| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
We sought to clarify the effect of nitrous oxide (N2O) on the immediate responses of cerebral vasculature to sudden changes in arterial carbon dioxide tension in healthy humans. By use of a transcranial Doppler ultrasonography, blood flow velocity in the middle cerebral artery (VMCA) was measured during a step increase followed by a step decrease in end-tidal CO2 tension (PETCO2) between normo- and hypercapnia while subjects inspired gas mixtures containing 70% O2 + 30% N2 (control) and 70% O2 + 30% N2O (N2O) separately. During the control condition, both step increase and decrease in PETCO2 produced rapid exponential changes in VMCA. An increase in VMCA produced by the step increase in PETCO2 was smaller (P < 0.001) and slower (P < 0.001) than a decrease in VMCA induced by the step decrease in PETCO2. These general features of the dynamic cerebrovascular response were not affected by substitution of N2O for N2 in the inspired gases although N2O increased baseline VMCA by 15% (P < 0.001) compared with the control condition. We conclude that N2O in itself does not affect the dynamic cerebrovascular response to arterial CO2 changes, although it produces static mild cerebral vasodilation. Implications: This study suggests that nitrous oxide does not affect the dynamic cerebrovascular reactivity to acute arterial carbon dioxide (CO2) changes, i.e., exponential changes in cerebral blood flow in response to step changes in alveolar CO2 tension, although it does produce a mild increase in normocapnic cerebral blood flow velocity.
Many studies have indicated that nitrous oxide (N2O), when combined with other anesthetics, affects cerebrovascular responses to hypo- or hypercapnia (15); however, only a few investigators have studied the effect of N2O as a sole anesthetic on the hypocapnic response of cerebral circulation (6,7). Despite the popular use of N2O and the clinical importance of hypercapnia on cerebral circulation, no study has investigated the effect of N2O alone on cerebral vascular responses to hypercapnia. Furthermore, previous studies have examined the static response of cerebral vessels to changes in PaCO2. However, anesthetized patients may encounter sudden changes in PaCO2 that could produce unexpected transient cerebral vascular responses. We therefore investigated the effect of N2O on the dynamic response of cerebral circulation to hypercapnia in humans. To determine temporal variations in cerebral blood flow, we used continuous measurement of blood flow velocity of the middle cerebral artery (MCA) using a transcranial Doppler ultrasonography. A mathematical model was adopted which allowed for quantitative interpretation of the dynamic response of cerebral blood flow (8).
Twenty-four healthy volunteers (20 men and 4 women, aged 2334 [mean 28.2] years; height 169 ± 8 [mean ± SD] cm; weight 61.0 ± 5.8 kg) took part in the study. Requirements were explained fully to all participants in writing and verbally, and each gave informed consent before participating in the study. The research was approved by the institutional ethics committee. Participants were not taking any medication, and none had a history of cardiovascular, cerebrovascular, or respiratory disease. A 2-MHz pulsed Doppler ultrasound system (PCDop 842; SciMed, Bristol, United Kingdom) was used to measure back-scattered Doppler signals from the right or left MCA. The Doppler signals were transformed to the maximal and weighted mean blood flow velocities, and the mean velocity (VMCA) was stored on a computer for off-line analysis. The MCA volume blood flow is the product of VMCA and the cross sectional area of MCA (8,9). Therefore, VMCA may not necessarily represent the volume blood flow in particular when MCA exhibits the temporal changes in the cross sectional area. However, Poulin et al. (8) showed that, during CO2 loaded breathing with PETCO2 variations similar to our study, changes in the MCA cross-sectional area were negligible compared with those in the VMCA. We therefore assumed that VMCA could represent changes in MCA volume blood flow. VMCA was identified by an insonation pathway through the right or left temporal window using a standard search technique (8). Small movements of the probe can cause some VMCA changes, which can be interpreted erroneously. To avoid this problem, extreme care was taken to identify the center of MCA where the signal was maximized and to attach the probe securely by a custom-made headband. The average insonation depth (the distance from the probe to the start of the Doppler sample volume for detecting signals from the MCA) was 5.1 ± 0.5 cm. Subjects were requested not to take foods or caffeine-containing beverages within 4 h before their testing sessions. Subjects were resting in a supine position on an operating table in a quiet room with the temperature maintained at 25°C. The usual monitoring was used. Each subject breathed spontaneously through a face mask fitted snugly to the face to ensure no leakage of inspired or expired gases from the breathing circuit. The face mask was connected to a nonrebreathing respiratory circuit. Expired gas was drawn continuously via a catheter placed inside the nasal cavity to an infrared anesthetic gas monitor (Normocap 200 oxy; Datex, Helsinki, Finland) to measure fractions of O2, CO2, and N2O in