INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

People now climb to the summit of Mount Everest (8,848 m) without supplementary oxygen, although inspired partial pressure of oxygen, near 40 mm Hg, suggests severe hypoxemia. In such conditions, human exercise capacity is limited, presumably owing to the effects of hypoxia but also to some environmental factors such as ambient temperature, humidity, stress, or the degree of activity. Experiments in a hypobaric chamber permit assessment of the role of hypobaric hypoxia alone. Operation Everest III (Comex '97) is the third simulated climb of Mount Everest in a hypobaric chamber. Previous operations Everest I and Everest II were conducted in the United States in 1946 and 1985. During those experiments, special attention was given to circulatory and cardiac functions because of their major role in physiological adjustments to both hypoxia and effort and their contribution to the decreased exercise capacity at high altitude. During Operation Everest I, heart size was determined in four volunteers using serial roentgenograms and a slight decrease was reported (1). During Everest II, left ventricle (LV) systolic function and pulmonary artery pressure (Ppa) were assessed in seven subjects, using both cardiac catheterization (2) and two-dimensional echocardiography (3) with the results of a preservation of LV systolic function as well as an increase in Ppa. During the present Operation Everest III (Comex '97), we performed a complete hemodynamic study including systolic and diastolic LV function and Ppa measurement using a totally noninvasive technique. We used a combination of M-mode echocardiography and two- dimensional (2D) echocardiography with continuous and pulsed Doppler. Indeed, echocardiography allows the measurement of cardiac diameters and LV systolic function; Doppler permits the measurement of systolic Ppa and cardiac output as well as an estimation of LV filling and LV pressure.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Operation Everest III (Comex `97)

Operation Everest III (Comex '97) took place in the COMEX S.A hyperbaric center, Marseilles, France, from March 10 to May 6, 1997. Eight subjects participated in the study. They were eight male volunteers age between 23 and 37 yr, mean weight 74.3 ± 6.6 kg, mean height 180 ± 6 cm, mean body mass index 22.9 ± 1.5 kg/m². All of them had mountain climbing experience near 4,000 m, four had experience near 6,000 m, one previously reached an altitude of 8,760 m without supplementary oxygen. The hypobaric chamber consisted of a large cylindrical living chamber (length 8 m; 32 m³) which was connected to a spherical study chamber (diameter 5 m; 65 m³) where all experiments were done. Relative humidity was maintained between 30 and 60% and the temperature between 18 and 24° C. Preclimb tests were performed at sea level (SL) (March 10 to 16). Subjects were then brought up to the Observatoire Vallot, Mont Blanc, 4,350 m, 440 mm Hg, for a preacclimatization period of 7 d (March 25 to 31). Subjects entered in the hypobaric chamber on April 1 (Day 1), 24 h after being brought back from Mount Blanc to Marseilles. A 31-d ascent profile was followed (Figure 1), simulating ascent of Mount Everest. The preacclimatization period allowed a rapid entrance to the chamber at 4,500 m. The pressures were chosen as the real pressures for the latitude of Mount Everest up to 253 mm Hg for the altitude of 8,848 m (4). Breathing gas in the hypobaric chamber was air. All subjects reached altitude of 8,000 m (284 mm Hg), on April 26 (Day 26). Seven subjects were decompressed to the altitude 8,848 m (253 mm Hg) during the day on April 30 and May 1, but recompressed at night to a lower altitude. Subjects left the hypobaric chamber on May 2 (Day 32) after a 24-h recompression. During the hypobaric stay in the chamber, subjects were asked to have a moderate activity during the day on a treadmill or on a cycloergometer and to participate in all experiments.

View larger version (25K): [in this window] [in a new window]   Figure 1.   Decompression profile.

The ethics committee of the university hospital of Marseilles approved the protocol. Written informed consent was obtained from all subjects.

