HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Free All Versions of this Article: 98/3/1125    most recent 00903.2004v1 Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (8) Citing Articles via Google Scholar Google Scholar Articles by Talbot, N. P. Articles by Robbins, P. A. Search for Related Content PubMed PubMed Citation Articles by Talbot, N. P. Articles by Robbins, P. A.

HIGHLIGHTED TOPICS
Pulmonary Circulation and Hypoxia

Two temporal components within the human pulmonary vascular response to 2 h of isocapnic hypoxia

¹University Laboratory of Physiology, University of Oxford, Oxford, and ²School of Sport and Exercise Sciences, University of Birmingham, Edgbaston, Birmingham, United Kingdom

Submitted 19 August 2004 ; accepted in final form 10 November 2004

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION APPENDIX GRANTS ACKNOWLEDGMENTS REFERENCES

 
The time course of the pulmonary vascular response to hypoxia in humans has not been fully defined. In this investigation, study A was designed to assess the form of the increase in pulmonary vascular tone at the onset of hypoxia and to determine whether a steady plateau ensues over the following 20 min. Twelve volunteers were exposed twice to 5 min of isocapnic euoxia (end-tidal PO₂ = 100 Torr), 25 min of isocapnic hypoxia (end-tidal PO₂ = 50 Torr), and finally 5 min of isocapnic euoxia. Study B was designed to look for the onset of a slower pulmonary vascular response, and, if possible, to determine a latency for this process. Seven volunteers were exposed to 5 min of isocapnic euoxia, 105 min of isocapnic hypoxia, and finally 10 min of isocapnic euoxia. For both studies, control protocols consisting of isocapnic euoxia were undertaken. Doppler echocardiography was used to measure cardiac output and the maximum tricuspid pressure gradient during systole, and estimates of pulmonary vascular resistance were calculated. For study A, the initial response was well described by a monoexponential process with a time constant of 2.4 ± 0.7 min (mean ± SE). After this, there was a plateau phase lasting at least 20 min. In study B, a second slower phase was identified, with vascular tone beginning to rise again after a latency of 43 ± 5 min. These findings demonstrate the presence of two distinct phases of hypoxic pulmonary vasoconstriction, which may result from two distinct underlying processes.

hypoxic pulmonary vasoconstriction; pulmonary circulation; pulmonary vascular resistance; pulmonary blood flow; cardiac output

In the human, few studies have addressed the time course of acute HPV. Although it has been demonstrated that pulmonary vasoconstriction in response to bronchial occlusion is rapid in onset in the human [time constant () = 150 s], clear interpretation of these data is difficult because of the combined stimulus of hypoxia and hypercapnia (29). Other human studies have been complicated by anesthesia (7) or have not made frequent enough measurements to resolve changes occurring within minutes of the onset of hypoxia but have instead identified a slower component of HPV developing gradually over a period of hours (15).

The technique of Doppler echocardiography permits noninvasive measurement of pulmonary vascular hemodynamics (43). The computer-controlled fast gas-mixing system used in our laboratory (32) allows the alveolar partial pressure of oxygen to be altered rapidly and accurately without a concomitant change in the alveolar partial pressure of carbon dioxide. Together, these techniques were used to address the two primary aims of the present studies. Study A was designed to determine the speed of response of any rapid increase in pulmonary vascular tone at the onset of isocapnic hypoxia and to determine whether a steady plateau occurs during the following 20 min. Study B was designed to look for the onset of the slower component of HPV that has previously been identified in the human (15) and to determine whether a latency for this response could be identified. Our results provide evidence for two distinct phases of HPV that would seem likely to be associated with two separate mechanisms of constriction in response to hypoxia.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION APPENDIX GRANTS ACKNOWLEDGMENTS REFERENCES

 
Volunteers.   Twelve healthy volunteers (six women, six men, age 26.6 ± 2.8 yr, mean ± SD) participated in study A. Seven healthy volunteers (two women, five men, age 24.4 ± 3.4 yr) participated in study B. All volunteers visited the laboratory on at least one occasion before the first experiment to become familiar with the laboratory and to confirm that they were suitable for echocardiographic measurement of tricuspid regurgitation. In common with other studies (22, 44), these measurements were possible in 75% of our volunteers. Women were asked to participate only during the first 14 days of their menstrual cycle. Protocols were explained fully to all volunteers, and written consent was obtained from each volunteer on each experimental day. Volunteers were, however, naive as to the exact purpose of the experiment. Ethical permission was obtained from the Oxfordshire Research Ethics Committee.

Protocols for study A.   In all protocols of both study A and study B, isocapnic refers to the condition in which end-tidal partial pressure of carbon dioxide (PET_CO2) was maintained at 1–2 Torr above normal for each volunteer, as determined at the beginning of each day in the laboratory. Euoxia refers to the condition in which end-tidal partial pressure of oxygen (PET_O2) was maintained at 100 Torr. Hypoxia refers to the condition in which PET_O2 was maintained at 50 Torr.

Study A consisted of two protocols, each of which lasted 35 min and was undertaken by each volunteer twice. In the control protocol, volunteers were exposed to 35 min of isocapnic euoxia. In the hypoxia protocol, volunteers were exposed to 5 min of isocapnic euoxia, followed by 25 min of isocapnic hypoxia, followed by a further 5 min of isocapnic euoxia.

