Pulmonary Gas Exchange

JOHN B. WEST and PETER D. WAGNER

Department of Medicine, University of California San Diego, La Jolla, California

	INTRODUCTION

TOP INTRODUCTION REFERENCES

Naturally, such a vast topic as pulmonary gas exchange cannot be treated in any balanced or comprehensive way in this mini-review. Rather, we have concentrated on a few highlights over the last 50 years that are best known to us, and we have noted the role of the National Institutes of Health (NIH) and, particularly, the National Heart, Lung, and Blood Institute (NHLBI).

Modern pulmonary gas exchange largely owes its origin to advances made during and shortly after World War II. As such, the topic is almost exactly as old as the National Heart Institute, which evolved into the NHLBI. Of course, the origins of our knowledge of pulmonary gas exchange go back very much further to Lavoisier and beyond (reviewed in this supplement by Severinghaus, Astrup and Murray, p. S114- S122). However, the exigencies of war, particularly the problems associated with flying at high altitude, resulted in a rapid acceleration of research that continued into the 1980s.

One of the most fertile groups was under Wallace Fenn at the University of Rochester. Two of the main protagonists, Hermann Rahn and Arthur Otis, described the improbable partnership as follows:

It may seem incongruous that a group of individuals with such diverse and unrelated interests (Leigh Chadwick was studying Drosophila flight, Hermann Rahn was developing a bioassay method in frogs for testing intermedin hormone of the pituitary, and Arthur Otis was studying the activation and inhibition of the enzyme tyrosinase in grasshopper eggs) could put aside these activities to participate and collaborate effectively in a project on pressure breathing. None of us had any previous training in human physiology (we were not certain which lung volume was, in the terminology of the day, complemental air and which was supplemental air) (1).

Such were the demands of war that in a few years this group had laid the foundations of modern pulmonary gas exchange and respiratory mechanics!

A major advance was the development of the oxygen-carbon dioxide diagram, and an early example is shown in Figure 1. This was initially developed to analyze the effects of hyperventilation on alveolar gas composition at high altitude (2). Note the remarkable ability of the diagram to relate at least six variables (PO₂, PCO₂, respiratory exchange ratio, arterial oxygen saturation, alveolar ventilation, and altitude). This work naturally led to the classic studies of ventilation-perfusion relationships, which later proved to be so important in understanding abnormal gas exchange in diseased lungs.

View larger version (33K): [in this window] [in a new window]   Figure 1.   An early oxygen-carbon dioxide diagram showing the respiratory exchange ratio (R) lines radiating from the inspired gas point for various altitudes, arterial oxygen saturation (HbO2 %) (top horizontal axis), alveolar ventilation assuming a CO2 production of 315 ml min1 (right vertical axis), and a line showing the average alveolar gas composition following acute exposure to high altitude. (Reprinted from Reference 2 by permission.)

Another very productive group at that time was led by Richard Riley. While he was at the U.S. Naval School of Aviation Medicine in Pensacola, Florida, he worked with Joseph Lilienthal on methods of measuring carbon monoxide levels in blood, because carbon monoxide poisoning was believed to be responsible for a number of aircraft accidents. Riley developed a bubble technique for measuring the partial pressures of oxygen and carbon dioxide in blood, and he went on to construct a scheme for understanding ventilation-perfusion inequality in diseased lungs, more or less in parallel with the Rochester group. An interesting historical vignette is that in 1948, exactly 50 years ago, Riley developed pulmonary tuberculosis (a common occurrence in pulmonary physicians at the time) but was allowed to rest at home. There he spent most of his time thinking about ventilation-perfusion relationships, arguing with Rahn by mail, and playing with his four-quadrant diagram. He afterwards wrote, "Never was enforced confinement given more profitably psychotherapy" (3).

After a period at Bellevue hospital in New York with Andre Cournand, Riley moved to Johns Hopkins in Baltimore in 1950. Hermann Rahn moved from Rochester to Buffalo in 1956, and the combined groups from Rochester, Buffalo, Bellevue, and Hopkins spawned a large number of postdoctoral fellows whose influence continues to the present day. It is interesting that the initial work of these groups was not supported by the Public Health Service. The Rochester/Buffalo team relied heavily on the Department of Defense, while the New York/Baltimore group had support from several sources, including the Life Insurance Medical Fund and the Commonwealth Fund. Later work, however, was supported by the NHLBI.