Echographic and Doppler Study

Technical considerations. The schedule of the study permitted 5 echographic observations of the eight subjects as follows: reference examination before any decompression, three examinations during the decompression corresponding to altitudes of 5,000 m (422 mm Hg; Days 4 and 5), 7,000 m (324 mm Hg; Day 17), 8,000 m (284 mm Hg; Day 26), and one examination 2 d after recompression (Day 33). The echographic-Doppler examinations were performed using a Diasonics Vingmed CFM 750 type ultrasonography machine (GE Ultrasound, Courtaboeuf, France). The ultrasonography unit was located outside the hypobaric chamber and connected to a combined 3.25-MHz imaging/2.5-MHz Doppler probe inside the chamber, via a lead-through into the hull. The explorations were carried out by an investigator (A.B.), decompressed to the pressure of the subjects within the chamber. In an effort to prevent altitude decompression sickness, he prebreathed 100% oxygen for 1 h before the decompression. Oxygen breathing was continued throughout the stay in the hypobaric chamber under a breathing cask to prevent adverse effects of acute hypoxia (Figure 2). This inside investigator was a cardiologist with considerable experience in echocardiography. Subjects were in the left lateral position, on an examination bed, located below a porthole. The inside investigator attempted echocardiography-Doppler examination with visual control on the outside monitor through the porthole. Indeed, a second investigator (F.M.) stayed outside the hypobaric chamber and positioned the monitor in such a way that the inside investigator could see it. The two investigators communicated via a microphone. The outside investigator also processed the echocardiogram using the Diasonics Vingmed computer system but also recorded the examinations on standard VHS videotape, to be reviewed later. The mean duration for each test was 20 min. Doppler recordings were performed at the end of normal expiration to eliminate the effects of respiration on the parameters studied. Sphygmomanometric blood pressure measurements on the right arm were obtained before each examination.

View larger version (135K): [in this window] [in a new window]   Figure 2.   The inside investigator is performing an echographic- Doppler examination. He is breathing 100% oxygen under an oxygen cask. He visually controls the position of the transducer on the monitor outside through the porthole, above the examination bed.

Echocardiographic parameters. LV systolic function. Left atrial (LA) diameter, left ventricle (LV) end-systolic and end-diastolic diameters (LVEDD, LVESD), right ventricle (RV) end-systolic and end-diastolic diameters (RVESD, RVEDD), and aortic diameter (Ao) were measured by M-mode echocardiography from the left parasternal short-axis view. Standard indices of global LV systolic performance were ejection fraction (EF) and LV percent fractional shortening (%FS). LV percent fractional shortening was taken as the ratio (LVEDD  LVESD)/LVEDD. LV ejection fraction was carried out using the Teicholz formula. Heart rate (HR) was recorded by echocardiogram and rate was averaged over 60 s. To measure , aortic cross-section area (ACSA) was first measured by 2D echocardiography from the left parasternal short-axis view at the level of the aortic root. The aortic systolic flow velocity time integral (VTI) was measured by computer-assisted determination from the pulsed-wave Doppler profile of aortic blood flow from the apical four-chamber view, allowing stroke volume (SV = Ao VTI × ACSA) and to be calculated ( = SV × HR).