The purpose of study A was to characterize the speed of any rapid pulmonary vascular response to isocapnic hypoxia and to identify whether a plateau of pulmonary vascular tone follows this rapid response. To allow almost continuous measurement of both cardiac output and an index of pulmonary vascular tone, both the control and hypoxia protocols were performed twice. This enabled Doppler measurements to focus on just cardiac output during one repeat of the protocol and on just pulmonary vascular responses during the other repeat of the protocol (see Data acquisition and analysis below). The two repeats of each protocol were undertaken within a 24-h period but not less than 4 h apart. The order of the protocols, and of echocardiographic measurements, was randomized.

Protocols for study B.   Study B consisted of two protocols, each of which lasted 120 min and was undertaken by each volunteer once. Protocols were separated by at least 5 days. In the control protocol, volunteers were exposed to 120 min of isocapnic euoxia. In the hypoxia protocol, they were exposed to 5 min of isocapnic euoxia, followed by 105 min of isocapnic hypoxia, followed by a further 10 min of isocapnic euoxia.

Study B was designed specifically to investigate the length of any plateau of pulmonary vascular tone identified in study A and to determine the relationship between any such plateau and the slower rise in pulmonary vascular tone that has previously been reported in humans (15). Doppler echocardiographic measurements were again made almost continuously throughout both protocols, but, in contrast to study A, both cardiac output and pulmonary vascular tone were assessed in a single protocol by switching between echocardiographic views sequentially. Cardiac output was measured approximately every 5–10 min (see Data acquisition and analysis below).

Control of end-tidal gases.   Volunteers were in the supine semi-left-lateral position throughout all protocols. In study A, they breathed on a mouthpiece with their noses occluded. In study B, they breathed through a face mask that allowed breathing only through the mouth. End-tidal gas was sampled continuously from a catheter within the mouthpiece or face mask and analyzed using a mass spectrometer. End-tidal profiles were generated using a computerized dynamic end-tidal forcing system, whereby actual end-tidal gas composition was recorded every breath and compared with desired values by a computer that controlled a fast gas-mixing system. Deviations of actual values from the desired values were used to modify inspired gas mixtures breath by breath by using an integral-proportional feedback control scheme. The control scheme has been described in detail previously (32). Inspired gases were heated and humidified. Respiratory volumes were measured by use of a turbine, and ventilation is reported BTPS.

Echocardiographic measurements.   Echocardiographic measurements were performed with a Hewlett-Packard Sonos 5500 ultrasound machine with an S4 two-dimensional transducer (2 MHz). Heart rate and respiratory waveform were also recorded. Throughout all protocols, volunteers were lying on a modified couch rolled slightly toward the left-lateral position, and echocardiographic measurements were recorded continuously to super-VHS video cassettes for subsequent offline analysis (see Data acquisition and analysis below).

Cardiac output.   For measurement of cardiac output, an apical five-chamber view was obtained using color Doppler echocardiography, and the Doppler beam was aligned with flow through the aortic valve during systole. The velocity profile of flow just below the aortic valve was recorded using pulsed-wave spectral mode (sweep speed 100 mm/s). An automated procedure was used to calculate the velocity-time integral (VTI) of the flow, which represents the distance traveled by blood in the left ventricular outflow tract during ventricular contraction. The diameter just below the aortic valve orifice was measured from a parasternal long-axis view, and the area at this site (A) was calculated. Heart rate (HR) was measured continuously with an ECG. Cardiac output () was then calculated by Eq. 1:

(1)

This technique has been shown in previous studies to correlate well with invasive measurement of cardiac output (8).

Tricuspid valve maximal pressure gradient.   In most people (22, 44), a small regurgitant jet of blood through the tricuspid valve is detectable during systole. Doppler echocardiography can be used to measure the velocity () with which this jet flows from the right ventricle into the right atrium (43). If the flow within the jet is regarded as steady, Bernoulli's equation (Eq. 2) can be used to relate the maximum pressure difference (P_max) to :

(2)

where is the density of blood. Because P_max is the difference between systolic pressure in the right ventricle and the relatively constant pressure in the right atrium, changes in P_max reflect changes in systolic pressure in the right ventricle (or main pulmonary artery). We have argued that P_max is an index of changes in pulmonary vascular tone arising from changes in the activity in the smooth muscle of pulmonary arteries and is relatively insensitive to changes in cardiac output (6, 16).

A standard technique was used for the measurement of P_max (43). Two-dimensional echocardiography allowed visualization of the tricuspid valve in an apical four-chamber view, and color Doppler format allowed detection of the regurgitant jet. The Doppler beam was aligned with the jet, and continuous-wave spectral mode (sweep speed 50 mm/s) was used to record its velocity profile. During analysis, the maximal velocity of the jet was measured with an electronic caliper, and P_max was calculated automatically by the echocardiography machine.

Data acquisition and analysis.   Throughout all protocols, echocardiographic images were recorded continuously to super-VHS video tapes. During subsequent offline analysis, tricuspid regurgitation and aortic flow velocity profiles were deemed to be suitable for inclusion only if a full profile was clearly visible, and to minimize any effect of changes in lung volume on P_max, velocity profiles were suitable only if they occurred at, or near, end expiration.

In study A, the hypoxia and control protocols were each repeated twice. During one repeat of each protocol, only images used for determination of P_max were obtained and recorded. During the other repeat, only images used for the determination of cardiac output were obtained and recorded. During subsequent offline analysis, a single value for P_max or cardiac output was generated for each volunteer every 30 s by averaging the first three to five suitable velocity profiles recorded during each 30-s period. A single value for cardiac output was generated for each volunteer every 60 s, by averaging the first three to five suitable velocity profiles recorded during each 60-s period.