	VENTILATION-PERFUSION INEQUALITY

The pioneering advances by the groups of Fenn, Rahn, and Otis on the one hand, and Riley, Cournand, and their colleagues on the other, greatly improved our understanding of the abnormalities of pulmonary gas exchange in lung disease. In particular, the Riley analysis allowed the diseased lung to be divided into three imaginary compartments. One was "ideal" in the sense that gas exchange was optimal, another had unperfused but ventilated alveoli (dead space), and the third had unventilated but perfused alveoli (shunt). Many valuable insights into disordered lung function resulted from this analysis.

However, it was recognized from the beginning that real lungs contain some sort of distribution of ventilation-perfusion ratios, and that whereas two or three compartment models clarified some aspects of impaired gas exchange, these models must be very remote from reality. But the apparently insuperable difficulty of dealing with distributions of ventilation-perfusion ratios, involving the nonlinear, interdependent oxygen and carbon dioxide dissociation curves, stifled progress in this direction for a long time.

The breakthrough came with the application of computer numerical analysis for analyzing the behavior of distributions. Kelman (4) and Olszowka and Farhi (7) introduced numerical procedures for describing the oxygen and carbon dioxide dissociation curves and their interactions, and these were then used to describe the gas exchange behavior as ventilation-perfusion inequality was imposed on computer models of the lung (8).

At about this time, Lenfant and Okubo (9) made a large leap forward in thinking about ventilation-perfusion ratio (A/) distribution by developing a method that allowed a continuous A/ distribution to be determined experimentally. This freed the pulmonary community from the confines of an unrealistic three-compartment approach to what was recognized to be a distributed function over a broad A/ ratio range from zero to infinity with potentially all values in between. What Lenfant and Okubo did was to record arterial P O₂ responses to an N₂ washout as the inspired O₂ concentration was changed from ambient to 100%. The profile of arterial PO₂ change depended on the distribution, which could then be inferred as a continuous relationship by computer- assisted (and computer-dependent), Laplace transformation procedures. Unfortunately, the method never caught on even as a research tool, possibly because at the time it was complex in its methodology and mathematical basis. Computers were just becoming available for biological applications, and most physicians were not enamored of quantitative techniques, a trait that seems to persist today. But there were also some physiological questions about the method. Briscoe and colleagues (10) had shown that as FI_O2 was increased from ambient, poorly ventilated alveoli could collapse and thus become transformed into shunts. The phenomenon of hypoxic pulmonary vasoconstriction was also known (11), and thus even if poorly ventilated areas (with low alveolar PO₂ breathing air) did not collapse while breathing 100% O₂, their blood flow would likely increase as hypoxia was relieved by the high inspired O₂ concentration. Thus, the method of Lenfant and Okubo could lead to changes in the A/ distribution in the course of its application.

These drawbacks and the intrinsic simplicity of nonrespiratory (inert) gases that obeyed Henry's law (concentration linearly related to partial pressure, unlike the relationships for O ₂ and CO₂) caused some early workers to examine the potential for using inert gases as a means of studying the A/ distribution. After the monumental essay of Seymour Kety, published in 1951 in Pharmacological Reviews (12), which addressed inert gas exchange in the lungs and tissues, the equations governing their exchange were well known. While these equations are in fact based on precisely the same principles and assumptions as those developed for O₂ and CO₂ by Rahn and Fenn (13) and by Riley and Cournand (14, 15), mentioned previously, their form was much simpler because of adherence to Henry's law. But what was far more important, the lungs could be made to exchange any number of inert gases at the same time, and in very low concentrations that enabled their use as tracer molecules that did not disturb the respiratory gas exchange function of the lungs. They could, of course, be applied at any desired FI_O2, and this in fact might allow the proposed weaknesses of the Lenfant approach to be directly assessed, by comparing results breathing air and breathing pure O₂.