Statistical Analysis

Data are expressed as mean ± SEM. Statistical tests were run on SAS software (Cary, NC). The cohorts for comparison consisted of the eight subjects at SL, 5,000 m, 7,000 m, 8,000 m and after the return at sea level (RSL). Comparison between cohorts of continuous variables having a normal distribution was carried out with parametric analysis of variance (repeated measures ANOVA); comparison of dichotomous variables was done with Tukey's test. In the case of cohorts of variables not having a normal distribution, comparisons were performed with nonparametric multivariate analysis (Friedman's test); comparison of dichotomous variables was done with Dunn's test. A p  0.05 was considered significant.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The modifications observed on cardiac diameters are reported in Table 1. On simulated ascent to altitude, LA and LVEDD decreased regularly. Ao was reduced at all altitudes but reduction reached statistical significance only at 7,000 m. LVESD and RVESD were reduced at 5,000 m (p < 0.05). RVEDD did not show any change. RV/LV end-diastolic diameters ratio (RV/LV) increased by 20% at 8,000 m, without reaching significance (p = 0.058). HR on all altitudes was significantly higher than at SL (Table 2); HR at 8,000 m was 41% higher than at SL. Mean values for systolic and diastolic blood pressure remained unchanged during the whole decompression (Table 2). A reduction in aortic VTI was observed at all altitudes, corresponding to a significant decrease of SV at 5,000 m. SV reduction was between SL and 5,000 m, without further changes between 5,000 m and 8,000 m. In total, remained unchanged throughout the whole decompression (p = 0.109) (Table 2). As indices of LV contractility, EF and %SF remained the same or slightly increased (not significant [NS]) in all subjects (Table 3). Study of the LV filling patterns was as follows (Tables 4 and 5): the mitral flow showed a decreased mean peak E velocity with a significant difference at 5,000 m and 8,000 m, although mean peak A velocity increased, with a significant difference at 7,000 m, and decreased thereafter, without reaching the SL value. As a result, the mean E/A ratio regularly decreased, from 2.16 at SL to 1.13 at 8,000 m (p < 0.05). Pulmonary venous flow velocities showed a decrease of mean peak D velocity at 5,000 m (p < 0.05) and 8,000 m (p < 0.05), a decrease of mean peak S velocity at 7,000 m (p < 0.05), and a reduction of the D/S ratio (Table 5). The IVRT remained unchanged at all altitudes.

As indices of LV filling pressures, the duration of the mitral A flow did not change, although the duration of the PV atrial reversal flow decreased from 100 ms at SL to 85 ms at 8,000 m (p = 0.096; NS). The duration of the PV atrial reversal flow always remained shorter than the duration of the mitral A wave. The ratio mitral A wave duration/PV atrial reversal flow duration remained unchanged at all altitudes (Table 6). Furthermore, the systolic fraction of pulmonary venous forward flow VTI remained unchanged (p = 0.153). RV/RAg increased from SL to 8,000 m (p < 0.05). As we did not measure right atrial pressure, and as it was not possible to consider it as a constant value throughout the decompression period, we expressed our results as RV/RAg, without calculation of systolic Ppa (Table 6).

RSL

After the RSL, HR decreased and reached the initial values, as well as SV; remained unchanged. All cardiac diameters (LA, LVEDD, LVESD, RVESD, Ao) returned to their initial values. Transmitral and pulmonary venous flow profiles changed back toward the initial pattern. RV/RAg fell but without reaching the initial SL value.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

During this simulated climb of Mount Everest in a hypobaric chamber, we obtained satisfactory and complete echocardiographic and Doppler examinations in all subjects at all altitudes. We used a technique with the echographic machine located outside the hypobaric chamber. We used a transducer extension cord passed through a specific air-lock connection. The transducer was brought into the hypobaric study chamber on the examinations days by an investigator and was the only material that underwent pressure variations. These conditions were first imposed because of the lack of space in the study chamber. This method also does not expose the ultrasonography unit to numerous and substantial variations of pressure, as the study chamber was decompressed and compressed every day to bring scientific investigators to the same simulated altitude imposed on the subjects. We previously used this technique with success during a COMEX saturation dive operation (7).

