| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Section of Environmental Physiology, Department of Physiology and Pharmacology, Karolinska Institutet, Stockholm, Sweden
Submitted 10 November 2004 ; accepted in final form 15 February 2005
 |
ABSTRACT |
|---|
 TOP ABSTRACT MATERIALS AND METHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
human physiology; cardiogenic oscillations; phase 4
A logical experimental approach to the question of nongravity-dependent heterogeneity is to perform experiments in microgravity (0 G) (20, 35, 39). Glenny et al. (20) injected fluorescent microspheres into the lung circulation of pigs during parabolic flight and then determined their topographical distribution in the excised lungs. So far, no methods are available to obtain corresponding information in humans at 0 G. Thus, in two conceptually important human studies, a more indirect technique has been used (35, 39): gas distribution in the lungs is first rendered more homogeneous, and CO2 concentration is lowered by a hyperventilation maneuver, then O2 and/or CO2 are allowed to be exchanged in the lungs during breath holding, and finally the gas is slowly expired. During the slow expiration period, heart-synchronous sequential emptying of lung units during phase III (alveolar phase) and airway closure during phase IV of the expirogram permit the detection of differences in O2 or CO2 concentrations between lung units as cardiogenic oscillations (COS) (phase III, COS O2, COS CO2) and phase IV phenomena [closing volume and phase IV amplitude (P4)]. A critical assumption in this design is that the concentration differences underlying COS and phase IV arise mainly during breath holding.
Michels et al. (35) found residual COS O2 and COS CO2 in transient 0 G, and Prisk et al. (39) found residual COS CO2 in sustained 0 G. They concluded that at 0 G there is still perfusion heterogeneity in the lungs. However, they did not determine to what extent any residual heterogeneity of O2 or CO2 distribution within the lungs after the initial hyperventilation would have influenced their results. Also, they did not discuss the potential influence of the gravity level during the post-breath-hold expiration: e.g., differences in stroke volume between gravity levels could influence the degree of mechanical agitation of the lung tissue by cardiac pumping, thereby modifying COS amplitudes in a manner not related to perfusion heterogeneity.
COS are determined mostly by intraregional, i.e., small-scale, gas concentration differences, and reports are contradictory about an added influence of interregional, i.e., apical-to- basal differences. If present, this influence is likely small (3, 32). In contrast, end-expiratory phenomena (phase IV) have been shown to be specific for interregional gas concentration differences (3, 32).
In the present paper, we present together two series of experiments performed with slightly overlapping subject populations. We estimated the heterogeneity of pulmonary perfusion during short-term 0 G and hypergravity with a further development of the technique described by Michels et al. (35) and Prisk et al. (39). In the first series of experiments, we studied sequential combinations of gravity that addressed mainly intraregional perfusion differences, whereas in the second series other combinations enabled us to address interregional perfusion differences as well.
We hypothesized that 1) signs of intraregional perfusion differences would be more in keeping with topographical studies in animals [80% of nongravitational heterogeneity (20)] than with previous estimates from humans [signs of nongravitational heterogeneity 20–50% of normal gravity (1 G) control (35, 39)]; 2) there would be signs of significant interregional perfusion differences at 0 G, which could only be detected if they were first allowed to develop at 0 G and then assessed in the presence of gravity.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
The sequence of a typical parabola was as follows (Fig. 1): starting from steady, normal horizontal flight, the aircraft produced a 1.8-G-load factor by nosing up to 45° and climbing to 7,600 m (23,000 ft.) over a period of 
20 s (the pull-up period). The thrust of the engine was then reduced to the point where it precisely overcame the aerodynamic drag (the injection period). The 0-G period lasted between 20 and 25 s and was followed by a 1.7-G pull-out period, symmetrical to the pull-up period, where the aircraft nosed down and was brought back to horizontal. The time that elapsed between two successive parabolas was 2 min, and a typical flight session consisted of 30 such parabolas. The average cabin pressure in flight was 80% of normal atmospheric pressure at sea level.
|
The subjects tested were healthy nonsmokers except for one subject in the first series of experiments who was an occasional smoker. This person did not smoke the day before the flight. All subjects had undergone extensive medical examination and were declared fit for parabolic flights by qualified medical authorities of their respective countries of origin. Preflight examination varied between nationalities. As a minimum, there was lung auscultation. The design of the studies was approved by the Ethics Committee of Karolinska Institutet in Stockholm, Sweden, the Comité Consultatif pour la Protection des Personnes dans la Recherche Biomédicale de Bordeaux, and the safety department of the Centre d'Essais en Vol at Bordeaux Airport. Written, informed consent was obtained from each subject.
