Journal of Anaesthesia 1994; 72: CLINICAL INVESTIGATIONS Effect of low-dose enflurane on the ventilatory response to hypoxia in humans B. NAGYOVA, K. L. DORRINGTON AND P. A. ROBBINS SUMMARY To investigate the effects of enflurane on the control of breathing we have studied the ventilatory responses to isocapnic hypoxia in 12 adults with and without sedation with enflurane. Design 1 consisted of three steps into hypoxia (PE' O2 = 6.7 kpa), each lasting 3 min, separated by periods ofeuoxia lasting 5 min (PE'O 2 = 13.3 kpa). Design 1 was repeated four times in each subject on the same day in random order: with carrier gas (control) and with.4 MAC,.7 MAC and.13 MAC of endtidal enflurane concentrations. Design 2 consisted of 2-min exposures to hypoxia with and without.7 MAC of enflurane. Each exposure was preceded and followed by 5 min of euoxia. End-tidal?CO 2 was held constant at kpa greater than the resting level throughout both designs. Mean (SEM) ventilatory responses to hypoxia for design 1 were: 8.2 (1.3) litre min-' (control), 6.6 (1.4) litre min-' (.4 MAC). 5.7 (1.1) litre min- 1 (.7 MAC) and 3.7 (.5) litre min-' (.13 MAC) (P <.1). For design 2. enflurane produced a 15% reduction in resting ventilation (P <.1), a 4% decrease in the acute ventilatory response to hypoxia (P <.1) and a 32% reduction in ventilatory decline (ns) which occurred during sustained hypoxia. (Br. J. Anaesth. 1994; 72: ) KEY WORDS Anaesthetics, volatile: enflurane. Ventilation hypoxic response. In 1978, Knill and Gelb [1] observed that the ventilatory response to hypoxia in humans was reduced markedly by a subanaesthetic dose (.1 MAC) of halothane. This observation raised the possibility that during and after inhalation anaesthesia patients are lacking, fully or partially, the normal protective chemoreflex response to small inspired concentrations of oxygen and hypoxaemia arising from other causes. Similar findings were published subsequently by the same research group for enflurane [2] and isoflurane [3]. Comparison of the ventilatory responses to hypoxia, administration of doxapram and administration of carbon dioxide led to the conclusion that anaesthetic vapours have a selective depressant effect on the peripheral chemoreflex [1]. It is now accepted widely that the human ventilatory response to hypoxia is extremely sensitive to the commonly administered inhalation agents [4, 5]. In respect of isoflurane, the findings of Knill and co-workers have recently been questioned. Temp, Henson and Ward [6] found that.1 MAC of isoflurane did not affect the ventilatory response to sustained eucapnic ( kpa greater than resting end-tidal Pco 2 ) hypoxia in adults. Similarly, Sjogren and colleagues [7] were unable to show an effect on the ventilatory response to a step decrease in end-tidal Po 2 during exposure to.9 MAC of isoflurane in humans. In view of the continuing widespread use of enflurane, we decided to examine further the effect of enflurane on the ventilatory response to hypoxia. The aims of our study were to measure a doseresponse relationship for the acute ventilatory response to hypoxia for enflurane in adult humans and to measure the effects of enflurane sedation on a period of sustained isocapnic hypoxia. PATIENTS AND METHODS The study was approved by the Central Oxford Research Ethics Committee. Twelve healthy adults were studied (eight males, mean age 23.7 (range 2-^) yr, height 1.8 ( ) m and weight 71.6 (54-87) kg). All subjects received a written and verbal description of the experiments before they gave their consent. Before each session, all were asked to refrain from food and drink for at least 6 h. It is known that sleep depresses the acute hypoxic response by 3 6% depending on the sleep stage [8]. Therefore, subjects were required to watch television to minimize differences in the level of consciousness between control and enflurane experiments. During experiments, subjects were seated in a chair. The first 2 min of each study was an equilibration period, when subjects breathed air, or air with enflurane, via a face mask with a totalflowof at least 66 litre min" 1. After the equilibration period, subjects breathed through a mouthpiece from a steady stream of gas delivered from a gas-mixing system at a flow of 66 litre min" 1. A pulse oximeter was used to monitor heart rate and arterial oxygen saturation throughout. The mouthpiece was connected in series with a turbine volume measuring device [9] and pneumotachograph for measuring respiratory volumes andflows.gas B. NAGYOVA, M.D., K. L. DORRINGTON, D.PHIL., D.M., F.R.C.A., P. A. ROBBINS, B.M., B.CH., D.PHIL., University Laboratory of Physiology, Parks Road, Oxford OX1 3PT. Accepted for Publication: November 1, Correspondence to K.L.D.