In study B, the hypoxia and control protocols were performed only once for each volunteer. During most of the protocol, images for determination of P_max were recorded, but every 5–10 min the position of the transducer was altered for 60 s to record images used for the determination of cardiac output. During subsequent offline analysis, a single value for P_max was generated every 60 s as described above, except during the periods when cardiac output was being recorded. A single value for cardiac output was generated for each 60-s measurement period, as described above.

Validation of P_max as an estimate of changes in systolic pulmonary artery pressure.   If right atrial pressure is assumed to remain relatively constant during hypoxia (19), changes in P_max should reflect changes in systolic pulmonary artery pressure (SPAP). In a previous study in our laboratory (15), we used pulmonary artery catheters to study the change in SPAP during 8 h of isocapnic hypoxia in five healthy volunteers. Volunteers were exposed in a purpose-built chamber to 8 h of isocapnic hypoxia (PET_O2 = 50 Torr), followed by 2 h of isocapnic euoxia (PET_O2 = 100 Torr). Measurements of SPAP were made at 30- to 120-min intervals using pulmonary artery catheters. The changes in SPAP are shown in Fig. 1A.

View larger version (11K): [in this window] [in a new window]   Fig. 1. Changes in systolic pulmonary artery pressure (SPAP), measured by right heart catheterization (A, n = 5) and by Doppler echocardiography [maximum pressure difference across tricuspid valve during systole (Pmax); B, n = 8] in healthy humans. Data in A are taken from Fig. 5 of Dorrington et al. (15).

To assess the extent of any error associated with the measurement of P_max, five volunteers from study B were chosen at random. Ten tricuspid valve regurgitation velocity profiles were selected at random from within designated periods distributed throughout the control protocol of each volunteer, and a further 10 profiles were similarly selected from the hypoxia protocol of each volunteer. This provided a total of 100 profiles over a wide range of P_max values. Each profile was then measured twice by the investigator responsible for the analysis of the data in the present study and once by another member of the laboratory trained in echocardiography. The intraobserver error for measurements of a single profile was 0.9 ± 0.1 mmHg (mean ± SE). The interobserver error for measurements of a single profile was 1.3 ± 0.1 mmHg.

Calculation of pulmonary vascular resistance.   It is not possible to calculate pulmonary vascular resistance (PVR) directly from our Doppler echocardiographic measurements without making some additional assumptions. The assumptions made for the calculation of PVR in the present study were made on the basis of data gathered by pulmonary artery catheterization of healthy humans. Data from the baseline (sea level) protocols of the "Operation Everest II" study of Groves et al. (19) were used to provide mean values for right atrial pressure (RAP) and pulmonary artery wedge pressure (PAWP, commonly taken to be equal to left atrial pressure). For both RAP and PAWP, Groves et al. provide evidence that these variables are unchanged during acute hypoxia at sea level, and we have therefore used the mean of values during hypoxia and euoxia for our calculations. Data from the study of Dorrington et al. (15) were used to relate peak SPAP to mean pulmonary artery pressure (MPAP), as described below.

PVR is defined by the following equation:

(3)

Measurement of cardiac output was made directly as described above, but values for MPAP have to be calculated from P_max. First, at each time point, SPAP was calculated from P_max by using the following equation:

(4)

MPAP can in turn be estimated from SPAP. Figure 2 shows the relationship between these parameters in six healthy human volunteers studied during isocapnic hypoxia using pulmonary artery catheters (15). The slope of this relationship, m, was used to estimate MPAP from SPAP at each time point:

(5)

Equations 3, 4, and 5 may be combined to yield:

(6)

This expression shows the estimate of PVR in terms of P_max and cardiac output, as estimated by Doppler echocardiography, a constant of proportionality between SPAP and MPAP, as provided by the data of Dorrington et al. (15), and RAP and PAWP, as provided by the data of Groves et al. (19).

View larger version (12K): [in this window] [in a new window]   Fig. 2. Relationship between mean pulmonary artery pressure (MPAP) and SPAP in 6 awake healthy humans studied during isocapnic hypoxia by using pulmonary artery catheters. Data () are taken from Fig. 5 of Dorrington et al. (15). Dashed line represents a linear regression through the data.

Correcting pulmonary vascular resistance for changes in cardiac output.   During hypoxia, cardiac output does not remain constant, and changes in pulmonary blood flow per se alter the resistance of the human pulmonary circulation (10, 19). The relationship between cardiac output and the pulmonary vascular pressure gradient (MPAP – PAWP) at sea level was determined by Groves et al. (19). The pressure gradient was seen to increase by a mean of just 0.24 ± 0.07 mmHg for each liter per minute increase in cardiac output, suggesting that, as cardiac output rises, PVR falls. Assuming that this relationship holds for the data of our study, we calculated the change in PVR that would have occurred during hypoxia in the absence of a change in cardiac output (this we denote PVR_c, to indicate that the value for PVR has been corrected, Eq. 7):

(7)

where _B is the mean cardiac output during the first 5 min of each protocol, and is the difference between _B and at each time point. To distinguish the raw uncorrected value for PVR from PVR_c, we subsequently denote this as PVR_r.

Modeling of pulmonary vascular responses.   To describe fully the pulmonary vascular responses to hypoxia observed in studies A and B, mathematical models were fitted to the responses observed in P_max, PVR_r, and PVR_c. Single-compartment models were fitted to the data of study A assuming a single "fast-up" process after the onset of hypoxia and a single "fast-down" process after the offset of hypoxia:

(8)

(9)

where P represents the index of pulmonary vascular tone being modeled (P_max, PVR_r, or PVR_c), y_fu and y_fd are the fast-up and fast-down responses, respectively, y_c is the baseline value for P before the onset of hypoxia, ₁ is the predicted steady-state value for P to be attained by the fast-down process, g_fu and g_fd are the gains and _fu and _fd the time constants of each response, and x(t) represents the input stimulus (zero during euoxia, one during hypoxia) at time t. These equations were solved to provide a set of difference equations and then fitted to the data as described in the APPENDIX.

