The absorption of nitrous oxide (N2O) during unidirectional flow was compared with the rate of uptake of nitric oxide (NO). At flow rates of 10, 20, and 60 ml/min from one nostril to the other, with the soft palate closed, the N2O reached a steady-state rate of absorption in 5–15 min. The mean superficial capillary blood flow (n = 5) calculated from solubility and the steady-state rate of N2O absorption ranged from 13.3 to 15.9 ml/min. The relation between absorption of N2O in the nose and capillary blood flow fits a ventilation-perfusion model used by others to describe uptake of inert, soluble gases in the rat nose. By contrast, the rate of uptake of NO gas, which is chemically reactive, is 25–31 times as great as predicted by just its blood-to-air partition coefficient. Exogenous NO (16.9 parts/million) did not induce nasal vasodilation as measured with laser Doppler and N2O absorption methods. The difference between the measured rate of uptake of NO and the rate of uptake attributable to its partition coefficient in blood at the rate of blood flow calculated from N2O uptake is probably due to chemical reaction of NO in mucous secretions, nasal tissues, and capillary blood.
- nasal uptake models
- quantitative absorption
the nose protects the delicate structures of the lower respiratory tract by absorbing harmful inhaled gases. The efficiency of this action is due to the ability of the rich vasculature of the nose to remove gases deposited in the nasal mucosa and underlying tissues (1).
Morris et al. (17-21) reported, in rats, hamsters, and guinea pigs, that inert gases, which are absorbed according to their blood-to-air partition coefficients, attained a steady-state rate of absorption between the nasal gas, tissues, and blood while they were passing through the nose and upper respiratory tract. Their work indicated that a simple ventilation-perfusion model can reasonably describe uptake of some nonmetabolized vapors in the noses of some species (21). This remained to be shown in the human nose.
Development of a gaseous uptake model of the human nose would be of value. Toxicologists are interested in predicting doses of inhaled gases delivered to specific tissues such as bronchial walls and alveolar membranes. Landahl and Herrmann (15) demonstrated that retention of certain gases in the human nose reached levels as high as 83%, yet Medinsky et al. (16) point out that many inhalation studies have neglected the effect of the nose in reducing gas uptake in the remainder of the respiratory tract. The effect of the nose on gases passing through it should not be excluded from a comprehensive model of overall respiratory tract uptake of inhaled gases. Despite this, experimental work in humans has been limited.
Nitric oxide (NO) is a highly reactive gas with a myriad of functions in the body. It is a vasodilator and neurotransmitter, and it appears to play roles in the regulation of immunity and inflammation (14). A previous study on production and absorption of NO in the human nose has been performed in this laboratory (7). A question arose as to how much NO was taken up due to reaction in the tissues and blood compared with the uptake due to its partition coefficient (a value related to its solubility).
The present experiments compare the nasal absorption of the inert gas nitrous oxide (N2O) with the nasal uptake of NO. The main purpose of the experiment was to determine the fraction of NO uptake that was due to solubility during its removal by the bloodstream, as calculated by its partition coefficient and the rate of blood flow with which it came in contact. The total rate of NO uptake minus the rate attributable to simple solubility represents the amount of NO undergoing chemical reaction in the nasal tissues. Furthermore, experimental confirmation was sought for the application of a simple ventilation-perfusion model of inert-gas uptake in the human nose, as used by Morris et al. (17-21) on small mammals, and traceable back to a model of gas exchange in human lungs as used by Henderson and Haggard (10).
Experimental design. The experiment was designed to study the rate of absorption of N2O in the nose during quasi steady-state uptake conditions at nasal gas flow rates of ∼10, 20, and 60 ml/min. A comparison between N2O uptake and previously published values for NO uptake (7) was made.
Subjects. Volunteers with prior experience in respiratory physiology were used, because the breathing maneuver required voluntary closure of the soft palate for periods of 10–15 min.