As usual, in young people, we documented a tricuspid regurgitation flow in all our subjects, from which we obtained a satisfactory measurement of RV/RAg. We observed a progressive and constant increase of RV/RAg, reaching 40.1 mm Hg at 8,000 m, which reflects a significant elevation of the systolic Ppa. However, systolic Ppa could not be calculated in the absence of a direct measurement of RA pressure. These results are concordant with those reported during Operation Everest II, where mean Ppa (), measured at cardiac catheterization, rose from 15 ± 0.9 mm Hg at SL to 33 ± 3 mm Hg at 8,848 m (8). This pulmonary arterial hypertension secondary to hypoxia has been well known for many years. It appears as soon as alveolar oxygen pressure (Pa_O2) drops below 60 mm Hg, corresponding to an altitude of 3,000 m (9). The increase in pulmonary vascular resistance (PVR) has been evaluated to 50% when the Pa_O2 decreases to 50 mm Hg (9). Its determinants are both the alveolar and arterial partial pressure of oxygen (Pa_O2). During EVEREST III, mean Pa_O2 measured at 8,000 m was 37 ± 4.6 mm Hg and 33.6 ± 3.7 mm Hg at 8,848 m (extremes: 29 to 40). The increased levels of PVR are the result of constriction of precapillary arteries (10), associated, when exposure is prolonged, to histological changes (11). After prolonged exposure to high altitude, normal levels of Ppa may only be achieved after residence at low altitude for 6 wk (12). In our study, we observed, 2 d after RSL, a reduction of the RV/RA gradient compared with its value at 8,000 m (NS) but which remained increased compared with the initial value (NS). This suggests that the remodeling of the pulmonary arteries that occurred at high altitude was not resolved after 2 d of reoxygenation. The increased levels of PVR may lead to an increased RV end-diastolic pressure, and, when exposure is prolonged, to RV hypertrophy and dilatation (13, 14). In the present work, the association of an elevation in RV afterload and a reduction in RV preload may explain the lack of modification of RV end-diastolic diameter. However, the slight elevation of the RV/LV ratio, even if not reaching significance, is in favor of a relative increase of RV compared with LV volume.

The second important result in our study is the reduction in LV preload, evidenced by the reduction in LV end-diastolic and end-systolic diameters and by the reduction in LA diameter at all simulated high altitudes. This is associated with a reduction in RV preload, as seen on the reduction of the RV end-diastolic diameter at 5,000 m. This reduction in cardiac filling is consistent with previous observations. During Everest II, echocardiographic examination showed a reduction of LV end-diastolic, end-systolic, and stroke volume (3) associated with decreased cardiac filling pressures, RA pressure, and wedge pressure, measured by cardiac catheterization (2). This reduction in cardiac preload may be partly due to a contraction of plasma volume, as suggested by the following elements: all subjects had progressive weight loss during the decompression, with an average loss of 5.4 kg (extremes 2.9 to 10). All subjects experienced a dramatic decrease of daily diuresis, associated with an increased hemoglobin level up to a mean value of 18 g/ dl at 8,848 m (extremes 16 to 20). Those observations are in accordance with other reports from experiments in the field or in hypobaric chamber. Fowles and Hultgren (15), during a high-altitude expedition reported an increase of mean venous hematocrit from 45% to 50% 48 h after arrival at 3,650 m, suggesting a decrease in plasma volume of 500 to 700 ml. During Everest II, subjects presented an average weight loss of 7.56 kg with an increment in total serum protein concentration (3). When exposure to altitude is prolonged, plasma volume continues to contract (16, 17).

From our observations, we demonstrated that cardiac contractility remained normal during exposure to altitude-induced hypoxia with preservation of LV ejection fraction and LV percent fractional shortening. The same observations were made during acute or subacute hypoxia, in hypobaric chamber during Everest II (2, 3), during a high-altitude expedition (15), or when breathing gas mixture with reduced oxygen content (18, 19). Preservation of LV contractility occurred despite major hypoxemia (20), as mean Pa_O2 at 8,000 m measured in our subjects was 37 mm Hg. In our study, we did not show any change in LV afterload. Mean values for systolic and diastolic blood pressure remained unchanged. Similarly, during Everest II, mean arterial blood pressure () did not increase nor did the systemic vascular resistance (SVR), calculated from data at cardiac catheterization (8). This is the result of a balance between two opposite effects: vasodilatation secondary to profound hypoxia and increase in sympathetic activity (21), as suggested by increased HR. Most investigators (2, 3, 15, 22- 24) have reported this altitude-induced tachycardia, caused by an increase in sympathetic activity, as evidenced by elevated plasma norepinephrine concentrations (25). During the present operation, mean plasma norepinephrine concentration measured at the end of the hypobaric period appeared 400% higher than the initial values (unpublished data). However, as sympathetic tone remains high, HR stabilized with prolonged altitude exposure. The increased HR at high altitude is accompanied, in our subjects, by a decreased SV. Reduction of SV at high altitude has been reported many times before (2, 3, 15, 22). The reduction of SV could be because of reduced cardiac filling or reduced contractility. Our findings imply that reduced ventricular filling may be the cause. This decrease in SV was offset by the increase in HR, such that remained unchanged.