The first series of tests was performed by twelve subjects, six men and six women, whose age, weight, and height [means (SD)] were 35 (9.3) yr, 64.5 (4.0) kg, and 169.5 (10) cm, respectively. The second series of tests was performed by eight subjects, three men and five women, with ages, weights, and heights of 37.4 (5.1) yr, 63.5 (15.6) kg, and 171.9 (14.4) cm, respectively. Four subjects were common to both studies.
Equipment
The respiratory monitoring systems used in the two series of flights were further developments of a photoacoustic gas analyzer previously described by Clemensen et al. (14). Both had been developed by the European Space Agency for use in space, with the device tested in series 2 being a prototype designed to be on the International Space Station.
During the first series of flights, a dedicated respiratory monitoring system, the Advanced Respiratory Monitoring System (ARMS; Damec, Odense, Denmark) was used. This device is made up of two parallel analyzers, each containing two elements. The first element is an optical cell containing either three or two filters for infrared analysis of CO2, sulfur hexafluoride (SF6), and Freon 22 (CHClF2, R22) in one analyzer, and methane (CH4) and carbon monoxide (CO) in the other. The second element of each analyzer includes a paramagnetic O2 analyzer. We were unable to analyze data on CO, CH4, CO2, and R22 because of the interference of the acoustic noise in the cabin with certain frequencies used in the photoacoustic gas analyzer. Therefore, we adopted O2 as blood-soluble gas and SF6 as insoluble gas.
During the second series of flights, we used a similar system but with a new generation of gas analyzers [pulmonary function system (PFS)] (Damec). This system is a further development of both the photoacoustic gas analyzer described by Clemensen et al. (14) and of the ARMS. Gases were analyzed by two pairs of analyzers operating in parallel. In each of these pairs, the first analyzer is an optical cell containing filters for infrared analysis of either CO2, SF6, and R22, or of CO and CH4. The second analyzer in each pair is designed for ultraviolet measurement of O2. The signal-to-noise ratio of this system is better than 1,000:1 on the ground but deteriorated during the gravity transitions in flight with respect to all of these gases except O2 and SF6.
The ARMS and the PFS devices had dynamic responses (10–90% rise time) of 130 and 120 ms, respectively. The time required for transit of the gas from the subject to the analyzer was, on average, 770 ms for the ARMS and 1.1 s for the PFS, being 1.1 s on the ground and 1.2 s in flight. This transit time was corrected for by shifting the gas concentration data in time by an amount equal to the delay between a flow spike of gas containing CO2 and the midpoint of the rise in the gas concentration data (5).
Both the ARMS and the PFS included a sensor for absolute ambient pressure and provided a signal output proportional to the fractional concentration of the analyzed gas components. Devices were connected to a rotary valve unit in which the subject could either breathe air or rebreathe gas from a 4-liter bag. For monitoring inspiratory and expiratory flows, a bidirectional, differential-pressure pneumotachograph (Damec) was attached to a filter connected to a wide-bore mouthpiece.
Gases were sampled through a 2-m-long Naphion tube (Perma Pure, Toms River, NJ) at a rate of 200 ml/min. The flowmeter was calibrated at the appropriate ambient pressure with a 3-liter syringe (Hans Rudolph, series 5530, Kansas City, MO), as described by Yeh et al. (46), on the ground and with a zero offset and gain setting step inflight. Gas analyzers were calibrated with two mixtures of known composition (AGA Gas, Bottrop, Germany).
Three ECG electrodes (Blue Sensor) were placed on the subject's chest for monitoring heart rate.
In the ARMS, data were sampled and stored digitally at a sampling frequency of 33.3 Hz except for the ECG and flow recordings, which were sampled at 200 Hz. In the PFS, all data were stored at a sampling frequency of 200 Hz.
Procedures
Subjects ate a light breakfast on the flight day, 3 h before takeoff, and were premedicated with an antiemetic drug (0.4 mg of Scopolamine and 5 mg of Dexamin) 1 h before departure. Nonetheless, two of the subjects suffered from motion sickness during flight series 2 and were unable to perform the tests. They have not been taken into account here. Two subjects in each series suffered from light nausea but could complete the tests. The remaining subjects tolerated the parabolas without problems.