2 51 BRITISH JOURNAL OF ANAESTHESIA was sampled at the mouth at a flow of 5 ml min~' and analysed for Pco 2, Po 2 and enflurane by a mass spectrometer. All experimental variables were recorded by an IBM PC AT computer. This computer executed a peak-picking program in real time to determine end-tidal Pco 2 (PE' C O 2 ) and Po 2 (PE' O2 ), together with inspiratory and expiratory volumes and durations. The breath-by-breath end-tidal values were passed to a second computer, which compared the actual end-tidal values with desired values and adjusted the inspiratory gas mixture by controlling the fast gas-mixing system to maintain the desired end-tidal values despite changes in ventilation. Details of the dynamic end-tidal forcing technique and gas-mixing system have been described elsewhere [9, 1]. During the 2-min equilibration period, the inspired concentration of enflurane was measured from the face mask by mass spectrometry and adjusted to one of four values:,.8%,.17%,.34%. The mass spectrometer (Airspec 3, Airspec Ltd, Biggin Hill, Kent) was calibrated using a standard gas mixture of.43% enflurane in air (British Oxygen Company, London; enflurane concentration certified to ±5%). The mass number used for enflurane was 51. During the subsequent attachment of the subject to the mouthpiece, the mass spectrometer was used to measure inspired and expired enflurane concentrations. During this period the end-expired enflurane concentration was held constant by manual adjustment of the inspired concentration to maintain the end-expired concentration at the value achieved following the period of equilibration with the face mask. This value was always approximately 7% of the inspired concentration. The dose-response relationship is presented as a function of end-expired enflurane concentration expressed in units of MAC. The 12 subjects undertook both of the following tests in random order. Test 1 was to determine the dose-response curve for the effect of small concentrations of enflurane on the ventilatory response to acute hypoxia. The duration of each experimental period was 27 min. End-tidal Po 2 was held at 13.3 kpa for 5 min and then there were a total of three separate exposures to 3 min of hypoxia (PE'O 2 = 6.7 kpa), separated by 5-min periods of euoxia (PE'O 2 = 13.3 kpa). After the last of the three hypoxic exposures, there was a final 3-min period of euoxia (fig. 1). PE' C O 2 wa s held constant at kpa greater than the subject's resting level throughout. This test was performed with carrier gas (control, %) and three different sedative concentrations of enflurane (inspired concentrations of approximately.8 %,.17 % and.34 %) in random order in each of the subjects on the same day. After each repeat of the test, there was a 2-min period of breathing room air before the subject was equilibrated with the next dose of enflurane. Test 2 was to determine the effect of small inspired doses of enflurane (inspired concentration of approximately.17%) on the ventilatory responses to sustained hypoxia. PE'O 2 was held at 13.3 kpa for the first 5 min, then at 6.7 kpa for the next 2 min and returned to 13.3 kpa for the final 5-min recovery period. PE' C O 2 wa s held at kpa greater than the resting value throughout the experiment. This test was performed on each of the 12 subjects under both of the pharmacological conditions in random order on the same day. Data for each run for each test were averaged over 6-s intervals. In order to obtain the dose-response curve for the acute hypoxic response from the data from test 1, we used the ventilation during the last 1 min of euoxia before each of three hypoxia steps ( VA) and the ventilation in the third minute of each hypoxic period (FB) for each concentration of enflurane (%,.8%,.17% and.34%). The ventilatory response for each step into hypoxia was calculated as the difference between V& and VA. The statistical significance of any difference between the four concentrations of enflurane was assessed by analysis of variance using Minitab release 7, run within MS DOS 5.1 on a Dell 32 LX PC. Subsequent analyses used two-tailed Student's paired t test. For test 2, we used four particular 1-min periods to characterize the ventilatory responses to sustained hypoxia, which are illustrated in figure 3. These four periods were: ventilation during the last 1 min before the step into hypoxia (V u pre-hypoxic ventilation), peak ventilation during the first 5 min of the hypoxic period (V 2, peak ventilation), ventilation during the last 1 min of hypoxia (V 1} depressed ventilation) and minimum ventilation during the last 5 min of the euoxic period (V 4) post-hypoxic ventilation). The on-response (or acute hypoxic ventilatory