	MULTIPLE INERT GAS ELIMINATION TECHNIQUE (MIDGET)

Farhi and Yokoyama (16) were the first to use a mixture of inert gases to assess A/ distribution in the lung. However, their conceptual approach paralleled that of Rahn and Fenn and of Riley and Cournandusing a virtual model of two A/ compartments upon which to base interpretation of their inert gas exchange data. They did not attempt to come up with a continuous A/ distribution. The work of Farhi and Yokoyama, however, explicitly reinforced a fundamental concept that had earlier been noted by Kety: that the solubility of the inert gases used in their method was critical. For example, if one wished to know how much shunt was present, and whether there were also areas of low A/ that were ventilated, even if poorly, one had to make the lungs exchange a poorly soluble gas. Symmetrically, to distinguish dead space (unperfused lung) from areas of high A/ ratio that were perfused, albeit poorly, required a gas of very high solubility. Thus, it became evident that if one wished to assess A/ mismatch using inert gases, the choice of which gases to use was key. This is an important point to understand. Kety (12) showed that if an inert gas of partition coefficient  was being eliminated by the lungs in a steady state, the fractional retention (R) of the gas in the arterial blood (i.e., the fraction not eliminated) is a simple function of  and of the A/ ratio of the (homogeneous) lung:

(1)

Table 1 shows the values of R in the presence of a shunt (A/ = 0) and in the presence of a low A/ ratio unit (A/ = 0.01) for gases of low, medium, and high partition coefficient, . The corresponding calculations for a high A/ unit and the dead space (infinite A/) are also shown. For the low solubility gas (  = 0.005), the retentions are very different in the low A/ unit (R = 0.33) compared to the unventilated unit (R = 1.0). Thus, measured values of R for such a gas will provide the basis for separating the two kinds of lung units. However, even the medium solubility gas shows what in the laboratory would be immeasurably small retention differences between the two units (R = 0.99 and 1.00, respectively), and the high-solubility gas shows no differences in R to two decimal places (in the laboratory, the coefficient of variation in measurements of retention is typically about 0.02-0.03). The opposite holds for separating dead space from areas of high A/. As shown in the table, retention values computed from Equation 1 show that only the high solubility gas (  = 500) gives measurable differences between the two units. It should therefore be evident that the ability to resolve among lung units of neighboring A/ ratios depends on the choice of inert gas and its solubility. To resolve the whole physiologic range of interest requires the use of many inert gases whose solubilities span the range of interest of the A/ domain.

What was realized in San Diego in the early 1970s (17) was that the relationship between retention, solubility, and A/ ratio was a very smooth function over the whole range: calculus can be used to show that R as a function of  (from Equation 1) is monotonic with no zero derivatives, or, in other words, very smooth. While the ideal (if unrealistic) experimental approach would have been to expose the lung to an infinite number of inert gases of all solubilities, the very smooth nature of the retention-solubility relationship implied by Equation 1 meant that with a few gases of well-chosen solubility, the entire retention-solubility curve could be estimated with some accuracy. After much development, a mixture of six gases was selected as affording the optimal balance between sufficient data and experimental feasibility. The six partition coefficients were chosen to cover a very wide rangefrom 0.005 (sulfurhexafluoride) to 300 (acetone). That is a range of 60,000, and allows a corresponding five-decade range of A/ from about 0.005 to 100 to be resolved.

The basic data set required for the multiple inert gas elimination technique (MIGET) was, therefore, the retention-solubility curve (as estimated from the retention values of each of six gases). This corresponds to the basic data set of Lenfant's methodthe curve relating arterial P O₂ to that of inspired gas. Consequently, similar mathematical methods could be proposed for converting the input data set to the desired frequency function showing the quantitative allocation of ventilation and blood flow to the range of A/ ratios potentially present in the lung from zero to infinity. What was initially used was a simple, numerically intensive, inerative process based on a least-squares minimization principle: after some initial starting guess at the A/ distribution, compartmental blood flows were changed minutely to achieve successively better approximations between the input retention data and the retentions calculated at each iteration from the blood flow distribution (according to Equation 1). Although this method was slow, was in theory if not in practice dependent on the initial guess, and suffered from the unsolvable problem of specifying objective criteria for acceptance of a result and ending the successive approximation process, model testing of the algorithm proved its worth. A whole host of different A/ distributions were used to generate theoretical retention data, and the algorithm was able to recover these, given A/ curves with encouraging fidelity (17). This model test was the basis of the very first (and NIH-supported) publication of the inert gas elimination technique in 1974. Grants HL-05931 and HL-13687 allowed this work to be done, with additional support from NASA.