Onboard the aircraft, the subjects sat upright in a passenger seat at an angle of 28° from the vertical, facing forward, and were restrained by a seatbelt fitted loosely to allow expansion of the chest and abdomen. A rebreathing bag was filled with 2.5–4 liters of a gas mixture consisting of either 25% O2, 1% SF6, and 0.3–5% of each of the gases CO, R22, and CH4 (series 1); 21% O2, 1.2% SF6, and 1.2% R22 (ground control tests, series 2); or 26% O2, 1.2% SF6, and 1.2% R22 (inflight tests, series 2). The balance was N2 in each mixture. The higher level of O2 in flight in series 2 was used to compensate for the lower cabin pressure. The volume of the rebreathing bag was 50% of the subject's estimated total lung capacity at normal gravity and was then slightly adapted according to the subject's preference.
After donning a nose-clip (Hans Rudolph), the subject took several normal breaths through the mouthpiece with the valve in the nonrebreathing mode. After expiring to residual volume, he/she switched the valve to rebreathe the entire volume of the bag four times, starting from residual volume, within a total period of 5 s (rebreathing). Thereafter, he/she emptied the bag, held his/her breath for 10 s (breath hold), and, finally, expired slowly to residual volume (expiration) (Fig. 2). Subjects were instructed to hold their breath with their glottis open if possible. Expiration was performed to the cabin at a flow rate adapted to each subject, which corresponded to the bag volume exhaled in 10 s. This flow rate was displayed on a screen, which allowed the subjects to control it.
|
The rebreathing-breath hold-expiration maneuver was performed on the ground (1 G ground, series 1 and 2), in flight at 1 G (1-G flight, series 2 only), and under four different combinations of hypergravity (1.7 or 1.8 G) and 0 G, i.e., rebreathing at 1.8 G (R)-breath holding in 0 G (b)-expiration in 0 G (e) (Rbe; series 1), rebreathing at 1.8 G (R)-breath holding at 1.8 G (B)-expiration in 0 G (e) (RBe; series 1), rebreathing at 0 G (r)-breath holding at 0 G (b)-expiration at 1.7 G (E) (rbE; series 2), and finally rebreathing at 0 G (r)-breath holding at 1.7 G (B)-expiration at 1.7 G (E) (rBE; series 2) (Fig. 1). For convenience, 1-G control and 1-G flight, when globally referred to, will be termed controls.
Criteria for acceptable test quality were proper timing with the different phases of the parabola, SF6 baseline at zero before the maneuver, speaking for complete clearance of the gases from the previous maneuver, complete bag emptying during the rebreathings, no airflow during breath holding apart from heart-synchronous flow oscillations compatible with open glottis, and a constant expiratory flow during the expiration period (i.e., ±10% of the target flow for each subject).
Data Analysis
Expiratory tracings were analyzed with a proprietary interactive graphic analysis program developed with the LabVIEW software (National Instruments, 1998, Austin, TX). Tracings were normalized by setting the mean concentrations of SF6 and O2 during the fifth inspiration of the rebreathing period to 100%. This was done because we wanted to correct for variability between subjects in the lung volume-dependent dilution of inert gas and thus allow direct comparison of the expiration parameters. The expirations for O2 and SF6 were then plotted as a function of expired volume (expirogram). Furthermore, to correct for incomplete equilibration of lung gases after rebreathing, the normalized tracings for the blood-soluble gas O2 were divided by the normalized tracings for the nonsoluble gas SF6. The normalized O2/normalized SF6 (O2/SF6) tracings thus obtained were plotted as an expirogram (Fig. 3). Henceforth, the terms O2, SF6, and O2/SF6 will refer to normalized values. In the three different types of expirograms, the beginning and end of the alveolar plateau (phase III) were identified visually and marked. The slope of phase III thus identified was determined for each plotted curve using linear regression and corrected for by subtraction. The software defined phase III oscillations as cardiogenic when they occurred between two R peaks in the ECG recording made simultaneously. Phase IV was defined as the first convincing departure of the tracing from the slope of phase III, and closing volume was defined as the point at which the expired gas concentration deviated from the phase III slope (1, 7).
|
The data for each subject and set of conditions were pooled, and individual medians were calculated in each condition. Finally, a single mean group value for each condition was computed.
Statistical Analysis
Where applicable, data were analyzed with a one-way repeated-measures ANOVA (STATISTICA 6, Statsoft, Tulsa, OK). The experimental design included the factor "condition," (control, Rbe, and RBe for series 1 and ground control, 1-G inflight, rbE, and rBE for series 2), three levels for series 1, and four levels for series 2.