In study B, the changes in P after the onset of hypoxia were described by using a two-compartment model, assuming an initial fast-up component of the response, plus a secondary "slow-up" component that began only after some pure delay (Eq. 10). A similar theoretical model was employed for the offset of hypoxia, assuming a fast-down component and a "slow-down" component (Eq. 11).

(10)

(11)

where y_su and y_sd are the slow-up and slow-down components, respectively, and x(t – d_su) represents the input stimulus at time t delayed by the pure delay of the slow-up component, d_su. These equations were solved and fitted as described in the APPENDIX. However, because of the nature of the input stimulus, not all model parameters could be identified for the offset response (see APPENDIX), and values for g_fd, g_sd, and _sd are not reported.

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION APPENDIX GRANTS ACKNOWLEDGMENTS REFERENCES

 
Gas control in studies A and B.   Figure 3 shows values of inspired PCO₂ (PI_CO2) and PO₂ (PI_O2), and PET_CO2 and PET_O2 for each protocol of study A. Separate plots show gas control during the measurements of cardiac output and of P_max in each protocol and demonstrate that near-identical conditions were achieved in the two cases. Figure 3 also shows gas control for each protocol of study B. In both studies, PET_CO2 was almost constant throughout each protocol, steps between euoxia and hypoxia were complete within 45 s, and during both euoxia and hypoxia PET_O2 was almost constant at the desired level.

View larger version (33K): [in this window] [in a new window]   Fig. 3. Inspired and end-tidal PO2 (PIO2, triangles, and PETO2, diamonds, respectively), and PCO2 (PICO2, circles, and PETCO2, squares, respectively) during each protocol of study A (top and middle) and study B (bottom). Open symbols, values during euoxia; closed symbols, values during hypoxia. Values are means ± SE.

View larger version (29K): [in this window] [in a new window]   Fig. 4. Ventilation (A), heart rate (B), stroke volume (C), and cardiac output (D), during control (circles) and hypoxia (triangles) protocols of study A. Open symbols, values during euoxia; closed symbols, values during hypoxia (25 min, PETO2 = 50 Torr). PETCO2 was constant at 1–2 Torr above each volunteer's normal value throughout both protocols. Ventilation is given at BTPS. Values are means ± SE.

View larger version (21K): [in this window] [in a new window]   Fig. 5. Ventilation (A), heart rate (B), stroke volume (C), and cardiac output (D) during control (circles) and hypoxia (triangles) protocols of study B. Open symbols, values during euoxia; closed symbols, values during hypoxia (105 min, PETO2 = 50 Torr). PETCO2 was constant at 1–2 Torr above each volunteer's normal value throughout both protocols. Ventilation is given at BTPS. Values are means ± SE.

View larger version (28K): [in this window] [in a new window]   Fig. 6. Pulmonary vascular responses to hypoxia in study A. A: changes in Pmax for a single volunteer (left, ) and the mean of all volunteers (right, ). Model fits to Pmax values are shown in A as solid lines. B: changes in raw (PVRr; triangles) and corrected (PVRc; circles) pulmonary vascular resistance for a single volunteer (left, open symbols) and the mean of all volunteers (right, closed symbols). Model fits for PVRr and PVRc in B are shown as dashed and solid lines, respectively.

View this table: [in this window] [in a new window]   Table 1. Study A: Parameter values obtained for the model of the pulmonary vascular response to hypoxia for the 3 indexes of response (Pmax, PVRr, or PVRc)

View larger version (26K): [in this window] [in a new window]   Fig. 7. Changes in Pmax (A), PVRr (B), and PVRc (C) during control (circles) and hypoxia (triangles) protocols of study A. Open symbols, values during euoxia; closed symbols, values during hypoxia (25 min, PETO2 = 50 Torr). PETCO2 was constant at 1–2 Torr above each volunteer's normal value throughout both protocols. Values are means ± SE.

View larger version (20K): [in this window] [in a new window]   Fig. 8. Changes in Pmax during hypoxia in study B. A–G: changes in Pmax for each individual volunteer (). Model fits to Pmax values for each individual are shown as solid lines. H: mean of all individual Pmax values (diamonds), with the mean of the individual models shown as a solid line.

View larger version (25K): [in this window] [in a new window]   Fig. 9. Changes in PVRr and PVRc during hypoxia in study B. A–G: changes in PVRr () and PVRc () for each individual volunteer. Model fits to PVRr and PVRc values for each individual are shown as dashed and solid lines, respectively. H: mean of all individual PVRr (triangles) and PVRc (diamonds) values, with the mean of the individual PVRr and PVRc models shown as dashed and solid lines, respectively.

View this table: [in this window] [in a new window]   Table 2. Study B: Parameter values obtained for the model of the pulmonary vascular response to hypoxia for the 3 indexes of response (Pmax, PVRr, or PVRc)

Figure 10 shows a comparison of average pulmonary vascular responses in the hypoxia and control protocols of study B. For each index of pulmonary vascular tone, a clear response to hypoxia is evident. In each case, in contrast to study A, pulmonary vascular tone can be seen not to return to control values within 10 min of the return to euoxia. This is confirmed in the case of P_max and PVR_c by a significant difference between model parameters y_c and ₁ (P < 0.02 for each index, paired t-tests). This difference was not significant in the case of PVR_r (P > 0.5).