The next 5 years were characterized by turmoil. No one believed that an essentially continuous A/ curve could be obtained from a set of just six data points. This healthy skepticism achieved a number of objectives, although it consumed its fair share of the NIH budget in doing so. A number of workers pointed out limitations of the method (18, 19), and others proposed alternative analysis for the data (20). We were pushed into devising a more robust numerical approach (21) and extensively used the engineering technique of linear programming to define the information content of the MIGET (22, 23). Limits placed on the method due to the finite number of tracer gases and additional limits from the fact that measurements inevitably had some experimental error were described, and a number of approaches were in the end worked out to clear the air. It is fair to say that the depth of statistical treatment that ended up being applied to the problem remains quite unprecedented in respiratory biology, and is rather amusing to the authors in this era of molecular biology that often seems not to stoop to even basic statistical analysis of the smudgy black data.

Once the dust had settled, users of the method examined a wide variety of circumstances in health and disease. The decade of the 1980s saw a large number of descriptions of the A/ abnormalities in adults and children, in awake and anesthetized states, at rest and during exercise, at sea level and at altitude, breathing air and O ₂, in man and in animals, and in most chronic cardiopulmonary disease states. To more than perfunctorily recite the findings of all is completely beyond the scope of this chapter. A common thread, however, has been that despite the mathematical limitations on the method (which amount to issues of mostly fine discrimination of the A/ pattern), its application has given us considerable insight into the workings of the normal and abnormal lung. For example, it remains the only approach that can resolve not only the A/ domain, but also indicate the extent of alveolar-capillary O ₂ diffusion limitation when present. This has shown that normal subjects at altitude are limited far more by such diffusion limitation than by A/ mismatch, although the latter can be very abnormal also (24). At the altitude equivalent of the Mount Everest summit, even at the very low exercise intensity sustainable, diffusion limitation causes arterial O₂ content to be some 4 ml/dl lower that would otherwise occur. Patients with emphysema do not show diffusion limitation on exercise, but those with interstitial fibrosis do (25). These and other diseases have shown characteristic patterns of A/ inequality that provide insight into the underlying pathophys iologic processes.

MIGET has also allowed the patient by patient partitioning of hypoxemia into causes due to intrapulmonary events disturbing A/ relationships versus those due to extrapulmonary factors, such as cardiac output, F I_O2 and O₂ consumption. For example, in most patients with asthma, arterial PO₂ is reduced little, yet there is often marked A/ inequality present (26). The usually elevated cardiac output affords the explanation: despite the inequality, the high cardiac output maintains a high mixed venous P O₂, and this is how the arterial PO₂ stays nearly normal. Quite the opposite occurs in acute heart failure following myocardial infarction: hypoxemia is often severe, yet the degree of A/ mismatch may be no greater than in asthma (27). The greater fall in arterial P O₂ is explained by the very low cardiac output that causes mixed venous PO₂ to fall precipitously below normal values. Another broad conclusion is that in the obstructive diseases (asthma and chronic obstructive pulmonary disease) there is a very poor relationship, if any, between the severity of A/ mismatch and the amount of airways obstruction, even in the same patient over time (28). This has implications for clinical care both in terms of what to monitor and what constitutes remission or improvement. Yet another overarching observation has been that in chronic lung diseases, A/ mismatch without significant shunt is the rule. On the other hand, in acute lung disease, the converse seems to be seen: shunts are prominent and A/ mismatch less in evidence. Both acute and chronic diseases may have obstruction of airways from many causes, but presumably the location of such abnormalities is important. Thus, in acute disease with very peripheral obstruction, including alveolar filling with debris or fluid, gas cannot reach the capillary surface, whereas in more proximal obstruction, collateral ventilation prevents the development of shunts. The interested reader will find documentation of these and many other findings made using MIGET in both the primary peer-reviewed literature cited previously and in book chapters and reviews (29).

Much of the clinical application of the MIGET has been accomplished in Barcelona, Spain. This reveals an interesting and possibly unique way in which the NIH has been involved in scientific development beyond the United States. In 1980, Dr. Roberto Rodriguez-Roisin came as a postdoctoral fellow to work with the authors using the MIGET. On his return to Spain he took the technique with him and began clinical research. In 1984 he noticed that the NIH was offering a unique opportunity for research development in Spain via a special grant mechanism that brought researchers from the United States and Spain together. This allowed the U.S. members to assist their Spanish colleagues in developing a research program. Together we applied for these funds (CCA 8309185), and the rest is, as they say, history. In difficult circumstances, the Spanish group built a strong clinical research program, initially based mostly on the MIGET. Their senior members have since become leaders in the European respiratory community, and now their program, of course expanded well beyond gas exchange, is attracting fellows, especially from the Spanish-speaking world of South America.