The requirements for ANOVA were fulfilled for all parameters except for phase IV phenomena for O2 in series 1 and COS O2/SF6 in series 2. Values for closing volume and O2 had not sufficient variance. In this particular case, a t-test was applied for pairwise comparisons between conditions. Moreover, sphericity was not verified for COS O2/SF6. The critical F value for this parameter was therefore corrected with Greenhouse-Geisser and Huyn-Feldt epsilons.
A P value of <0.05 was considered significant. In all cases where such significance was observed, a Tukey's honestly significant difference post hoc test was performed for all pairwise comparisons.
|
RESULTS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
|
|
|
|
COS SF6 (Table 1) was similar in the two inflight conditions and was 50% larger in Rbe and RBe than in control. There was no difference between Rbe and RBe. COS O2 and COS O2/SF6 (Table 1, Fig. 4) differed between conditions and were lower in Rbe (breath holding at 0 G) than during RBe (breath holding at 1.8 G). Moreover, COS O2 was lower in Rbe than at control (breath holding at 1 G), and COS O2/SF6 tended to have the same result (P = 0.051). COS O2 was 26–44% lower than corresponding values for O2/SF6. There were no differences in COS O2 and COS O2/SF6 between control and RBe.
Closing volume and P4 for O2, SF6, and O2/SF6 (Table 1) differed between conditions, and values were always larger during control than inflight, where expirations took place at 0 G. In fact, no significant closing volumes and P4 were found inflight, since their 95% confidence interval encompassed zero.
Series 2
COS O2, COS SF6 (Table 2), and COS O2/SF6 (Table 2, Fig. 5) all differed during the various conditions tested. The values for COS O2 and COS O2/SF6 during rbE were lower than under all of the other conditions. Moreover, COS O2/SF6 was lower during controls than during rBE. In general, the means for COS O2 were 13–50% lower than the corresponding values for O2/SF6.
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
Gas Equilibration During Rebreathing
Even though Verbanck et al. (43) have shown that 98% of the washin of an insoluble indicator gas is completed after five rebreathings with a bag volume of 1.8–2.2 liters and we used larger volumes than these, the much higher frequency and respiratory flow of our rebreathing maneuver prevented the complete equilibration of rebreathing gases between the bag and the lungs. This was shown on the tracings of the insoluble gas SF6 by the existence of COS, speaking for intraregional heterogeneity (3, 32) of ventilation, and the presence of phase IV phenomena, speaking for interregional heterogeneity (3, 32) of ventilation, during control. Moreover, all of these parameters for SF6 varied with condition, showing that ventilation distribution was not entirely comparable at all gravity levels. For example, we found that COS SF6 increased with rebreathing at 1.8 G, probably because of difficulty to rebreathe due to the increased weight of the chest and of increased airway closure (22).
Influence of Incomplete Gas Equilibration on O2 Data
In these studies, it is not possible to ascertain solely by analyzing O2 expirograms whether time-dependent variations of O2 concentration were the result of residual heterogeneous distribution of O2 after the initial rebreathing, heterogeneous perfusion and O2 uptake during breath holding, or a combination of both.
A comparison between COS O2 and COS O2/SF6 in series 1 (Table 1, Fig. 4) shows 24–43% lower COS O2 values in all conditions. Moreover, during control, P4 for O2/SF6 was 35% larger than for O2. The same phenomenon is observed in series 2. In summary, a measurement of O2 parameters alone would have ignored incomplete gas equilibration and underestimated signs of perfusion heterogeneity quite substantially.
Influence of Cardiac Activity on COS Amplitudes During Expiration
The mechanisms by which cardiac mechanical activity generates COS of expired gas concentrations are not completely understood. COS were first described by Dahlström et al. (15) and Fowler and Read (18). Dahlström et al. studied COS N2 during a single-breath washout maneuver in cadaver lungs and associated them with pulsations in the pulmonary vessels created with a rubber bulb filled with saline (15). They found that the expansion of the pulmonary vascular bed resulted in addition of gas to the expirate from alveoli with a relatively low ventilatory gas turnover. On the other hand, Arieli and colleagues suggested that COS of expired gas concentrations result from differential emptying rates from lung units with differing time constants in their response to pressure waves generated by the heart (2) or different locations in relation to a pulsating artery (4).
External cardiac volume changes are small during the cardiac cycle (27) and not likely to have any large impact on the rate of alveolar emptying. In fact, volume changes are mostly internal and reciprocal between the atria and ventricles (26, 27, 41). For instance, during the rapid ejection period of the ventricles, atrial pressure decreases, the atrioventricular groove moves down toward the apex, and blood flows into the atria ("ventricular systolic suction" phenomenon) (9).
