Pro: Hypoxic Pulmonary Vasoconstriction Is a Limiting Factor of Exercise at High Altitude

Hypoxic pulmonary vasoconstriction may limit aerobic exercise capacity at high altitude by two different mechanisms, which both affect the convectional transport of oxygen to the exercising muscles. The first is a decrease in arterial oxygen content because of an alteration in pulmonary gas exchange. The second is a decrease in maximum cardiac output because of an excessive right ventricular afterload.

The function of hypoxic vasoconstriction is to improve gas exchange of inhomogeneous lungs by redirecting blood flow to the better oxygenated lung regions (Brimioulle et al, 1996). A hypoxic pressor response in homogeneous lungs would simply increase pulmonary artery pressure without redistribution of flow. However, normal lungs are not homogeneous, and gravity imposes a vertical gradient of perfusion that is only partially corrected by active PO₂-dependent tone in the most dependent lung regions. Low O₂ breathing increases mean pulmonary artery pressure (Ppa_m), which further counteracts the effects of gravity and makes the distribution of perfusion more homogeneous with no significant change in the distribution of ventilation (Dawson, 1968, Dawson, 1972). This improves lung diffusing capacity (Dehnert et al, 2010) and ventilation/perfusion (V_A/Q) matching (Mélot et al, 1987; Sylvester et al, 1981). The impact of it on arterial blood gases is difficult to detect as the normal alveolar-arterial PO2 gradient (A-aPO2) is already very small and decreases with inspiratory PO₂ because of the shape of the hemoglobin disscociation curve ((Farhi and Rahn, 1955). The improvement in V_A/Q matching induced by global hypoxic vasoconstriction can be demonstrated by the multiple inert gas elimination method (MIGET), but only if either cardiac output remains naturally unchanged (Sylvester et al, 1981) or is normalized by mathematical manipulation (Mélot et al, 1987).

However, exercise increases A-aPO₂, and this is more pronounced in hypoxia than in normoxia (Lilienthal et al, 1946). Hypoxic exercise is associated with a deterioration in V_A/Q matching, as shown by measurements of nitrogen gradients (Haab et al, 1969) or MIGET-derived V_A/Q distributions (Hammond et al, 1986; Wagner et al, 1987). Whether hypoxia is normobaric or hypobaric is irrelevant (Hammond et al, 1986). There may also be a contribution of diffusion limitation at the highest levels of exercise in normoxia, and at lower levels of exercise or even at rest in hypoxia (Farhi and Rahn, 1955; Hammond et al, 1986; Wagner et al, 1987).

Why does hypoxic vasoconstriction not prevent exercise-induced deterioration in V_A/Q matching? A first answer is that hypoxic vasoconstriction is only moderately efficient to redirect blood flow and therefore easily overwhelmed by an increase in cardiac output (Mélot et al, 1987). A second answer is that hypoxic vasoconstriction, together with increased pulmonary blood flow, may be associated with increased pulmonary capillary pressure (Pcp), resulting in interstitial lung edema. The recovered patterns of V_A/Q distributions with increased A-aPO₂ during and after hypoxic exercise are consistent with this explanation (Eldridge et al, 2006; Hammond et al, 1986; Wagner et al, 1987). Exercise studies at high altitudes have shown that the dispersion of blood flow as an index of V_A/Q in homogeneity was correlated to Ppa_m, not to cardiac output (Wagner et al, 1987), confirming vascular pressure as a main determinant of deteriorated gas exchange. Bronchoalveolar lavage fluid sampled after bouts of strenuous hypoxic exercise contains red blood cells and protein, suggesting ongoing capillary leak (Eldridge et al, 2006).