Apparatus. Different gas flow rates through the nose were produced by a variable-speed roller pump (Masterflex model 7518, Cole-Parmer Instrument Company, Niles, IL) and Tygon tubing (5 mm ID). The flow rate was calibrated by using a ground glass syringe. The N2O mixture [concentration of N2O entering the nose ( ) ∼25% N2O, remainder air] was delivered to the nostril from a Teflon gas-collection bag connected to a nasal fitting placed snugly against the nostril to supply the . N2O concentration in the bag was analyzed before and after each trial. A length of Nafion (DuPont) tubing removed water vapor from the sample gas before analysis.
The other nostril was connected to a nasal olive consisting of a ground glass bulb (Pyrex ground joint ball, 18 mm OD, 9 mm ID, Corning Glass Works, Corning, NY) placed tightly against the nostril. The gas was drawn from it through a Tygon tubing into a roller pump and to glass syringes used to collect samples. N2O gas samples were collected periodically in 10-ml ground glass syringes (Popper & Sons, New Hyde Park, NY) and then analyzed by infrared absorption with an Ohmeda model 5250 respiratory gas monitor (RGM). The electrical output from the gas analyzer was recorded on a strip-chart recorder (model BD41, Kipp and Zonen).
The Ohmeda RGM was calibrated by using serial dilutions of 100% N2O obtained from a United States Pharmacopoeia tank (BOC Gases) by mixing a known volume of the 100% N2O from the tank with known volumes of air.
Procedure. On arrival, the subject was seated in front of the apparatus and instructed in the procedure. The subject was shown how to seal the fittings tightly against the nostrils. Instruction was given on closing the soft palate while breathing through the mouth. Closure of the soft palate was aided by creating pressure in the mouth by either breathing through pursed lips or by producing plosive consonants (“papa” on expiration, “mama” on inspiration).
The subject placed the fittings tightly against the nostrils, and the soft palate was shut. Gas was drawn from the gas bag through the nostrils at flow rates of ∼10, 20, and 60 ml/min. Outward leaks at the nostrils were detected by holding a strip of tissue paper in front of the nose. Both outward and inward leaks at this site moved the tissue paper and signaled a leak. Inward leaks, produced by the opening of the soft palate, resulted in a dramatic drop in N2O concentration leaving the nose ( ). When a leak was detected, the subject was notified. If the subject was unable to correct the leak, the trial was stopped and repeated after a rest of at least 15 min.
The 10-ml gas samples were collected once a minute in glass syringes and analyzed immediately in the Ohmeda RGM. A steady-state of absorption was indicated when the concentration did not change for three successive samples. Then either the run was halted and the subject rested or the run was continued to confirm that a steady state had been reached. During the rest period, the bag was recalibrated for N2O concentration, and the roller pump was reset and calibrated at a different flow rate. When the subject had rested sufficiently, the next trial was begun. The runs were repeated until the results were reproducible.
Analysis of data. The experiment was designed to study steady-state absorption of N2O in the human nose and compare it with the uptake of NO. Once a steady state is reached, the amount of N2O being absorbed, as calculated from a ventilation equation, is equal to the amount removed by the blood, as calculated by using the Fick equation. A blood flow was calculated by each of two sets of equations combining the ventilation equation with the Fick equation. Set 1is based on upper respiratory tract absorption of gases in small mammals as done by Morris et al. (17-21). Their study suggested that the absorption of nonmetabolized gases in the upper respiratory tract of small mammals follows a ventilation-perfusion relationship. We analyzed our data in the same manner as did Morris et al. to determine whether this was also true for humans. From this analysis it became possible to calculate a value for superficial capillary blood flow by using their set of equations that employ terms for the deposited fraction of gas (D), nondeposited fraction of gas (N), and the reciprocal of gas flow through the nose (1/V˙). The equations of Morris et al. are listed in the .