response, AHVR) was calculated as the difference between peak ventilation and pre-hypoxic ventilation {V 2 Vi), and the magnitude of the off-response was calculated as the difference in ventilation between the depressed ventilation and post-hypoxic ventilation (V 3 V 4 ). The absolute magnitude of the decline in ventilation during sustained hypoxia (HVD) was obtained by calculating the difference between peak ventilation and depressed ventilation (V 2 F 3 ). The significance of any differences between enflurane and control were assessed using Student's two-tailed paired t tests. RESULTS All 12 subjects completed both tests without any problems. Mean values for inspired and end-tidal anaesthetic concentrations are shown in table I. MAC for enflurane has been taken to be 1.7% [11]. TABLE I. Mean (SEM) values/or inspired (%) and end-tidal {% and MAC) concentrations of enflurane during the two tests Target (%) Test Test 2.17 Inspired (%).92 (.6).168(.7).31(.9).172(.3) End-tidal (%).66 (.4).116(.6) 214(.7) 123 (.5) End-tidal (MAC).39 (.3).68 (.4).126(.4).73 (.3)

3 ENFLURANE AND HYPOXIC VENTILATORY RESPONSE , 15- c 15 H E VB,VB VB ~ 1- 'c E CD H 3-15-, U Time (min) FIG. 1. Test 1. Ventilation, end-tidal carbon dioxide (PE'CO 2 ) anc > end-tidal oxygen partial pressure (PE'O 2 ) f r subject No. 937 at the four target concentrations of inspired enflurane (O = %; =.8 %; V =.17 %; A =.34 %). The data shown are typical of those for all subjects. Vk and VB indicate ventilations referred to in the text. TABLE II. Test 1. Acute hypoxic ventilatory responses (litre min ') for each subject for the different concentrations of enflurane Subject No Mean SEM Control Enflurane concn (%) Test 1. Figure 1 demonstrates the end-tidal gas profiles and ventilatory responses from test 1 for the four different doses of enflurane for subject No Control of end-tidal Pco 2 and Po 2 was similar during each of the four doses of enflurane. The ventilatory Enflurane concn (MAC).2 FIG. 2. Test 1. Dose-response relation (mean, SEM) for the acute hypoxic ventilatory response for all 12 subjects. The increase in ventilation at the onset of hypoxia is plotted as a function of the MAC of enflurane. responses were brisk with, in general, little difference between the second and third minute of hypoxia. The responses during enflurane appeared smaller than the control response. The average hypoxic ventilatory responses for the different concentrations of enflurane for each of the 12 subjects are shown in table II. The mean ventilatory response to hypoxia for all subjects declined with increasing doses of enflurane and this is plotted as a dose-response curve in figure 2. Analysis of variance of the individual hypoxic responses by subject and enflurane dose, together with an interactive term, revealed a highly significant effect of dose on the acute hypoxic response (P <.1). Analysis using two-tailed Student's paired t tests revealed that the difference in acute hypoxic response between control and the smallest dose of enflurane (.8%) did not reach statistical significant (P =.77), but that the differences between control and the larger doses of enflurane were highly significant (.17% enflurane: P =.14;.34% enflurane: P =.3). Test 2. Figure 3 illustrates the response of one subject to a single sustained period of hypoxia with and without enflurane and figure 4 illustrates mean ventilations for all 12 subjects combined for the two conditions. Also shown are the end-tidal gas profiles for each experimental condition. PE' OI reached desired levels at the beginning of the hypoxic period and remained constant throughout, although the PE'Q 2 profile was rather more variable during the transient out of hypoxia. PE' CO2 remained constant in all subjects throughout the experiments. Resting ventilation, AHVR, HVD and off-responses for the individual subjects and means for those variables are shown in table III. There was a significant decline in euoxic ventilation (P =.18) and the magnitude of the on-transient (P =.21) and off-transient (P =.14) when enflurane responses were compared with control responses. There were no significant differences in the magnitude of the HVD (P =.12) using a two-tailed paired t test. Using the experimentally determined