At the same time, during systole at rest, 60–70 ml of blood (stroke volume) are ejected from the left ventricle and thus leave the thorax (9). The amount of blood ejected from the left ventricle during systole is simultaneously partially compensated for by right ventricular systolic suction, the remainder coming to the right atria during diastole (9). Thus there is probably a temporary systolic negative blood balance in the thorax, which, however, is smaller than the left ventricular stroke volume. Moreover, Wessale et al. (44) have shown that during the postexpiratory pause of a normal breath in the dog, for each 10 ml of left ventricular ejection there is 2–5 ml of air inflow in the trachea. If this relationship were the same in humans, one would expect the volume of air temporarily withheld from the expiratory flow during each systole and then expelled during each diastole to be 14–35 ml. If such a pulsatile volume would come primarily from lung units with short mechanical time constants, then COS of expired gas concentrations would reflect concentration differences between lung units with short and long mechanical time constants.
It has been shown that COS of expired gases are representative mostly of intraregional differences in the concentrations of these gases (3, 32). The possible participation of interregional (apical to basal) differences to COS has been studied by Laviolette and Cormier (32) and Arieli et al. (3), who showed that COS are not generated from gas concentration differences between apical and basal parts of the lung. This result speaks strongly for a small, if not totally absent, effect of interregional gas differences on COS amplitude.
Intraregional Perfusion Heterogeneity
Although the present indirect approach does not permit any analysis of a topographical nature, nor of the mechanisms for nongravitational lung perfusion heterogeneity, it is of interest to compare our results with more direct measurements of lung perfusion distribution.
Hakim et al. (24, 25) studied isogravity slices of dog lungs using single-photon computed tomography and described an "eggshell model," i.e., concentric oval territories with the innermost central area representing the highest flow. This distribution has not been found by other groups and remains controversial. Melsom et al. (34) observed an 40% gravity-independent heterogeneity in sheep lungs within horizontal slices in both inner and outer layers. Comparable numbers have been estimated by electron-beam computerized tomography in dogs and humans (13, 31). Findings on pigs suggest that, if there were any central to peripheral gradient in blood flow, it would likely be low (20).
In the dog, it seems that the geometry of the pulmonary vascular tree is an important determinant of regional blood flow (6). In baboons, animals whose regional blood flow gradient might more closely resemble that of humans, measurements performed after prostacyclin infusion have shown that the regulation of pulmonary vascular tone is not responsible for inequalities of pulmonary perfusion during normoxia (21).
In summary, previous data on the topographical distribution of lung perfusion indicate that there is substantial nongravity-dependent heterogeneity. This heterogeneity is probably related to vascular structure rather than to active control.
Invasive imaging techniques have been used in humans in different gravity conditions, but the results are difficult to interpret and cannot be compared with our results: injections took place either in an undistorted (0 G) or in an overdistorted lung (4 G, 8 G), whereas scanning was performed at 1 G. Thus what was measured was likely the distribution of pulmonary perfusion per vascular unit and not the spatial distribution in micro- and hypergravity per unit lung volume (29, 42). Mostly due to mass and volume restrictions, high-resolution imaging techniques to provide three-dimensional topographical data are presently not suitable for inflight use.
In series 1, we found that COS O2/SF6 values were lower in Rbe than in RBe. The comparison between these two conditions is straightforward because they differ mainly by the gravity level during breath holding. COS O2/SF6 tended also to be lower in Rbe than at 1 G. There, the comparison is less obvious than between the two inflight conditions, because Rbe and 1 G differ in the gravity level during both breath holding and expiration. However, a plot of the data from the three conditions together (Fig. 4) suggests a linear relationship between the gravity level during breath holding and the amplitude of COS O2/SF6. A logical conclusion of this is that the gravity level during the expiration period does not influence the outcome of COS O2/SF6 greatly and that the major influencing factor is the gravity level during the breath-holding period.
In series 2, we were able to perform control tests both on the ground and in flight. There were no significant differences between any of parameters obtained from the O2/SF6 expirograms recorded during these conditions. This observation demonstrates that the differences in total ambient pressure, gas density, and partial pressure of O2 on the ground and during 1 G in flight exerted no impact on COS and phase IV phenomena. This in turn validates the conclusions drawn from the series 1 experiments, which are based partially on comparisons between ground and inflight conditions.