We wrote set 2 based on N2O into the nose minus N2O out of the nose. The new set of terms is as follows: , , mean concentration of N2O in the nose ( ), gas flow into the nose (V˙n I), gas flow out of the nose (V˙n E), Ostwald solubility or blood-to-gas partition coefficient for volume of N2O (btps) equilibrated with blood at 37°C (λ = 0.47), and nasal superficial capillary blood flow (Q˙n sc). , , and are converted from fractions of dry gas tobtps in the calculations through multiplication by (B − 47)/B, where B is equal to the barometric pressure.V˙n EandV˙n Iare converted from atps tobtps in the calculations through multiplication by [(273 + 37)/(273 + Tam)] × [(760 − Ph 2 o 24°C)/(760 − Ph 2 o 37°C)], where Tam is the ambient temperature and Ph 2 ois the partial pressure of water vapor at the specified temperature. The amount of N2O that flows into the nose per minute equals ×V˙n I, and the amount of N2O that flows out of the nose per minute equals ×V˙n E. The amount of gas absorbed per minute is the amount of N2O into the nose minus the amount of N2O out of the nose per minute.
Setting the extraction from gas flow equal to absorption by blood flow Equation 1Since the rate of absorption of N2O by the blood flow (Q˙n sc× × ) causes the rate of gas flow leaving the nose to be less than the rate of gas flow entering the nose (V˙n I− V˙n E) Equation 2CombiningEq. 2 with Eq.1 by subtraction of Eq.2 from Eq.1 gives Clearing terms Solving forV˙n I Equation 3SolvingEq. 2 for blood flow Equation 4
Subjects. Nine subjects attempted the experiment. Four subjects failed to complete it because they were unable to maintain closure of their soft palate. Five subjects, four male and one female, were successful. They ranged in age from 23 to 73 yr, with a mean weight of 73 kg and a mean height of 178 cm.
Calibration of N2O analyzer.
Calibration of the Ohmeda 5250 RGM was linear from 0 to 50% N2O concentration, but there was a slight deviation from linearity at concentrations between 50 and 100% N2O.
Data analyzed by using new equations (set 2).
Table 1 listsV˙n E, , , , and the calculatedQ˙n scfor each subject. Group means and SD and SE values are also listed. An example of the calculation ofQ˙n scfrom the mean experimental values obtained on five subjects,V˙n Eset at 10 ml/min (Table 1), is as follows. At a meanV˙n Eof 11.2 ml (btps)/min, of 26.7% (stpd) (25.1%btps), and of 17.5% (stpd) (16.4% btps), where = ( + )/2, or 22.1% (stpd) (20.8%btps),V˙n Iis calculated by substitution of these values into Eq.3 (see methods):V˙n I=V˙n E(1 − )/(1 − ). Data analyzed by using Morris et al. (17-21) equations (set 1). Table2 shows the mean D and N fractions at each flow rate. Figure 1 shows the linear relationship of D/N to 1/V˙n Iand 1/V˙n E. The equations of Morris et al. (17-21) consider flow into the nose and out of the nose to be equal. At the slow flow rates we utilized, the amount of gas absorbed in the nose causes the gas flow rate into the nose to be greater than the gas flow rate out of the nose. Figure 1shows the divergence between the lines of ratios of D/N fractions graphed against the reciprocal of flow into the nose, 1/V˙n I,and out of the nose, 1/V˙n E, illustrating this point.V˙n Eis utilized for the calculation of blood flow with this set of equations because it is a measured value, whereasV˙n Iis a calculated value. The slope of D/N vs. 1/V˙n Eis 7.86. When the slope is divided by the blood-to-gas partition coefficient of N2O (0.47), a blood flow for all subjects, when using the equations of Morris et al. (set 1), is calculated to be 16.7 ml/min, a number very close to the value 14.6 ml/min calculated from the equations in set 2. The blood flow calculated from either set 1 orset 2 does not represent a total physiological flow but, rather, the blood flow that is in diffusion equilibrium with the airstream.
Exponential rise of N2O concentration.
The 10-ml samples of gas out of the nose, collected once a minute, showed a rise in from a low concentration toward an asymptote that was less than the . The asymptotic level of N2O concentration was continued as long as the subject maintained closure of the soft palate. When the is graphed against time, the curve of best fit follows the equation for an exponential rise approaching a limit. The equation for the exponential rise is y =a (1 − ) + c. The variablea represents the difference between inspired N2O concentration attime = 0 and N2O concentration at time infinity, b represents the rate constant, and c is the at time = 0.