the expired gas. After obtaining steady state of respiratory and circulatory conditions, each subject started to breathe a gas mixture of either 70% O2 + 30% N2 (control) or 70% O2 + 30% N2O (N2O). After 15 min of accommodation period to the gas mixture, a series of measurements was started. After 3 min of quiet breathing with the subjects' natural PETCO2 (prehypercapnic period), hypercapnia was induced suddenly by supplementing CO2 to the inspiratory gas mixture. By closely observing the respired CO2 tension on the gas monitor, the supplementing dose of CO2 was manually adjusted breath by breath, such that PETCO2 increased to 50 mm Hg in a stepwise manner (in a few breaths) and was maintained at approximately 50 mm Hg for a subsequent 5 min. After the 5-min hypercapnic period, CO2 supplementation was terminated immediately, and subjects were left to hyperventilate, then to return to normoventilation for 3 min (recovery period), resulting in a quasi-stepwise PETCO2 decrease to the prehypercapnic level in several breaths. After the measurements were completed, inspiratory gas was switched to the other mixed gas (control to N2O, or N2O to control gas mixture). After 15 min of accommodation period to the new gas, the second series of measurements was performed. The order of administration of two gas mixtures was randomized; 12 subjects were given control gas first, and the other 12 were given N2O gas first. A relatively low concentration (30%) of N2O was used because cooperation of the subjects was required to control spontaneous breathing to obtain the aforementioned stepwise variations in PETCO2. Thus, a series of measurement protocols consisted of a 3-min prehypercapnic period with the stable subject's natural PETCO2, a sudden increase in PETCO2, a 5-min hypercapnic period with PETCO2 50 mm Hg, a sudden decrease in PETCO2, and a 3-min recovery period with the prehypercapnic PETCO2 (Fig. 1 ). Subjects were encouraged to regulate spontaneous breathing according to quiet and frequent requests from the investigators. Respiratory rate was adjusted at 16 breaths per min throughout the measurement by use of a metronome.
To align interindividual variance in the absolute values in VMCA, VMCA was normalized by dividing it by the mean value in the prehypercapnic period during control gas breathing. Thus, VMCA is expressed as relative values to the prehypercapnic mean value of control gas breathing.
A graphical representation of our modeling process is shown in Figure 1. A simple dynamic model described by Poulin et al. (8) was used, which is written in the form:
(sec) is a time constant, G (mm Hg-1) is a gain, and Td (sec) is a pure time delay. VMCA* and PETCO2* (mm Hg) are their respective steady-state values before a step change is undertaken. The change in steady-state VMCA is assumed to be proportional to the change in PETCO2. Such a dynamic model produces an exponential output (VMCA) for a step input (PETCO2).
To allow for asymmetry between the VMCA response to a step increase (on-response) and to a step decrease (off-response) in PETCO2, separate variable values were estimated for the on- and off-responses. This resulted in five variables for estimation, namely, gains for the on- and off-responses (Gon, Goff), time constants for the on- and off-responses ( Model fitting for variable estimation was performed on the on- and off-responses separately. The on-response model was fitted to the data from duration containing the 3-min prehypercapnic period and the first 3-min hypercapnic period with a step PETCO2 increase in the middle. The off-response model was fitted to the data from duration containing the last 3-min hypercapnic period and the 3-min recovery period with a step PETCO2 decrease in the middle. The best fit models which minimized the sum of square of residuals between the data and model were computed using a grid search technique (8). Statistical comparisons were performed using the Mann-Whitney U-test. P < 0.05 was considered significant.
All subjects had normal mean blood pressure (82 ± 8 mm Hg) at the beginning of the experiment. In all subjects, variations of blood pressure remained within a range of the value at the beginning of the experiment ±10 mm Hg. Heart rate variations also remained within a range of the value at the beginning of the experiment ±15/min throughout the experiment. Inspiration of N2O in O2 increased VMCA by 15% ± 16% (P < 0.001) as compared with VMCA during inspiration of N2 in O2. Figure 2 A shows the responses of VMCA to step changes in PETCO2 in a representative subject. Exponential contours in the VMCA response curve to step changes in PETCO2 were observed both in control and N2O gas breathing. Figure 2B delineates the models best fitted to the data shown in Figure 2A. Model fitting performance was good in all the subjects and the model was able to track the dynamic changes of VMCA in both gas mixture breathing.
Table 1 shows comparisons of the model variables between the two gas groups, and between the on- and off-responses. No variable differed significantly between the two groups. However, on were two times greater than off in both control (P < 0.001) and N2O (P < 0.01) groups, indicating slower on-response than off-response. Gon was smaller than Goff in both control (P < 0.001) and N2O (P < 0.001) groups, implying smaller on-response than off-response.