Finally, it is interesting in the context of this semi-historical review, to look at two extreme environments where pulmonary gas exchange has recently been studied for the first time. The first is the highest point on Earth, and the second is a hundred miles or so above this in space.

	GAS EXCHANGE AT EXTREME ALTITUDE

Since the beginning of the century, there has been an increasing interest in the extraordinary challenges posed to the human respiratory system by extreme altitudes. Following the Pike's Peak expedition of 1911, Douglas and colleagues (35) argued that the only way that humans could survive at extreme altitude was by having the lungs secrete oxygen. In 1924 Norton climbed within 300 m of the Everest summit without added oxygen, but the first "oxygenless" ascent was not until 54 years later in 1978 when Messner and Habeler astounded many physiologists by their feat. During the Silver Hut expedition of 1960-1961, Pugh (36) showed how accurate, though relatively simple, measurements of human physiology could be made in the field at extreme altitudes. The possibility, therefore, arose of making the first physiologic measurements on the summit of Mount Everest, and this ambitious aim was realized by the American Medical Research Expedition to Everest (AMREE) in 1981.

The results from the Silver Hut expedition and a subsequent theoretical analysis (37) had shown that the critical factors for humans to reach the summit without supplementary oxygen were the barometric pressure, extent of the hyperventilation, and the maximal oxygen uptake. All three measurements were successfully obtained during AMREE (38). The first direct measurement of barometric pressure on the summit gave a value of 253 mm Hg which, as had been predicted by Pugh and others, was far higher than given by the standard atmosphere. This model had previously been inappropriately used by a number of high-altitude physiologists, and the fact that the actual pressure exceeded the model pressure by some 17 mm Hg went a long way toward explaining how an "oxygenless" ascent was possible. The degree of hyperventilation was also much greater than expected. Christopher Pizzo, M.D., who collected his alveolar gas on the summit, had an alveolar P CO₂ between 7 and 8 mm Hg, and the mean value of 27 measurements on four subjects at an altitude of 8,050 m was 11.0 mm Hg. These extremely low values emphasize the extraordinary degree of hyperventilation necessary to reach these extreme altitudes. Finally, the maximal oxygen consumption was measured, not on the summit itself but at 6,300 m on extremely well acclimatized subjects breathing the same inspired PO₂ as on the summit. The result was about 1 L/min, a miserably low value, but just enough to explain the "oxygenless" ascent of Messner and Habeler. Four years later, a simulated ascent of Mount Everest in a low-pressure chamber, Operation Everest II, added a great deal of information, especially on the pulmonary circulation, pulmonary gas exchange, and changes in skeletal muscle obtained by biopsy (39).

The role of the NHLBI in the support of AMREE has some entertaining historical overtones. An early feasibility study was made possible by a small contract, which owed much to the support of Dr. Claude Lenfant, then director of the Lung Division. However, when the main proposal was reviewed by the Applied Physiology and Orthopedics (sic) Study Section in February 1979, three of the members voted disapproval mainly on the grounds of the hazards of the expedition and the low probability that measurements would be obtained on the summit. Since any disapproval vote has a priority score of 5, the normalized score of the Study Section was only 248. Of course, in today's climate such a score would be a death knell. However, back in 1979 the score was sufficient to allow funding to squeak through. Those were the good old days! It is a fact if the proposal had not been funded at that time, the expedition would certainly have been abandoned, because it was too late to resubmit for the time slot on Mount Everest assigned to us.

	GAS EXCHANGE DURING SPACE FLIGHT

The other unusual environment where extensive measurements of pulmonary gas exchange have now been carried out is space. Following some preliminary studies using parabolic profile flights giving only 20-25 s of weightlessness, extensive studies have now been carried out on the 9-14-d flights of Spacelabs SLS-1, SLS-2, and LMS, and more measurements are scheduled for the Neurolab Spacelab in April 1998. At the outset, it was argued that because the lung is exquisitely sensitive to gravity, which normally causes regional differences of blood flow, ventilation, gas exchange, alveolar size, intrapleural pressure, and parenchymal stress, substantial changes of pulmonary function would be expected in the microgravity environment, and indeed these have been found. In addition, recent experience on the Russian space station, Mir, shows that the respiratory system is one of the most vulnerable of the body systems in the space environment. The occurrence of a fire that filled the spacecraft with smoke, failure in the oxygen-generating systems, and improper function of the carbon dioxide scrubbers has meant that the lungs of the crew members have been exposed to a potpourri of hazards.