Figure 2 shows an example of this in one subject. The equation for the exponential curve has anR value of 0.97 for this set of data. Similar trends and correlations were seen at the flow rates of ∼10, 20, and 60 ml/min.
Effect of amount of N2O absorption on calculations.
The absorption of N2O in the nose has several effects on the flow and concentrations measured. Table 2 shows thatV˙n I, calculated byEq. 3 in methods, is slightly greater than the V˙n E. The amount of this difference represents the rate of N2O absorption in milliliters (btps) per minute, as listed under the rate of N2O absorption ( ) in Table 2. There is a slight increase in with gas flow (P < 0.01). An increase in with gas flow should result in the calculated value for blood flow increasing (Q˙n sc =V˙ / × ). The increase inV˙ with gas flow is accompanied by an increase in with gas flow, which offsets the rise inV˙ . This explains why the calculation ofQ˙n sc is not significantly affected by the rate of gas flow (see below).
Effect of gas flow on calculated blood flow. Calculated blood flow was graphed against nasal gas flow through the nose to determine whether there was a trend correlating the two. Figure 3 shows the mean and SE for the five subjects at the three gas flow rates. Because the slope of the regression line of blood flow on gas flow was not significantly different from zero (P = 0.47), there was no significant correlation.
Additional refinements in calculation of blood flow. A third set of equations for calculation of blood flow was suggested by one of the anonymous reviewers. It consisted of two differential equations expressing the local rate of change of gas volume due to absorption by blood flow and the rate of change of nasal gas flow and concentration owing to this absorption. These two differential equations were combined and integrated between entrance and exit of the nose and to solve for blood flow. We calculated blood flow by using this third set of equations by substitution of values for N2O concentration entering and leaving the nose and gas flow leaving the nose. The blood flow in Table 2, instead of being 13.3, 14.5, and 15.9 ml/min at 10, 20, and 60 ml/min of nominal gas flow, would become 18.6, 19.8, and 21.9 ml/min, respectively. However, we are not yet sufficiently sure of this method of calculation to adopt these figures at this time.
NO is a biologically reactive gas. It is ∼1/10th as soluble in blood as N2O is (0.04–0.05 vs. 0.47). When NO is studied, however, an absorption coefficient (A NO) is found to be 18 nl/min per part per million (ppm), or milliliters per minute (7). Let us now compare an apparent partition coefficient for NO ( ) with its actual partition coefficient (λNO). Mathematically,A NO = ×Q˙, so =A NO/Q˙.A NO divided byQ˙n sc(the blood flow obtained from the N2O uptake data) is 18/14.6, or 1.23. Therefore, based on divided by λNO, the absorption of NO is 1.23/0.04 or 1.23/0.05. Therefore, the uptake of NO is 25–31 times as great as can be predicted due to removal by solubility in the blood flow.
The possibility existed that gaseous NO might have induced local vasodilation and, therefore, increased its own uptake. Two short experiments were performed to test this hypothesis. A mixture of NO and N2O was passed through the nose. The uptake of N2O was compared with our previous uptake data in the absence of increased NO levels. N2O uptake was not increased in the presence of elevated NO (16.9 ppm). We then measured nasal perfusion by using a laser Doppler method with and without NO in the air. No effect was seen with the addition of NO (16.9 ppm). However, a statement cannot be made about gaseous NO inducing vasodilation in the nose. NO is produced in this cavity, and “clean air” entering the nose contains NO after it has been passed through the nose. Therefore, it is possible that absorption of this endogenously produced gas may induce vasodilation. It can only be said that further gaseous NO did not appear to have an effect on nasal vasodilation. On the basis of our comparisons, we, therefore, conclude that the main uptake of NO in the nose may be attributed to various reactions, which it undergoes in the tissue, blood, or mucous secretions, rather than to a vasodilatory effect.