Our finding with respect to COS O2/SF6 in series 2 confirms that of series 1 of 80% of gravity-independent indexes of intraregional heterogeneity in pulmonary perfusion (72% in series 2) and does not contradict the possibility of a linear relationship with regard to the level of gravity during breath holding (Fig. 5). The only discrepancy between the two series of experiments is that the COS O2/SF6 values observed in series 2 were generally lower. This is probably a random occurrence and unlikely an effect of age because the average age of the subjects studied was comparable in both studies (35 and 37 yr, respectively).
Our data confirm and extend the observations made by Michels et al. (35) and Prisk et al. (39), although the decrease of COS of blood-soluble gas (O2 and CO2) observed by these teams at 0 G was much larger than our findings on COS O2/SF6. Michels et al. (35) report data from one subject in parabolic flight who performed breath holdings at 0 and 1 G and expirations at 0 G. COS O2 values were 77–84% lower than the 1-G inflight control. Prisk et al. (39), who studied subjects performing both breath holding and expiration at either 1 G standing or sustained 0 G, did not record COS O2 but observed that COS CO2 was 44% lower at 0 G than at 1-G control. In the present study, we found a 19% decrease of COS O2/SF6 between 1 G and breath holding at 0 G. When comparing COS O2/SF6 between Rbe and RBe, we found 29% lower COS O2/SF6 with breath holding at 0 G. These data speak for decreased but still existing intraregional perfusion heterogeneity at 0 G and probably increased heterogeneity at 1.8 G. The larger indexes of residual heterogeneity of perfusion at 0 G found in the present study, as reflected by COS O2/SF6, could be explained by an absence of correction for incomplete gas mixing after rebreathing in the other studies, which, as discussed previously, underestimates COS O2 to different extents depending on the gravity level during rebreathing. The difference could also be caused by the use of CO2 as blood-soluble gas instead of O2, as discussed in Benefits and Limitations of the Study Design below.
Origin and Significance of Phase IV Phenomena
At the final stage of expiration, a deviation in the phase III slope, referred to as phase IV, can be observed in the expirogram, a phenomenon due to airflow limitation in certain parts of the lung (38). This limitation leads to closure of airways in pulmonary units where the pleural pressure becomes greater than the alveolar pressure (12, 16) as expiration proceeds and lung volume gets closer to residual volume (10, 36). Phase IV phenomena are representative of interregional differences in pulmonary gas concentrations (32) such that the greater the basal-to-apical differences, the greater the value of P4 (16).
The occurrence of sequential emptying, together with these differences in gas concentrations, is a prerequisite for the appearance of phase IV phenomena: if all pulmonary units should empty at the same time, differences in gas concentrations would not appear in the expirate because of gas mixing between the different units. In the presence of the pleural pressure gradients induced by gravity, basal regions of the lung are relatively more deflated at the end of expiration due to the overlying weight of the rest of the lung (37). As a consequence, airway closure is not homogeneous in the lung; the highest (least negative) pleural pressures being in the dependent parts, a basal-to-apical sequential emptying occurs. Furthermore, airway closure is influenced by bronchomotor tone because induced bronchoconstriction decreases P4 and closing volume in the seated subject (28) and increases regional residual volume (16). Moreover, closing volume is said to be sensitive to elastic recoil (28). P4 is thought to be influenced by alterations in interregional gas content rather than by changes in sequencing (28).
Effects of Gravity on Phase IV Phenomena
In series 2, hypergravity was used as a tool to produce sequential emptying to detect possible differences in the ratio of O2/SF6 concentrations between pulmonary units. As depicted in Figs. 6 and 7, the magnitude of phase IV phenomena increase as the level of gravity present during breath holding increases. P4 would be expected to exert such behavior, but this is not necessarily the case for closing volume, since the extent of airway closure must primarily depend on the gravity level during expiration. The observation that closing volume increases in parallel with P4 probably reflects the fact that airway closure is easier to detect when the interregional differences in gas content are large, such as after breath holding at 1.7 G (rBE).
Phase IV Phenomena as Indicators of Interregional Heterogeneity of Pulmonary Perfusion
Phase IV phenomena are usually calculated during single-breath washout maneuvers from tracings of inert insoluble gases and are employed to determine the extent of airway closure (10, 16, 36). The principle underlying the single-breath washout maneuver involves dilution of a resident gas (classically N2) by inspiration of one vital capacity of another gas (classically O2), resulting in temporal variations in the N2 concentrations in the expired air that reflect heterogeneities in ventilation. On the other hand, if ventilation is rendered homogeneous and a blood-soluble gas is monitored, recording of phase IV phenomena after a period of breath holding could be used to assess interregional heterogeneity in pulmonary perfusion. To date, only a few groups have employed phase IV phenomena for this purpose (35, 39, 40).