The major changes in pulmonary functions in sustained microgravity can be summarized as follows (40). All comparisons are with the standing posture in normal gravity. The area of gas exchange surface of the lung is greatly increased, as measured from the diffusing capacity for carbon monoxide of the alveolar membrane (D_M). Interestingly, the increase is much more than would be expected from the simultaneous increase in volume of the pulmonary capillary blood. Cardiac stroke volume is increased despite, paradoxically, a reduction in central venous pressure and circulating blood volume. Pulmonary blood flow and ventilation become more uniform, as expected, but substantial inequality remains, demonstrating the importance of nongravitational factors. Total ventilation is decreased but, because alveolar deadspace is smaller, alveolar ventilation is unchanged.

A surprising finding was that the amount of ventilatory inequality during normal breathing was unchanged by microgravity, showing that this is not gravitational in origin. Even more surprising was that the closing volume of the lung (measured with an argon bolus in a single-breath nitrogen washout) was unchanged on average, indicating that gravitation compression of lower zone airways was not responsible. Residual volume was decreased, possibly because uniform alveolar expansion allowed the apical alveoli to reach a smaller size during a full expiration. Functional residual capacity was intermediate between the volumes measured standing and supine in normal gravity.

Perhaps most surprising of all was a change in behavior of two inhaled gases of very different molecular weights, helium (He) and sulfurhexafluoride (SF₆). These have very different diffusion rates, but since diffusion is mass, not gravity-dependent, their behavior was not expected to change. On the contrary, whereas in normal gravity the slope of the single-breath plateau for SF₆ always exceeds that for He (because of the latter's higher diffusion rate), in microgravity the plateau for SF₆ had the same slope as for He, and after a period of 10 s of breathholding, the SF₆ slope was actually less than that for He (Figure 2) (41). These very provocative results have so far defied explanation.

View larger version (15K): [in this window] [in a new window]   Figure 2.   Slopes of alveolar plateaus from single-breath tests for helium (He) (closed circles) and sulfur hexafluoride (SF6) (open squares) for subjects standing in normal gravity (1G) and in microgravity (µG). Note that with no breathholding, the SF6 slope was greater than that for He in 1G, but the difference disappeared in µG. When the test was performed in µG with 10 s of breathholding, the SF6 slope was significantly less than that for He (p < 0.05). The mechanism of these changes is a total mystery at the present time. (Reprinted from Reference 41 by permission.)

Most of this work has been supported by NASA rather than the NIH. However, the upcoming Neurolab flight is jointly supported by both of these two giant agencies, and some amusing areas of friction have emerged. When the NIH Study Section saw our travel budget for Neurolab, terms such as "preposterous" and "totally unacceptable" were tossed around. The fact is that a flight project, which involves many visits to the Johnson Space Center in Houston and extensive periods of crew training, is totally unlike an NIH study. For example, our present annual budget for travel for Neurolab exceeds $30,000. Try putting that into an RO1 proposal!

We cannot resist the temptation of adding a final postscript on the topic of pulmonary gas exchange. The present movement away from integrative physiology means that a whole generation of young pulmonologists are poorly informed on this topic. This is regrettable because many important problems remain unanswered, and a sound understanding of gas exchange is essential in many clinical environments. To cite just one example, many of the factors influencing pulmonary gas exchange in the intensive care environment are poorly understood, and it is depressing to see young faculty members in such units in university hospitals twiddling the knobs of the ventilators with so little understanding of the basic physiology. Presumably the pendulum will swing back eventually, and indeed there are signs that this process may have begun.

	Footnotes

Correspondence and requests for reprints should be addressed to John B. West, M.D., Ph.D., UCSD Department of Medicine 0623A, 9500 Gilman Drive, La Jolla, CA 92093-0623. E-mail: jwest{at}ucsd.edu

	References

TOP INTRODUCTION REFERENCES

1. Otis, A. B., and H. Rahn. 1980. Development of concepts in Rochester, New York, in the 1940s. In J. B. West, editor. Pulmonary Gas Exchange, Vol. 1. Ventilation, Blood Flow, and Diffusion. Academic Press, New York. 33-66.