View larger version (25K): [in this window] [in a new window]   Fig. 10. Changes in Pmax (A), PVRr (B), and PVRc (C) during control (circles) and hypoxia (triangles) protocols of study B. Open symbols, values during euoxia; closed symbols, values during hypoxia (105 min, PETO2 = 50 Torr). PETCO2 was constant at 1–2 Torr above each volunteer's normal value throughout both protocols. Values are means ± SE.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION APPENDIX GRANTS ACKNOWLEDGMENTS REFERENCES

P_max as a measurement of systolic pulmonary artery pressure.   Doppler echocardiography has been used in several previously published studies to measure P_max as an index of the pulmonary vascular response to hypoxia (6, 18, 20, 31). However, we have also performed our own validation of changes in P_max as a measure of SPAP in healthy humans by comparing the changes in P_max and SPAP in two identical 8-h exposures to hypoxia. Despite the use of different groups of healthy volunteers in the two exposures, time-aligned changes in P_max and SPAP correlated closely (r = 0.91, P < 0.01).

Because of differences in posture in the two protocols of our own validation study, comparison of absolute values measured invasively and noninvasively are of little value. However, our baseline values of 20–21 mmHg for P_max compare well with values from invasive studies of supine subjects in the literature. In a single group of 11 healthy volunteers, Lagerlöf and Werkö (25) recorded values of 2.8 and 23 mmHg for RAP and SPAP, respectively, giving a predicted value of 20.2 mmHg for P_max. These values fall within the normal range of –2 to 10 mmHg for RAP and 11 to 30 mmHg for SPAP, on the basis of a broad literature search (3).

Numerous studies have previously established the use of Doppler echocardiography as an accurate means of measuring SPAP. In 1984, Yock and Popp (43) performed the first such study, in which Doppler measurements and catheter values for SPAP correlated well (r = 0.93) in a single group of 54 patients with tricuspid regurgitation. More recently, a study performed in 28 volunteers at high altitude demonstrated that Doppler and invasive measurements also correlated well (r = 0.89) when pulmonary vascular tone is elevated (2).

Indexes of pulmonary vascular tone.   The effect of hypoxia on vascular smooth muscle is clearly characterized better by studying its effects on the whole of the tension-length relationship rather than by studying its effect on any single parameter. In the whole lung, however, this becomes complicated. One strategy has been to use changes in the pressure-flow relationship to quantify pulmonary vascular responses to hypoxia. In dogs at rest, plots of pulmonary vascular pressure against cardiac output have been obtained by graded constriction of the thoracic vena cava (27). In humans, similar plots can be produced using exercise to vary blood flow (19, 20). Neither of these approaches has been ideal, however, because both constriction of the vena cava and exercise will induce substantial changes in the PCO₂ and PO₂ of the mixed venous blood with which the lungs are then perfused. In the present study, it was not possible to look at the evolution of the pressure-flow relationship over time, and effectively the data provide just one point on this relationship for each of the times at which a measurement was made.

A second complication is that the measurements of the present study were not made under conditions of constant flow because cardiac output increases during hypoxia. We have argued previously that P_max provides a good index of changes in pulmonary vascular tone in normal individuals despite these variations in cardiac output (6, 16). However, by making some assumptions, it is also possible to calculate PVR from P_max. This has been done to derive both a "raw" PVR and, using data from the study of Groves et al. (19), a "corrected" PVR that attempts to present the PVR that would have prevailed had cardiac output remained constant. In a sense, changes in PVR_r and PVR_c represent two extreme possibilities for the assessment of the effects of hypoxia on the pulmonary circulation. PVR_r ignores the reduction in PVR that will be caused by the increase in cardiac output during hypoxia and is likely therefore to underestimate the effect of hypoxia on smooth muscle tone. PVR_c corrects for the effect of cardiac output on the basis of the relationship between pulmonary artery pressure and cardiac output for healthy individuals during euoxia. This relationship may become steeper during hypoxia, in which case PVR_c would overestimate the effects of hypoxia on smooth muscle tone. It seems most likely that the true effects of hypoxia on PVR at constant cardiac output lie somewhere between these two extremes. Most crucially for our study, however, the principal findings of the present study are essentially independent of the index that is used to represent pulmonary vascular tone.

Rapid phase of the pulmonary vascular response to hypoxia.   The literature relating to the time course of the pulmonary vascular response to acute hypoxia appears confusing and contradictory. In isolated buffer-perfused lungs and isolated pulmonary arteries from many species, HPV is commonly divided into two phases. Phase 1 is a marked transient constriction that occurs within seconds of the onset of hypoxia, reaches a maximum within minutes, and then rapidly reverses. This is superimposed on phase 2, a more modest but sustained constriction that begins within minutes and appears to reach a plateau within 40 min (12, 26, 40). In contrast, in vivo animal studies and those in isolated blood-perfused lungs usually report a constriction that occurs within seconds and reaches a sustained maximum within minutes (5, 14, 24, 28). The present study confirms in awake humans that the acute (<30 min) pulmonary vascular response to isocapnic hypoxia is monophasic and does not consist of two phases analogous to those of in vitro preparations but suggests that most previous studies have not been of sufficient duration to identify an additional slower component of HPV, discussed below.