Series 1. Although the phase IV results from series 1 are difficult to interpret because of the absence of a gravity-induced sequential emptying during expiration at 0 G, we felt that it was necessary to present them here, first for completeness but also for comparison with the findings of series 2. There was a relatively large scatter in closing volume and P4 (phase IV phenomena) data at 1 G. This was expected since several studies have shown that closing volume has a large inter- and intraindividual variability (7, 11, 33). We found no similar studies on P4, but it is not surprising that it shows the same variability as closing volume because both are components of phase IV. With expiration at 0 G, as in Rbe and RBe, phase IV phenomena for SF6 and O2/SF6 did not differ from zero (zero was within the 95% confidence interval). Indeed, phase IV phenomena for O2/SF6 would be less evident with expiration at 0 G than at control because of the absence of gravity-induced airway closure. Thus the series 1 data on phase IV phenomena may very well have underestimated existing interregional O2/SF6 concentration differences compared with 1 G, since airway closure partly depends on gravity-induced pleural pressure gradients (23).
Series 2. Our finding of reduced absolute values of P4 for O2/SF6 (Table 2, Fig. 7) in connection with rbE is in agreement with the observations of Prisk et al. (39) on P4 for CO2 recorded during exposure to sustained 0 G after a maneuver comparable to the one we used in the present study. Although not statistically significant, the closing volume for O2/SF6 observed here (Table 2, Fig. 6) after breath holding at 1.7 G was greater than at 1 G. This trend is consistent with the measurements performed by Rohdin et al. (40) after a comparable maneuver at 2 G, although our closing volumes were generally smaller than those they reported. The difference could be explained by the approximately fourfold longer exposure to hypergravity utilized by Rohdin et al. The fact that the P4 value for O2/SF6 was greater at 1.7 G than at 1 G is also in agreement with the findings of these investigators. They proposed that hypergravity-induced vascular engorgement of dependent pulmonary regions reduces airway dimensions, thereby promoting airway closure. Such vascular engorgement would also prevent pulmonary units with a relatively high degree of perfusion from contributing to the expirate, probably more so at 3 G than at 2 G (40), thus masking part of the heterogeneity in perfusion. Considered together, these observations indicate an increase in the interregional heterogeneity of perfusion under conditions of hypergravity, in agreement with the topographical xenon measurements performed by Glaister (19).
Relative Impact of Breath Holding vs. Expiration on P4
The P4 values associated with rbE and rBE can be considered to reflect events occurring during both breath holding and expiration. Thus, although differences in the ratio of O2/SF6 concentrations in connection with rBE develop at 1.7 G over a period of 20 s, i.e., before they appear as P4, in the case of rbE these differences arise in response to first 10 s of exposure to 0 G followed by 10 s at 1.7 G. Comparison of the P4 values obtained for O2/SF6 during rbE and rBE reveals that the latter were four times larger. In the theoretical case, where interregional differences in perfusion disappear totally during breath holding at 0 G, the P4 value associated with rbE would reflect differences in perfusion caused exclusively by exposure to 1.7 G during expiration. These differences would only become evident toward the end of expiration as a consequence of gravity-induced sequential emptying.
The observation that the value of P4 at 1 G is approximately one-half of that observed during rBE suggests that interregional differences in the ratio of O2/SF6 concentrations that develop during breath holding and expiration are linearly proportional to the level of gravity when the exposure time to each level is the same. From this, it does not necessarily follow that, for the same gravity level, O2/SF6 differences develop as a linear function of time, but this would be a reasonable assumption. Making this assumption, the P4 value during rbE would be expected to be equal to that observed at 1 G and one-half of that associated with rBE. However, P4 during rbE is 25% of that during rBE, i.e., approximately one-half of what would be expected.
A tentative explanation for this would be that a "reversed difference in the ratio of O2/SF6 concentrations" develops during the period of rbE at 0 G. Such a reversed difference would indicate more extensive perfusion of pulmonary units that are not prone to airway closure during the subsequent expiration at 1.7 G (i.e., apical units) than of units more prone to such closure (i.e., basal units). In other words, we speculate that the fact that the P4 value for O2/SF6 in connection with rbE is lower than expected indicates the presence of a "negative" P4 at the end of breath holding at 0 G, which partially offsets the positive P4 value that is likely to develop during expiration at 1.7 G. An alternative explanation is that a much more severe degree of airway closure develops in rBE than in rbE due to the twice longer hypergravity exposure. The generally larger closing volumes in rBE compared with rbE are compatible with such a mechanism. In such a case, the lower than expected P4 in rbE would not necessarily be a sign of less interregional O2/SF6 differences than in rBE and thus not be a sign of negative P4 at the end of the breath-holding period.