4 512 BRITISH JOURNAL OF ANAESTHESIA 5 -I 3-, -4- c E 3- a> M 2- UJ ^ 1- Jb o «f T O'^DCPQQ 3 n E 2- E I 3-15-, * 12 " ^ 11- ~~1-1? 9- O ~ Time (min) FIG. 3. Test 2. Ventilation, end-tidal carbon dioxide (PE' CO,) and end-tidal oxygen partial pressure (PE'O 2 ) f r subject No. 933 at the two target concentrations of inspired enflurane (O = %j =.17%). The data shown are typical of those for all subjects. V x, V 2, V t and V^ indicate ventilations referred to in the text. TABLE III. Test 2. Values for euoxic ventilation (V), on-response (On), HVD and off-response (Off) (all litre mm' 1 ) for each subject during control conditions (subscript C) and enflurane sedation (subscript E) Subject No Mean SEM Vic On c On E HVD C HVD B Off c Off E value for the standard error, power analysis suggests that there should be an 8% or greater chance of detecting a difference of 5.3 litre min" 1 or greater in the values of HVD between control and enflurane sedation. (This value decreases to 4 litre min~' if a 3-15-, 14- T3- ^ 11- ~ s- o ~ Time (min) FIG. 4. Test 2. Mean values for ventilation, end-tidal carbon dioxide (PE'CO 2 ) an d end-tidal oxygen partial pressure (PE'O 2 ) f r a " 12 subjects. one-tailed test is used.) No significant differences in the ratio HVD/AHVR were detected between enflurane sedation and control. DISCUSSION The experiments presented here demonstrate that sedative doses of enflurane have a significant and substantial effect on the ventilatory response to hypoxia in humans. This finding is important in the context of the current uncertainty surrounding the effects of inhalation anaesthetic agents on the ventilatory response to hypoxia. In their pioneering work, Knill and colleagues [1-3] found that halothane, enflurane and isoflurane snowed very similar depressant effects on the ventilatory response to hypoxia. Typically, a sedative dose of.1 MAC reduced the response by 6% and an anaesthetic dose of 1.1 MAC eliminated the response entirely. The older, more soluble, anaesthetic vapours, methoxyflurane and diethyl ether, were found to have a less striking effect [12]. After criticism of aspects of the techniques which had been used, Knill commented in 1991 on how unsatisfactory it was that none of the findings of these studies had been corroborated or refuted by investigators other than in his group at the University of Western Ontario, and encouraged further study [13].

5 ENFLURANE AND HYPOXIC VENTILATORY RESPONSE 513 Subsequently, Temp, Henson and Ward [6] studied eight adult human subjects during inhalation of.1 MAC of isoflurane and concluded from their data that this concentration of isoflurane does not affect the ventilatory response to acute or sustained eucapnic (end-tidal Pco kpa greater than the resting level) hypoxia, although they did find an effect of isoflurane sedation on ventilation after 2 min of hypoxia with hypercapnia (end-tidal Pco kpa). The 95% confidence interval for their measurement of the ratio of acute hypoxic responses with and without isoflurane was , with a mean of.99. Noting that Knill and colleagues obtained a ratio of.42 from a small sample of five subjects, the two papers can be seen to be consistent with a 2-3% reduction in response [3, 6]. However, there remains a possibility that the different results obtained by these workers for sedation with isoflurane reflect a fundamental difference in their techniques for measuring the ventilatory response to hypoxia. In the experiments of Knill and colleagues, hypoxia was introduced gradually from hyperoxia over 8-1 min, with eucapnia being maintained by manual control. In the experiments of Temp's group, hypoxia was introduced abruptly from an endexpired Po 2 close to normal, with eucapnia being maintained using an automated end-tidal forcing system. However, it has become clear that the ventilatory response to hypoxia during exposures over periods of about 3 min consists of two components and that these become apparent only when hypoxia is introduced rapidly, within approximately 1 min [14]. After such an acute exposure, most subjects show a rapid increase in ventilation which reaches a maximum within approximately 3 min (the acute hypoxic ventilatory response, AHVR), followed by a decline in ventilation (the hypoxic ventilatory decline, HVD) over approximately 3 min. The control response in figure 4 is typical. It follows that experiments such as those of Knill and colleagues, which introduce hypoxia gradually, observe effects of hypoxia relating to both the AHVR and HVD in a manner which may be difficult to interpret. Consequently, we thought it important to establish if a similar degree of change in the ventilatory response occurs for enflurane sedation after a step change into hypoxia. In the enflurane study by Knill, Manninen and Clement [2], 1 subjects were studied. Baseline normoxic ventilation was reduced by enflurane at.1 MAC from 8. to 6.9 litre min" 1, a significant reduction of 14%. In this study (test 2), resting ventilation was reduced by 15% at.73 MAC (table III). Knill, Manninen and Clement [2] found a significant reduction of 55 % in the ventilatory response to hypoxia. Linear interpolation