Previous evaluation of LV filling pattern has not been made in humans at extreme altitudes. During Operation Everest III, we studied, as indices of LV filling patterns, transmitral flow velocities, pulmonary venous flow velocities obtained by pulsed Doppler, and measurement of the isovolumetric relaxation time. In normal young adults, LV elastic recoil is vigorous and myocardial relaxation is swift, so most filling is completed during early diastole with only a small contribution of filling during atrial contraction. This results in an elevated transmitral E velocity peak, a smaller transmitral A velocity peak, and an E/A greater than 1, associated with a large pulmonary venous D velocity peak and a venous pulmonary D/S ratio greater than 1 (5). This is the exact diastolic filling pattern we observed in our subjects at SL. With increasing simulated altitude, transmitral diastolic filling patterns were modified, with a gradual decrease in E velocity, a gradual increase in A velocity resulting in a decreased E/A ratio. This was associated with similar changes in PV flow velocities with a reduced pulmonary venous D velocity and a decreased D/S ratio. Those filling patterns demonstrated a contribution of the atrial contraction to LV filling which became more important at altitude. Several factors may be responsible for those modifications. Tachycardia shortened diastolic filling period and atrial contraction may occur before the early filling is completed; the A velocity will then be higher than it would be if the HR were slower. The decreased LV preload, as seen previously, by reducing the pressure gradient from the left atrium to the left ventricle may decrease early filling (5, 26, 27) and prolonged IVRT. Our subjects presented a prolonged IVRT, which did not reach significance (p = 0.35). However, as IVRT is also known to shorten when HR increases, prolongation of IVRT secondary to preload reduction may have been offset, in our subjects, by reduction due to tachycardia. Diastolic ventricular interdependence should also play a role. The elevation of Ppa led to increased RV pressure loads. The increased RV pressure loads restricted early left ventricle filling and prolonged IVRT in relationship with geometric interaction through the interventricular septum, secondary to the decrease of the pressure gradient from RV to LV (28). At a minimum, LV relaxation alteration due to hypoxia itself must be considered. Hypoxia induced impairment in LV relaxation by increasing the diastolic chamber stiffness because of reduced energy supply (31). This phenomenon, reversible when the normal oxygen supply-demand is restored, is caused by an incomplete ventricular relaxation (31, 32). A recent echocardiography-Doppler study in healthy subjects breathing gas mixture with reduced oxygen content (fraction of inspired oxygen [FI_O2 = 0.14]) for a few minutes (acute normobaric hypoxia) evoked a reduced LV diastolic function, although LV systolic function is well preserved (19). Whatever the mechanism of the LV filling pattern modifications, they are not accompanied, in our subjects, by any increase in LV filling pressure. At altitude, PV atrial reversal flow remained always shorter than mitral A duration without any modification of the ratio which always remained greater than 1. This absence of elevation of LV filling pressure is consistent with measurements made during Everest II, where wedge pressure measured by cardiac catheterization was not elevated at altitude (2).

In conclusion, this echocardiographic-Doppler study is the first complete and noninvasive hemodynamic study of subjects at barometric pressure equivalent to 8,000 m. We confirm the progressive and constant increase in pulmonary pressures with progressive and prolonged hypoxia, which persists for some days after RSL. We also confirm the preservation of LV contractility at altitude. is not modified; the decrease in SV, owing to reduced preload, is offset by the increase in HR, secondary to increased sympathetic activity. To our knowledge, this is the first study of LV diastolic filling during prolonged altitude-induced hypoxia. We demonstrated a modified LV filling pattern, with a decreased early filling and a greater contribution of the atrial contraction, without any elevation of LV end- diastolic pressure. The modification of LV filling may be an adaptative response to tachycardia or reduced preload or may be secondary to impairment in LV relaxation caused by ventricular interdependence or hypoxia itself. It is likely that there is not a single mechanism but a combination of many. Further experiments are needed to clarify this point.