The uptake of NO in the nose is due partly to solubility but mainly to other factors, which are as follows. Reaction with the mucous secretions of the nasal mucosa cannot be discounted, but no studies have been done to assess this. Absorption studies have shown that tissue reactivity can enhance absorption of gases (17), but data on NO are limited. Evidence exists to suggest that NO is oxidized to nitrite in the presence of water and to nitrate in the presence of oxyhemoproteins in aqueous solutions containing oxygen (11). NO has a strong affinity for hemoglobin, and much of its uptake may be due to this reaction. Carlsen and Comroe (3) reported that the affinity of NO for hemoglobin in solution is 400,000 times that of oxygen for hemoglobin and observed that the reaction is extremely rapid. These factors, as well as others, are probably all partly responsible for NO absorption.
We performed earlier studies in which gas was allowed to flow through the nose for several minutes, and we found that the uptake of NO was not affected by the velocity of gas flow. The absorption coefficientA calculated in that study was found to be fairly constant despite a wide range of flows (8–347 ml/min). The human nose has asymmetric patterns of airflow (22). By varying the airflow velocities, it was hoped that different areas of the nose would be ventilated. Absorption would then be changed by varying the surface area exposed to the gas or boundary layer thickness above the tissue surface. No change was observed. This indicates that either all portions of the nose were ventilated or that the airflow patterns did not change over the range of flows used.
N2O uptake measured at gas flows of ∼10, 20, and 60 ml/min when analyzed by our method (equations inset 2) yielded mean calculated blood flows of 13.3, 14.5, and 15.9 ml/min, respectively. These values are not significantly different from each other. From this we conclude that the ventilation-perfusion model of N2O nasal uptake is valid over this range of unidirectional flows and yields values that may be used to compare or contrast with those obtained from steady-state measurements of NO at these flows.
The calculated blood flow does not represent the total nasal blood flow. Diffusion limitations occur in tissues deeper than 30 μm (7,12, 13). The thickness of the nasal mucous membrane varies from 0.3 mm to ∼5 mm in depth (4). Cauna and Hinderer (4) reported fenestrated subepithelial capillaries that by inspection appear to occur in depths of 30 μm. In general, the capillaries are closer to the surface than are the venous sinusoids that lie deeper in the tissue (23). The calculated blood flow may be representative of the flow through these superficial capillary vessels, which are responsible for most of the removal of inhaled gases from the nasal tissues near the surface. The porosity of the endothelial basement membrane and the fenestrations of the capillaries probably facilitate gaseous diffusion (4) and possibly increase the depth that inhaled gases can reach.
When ∼25% N2O in air was passed through the nose at each of the flow rates used, an exponential rise in N2O concentration approached a plateau between 5 and 15 min. The asymptote reached by the increase in concentration was lower than the N2O concentration that was introduced into the nose. This asymptote is at a steady-state level. The difference between the introduced concentration of N2O and the steady-state level of N2O leaving the nose reflects the amount of N2O that is removed by the blood flow. At each rate of gas flow, the blood flow calculated from N2O uptake was about the same. We conclude that this indicates that a steady state of absorption had been reached. Our equations useV˙n Eto calculateV˙n I. The volume of gas absorbed per minute causesV˙n Ito be greater thanV˙n Eto maintain the constant rate of gas flow into the analyzer. At rapid rates of nasal gas flow, the rate of N2O absorption is so small compared with the rate of gas flow that it need not be added to the rate of nasal outflow to calculate the rate of inflow for purposes of using the equations in set 1. However, this correction is necessary when slow rates of gas flow are studied, and the rate of N2O uptake is calculated by using equations in set 2.
Conclusions from ventilation-perfusion data. Morris et al. (17-21) studied the upper respiratory tract gas absorption in various animals. They found that the nasal tissues rapidly reach a steady state of absorption. The evidence for this was that, when the deposited fraction of gas divided by the nondeposited fraction of gas was plotted against the reciprocal of gas flow through the nose, the resulting relationship was linear. They stated that nasal uptake of some nonmetabolized vapors under unidirectional inspiratory flow conditions can be described by a ventilation-perfusion model used by others to describe uptake in the lung (2, 10, 12, 13). The results obtained by using the model were found to be vapor specific and species specific when applied to the nose (17, 20).