Despite a huge literature in this area, relating mainly to studies in animals, the mechanisms underlying the rapid component of HPV are not well understood. We are unable to extensively review this literature here, but numerous others have recently done so (1, 4, 17, 38). In human studies, some degree of HPV has been shown to persist in lung transplant recipients (34) and in isolated pulmonary arteries (11, 30, 35), suggesting that local factors are probably of primary importance. There are, however, some reports in animals for a role of systemic chemoreceptors and the autonomic nervous system in HPV. In anesthetized dogs, for example, an immediate rise in pulmonary artery pressure during hypoxia can be markedly reduced by bilateral cervical and upper thoracic sympathectomy (23) or by denervation of the carotid and aortic bodies (5). To the best of our knowledge, the role of the autonomic nervous system in HPV has rarely been studied in humans.

Gradual phase of the pulmonary vascular response to hypoxia.   A recent study in mouse isolated lung demonstrated that pulmonary artery pressure rises progressively throughout a 2-h exposure to hypoxia (39). A similarly slow time course for HPV has previously been described both in anesthetized rabbits and in awake humans (13, 15, 37). These previous human experiments lacked the time resolution necessary to distinguish any rapid and slower components of the pulmonary vascular response during the first 60 min of hypoxia, but the findings of the present study demonstrate that, in humans, a period of 45 min separates the onset of these two components.

We can only speculate about the nature of the lag between the onset of hypoxia and the onset of slow HPV, but it may represent the time needed for changes to occur in the release or de novo production of one or more vasoactive substances. The importance of hypoxia-regulated gene transcription in the pulmonary vascular response to chronic hypoxia is well established (46), but a possible role for rapid changes in gene expression during hypoxia has been raised by studies on hypoxia-inducible factor-1 (HIF-1) and early growth response-1. In healthy humans, infusion of desferrioxamine, an iron chelator that stabilizes HIF-1, acts by 30 min to induce a small degree of pulmonary vasoconstriction (6), and in isolated pulmonary vascular smooth muscle cells from the ferret, HIF-1 protein was raised by hypoxia at 1 h into an 8-h study, the earliest time at which a measurement was made (45). In a study of early growth response-1 transcripts from the lungs of mice exposed intact to 6% oxygen, a 20-fold increase was observed after a 30-min exposure to hypoxia, the earliest time point in a 48-h study (42). In addition, cycloheximide, a nonspecific inhibitor of protein synthesis, has previously been reported to inhibit HPV in isolated lung (36), and, more recently, an unidentified pulmonary vasoconstrictor has been isolated from the perfusate leaving isolated rat lung during 2 h of hypoxia (33).

Respiratory and systemic vascular responses to hypoxia.   It is of note that the changes in heart rate and ventilation in study B follow a similar pattern to the changes in pulmonary vascular tone, with a rapid increase in both cardiac output and ventilation occurring within minutes of the onset of hypoxia, followed by a more gradual rise beginning after a delay of 35–40 min. Similar slow changes have been described previously; the progressive rise in heart rate during 8 h of isocapnic hypoxia has been ascribed to a gradual withdrawal of parasympathetic tone to the heart (9), and the gradual increase in ventilation has been attributed to a gradual rise in the sensitivity of the carotid body to hypoxia (21). Further experiments are required to investigate the possibility that common cellular mechanisms are involved in several components of the cardiorespiratory response to hypoxia and underlie the common time course of some of these responses.

In summary, we have identified more than one component in the time course of the human pulmonary vascular response to 2 h of isocapnic hypoxia. Our findings may help to reconcile the apparently contradictory findings of some previous experiments and suggest that more than one mechanism might be involved in human HPV.

	APPENDIX

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION APPENDIX GRANTS ACKNOWLEDGMENTS REFERENCES

 
The single-compartment model used to describe the changes in P_max, PVR_r, and PVR_c in study A (Eqs. 8 and 9; see MATERIALS AND METHODS) was integrated over 30-s periods, assuming a constant input during each period, to yield a set of difference equations:

(12)

(13)

where t is the time period over which data were averaged (30 s), and x₁ = (x_i + x_i+1)/2.

The two-compartment model used to describe the changes in P_max, PVR_r, and PVR_c in study B (Eqs. 10 and 11; see MATERIALS AND METHODS) was integrated over 1-min periods, assuming a constant input stimulus within each period, to yield a set of difference equations. For the onset of hypoxia:

(14)

(15)

(16)

where x₁ = (x_i + x_i+1)/2, x₂ = (x+ x)/2, and t is the time period over which data were averaged (1 min).

For the offset of hypoxia, because of the nature of the input stimulus, not all model parameters could be identified in study B. In particular, the slow-down component of the offset response could not be identified because of a lack of data. Hence, for the offset of hypoxia, a difference equation containing only the fast component was employed:

(17)

(18)

Furthermore, the lack of a steady state of P_max, PVR_r, or PVR_c by the end of the exposure to hypoxia in study B, with the resultant difficulty of separately identifying g_su and _su, made the identification of g_fd unreliable, and this parameter was therefore not estimated directly in study B.

The parameters for both study A and study B were estimated by nonlinear least squares regression using MATLAB (version 6.1.0).

	GRANTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION APPENDIX GRANTS ACKNOWLEDGMENTS REFERENCES

 
This research was funded by the Wellcome Trust.

	ACKNOWLEDGMENTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION APPENDIX GRANTS ACKNOWLEDGMENTS REFERENCES

 
We thank D. F. O'Connor for skilled technical assistance.