Topographical Distribution of Interregional Pulmonary Perfusion at 0 G
The few available studies on the distribution of pulmonary perfusion at 0 G (20, 42) have not taken into account both of the gravity-dependent factors that influence this distribution, i.e., tissue distortion and the redistribution of flow with respect to the distorted vascular network. Some information has been provided by Stone et al. (42), who injected radioactive macroaggregates into sitting subjects, both at 1 G and during short-term exposure to 0 G, and then recorded scintigrams in the supine subjects on the ground. In this manner, the distribution of perfusion per unit of pulmonary volume at 1 and 0 G was examined with the same degree of tissue distortion after both gravity conditions. Compared with 1-G supine measurements, there was a 42% increase in perfusion in the upper 4.5 cm of the lung at 0 G, whereas perfusion in the middle part of the lung decreased by 15%.
The findings of Glenny et al. (20) on the distribution of pulmonary perfusion in pigs point toward a similar conclusion: following the injection of microspheres, these investigators examined vertical perfusion gradients in supine and prone animals at 0, 1, and 1.8 G. The data were subsequently gathered at 1 G after drying the lungs out and then inflating them to a pressure of 25 cmH2O, i.e., with the same degree of tissue distortion regardless of the level of gravity present during microsphere injection. The flow was redistributed from dependent to nondependent pulmonary regions at 0 G compared with 1 and 1.8 G.
Thus both of these topographical studies report that perfusion per unit alveolar tissue increases in the apical parts of the lung at 0 G. Although these reports provide no proof of the existence of a reversed perfusion gradient at 0 G, they do not refute the possibility that it exists.
Benefits and Limitations of the Study Design
We do not feel that the fact that we were unable to use CO2 as blood-soluble gas had a negative effect on data interpretation. On the contrary, we think that O2 is a better choice: within the range of partial pressures occurring in the present experiments (O2 above 75 Torr, 10 kPa), the dissociation curves for O2 and CO2 are approximately linear; the major difference between them is that the CO2 curve has a slope of the order of 50 ml·l–1·kPa–1, whereas the O2 curve is practically horizontal. Therefore, with assumed constant mixed venous contents of O2 and CO2 during a 20-s period (10-s breath holding + 10-s expiration), the rising alveolar and arterial CO2 will result in a gradual reduction of the arteriovenous difference in CO2 concentration, whereas the difference in O2 will be constant (or near so) (17). As a result, the rate of CO2 accumulation during the breath hold + expiration period will not be a constant function of time but will be larger in the beginning and then fall gradually. In contrast, the rate of O2 uptake (and fall in alveolar partial pressure of O2) will be an approximately linear function of time (8), which is an advantage in the present data interpretation.
Conclusions
Our data show that there is considerable nongravity-dependent intraregional heterogeneity of perfusion in the human lung. Compared with moderate hypergravity (1.8 G), 70% of the indexes of heterogeneity of O2 uptake between lung units during breath holding remains in short-term 0 G. Compared with 1 G, the corresponding value is estimated to be 80%.
We also found signs of residual interregional differences in pulmonary perfusion at 0 G. Thus some 50% of the indexes of interregional heterogeneity in perfusion at 1 G appears to be independent of gravity. These findings are, however, somewhat confounded by the necessity of obtaining gas samples from the lungs during expiration at 1.7 G to achieve sequential emptying of pulmonary units. In light of this factor, we speculate that interregional differences in perfusion do exist at 0 G but are reversed compared with those present at 1 G. Direct inflight measurements employing topographical imaging techniques with high resolution are required to confirm or refute this hypothesis.
|
GRANTS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
|
FOOTNOTES |
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
|
REFERENCES |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  R. Glenny Counterpoint: Gravity is not the major factor determining the distribution of blood flow in the healthy human lung J Appl Physiol, May 1, 2008; 104(5): 1533 - 1535. [Full Text] [PDF]
|
|
|
|
|
|
S. Montmerle and D. Linnarsson Effects of gravity and blood volume shifts on cardiogenic oscillations in respired gas J Appl Physiol, September 1, 2005; 99(3): 931 - 936. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
) obtained during an RBe sequence (see 
Significantly different from RBe (P < 0.05).
Significantly different from rBE (P < 0.05).