2. Fenn, W. O., H. Rahn, and A. B. Otis. 1946. A theoretical study of the composition of the alveolar air at altitude. Am. J. Physiol. 146: 637-653 .

3. Riley, R. L. 1980. Development of the three-compartment model for dealing with uneven distribution. In J. B. West, editor. Pulmonary Gas Exchange, Vol. 1. Ventilation, Blood Flow, and Diffusion. Academic Press, New York. 67-85.

4. Kelman, G. R.. 1966. Calculation of certain indices of cardio-pulmonary function, using a digital computer. Respir. Physiol. 1: 335-343 [Medline].

5. Kelman, G. R.. 1966. Digital computer subroutine for the conversion of oxygen tension into saturation. J. Appl. Physiol. 21: 1375-1376 [Free Full Text].

6. Kelman, G. R.. 1967. Digital computer procedure for the conversion of PCO₂ into blood CO₂ content. Respir. Physiol. 3: 111-115 [Medline].

7. Olszowka, A. J., and L. E. Farhi. 1968. A system of digital computer subroutines for blood gas calculations. Respir. Physiol. 4: 270-280 [Medline].

8. West, J. B.. 1969. Ventilation-perfusion inequality and overall gas exchange in computer models of the lung. Respir. Physiol. 7: 88-110 [Medline].

9. Lenfant, C., and T. Okubo. 1968. Distribution function of pulmonary blood flow and ventilation/perfusion ratio in man. J. Appl. Physiol. 24: 668-677 [Free Full Text].

10. Briscoe, W. A., E. M. Cree, S. Filler, H. E. Houssay, and A. Cournand. 1960. Lung volume, alveolar ventilation and perfusion inter-relationships in chronic pulmonary emphysema. J. Appl. Physiol. 15: 785-795 [Abstract/Free Full Text].

11. von Euler, U. S., and G. Liljestrand. 1947. Observations on the pulmonary arterial blood pressure in the cat. Acta Physiol. Scand. 12: 301-320 .

12. Kety, S. S.. 1951. The theory and applications of the exchange of inert gas at the lungs and tissues. Pharmacol. Rev. 3: 1-41 [Free Full Text].

13. Rahn, H., and W. O. Fenn. 1955. A Graphical Analysis of the Respiratory Gas Exchange. American Physiological Society, Baltimore.

14. Riley, R. L., and A. Cournand. 1949. `Ideal' alveolar air and the analysis of ventilation-perfusion relationships in the lungs. J. Appl. Physiol. 1: 825-847 [Free Full Text].

15. Riley, R. L., and A. Cournand. 1951. Analysis of factors affecting partial pressures of oxygen and carbon dioxide in gas and blood of lungs: theory. J. Appl. Physiol. 4: 77-101 [Free Full Text].

16. Farhi, L. E., and T. Yokoyama. 1967. Effects of ventilation/perfusion inequality on elimination of inert gases. Respir. Physiol. 3: 12-20 [Medline].

17. Wagner, P. D., H. A. Saltzman, and J. B. West. 1974. Measurement of continuous distributions of ventilation-perfusion ratios: theory. J. Appl. Physiol. 36: 533-537 [Free Full Text].

18. Jaliwala, S. A., R. E. Mates, and E. J. Klocke. 1975. An efficient optimization technique for recovering ventilation/perfusion distributions from inert gas data. J. Clin. Invest. 55: 188-192 .

19. Olszowka, A. J.. 1975. Can A/ distributions in the lung be recovered from inert gas retention data? Respir. Physiol. 25: 191-198 [Medline].

20. Hlastala, M. P., and H. T. Robertson. 1978. Inert gas elimination characteristics of the normal and abnormal lung. J. Appl. Physiol. 44: 258-266 [Free Full Text].

21. Evans, J. W., and P. D. Wagner. 1977. Limits on A/ distribution from analysis of experimental gas elimination. J. Appl. Physiol. 42: 889-898 [Abstract/Free Full Text].

22. Ratner, E. R., and P. D. Wagner. 1982. Resolution of the multiple inert gas method for estimating A/ maldistribution. Respir. Physiol. 49: 293-313 [Medline].