	FOOTNOTES

 

Address for reprint requests and other correspondence: P. A. Robbins, University Laboratory of Physiology, Parks Road, Oxford, OX1 3PT, United Kingdom (E-mail: peter.robbins{at}physiol.ox.ac.uk)

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

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION APPENDIX GRANTS ACKNOWLEDGMENTS REFERENCES

 

1. Aaronson PI, Robertson TP, and Ward JP. Endothelium-derived mediators and hypoxic pulmonary vasoconstriction. Respir Physiol Neurobiol 132: 107–120, 2002.
2. Allemann Y, Sartori C, Lepori M, Pierre S, Melot C, Naeije R, Scherrer U, and Maggiorini M. Echocardiographic and invasive measurements of pulmonary artery pressure correlate closely at high altitude. Am J Physiol Heart Circ Physiol 279: H2013–H2016, 2000.
3. Altman PL and Dittmer DS. Respiration and circulation. In: Biological Handbooks. Bethesda, MD: Federation of American Societies for Experimental Biology, 1971.
4. Archer S and Michelakis E. The mechanism(s) of hypoxic pulmonary vasoconstriction: potassium channels, redox O₂ sensors, and controversies. News Physiol Sci 17: 131–137, 2002.
5. Aviado D, Ling J, and Schmidt C. Effects of anoxia on pulmonary circulation: reflex pulmonary vasoconstriction. Am J Physiol 189: 253–262, 1957.
6. Balanos GM, Dorrington KL, and Robbins PA. Desferrioxamine elevates pulmonary vascular resistance in humans: potential for involvement of HIF-1. J Appl Physiol 92: 2501–2507, 2002.
7. Carlsson AJ, Bindslev L, Santesson J, Gottlieb I, and Hedenstierna G. Hypoxic pulmonary vasoconstriction in the human lung: the effect of prolonged unilateral hypoxic challenge during anaesthesia. Acta Anaesthesiol Scand 29: 346–351, 1985.
8. Christie J, Sheldahl LM, Tristani FE, Sagar KB, Ptacin MJ, and Wann S. Determination of stroke volume and cardiac output during exercise: comparison of two-dimensional and Doppler echocardiography, Fick oximetry, and thermodilution. Circulation 76: 539–547, 1987.
9. Clar C, Dorrington K, Fatemian M, and Robbins P. Effects of 8 h of isocapnic hypoxia with and without muscarinic blockade on ventilation and heart rate in humans. Exp Physiol 86: 529–538, 2001.
10. Cournand A, Riley RL, Himmelstein A, and Austrian R. Pulmonary circulation and alveolar ventilation-perfusion relationships after pneumonectomy. J Thorac Surg 19: 80–116, 1950.
11. Demiryurek AT, Wadsworth RM, Kane KA, and Peacock AJ. The role of endothelium in hypoxic constriction of human pulmonary artery rings. Am Rev Respir Dis 147: 283–290, 1993.
12. Dipp M, Nye PC, and Evans AM. Hypoxic release of calcium from the sarcoplasmic reticulum of pulmonary artery smooth muscle. Am J Physiol Lung Cell Mol Physiol 281: L318–L325, 2001.
13. Dirken MN and Heemstra H. The adaptation of the lung circulation to the ventilation. Q J Exp Physiol 34: 213–226, 1948.
14. Domino KB, Chen L, Alexander CM, Williams JJ, Marshall C, and Marshall BE. Time course and responses of sustained hypoxic pulmonary vasoconstriction in the dog. Anesthesiology 60: 562–566, 1984.
15. Dorrington KL, Clar C, Young JD, Jonas M, Tansley JG, and Robbins PA. Time course of the human pulmonary vascular response to 8 hours of isocapnic hypoxia. Am J Physiol Heart Circ Physiol 273: H1126–H1134, 1997.
16. Dorrington KL and Talbot NP. Human pulmonary vascular responses to hypoxia and hypercapnia. Pflügers Arch 449: 1–15, 2004.
17. Evans AM and Dipp M. Hypoxic pulmonary vasoconstriction: cyclic adenosine diphosphate-ribose, smooth muscle Ca²⁺ stores and the endothelium. Respir Physiol Neurobiol 132: 3–15, 2002.
18. Ghofrani HA, Reichenberger F, Kohstall MG, Mrosek EH, Seeger T, Olschewski H, Seeger W, and Grimminger F. Sildenafil increased exercise capacity during hypoxia at low altitudes and at Mount Everest base camp: a randomized, double-blind, placebo-controlled crossover trial. Ann Intern Med 141: 169–177, 2004.
19. Groves BM, Reeves JT, Sutton JR, Wagner PD, Cymerman A, Malconian MK, Rock PB, Young PM, and Houston CS. Operation Everest II: elevated high-altitude pulmonary resistance unresponsive to oxygen. J Appl Physiol 63: 521–530, 1987.
20. Grünig E, Mereles D, Hildebrandt W, Swenson ER, Kübler W, Kuecherer H, and Bärtsch P. Stress Doppler echocardiography for identification of susceptibility to high altitude pulmonary edema. J Am Coll Cardiol 35: 980–987, 2000.
21. Howard LS and Robbins PA. Alterations in respiratory control during 8 h of isocapnic and poikilocapnic hypoxia in humans. J Appl Physiol 78: 1098–1107, 1995.
22. Jobic Y, Slama M, Tribouilloy C, Lan Cheong Wah L, Choquet D, Boschat J, Penther P, and Lesbre JP. Doppler echocardiographic evaluation of valve regurgitation in healthy volunteers. Br Heart J 69: 109–113, 1993.