Our data were analyzed in the same way that Morris et al. (17-21) analyzed their animal data. Experimentally, we observed a steady state of absorption in 5–15 min of unidirectional flow with the soft palate closed. When D/N vs. 1/V˙n Ewas plotted, a linear relationship was obtained. By comparison of our findings with those of Morris et al., it is logical to conclude that under unidirectional flow the absorption of N2O in the nose can be described by a simple ventilation-perfusion model. A blood flow of 16.7 ml/min was calculated by dividing the slope of the line by the partition coefficient of N2O. As explained above, the calculated blood flow cannot be assumed to represent total blood flow.
Our study has shown that a ventilation-perfusion model adequately describes nasal uptake of N2O during unidirectional gas flow. Although rats are commonly used in inhalation studies, more experimental work needs to be done in humans, because available evidence suggests that rats are a poor model for comparison with humans (6). More advanced models of uptake in the human nose are currently limited by the lack of quantitative data on depths and volumes of nasal tissue, lack of total blood flow measurements, influences of airflow patterns, enzymatic distribution, and the different characteristics of respiratory and olfactory mucosa. Whereas these parameters exist in animal models (9, 21), a complete model of human nasal absorption must incorporate all these factors and is beyond the scope of this paper. The data obtained here may aid in further efforts at modeling the uptake of gases in the human nose.
This method for determining the nasal uptake of gases is not without drawbacks. Some subjects were not able to close their soft palate. After fatigue of the soft palate during a successful trial, some subjects were unable to regain voluntary control of it the same day. Congestion in the nose sometimes resulted in partial blockages, which obstructed flow. The nares would then collapse under suction from the roller pump. On these days, the experiments ceased. A continuous sampling of the nasal gases would prove more efficient than the batch method utilized. The fixed sampling rate of the Ohmeda analyzer, 200 ml/min, did not permit this. The benefits of this method are that it is noninvasive, isolates the nasal cavity, and that gas flow rate can be varied.
In summary, our research shows that the absorption of NO cannot be explained by its solubility alone but that its uptake appears due to its reaction in either the mucous secretions, blood, or tissues of the nose. The absorption of N2O in the human nose fits a ventilation-perfusion model of absorption, similar to that of the lung. The N2O equilibrates with the tissues supplied byQ˙n scand reaches a steady state of absorption in a few minutes of unidirectional gas flow. This method of measuring the uptake of gases in the human nose seems valid for research purposes.
The authors acknowledge the generosity of the Yale-New Haven Hospital in lending us the Ohmeda model 5250 respiratory gas monitor.
Address for reprint requests: A. B. DuBois, c/o John B. Pierce Laboratory, 290 Congress Ave., New Haven, CT 06519.
This work has been supported in part by Grant HL-53630 from the National Heart, Lung, and Blood Institute.
- Copyright © 1998 the American Physiological Society
Derivation of Equations for Uptake of Soluble Gases by the Nose
Morris and Cavanagh (18, 19) provide equations for absorption of soluble gases, as follows Equation A1whereV˙i is inspiratory flow rate and L is blood-to-air partition coefficient at 37°C Equation A2where D is amount deposited divided by amount inspired Equation A3and Equation A4The authors (18, 19) indicate that these equations, which were derived from equations for the pulmonary deposition of nonreactive gases [see Henderson and Haggard (10)], may apply by analogy to uptake of soluble gases in the nose. Morris and Cavanagh (18, 19) conclude that “The successful application of a V-P model (i.e., ventilation-perfusion model) to URT [upper respiratory tract] gas deposition in the rat suggests the air-blood barrier in the URT of this species is sufficiently thin to allow equilibration of acetone and ethanol vapors between the inspired air and capillary blood.”
Our findings on absorption of N2O in the human nose support this viewpoint and extend it to humans. However, in humans, the nasal mucosa is thicker than it is in the rat. Therefore, equilibration in the human nose is probably limited to the superficial capillary blood flow.