on our dose-response relationship from test 1 (fig. 2) yields a 44% reduction in AHVR at.1 MAC. Linear extrapolation of AHVR from test 2 (table III) gives a second, independent estimate of the effect of enflurane on the AHVR. At.1 MAC there is a 55 % reduction in AHVR in exact agreement with the value of Knill, Manninen and Clement [2]. In our study, mean control AHVR measured in test 1 was 8.22 litre min" 1 (table II), yet the mean control AHVR measured in test 2 was 14.1 litre min" 1. This finding could be caused by averaging the AHVR from three steps of hypoxia in test 1, during which successive steps into hypoxia tended to produce weaker ventilatory responses. This is seen clearly for subject No. 937 in figure 1. In contrast, single steps into hypoxia were used in test 2 and these could have yielded a more vigorous response than the mean of the triple responses in test 1. In order to examine this possibility, we calculated the AHVR in test 1 for the first step into hypoxia only. This gave mean values for ventilation during the third minute of hypoxia of 1.29, 8.4, 5.68 and 4.68 litre min" 1, respectively, at,.39,.68 and.126 MAC of enflurane. Analysis of variance confirmed a significant effect (P =.1) of the order of hypoxic steps on ventilation during the third minute of hypoxia. Interpolation of the doseresponse data for the first step into hypoxia predicted a decrease in ventilation of 5% from control to.1 MAC of enflurane. Therefore, although averaging the three steps generates a dose response relationship (fig. 2) with smaller ventilations than those obtained in the first step into hypoxia, the proportionate effect of enflurane on ventilation is not attenuated markedly by using the averages. In this study, we have been unable to resolve the question of whether or not HVD is unaffected by sedation with enflurane, or indeed if it is diminished in the same proportion as AHVR. Both of these possibilities remain within the resolution of our study. It is clear, however, that HVD remains a feature of the ventilatory response to hypoxia, such that the general pattern of the biphasic response is retained in the presence of enflurane. In conclusion, we have confirmed and extended the finding that sedative doses of enflurane depress markedly the ventilatory response to hypoxia in humans. In view of the clinical significance of this observation, it is important that the effects of other anaesthetics are better documented. ACKNOWLEDGEMENTS This work was funded by the Wellcome Trust. Dr Dorrington is supported by the Dunhill Medical Trust. Dr Nagyova is supported by Balliol College, Oxford. REFERENCES 1. Knill RL, Gelb AW. Ventilatory responses to hypoxia and hypercapnia during halothane sedation and anesthesia in man. Anesthesiology 1978; 49: Knill RL, Manninen PH, Clement JL. Ventilation and chemoreflexes during enflurane sedation and anaesthesia in man. Canadian Anaesthetists Society Journal 1979; 29: Knill RL, Kieraszewicz HT, Dodgson BG. Chemical regulation of ventilation during isoflurane sedation and anaesthesia in humans. Canadian Anaesthetists Society Journal 1983; 3: Knill RL, Gelb AW. Peripheral chemoreceptors during anesthesia: Are the watchdogs sleeping? Anesthesiology 1982; 57: Nunn JF. Applied Respiratory Physiology, 3rd Edn. London: Butterworths, 1987; Temp JA, Henson LC, Ward DS. Does a subanaesthetic concentration of isoflurane blunt the ventilatory response to hypoxia? Anesthesiology 1992; 77:

6 514 BRITISH JOURNAL OF ANAESTHESIA 7. Sjogren D, Ebberyd A, Sollevi A, Lindahl E. Isoflurane anesthesia attenuates the awake response to hypercapnia but stimulates the rate response to hypoxia. Anesthesiology 1992; 77: A Douglas NJ, White DP, Weil JV, Pickett CK, Martin RJ, Hudgel DW, Zwillich CW. Hypoxic ventilatory response decreases during sleep in normal men. American Review of Respiratory Diseases 1982; 125: Howson MG, Khamnei S, Mclntyre ME, O'Connor DF, Robbins PA. A rapid computer-controlled binary gasmixing system for studies in respiratory control. Journal of Physiology (London) 1987; 394: 7P. 1. Robbins PA, Swanson GD, Howson MG. A predictioncorrection scheme for forcing alveolar gases along certain time courses. Journal of Applied Physiology 1982; 52: Hill DW. Physics Applied to Anaesthesia, 4th Edn. London: Butterworths, 198; Knill RL, Clement JL. Variable effects of anaesthetics on the ventilatory response to hypoxaemia in man. Canadian Anaesthetists Society Journal 1982; 29: Knill RL. Uncertainties about the ventilatory response to hypoxaemia. British Journal of Anaesthesia 1991; 67: Easton PA, Slykerman LJ, Anthonisen, NR. Ventilatory response to sustained hypoxia in normal adults. Journal of Applied Physiology 1986; 61: