Capillary Transit Time

As red blood cells pass through the capillary network during basal conditions, oxygen and hemoglobin combine with sufficient rapidity to completely saturate hemoglobin in about the first quarter of its transit. The results of a number of studies of the saturation of hemoglobin with oxygen, each using differing assumptions, have been reviewed by Wagner

[1]. Even with the wide range of variables used by the investigators, 0.25 seconds consistently appeared as a reason-

able bound for the minimum time required for red blood cell Po2 to increase from 40 to 100torr. With exercise, capillary transit times decrease as cardiac output increases, and more of the capillary length is utilized for oxygen uptake. Only if capillary transit time becomes too rapid, that is, less than 0.25 seconds, will red blood cells leave the capillaries without becoming fully saturated.

The predominant method for measuring pulmonary capillary transit time is based on the diffusing capacity of the lung for carbon monoxide to compute pulmonary capillary volume. Capillary transit time can then be calculated by dividing capillary blood volume by cardiac output measured concomitantly. For example, 75 mL (capillary volume) n 75mL/second (a 4.5L/minute resting cardiac output) = 1.00 second (capillary transit time). Based on this kind of calculation, it is believed that, in the resting human, red blood cells spend an average of 0.75 to 1.0 seconds in the pulmonary capillaries [1], well above the calculated 0.25-second minimum transit time. During maximal exercise, however, hypoxemia has been shown to be present in horses, in exercising humans at simulated high altitude, and even in normal humans exercising maximally at sea level. Too rapid a transit time for the fastest moving red cells is implicated in these findings. Naturally, not every red cell traverses the pulmonary capillaries at the mean time. Rather, there is a range of transit times distributed about the mean with some times significantly shorter than the mean, but the shape of that important distribution has escaped measurement with nearly all methods. During the past two decades, there has been sufficient development of intensified television cameras and computer video processing systems to permit pulmonary capillary transit times to be measured directly. Video fluorescence microscopic recordings of the passage of a fluorescent dye bolus through the subpleural pulmonary microcirculation have been used to obtain dye dilution curves from an arteriole and venule that shared a common capillary network. Initially, the experiments were restricted to calculation of the mean capillary transit time. Using this technique, rabbits have a more rapid mean transit time than dogs at basal cardiac outputs, (rabbits, 0.41 to 0.60 seconds and dogs, 1.37 to 2.0 seconds). These experiments also showed a substantial vertical gradient of transit times in the dog lung, 10 to 20 seconds in the upper lung and 1.5 to 2.0 seconds in the lower lung. These video microscopic measurements have been confirmed by Hogg and colleagues using independent techniques.

Although the measurements of mean capillary transit time have been important, the accurate measurement of the entire distribution of transit times, including the minimum and maximum, has eluded investigators. Recently, however, Presson et al. have made these measurements using a fluorescent plasma marker. As blood flow was increased in this study, both the skew and kurtosis of the curves changed so that the longest times shortened, and the curves approached the ideal shape, that is, a spike at 0.25 seconds such that all red blood cells would become saturated exactly at the downstream end of the capillaries. The more homogeneous transit time distribution (a higher, sharper peak with a narrower base) along with some perfusion of previously unperfused capillaries that increased capillary blood volume, both acted to prevent even the most rapidly moving plasma from crossing in less than 0.25 seconds.

In the latest report from Presson and colleagues [2], the transit time distribution of fluorescently labeled red blood cells was reported. In intact dogs, they found that fluorescently labeled plasma boluses required about half again as long to traverse the capillaries as red blood cells (Figure 1), probably because of a combination of factors including the plasma-skimming Fahraeus effect in perfused capillaries, and the possibility of plasma alone perfusing some capillaries. This means that data obtained from the fluorescently labeled plasma dilution curves overestimate red cell transit time. Nevertheless, at the highest cardiac outputs, capillary recruitment and increasing red blood cell transit time homogeneity combined, as they did for plasma capillary transit time, to prevent even the most rapidly transiting red blood cells from going below 0.25 seconds in the dog lung at a cardiac output increased fourfold over basal.

Distention

Distention of the capillaries, which increases the number of red blood cells in the gas exchanging vessels, has been studied extensively by Fung and Sobin [3] and colleagues. Much of their work has been based on careful measurements of capillary diameters filled under various pressure conditions by a polymer that, after having flowed into all the vessels, formed a cast of the capillary bed. The appearance of the extraordinarily dense capillary network embedded

Figure 1 In intact dogs, fluorescently labeled plasma boluses require about half again as long to traverse the capillaries as red blood cells. This causes the plasma dilution curves to overestimate red blood cell capillary transit times. The mean transit times (plasma = 1.4 seconds, red blood cell = 1.0 seconds), the minimum transit times (plasma = 0.6 seconds, red blood cell = 0.4 seconds), and the maximum transit times (plasma = 3.5 seconds, red blood cell = 2.5 seconds) were all significantly longer for plasma. At higher cardiac outputs, the curves become more narrow with higher peaks; the minimum transit time is < 0.25 seconds, even at very high cardiac outputs. (Reproduced with permission from Ref. [2].)

Figure 1 In intact dogs, fluorescently labeled plasma boluses require about half again as long to traverse the capillaries as red blood cells. This causes the plasma dilution curves to overestimate red blood cell capillary transit times. The mean transit times (plasma = 1.4 seconds, red blood cell = 1.0 seconds), the minimum transit times (plasma = 0.6 seconds, red blood cell = 0.4 seconds), and the maximum transit times (plasma = 3.5 seconds, red blood cell = 2.5 seconds) were all significantly longer for plasma. At higher cardiac outputs, the curves become more narrow with higher peaks; the minimum transit time is < 0.25 seconds, even at very high cardiac outputs. (Reproduced with permission from Ref. [2].)

in thin alveolar walls suggested that perfusion of the pulmonary capillaries functioned more like a sheet of flowing blood rather than flow through a network of tubes. Measurements of the polymer casts produced the relationship shown in Figure 2. In this preparation, once transmural capillary pressures exceeded alveolar pressure, the capillary bed opened suddenly to form a sheet about 4 mm thick. Additional pressure increases produced a linear increase in thickness. The sheet flow model permitted mathematical modeling of capillary hemodynamics, a significant advantage. The complexity of other models, composed of a seemingly infinite number of tubes nearly as wide as they were long, made it virtually impossible to manage the mathematical analysis of flow.

Other workers have also found that the capillaries distended as pressure increased. In their classic paper, Glazier et al. [4] studied isolated perfused greyhound lungs that were rapidly frozen. The lungs were frozen under three vascular pressure conditions: (1) where alveolar pressure was greater than pulmonary arterial pressure, which in turn, was greater than pulmonary venous pressure, and there was little or no flow (PA > Ppa > Ppv = zone 1); (2) where arterial pressure exceeded alveolar pressure, but alveolar pressure was greater than venous pressure (Ppa > PA > Ppv = zone 2); and (3) where arterial pressure exceeded venous pressure, and venous pressure was greater than alveolar pressure (Ppa > Ppv > PA = zone 3). In lungs frozen under zone 2 and zone 3 conditions, the capillaries distended linearly as vascular transmural pressure increased.

Figure 2 (Upper panel) The distensibility curve of the pulmonary capillaries in the cat lung. Transmural capillary pressure is shown on the x-axis and sheet thickness on the y-axis. The sheet remains empty until capillary pressure is ~ 0, then the sheet opens suddenly to a thickness of ~4 mm and distends linearly thereafter. (Lowerpanel) Theoretical capillary shapes are shown as transmural pressure increases. (Reproduced with permission from Ref. [3].)

Figure 2 (Upper panel) The distensibility curve of the pulmonary capillaries in the cat lung. Transmural capillary pressure is shown on the x-axis and sheet thickness on the y-axis. The sheet remains empty until capillary pressure is ~ 0, then the sheet opens suddenly to a thickness of ~4 mm and distends linearly thereafter. (Lowerpanel) Theoretical capillary shapes are shown as transmural pressure increases. (Reproduced with permission from Ref. [3].)

As a final note about distending capillaries, the situation becomes more complex when hematocrit is considered. In individual capillaries, hematocrit varies from moment to moment, as can readily be demonstrated by in vivo microscopy. As the capillaries distend sufficiently to admit red blood cells edge on, rouleau formations become possible. In that extreme, capillary hematocrit can reach very high values. One the other hand, large plasma gaps between red cells can reduce individual capillary segmental hematocrit to 0 percent momentarily. Other combinations of red cells and plasma cause the entire gamut of hematocrits to be run in a matter of seconds. These variations have been quantitated; their implications for gas exchange have recently been explored and suggest that hematocrit, red blood cell orientation, and the cross-sectional shape of the capillary may have an important effect on gas exchange.

Recruitment

At baseline, not all capillaries are perfused by red blood cells. As more capillaries become perfused by red blood cells, they are "recruited" and add to the gas exchange surface area. There are two models of capillary recruitment: segment-by-segment recruitment over a wide range of distending pressure, and the sheet flow model in which all capillaries suddenly become perfused when alveolar pressure is exceeded. This disparity between capillary recruitment models has been a difficult and important issue to resolve. The evidence supporting each idea appears solid, yet the physiological implications of each model are significantly different. Recent work by Godbey et al. [5] may have resolved the issue, because they found a way in which capillaries could be predictably recruited either segment by segment or as a sheet. The pattern of recruitment did not depend on capillary pressure or on alveolar pressure per se; rather, the recruitment pattern depended in a surprising way on the state of distention of the alveoli. Distended alveoli stretch capillaries into oval cross sections, which seemed to cause the individual segments to have a range of opening pressures. This, in turn, led to segment-by-segment recruitment. Whether alveolar distention was caused by positive airway pressure or by the pull of gravity on a lung suspended in the thorax by negative intrapleural pressure as in life (0 mmHg airway pressure), progressive capillary recruitment occurred over a wider range of capillary pressures as the alveoli enlarged. Further, the oval cross sections of the capillaries constrained the red blood cells to flow single file with their broad sides facing alveolar gas, an orientation favorable for gas exchange. Smaller, less distended alveoli permit capillaries to become circular; the "relaxed" capillaries in smaller alveoli opened suddenly as alveolar pressure was exceeded, that is, as a sheet. The circular cross sections permitted the red blood cells to flow in any orientation with respect to alveolar gas. However, many of these orientations are less favorable for gas exchange.

These experiments showed that each model of capillary recruitment was possible; the appropriate model depended on the amount of alveolar distention (Figure 3). The sheet flow model seems more predictive for the lower lung of large animals where alveoli are small, and in small animals in which the alveoli are relatively small in diameter. Segment-by-segment recruitment seems more appropriate in the upper lung of large animals where alveoli are large. That the state of alveolar distention could so dramatically affect capillary recruitment characteristics was unexpected and surprising. One explanation is based on the anatomical observation that connective tissue fibers are interlaced throughout the capillary network. When the lobes are inflated to low levels of distention (airway pressure = 2 mmHg), the connective tissue fibers are slack. Under these conditions, all of the capillaries could be readily perfused as microvascular pressure exceeds alveolar pressure. As the alveoli are distended, the connective tissue fibers are stretched. Individual fibers become progressively taut, but not equally so. As taut fibers cross each capillary, they tend to pinch the segments. The amount of pinching would be likely to vary among segments and lead to a range of opening pressures and therefore recruitment over a range of pressures. Further increases in alveolar distention would make these variations more pronounced. In this way, capillary recruitment on a segment-by-segment basis or on a sheet flow basis can be explained (Figure 3).

Although the work of Godbey explained how average patterns of recruitment (sheet versus segment-by-segment) were a function of the degree of alveolar distension, there was significant variability in the recruitment pattern among neighboring alveoli, all of which were inflated with the same pressure. Baumgartner and colleagues [6] designed a study to investigate this variability. In isolated, pump-perfused canine lung lobes, fields of six neighboring alveoli were recorded with video microscopy as pulmonary venous pressure was raised from 0 to 40 mmHg in 5-mmHg increments. The largest group of alveoli (42%) recruited gradually. Another group (33%) recruited suddenly (sheet flow). Half

Figure 3 In excised perfused dog lobes (panels A, B, and C), when alveolar pressure is low (A), the capillary bed opens suddenly and completely as alveolar pressure is exceeded by perfusion pressure; the capillaries are perfused as a sheet. As alveolar pressure is increased (B) and alveoli enlarge, capillary recruitment occurs more gradually. Once the alveoli are distended (C), recruitment is linear. Recruitment patterns are different in panel D. The lungs were inflated by negative intrapleural pressure and perfused in an intact canine thorax (alveolar pressure = 0mmHg). Observations were made on the uppermost lung where highly distended alveoli had capillary networks that were distended and recruited gradually on a segment-by-segment basis. (Reproduced with permission from Ref. [5].)

of the neighborhoods had at least one alveolus that paradoxically derecruited when pressure was increased, even though neighboring alveoli continued to recruit capillaries. At pulmonary venous pressures of 40 mmHg, 86 percent of the alveolar-capillary networks were not fully recruited. They

Figure 4 Average capillary recruitment for alveoli (n = 54) in six isolated, pump-perfused dog lung lobes. Although there were a variety of modes of recruitment among alveoli, the average recruitment was gradual. The arterio-venous pressure gradient at a venous pressure of zero was 8 mmHg, which decreased to 3 mmHg at a venous pressure of 40 mmHg, which suggests that capillary transmural pressure was close to venous pressure at the higher perfusion pressures. (Reproduced with permission from Ref. [6].)

Figure 4 Average capillary recruitment for alveoli (n = 54) in six isolated, pump-perfused dog lung lobes. Although there were a variety of modes of recruitment among alveoli, the average recruitment was gradual. The arterio-venous pressure gradient at a venous pressure of zero was 8 mmHg, which decreased to 3 mmHg at a venous pressure of 40 mmHg, which suggests that capillary transmural pressure was close to venous pressure at the higher perfusion pressures. (Reproduced with permission from Ref. [6].)

concluded that the pattern of recruitment among neighboring alveoli was complex, was not homogeneous, and may not have reached full recruitment, even during extreme pressures. Taken as a group, however, the capillaries recruited gradually (Figure 4).

Although these investigations have increased our knowledge about capillary recruitment, it was not known whether larger vessels in the circulation could be recruited. If arteri-oles or arteries could be recruited, then the recruitment of a single precapillary vessel would add significantly to capillary volume. Hanson et al. [7] investigated this issue using in vivo microscopy to study capillary segments and branches of arterioles and venules in the dog lung. With each vessel serving as its own control, observations were made during low and high vascular pressures. The only vessels to recruit were capillaries, indicating that, in the dog, recruitment was exclusively a capillary event (Figure 5). A similar conclusion was reached in the study of Warrell et al., who studied frozen dog lungs.

Much of the data discussed thus far have involved static or very nearly static conditions: injected vessels, frozen lungs, or vessels perfused at very low flow rates all sampled

Figure 5 Experiments in intact dogs in which pulmonary arterial pressure was raised either by airway hypoxia or by fluid loading with dextran. In both cases, pulmonary arterial pressure rose from 8 to 23 mmHg. The only vessels that recruited in response to these hemodynamic loads were capillaries, which showed that recruitment is a capillary event. (Reproduced with permission from Ref. [7].)

Figure 5 Experiments in intact dogs in which pulmonary arterial pressure was raised either by airway hypoxia or by fluid loading with dextran. In both cases, pulmonary arterial pressure rose from 8 to 23 mmHg. The only vessels that recruited in response to these hemodynamic loads were capillaries, which showed that recruitment is a capillary event. (Reproduced with permission from Ref. [7].)

at one point in time. As valuable as those data are, they lack the dynamic quality of perfusion that exists in the living lung. The complexity of the pulmonary microcirculation can only be measured by performing studies in vivo. Data from in vivo microscopy produced the first direct evidence for a recruitable reserve of pulmonary capillaries. Those classic observations were made by Wearn et al. [8], who studied the pulmonary microcirculation on the surface of the cat lung. Not only did they demonstrate that the capillary bed was only partially perfused by red blood cells at any given time, but they also showed that the perfusion pattern varied with time within a single alveolar wall. Wearn et al. speculated that the changes in the capillary perfusion pattern could be explained by changes in pulmonary blood flow or pressure, but in 1934, they lacked the techniques necessary to study these hemodynamic variables.

These important findings remained unconfirmed by direct observation for 60 years until Okada et al. [9] repeated Wearn's experiment using modern cardiovascular monitoring combined with in vivo microscopy of an intact canine preparation. During a 45-minute study period, 1-minute observations were made every 5 minute to determine which capillaries were perfused. On average, Okada et al. found about half of the capillaries to be perfused over the course of a total of nine observations, confirming that not all of the capillaries were perfused all of the time. The level of recruitment fluctuated, but the variations in perfusion did not correlate with the small fluctuations in pulmonary arterial pressure or cardiac output that occurred during the study. In most animals, the correlation was poor (r2 less than 0.06), which was a surprise, because previous work by this group [10] had shown that pulmonary arterial pressure and capillary recruitment were tightly correlated (Figure 6), at least when the pressure changes were large. These findings imply that capillary perfusion alterations were being influenced by factors more subtle than small variations in pulmonary arterial pressure or cardiac output.

Figure 6 In intact dogs, capillary recruitment was closely linked to pulmonary arterial pressure. Airway hypoxia was used to raise pulmonary arterial pressure. (Reproduced with permission from Ref. [9].)

Two characteristics emerged when the perfusion of individual capillary segments was studied in detail. First, of the perfused segments, more than half were perfused by red blood cells during at least eight of the nine observations. This finding showed that some capillary segments had hemodynamic characteristics that were sufficiently stable to cause them to be continuously or nearly continuously perfused. These capillaries generally were interconnected to form pathways across entire alveolar walls. Second, the remaining half of the segments were not perfused stably. Some were perfused only once or twice during the nine 1-minute observation periods. The frequency of perfusion of the unstably perfused segments was evenly dispersed among observations and resulted in considerable switching of flow among the segments within each alveolar network.

The complexity of the recruitment process is even more apparent when it is considered that capillary recruitment generally refers to perfusion of new segments by a combination of red cells and plasma. However, capillaries can be perfused (recruited) by plasma alone. Recent work by Con-haim and Rodenkirch [10a] has shown that capillaries in the rat lung even under zone 1 conditions have an estimated functional diameter of 1.7 mm, sufficiently large to admit plasma, but an impediment to red blood cell perfusion. Consistent with these findings, König et al. [10b] injected colloidal gold nanospheres (8 nm diameter) into the pulmonary circulation of rabbits. After 2minutes, the circulation was stopped, and the lungs were fixed and examined both by electron and light microscopy. They found that the entire pulmonary capillary bed was perfused by gold particles after 2 minutes, and in later experiments, complete dispersion occurred in 10seconds. These studies suggest that all of the capillaries were perfused by some of the plasma all of the time. This work suggests the following possibilities. The entirety of the capillary endothelium could act continuously on the plasma to perform its many metabolic functions. Red blood cell recruitment, however, would occur only when capillary pressure increased sufficiently to distend capillary segments enough for the passage of red cells.

At the outset, we pointed out that the lung is conceptualized in an overly simplistic way: a single-airway, single-alveolus, single-capillary model. Much evidence demonstrates that the pulmonary microcirculation behaves in a complex manner with perfusion rapidly switching between capillary segments and red blood cell velocity and hematocrit continually varying. As cardiac output increases with exercise, however, a number of changes occur that cause the microcirculation to behave more homogeneously. First, as capillary pressure increases, capillary recruitment approaches 100 percent. Presumably, capillary resistance becomes more uniform as segments that were closed open and distend toward the plateau of their compliance curve. These changes lead to a diminution of red blood cell velocities, and the heterogeneity of capillary transit times lessens. The complexity of perfusion of the pulmonary microcirculation during increasing exercise, therefore, decreases and the simplistic one-alveolus, one-capillary model becomes progressively more realistic.

Glossary

Capillary recruitment: perfusion of previously unperfused capillary Capillary transit time: time to vessel from arteriods to venules Pulmonary capillaries: gas exchange vessel in alveolar wall

References

1. Wagner, P. D. (1977). Diffusion and chemical reaction in pulmonary gas exchange. Physiol. Rev. 57, 257-312. Although somewhat dated, this superb review will provide excellent grounding in the pulmonary physiology of gas exchange.

2. Presson, R. G., Jr., Hanger, C. C., Godbey, P. S., Graham, J. A., Sidner, R. A., and Wagner, W. W., Jr. (1995). Distribution of pulmonary capillary red blood cell transit times. J. Appl. Physiol. 79, 382-388.

3. Fung, Y. C., and Sobin, S. S. (1972). Pulmonary alveolar blood flow. Circ. Res. 30, 470-490. These investigators did classic work showing that all of the pulmonary capillaries opened suddenly and completely once capillary pressure exceeded alveolar pressure, a phenomenon known as sheet flow. This work started a long-running controversy between the sheet flow school and the recruitment school, which was finally settled by the paper in Ref. [5].

4. Glazier, J. B., Hughes, J. M. B., Maloney, J. E., and West, J. B. (1969). Measurements of capillary dimensions and blood volume in rapidly frozen lungs. J. Appl. Physiol. 26, 65-76. This group led the way in showing that blood flow tended to perfused the base of the lung rather that the apex (gravitationally dependent versus independent). In this paper they show that more capillaries are perfused with blood (recruited) in the lower lung.

5. Godbey, P. S., Graham, J. A., Presson, R. P., Jr., Wagner, W. W., Jr., and Lloyd, T. C., Jr. (1995). Effect of capillary pressure and lung distention on capillary recruitment. J. Appl. Physiol. 79, 1142-1147.

6. Baumgartner, W. A., Jr., Jaryszak, E. M., Peterson, A. J., Presson, R. G., Jr., and Wagner, W. W., Jr. (2003). Heterogeneous capillary recruitment among adjoining alveoli. J. Appl. Physiol. 95, 469-476.

7. Hanson, W. L., Emhardt, J. D., Bartek, J. P., Latham, L. P., Checkley, L. L., Capen, R. L., and Wagner, W. W., Jr. (1989). The site of recruitment in the pulmonary microcirculation. J. Appl. Physiol. 66, 2079-2083.

8. Wearn, J. T., Ernstene, A. C., Bromer, A. W., Barr, J. S., German, W. J., and Zschiesche, L. J. (1934). The normal behavior of the pulmonary blood vessels with observations on the intermittence of the flow of blood in the arterioles and capillaries. Am. J. Physiol. 109, 236-256.

9. Okada, O., Presson, G. R., Jr., Godbey, P. S., Capen, R. L., and Wagner, W. W., Jr. (1994). Temporal capillary perfusion patterns in single alveolar walls of intact dogs. J. Appl. Physiol. 76, 380-386.

10. Wagner, W. W., Jr., Latham, L. P., and Capen, R. L. (1979). Capillary recruitment during airway hypoxia: Role of pulmonary arterial pressure. J. Appl. Physiol. 47, 383-387.

10a. Conhaim, R. L., and Rodenkirch, L. A. (1997). Estimated functional diameter of alveolar septal microvessels at the zone I—II border. Micro-

10b. König, M. F., Lucocy, J. M., and Weibel, E. R. (1993). Demonstration of pulmonary vascular perfusion by electron and light microscopy. J. Appl. Physiol. 75: 1870-1876.

11. Wagner, W. W., Jr., Latham, L. P., Hanson, W. L., Hofmeister, S. E., and Capen, R. L. (1986). Vertical gradient of pulmonary capillary transit times. J. Appl. Physiol. 61, 1270-1274. This group was the first to demonstrate that not only is there less capillary recruitment in the upper lung, but also the transit times are much longer than in the lower lung. This work was controversial until verified by independent techniques by other investigators.

Capsule Biographies

Dr. Wagner has been the Director of Research for the Anesthesia Department at Indiana University since 1985. His laboratory has focused on the pulmonary microcirculation using in vivo microscopy with the aim of understanding how blood flow is controlled at rest and during exercise. His work is supported by the NIH.

Dr. Presson has been a close collaborator with Dr. Wagner for the past 17 years. He has focused on mathematical analysis and modeling of pulmonary microcirculatory flow data. In addition he works with animal models of emphysema. He is a full-time practicing anesthesiologist. His work is supported by NIH.

Delicious Diabetic Recipes

Delicious Diabetic Recipes

This brilliant guide will teach you how to cook all those delicious recipes for people who have diabetes.

Get My Free Ebook


Post a comment