23. Kapitan, K. S., and P. D. Wagner. 1987. Linear programming analysis of A/ distributions: average distribution. J. Appl. Physiol. 62: 1356-1362 [Abstract/Free Full Text].

24. Wagner, P. D., J. R. Sutton, J. T. Reeves, A. Cymerman, B. M. Groves, and M. K. Malconian. 1987. Operation Everest II: pulmonary gas exchange during a simulated ascent of Mt. Everest. J. Appl. Physiol. 63: 2348-2359 [Abstract/Free Full Text].

25. Wagner, P. D. 1977. Ventilation-perfusion inequality and gas exchange during exercise in lung disease. In J. A. Dempsey and C. E. Reed, editors. Muscular Exercise and the Lung. University of Wisconsin Press, Madison, WI. 345-356.

26. Wagner, P. D., G. Hedenstierna, and G. Bylin. 1987. Ventilation-perfusion inequality in chronic asthma. Am. Rev. Respir. Dis. 136: 605-612 [Medline].

27. Bencowitz, H. Z., M. M. LeWinter, and P. D. Wagner. 1984. Effect of sodium nitroprusside on ventilation-perfusion mismatching in heart failure. J. Am. Coll. Cardiol. 4: 918-922 [Abstract].

28. Wagner, P. D., G. Hedenstierna, and R. Rodriguez-Roisin. 1996. Gas exchange, expiratory flow obstruction and the clinical spectrum of asthma. Eur. Respir. J. 9: 1278-1282 [Abstract].

29. Wagner, P. D. 1978. Measurement of the distribution of ventilation/perfusion ratios. In D. G. Davies and C. D. Barnes, editors. Regulation of Ventilation and Gas Exchange. Academic Press, New York. 217-260.

30. Wagner, P. D., and J. B. West. 1980. Ventilation-perfusion relationships. In J. B. West, editor. Ventilation, Blood Flow and Diffusion. Academic Press, New York. 219-262.

31. West, J. B., and P. D. Wagner. 1991. Ventilation-perfusion relationships. In R. G. Crystal, J. B. West, P. J. Barnes, N. S. Cherniack, and E. R. Weibel, editors. The Lung: Scientific Foundations. Raven Press, Ltd., New York. 1289-1305.

32. Wagner, P. D. 1991. Effects of COPD on gas exchange. In N. S. Cherniack, editor. Chronic Obstructive Pulmonary Disease. Harcourt Brace Jovanovich, Inc., Philadelphia. 73-79.

33. Wagner, P. D., and G. E. Gale. 1991. Ventilation-perfusion relationships. In B. J. Whipp and K. Wasserman, editors. Exercise: Pulmonary Physiology and Pathophysiology. Marcel Dekker, New York. 121- 142.

34. Wagner, P. D. 1997. Gas exchange. In A. R. Leff, editor. Asthma. Raven Press, New York. 1335-1347.

35. Douglas, C. G., J. S. Haldane, Y. Henderson, and E. C. Schneider. 1913. Physiological observations made on Pikes Peak, Colorado, with special reference to adaptation to low barometric pressures. Phil. Trans. R. Soc. Lond. Ser. B. 203: 185-381 .

36. Pugh, L. G.. 1962. Physiological and medical aspects of the Himalayan Scientific and Mountaineering Expedition, 1960-61. B.M.J. 2: 621-633 .

37. West, J. B., and P. Wagner. 1980. Predicted gas exchange on the summit of Mt. Everest. Respir. Physiol. 42: 1-16 [Medline].

38. West, J. B.. 1984. Human physiology at extreme altitudes on Mount Everest. Science 223: 784-788 [Abstract/Free Full Text].

39. Houston, C. S., J. R. Sutton, A. Cymerman, and J. T. Reeves. 1987. Operation Everest II: man at extreme altitude. J. Appl. Physiol. 63: 877-882 [Abstract/Free Full Text].

40. West, J. B., A. R. Elliott, H. J. Guy, and G. K. Prisk. 1997. Pulmonary function in space. J.A.M.A 277: 1957-1961 [Abstract/Free Full Text].

41. Prisk, G. K., H. J. Guy, A. R. Elliott, and J. B. West. 1996. Anomalous behavior of helium and sulfur hexafluoride in single breath washins during sustained microgravity. J. Appl. Physiol. 80: 1126-1132 [Abstract/Free Full Text].