23. Kazemi H, Bruecke PE, and Parsons EF. Role of the autonomic nervous system in the hypoxic response of the pulmonary vascular bed. Respir Physiol 15: 245–254, 1972.
24. Krebs MO, Boemke W, Simon S, Wenz M, and Kaczmarczyk G. Acute hypoxic pulmonary vasoconstriction in conscious dogs decreases renin and is unaffected by losartan. J Appl Physiol 86: 1914–1919, 1999.
25. Lagerlöf H and Werkö L. Studies on the circulation in man. II. Normal values for cardiac output and pressure in the right auricle, right ventricle and pulmonary artery. Acta Physiol Scand 16: 75–82, 1948.
26. Leach RM, Robertson TP, Twort CH, and Ward JP. Hypoxic vasoconstriction in rat pulmonary and mesenteric arteries. Am J Physiol Lung Cell Mol Physiol 266: L223–L231, 1994.
27. Lodato RF, Michael JR, and Murray PA. Multipoint pulmonary vascular pressure-cardiac output plots in conscious dogs. Am J Physiol Heart Circ Physiol 249: H351–H357, 1985.
28. Malik AB and Kidd BS. Time course of pulmonary vascular response to hypoxia in dogs. Am J Physiol 224: 1–6, 1973.
29. Morrell NW, Nijran KS, Biggs T, and Seed WA. Magnitude and time course of acute hypoxic pulmonary vasoconstriction in man. Respir Physiol 100: 271–281, 1995.
30. Ohe M, Ogata M, Katayose D, and Takishima T. Hypoxic contraction of pre-stretched human pulmonary artery. Respir Physiol 87: 105–114, 1992.
31. Peacock AJ, Challenor V, and Sutherland G. Estimation of pulmonary artery pressure by Doppler echocardiography in normal subjects made hypoxic. Respir Med 84: 335–337, 1990.
32. Robbins PA, Swanson GD, and Howson MG. A prediction-correction scheme for forcing alveolar gases along certain time courses. J Appl Physiol 52: 1353–1357, 1982.
33. Robertson TP, Ward JP, and Aaronson PI. Hypoxia induces the release of a pulmonary-selective, Ca²⁺-sensitizing, vasoconstrictor from the perfused rat lung. Cardiovasc Res 50: 145–150, 2001.
34. Robin ED, Theodore J, Burke CM, Oesterle SN, Fowler MB, Jamieson SW, Baldwin JC, Morris AJ, Hunt SA, Vankessel A, Stinson EB, and Shumway NE. Hypoxic pulmonary vasoconstriction persists in the human transplanted lung. Clin Sci (Lond) 72: 283–287, 1987.
35. Savineau JP, Gonzalez de la Fuente P, and Marthan R. Cellular mechanisms of hypoxia-induced contraction in human and rat pulmonary arteries. Respir Physiol 99: 191–198, 1995.
36. Smith RM, Brown TJ, Roach AG, Williams KI, and Woodward B. Evidence for endothelin involvement in the pulmonary vasoconstrictor response to systemic hypoxia in the isolated rat lung. J Pharmacol Exp Ther 283: 419–425, 1997.
37. Vejlstrup NG and Dorrington KL. Intense slow hypoxic pulmonary vasoconstriction in gas-filled and liquid-filled lungs: an in vivo study in the rabbit. Acta Physiol Scand 148: 305–313, 1993.
38. Waypa GB and Schumacker PT. O₂ sensing in hypoxic pulmonary vasoconstriction: the mitochondrial door re-opens. Respir Physiol Neurobiol 132: 81–91, 2002.
39. Weissmann N, Akkayagil E, Quanz K, Schermuly RT, Ghofrani HA, Fink L, Hanze J, Rose F, Seeger W, and Grimminger F. Basic features of hypoxic pulmonary vasoconstriction in mice. Respir Physiol Neurobiol 139: 191–202, 2004.
40. Weissmann N, Grimminger F, Walmrath D, and Seeger W. Hypoxic vasoconstriction in buffer-perfused rabbit lungs. Respir Physiol 100: 159–169, 1995.
41. Wiener CM and Sylvester JT. Effects of glucose on hypoxic vasoconstriction in isolated ferret lungs. J Appl Physiol 70: 439–446, 1991.
42. Yan SF, Lu J, Xu L, Zou YS, Tongers J, Kisiel W, Mackman N, Pinsky DJ, and Stern DM. Pulmonary expression of early growth response-1: biphasic time course and effect of oxygen concentration. J Appl Physiol 88: 2303–2309, 2000.
43. Yock PG and Popp RL. Noninvasive estimation of right ventricular systolic pressure by Doppler ultrasound in patients with tricuspid regurgitation. Circulation 70: 657–662, 1984.
44. Yoshida K, Yoshikawa J, Shakudo M, Akasaka T, Jyo Y, Takao S, Shiratori K, Koizumi K, Okumachi F, and Kato H. Color Doppler evaluation of valvular regurgitation in normal subjects. Circulation 78: 840–847, 1988.
45. Yu AY, Frid MG, Shimoda LA, Wiener CM, Stenmark K, and Semenza GL. Temporal, spatial, and oxygen-regulated expression of hypoxia-inducible factor-1 in the lung. Am J Physiol Lung Cell Mol Physiol 275: L818–L826, 1998.
46. Yu AY, Shimoda LA, Iyer NV, Huso DL, Sun X, McWilliams R, Beaty T, Sham JS, Wiener CM, Sylvester JT, and Semenza GL. Impaired physiological responses to chronic hypoxia in mice partially deficient for hypoxia-inducible factor 1. J Clin Invest 103: 691–696, 1999.

This article has been cited by other articles:

				C. D. Steinback and M. J. Poulin Cardiovascular and cerebrovascular responses to acute isocapnic and poikilocapnic hypoxia in humans J Appl Physiol, February 1, 2008; 104(2): 482 - 489. [Abstract] [Full Text] [PDF]