Inhalation of 30% N2O increased baseline VMCA by 15%, as compared with VMCA during inhalation of 30% N2 in O2. This is consistent with the results of previous studies which found that N2O caused a mild degree of cerebral vasodilation (10,11). Only one previous study performed precise step changes in PETCO2 while measuring continuously beat by beat an index of MCA blood flow in humans (8). That study reported that the shape of the curve relating the cerebral blood flow responses to step changes in PETCO2 was exponential. Only their technique allows for a quantitative description of the immediate responses of cerebral vasculature to sudden changes in PETCO2 in vivo. Using their technique, therefore, we intended to address the effect of N2O on the dynamic cardiovascular response to PETCO2. Our study exhibited essentially the same dynamic cerebrovascular response and model performance in control gas breathing. N2O did not induce any complicated or unexpected behavior in VMCA response, such as overshoots and damped oscillations, which were produced by sudden changes in arterial blood pressure (12).
The results of the variable estimation in control gas breathing obtained in our study were comparable to those of Poulin et al. (8). Their cerebrovascular response times (Td, Magnitudes of the responses (Gon and Goff) reported by Poulin et al. (8) and us are slightly larger than those of previous dynamic studies (14,16) and approximately 2 times larger than those of steady-state hypocapnia studies (6,7). This discrepancy may have been produced by sustained hypo- or hypercapnia in the previous studies, in which gradual adaptation of cerebral vascular regulation toward baseline levels may have occurred (16,17). Greater Goff than Gon may be strange because it implies that cerebral blood flow decreased below baseline (prehypercapnic) levels during the recovery period. Poulin et al. (8) attributed this greater Goff to a transient undershoot in the MCA blood flow observed immediately after a step decrease in PETCO2, which then diminished in several minutes. No significant effects on all the model variables were produced by N2O, which indicates that N2O (Fig. 3 ) does not affect the general features of the dynamic cerebrovascular response to arterial CO2 changes and involved mechanisms. The initial step of direct vasodilation may be the alteration of extracellular pH, which precedes mechanisms including nitric oxide-cyclic guanosine 3,5-monophosphate pathway, activation of KATP channels, and an effect of H+ on membrane fluxes of Ca2+ (17). Although it is unclear which mechanisms are responsible for the immediate responses to step changes in PETCO2 that we observed, our results indicate that N2O does not alter the mechanisms responsible for dynamic cerebrovascular reactivity.
We cannot be assured that our results still hold in cases in which higher concentrations of N2O are applied. However, we believe that the general property of the dynamic cerebrovascular response to a step change in PETCO2 is conserved among a variety of concentrations. Our belief is based on a previous article that showed that inhaled N2O at two concentrations of 30% and 60% produced a similar extent of static cerebral vasodilation (10). The literature regarding the effect of N2O on the cerebrovascular response to CO2 is somewhat confusing (17), presumably because of species differences, interactions with other anesthetics, or interventions (hypocapnia versus hypercapnia, and their level and duration). However, our study indicates that N2O in itself does not affect the dynamic cerebrovascular reactivity to acute arterial CO2 changes, although N2O produces a mild increase in baseline VMCA. This may imply that apart from a baseline increase in VMCA, N2O does not further increase risks arising from acute changes in cerebral vasculature produced by sudden changes in arterial CO2.
As described in the modeling section, the model used produces an exponential VMCA change for a step PETCO2 change. However, it was difficult for our manual adjustment technique to invoke precise step changes in PETCO2. Although the precise steps are not necessarily required for the modeling process, nonuniform step shapes of PETCO2 variations might affect the model-fitting performance. For example, a blunted step change in PETCO2 may produce a slow exponential change in VMCA, which would lead to greater values in time variables Td, on, and off. Furthermore, the PETCO2 baseline was not completely constant but was usually contaminated with slow and small oscillations because of respiratory unsteadiness particularly during N2O gas breathing. Therefore, we tested how the model-fitting performance was affected by the nature of the PETCO2 signal. First, three PETCO2 variations were simulated with various step-like shapes; the first showed a pure step, the second and third showed exponential PETCO2 changes with step (exponential) time constants of 5 and 10 s (approximately 78 and 1516 s of rise time), respectively. All PETCO2 data were contaminated with small and slow baseline oscillations, simulating spontaneous breathing instability. A representative set of model variable values obtained from one subject was substituted into Equations 1 and 2 to construct the simulated VMCA responses, to which noises were added to further simulate the sampling noises in transcranial Doppler ultrasonography measurements. The extents of the step bluntness and oscillations observed in our measurements were enclosed by the range of the simulation. Our mathematical model was fitted to these three sets of simulated data and the best-fit model variables were estimated. This process and the results are shown in Figure 5 and Table 2 . The model fitting was excellent in all three step responses. As indicated in Table 1, the estimation errors were very small and at most 10%, and no systematic difference was found in the model variable values between three simulations. These results may validate our manual breath-by-breath adjustment of PETCO2.
This work was supported by grant-in-aid 09671540 from the Ministry of Education and Science, Japan.
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|