The hypoxic pressor response varies from one subject to another, with no detectable increase in Ppa_m in up to 20% of normal subjects, but a considerable increase in Ppa_m up to 40 mmHg and more, in another 1%–5% of them (Hultgren et al, 1971; Naeije et al, 1982; Naeije et al, 1993; Maggiorini et al, 2001). Strong responders to hypoxia are prone to the development of high altitude pulmonary edema (HAPE) (Bartsch et al, 2005). The longitudinal distribution of resistances in the pulmonary circulation is not affected by hypoxia, with 60% of pulmonary vascular resistance upstream to the capillary–venous compartment (Maggiorini et al, 2001). Accordingly, effective pulmonary capillary pressure increases along with Ppa_m as predicted by the Gaar equation, which states that Pcp=0.6×(Ppa_m - Pla) + Pla, where Pla is left atrial pressure (Gaar et al, 1967). Very high pulmonary blood flow at exercise may decrease the arterial segment of pulmonary vascular resistance (Younes et al, 1987), so that Pcp increases more than predicted by the Gaar equation at a given flow-dependent increase in Ppa. Intense hypoxic vasoconstriction at rest may be associated with Pcp > 20 mmHg and accumulation of extravascular lung water in HAPE-susceptible subjects (Maggiorini et al, 2001). Hypoxic vasoconstriction combined with exercise is associated with a steep flow-dependent increase in Ppa_m which may be accompanied by an increase in Pla (Reeves et al, 1987) and thus a brisk increase in Pcp. The incidence of HAPE is increased by the combination of hypoxic exposure and exercise (Bartsch et al, 2005). An additional aggravating factor is the inhomogeneous nature of hypoxic vasoconstriction, which further increases regional capillary stress (Burnham et al, 2009; Hopkins et al, 2005) and may be a cause of rupture or “stress failure” of the pulmonary capillaries (West, 2004). These observations explain the high protein and red blood cell content of bronchoalveolar lavage fluid of resting subjects with HAPE (Swenson et al, 2002), as reported in athletes after strenuous hypoxic exercise (Eldridge et al, 2006).

Thus, hypoxic pulmonary vasoconstriction during exercise at high altitude does not limit hypoxemia, even when lungs become markedly inhomogeneous, but rather contributes to a worsening of hypoxemia by an alteration of V_A/Q matching that is the consequence of interstitial and eventual alveolar edema.

More speculative is the recently introduced notion that pulmonary artery pressure might limit maximal cardiac output because of an excessive right ventricular afterload. Pulmonary vascular resistance is normally low and decreases moderately at exercise. But with an average shallow slope of Ppa_m-Q relationships of approximately 1 mmHg/L/min in young adults, increasing to an average of 2.5 mmHg/L/min over several decades of life, Ppa_m may double or even triple at very high cardiac outputs (Reeves et al, 1989). This represents a considerable increase in afterload for the thin-walled flow-generating right ventricle (Haddad et al, 2008). A progressive increase in Ppa_m in pulmonary vascular diseases allows for right ventricular hypertrophy and adaptation of systolic function to afterload. Incomplete right ventriculo-arterial coupling is a cause of dyspnea–fatigue symptoms triggered by exercise in patients with pulmonary hypertension (Haddad et al, 2008). Repetitive exercise induces cardiac remodeling which includes the right ventricle, and it is generally believed that there is no right ventricular limitation to aerobic exercise capacity (La Gerche, 2011). However, the individual pulmonary vascular responses to exercise vary greatly, so that some healthy athletes may present with Ppa_m up to 30–40 mmHg or more at the highest levels of exercise or cardiac output (Argiento et al, 2010, Bidart et al, 2007). One of the mechanisms to limit this increase in Ppa_m is pulmonary vascular distension. The distensibility α of the pulmonary resistive vessels is normally 1.5%–2% increase in diameter per mmHg, but decreases with aging and/or hypoxia (Reeves et al, 2005), and may also vary from one subject to another (Argiento et al, 2010). These observations have fueled the hypothesis that highly performant athletes in normoxia need a phenotype of high “pulmonary vascular reserve” to accommodate very high pulmonary blood flows without excessive pulmonary hypertension. Those with steep Ppa_m-Q relationships and low α would be more limited in exercise capacity, and in the long-term exposed to a risk of arrythmogenic right ventricular cardiomyopathy (La Gerche, 2010; La Gerche, 2011)

Hypoxia is associated with steeper slopes of Ppa_m-Q relationships and decreased pulmonary vascular distensibility (Reeves et al, 2005), so that excessive right ventricular afterload is even more likely to occur during hypoxic exercise. Maximal cardiac output is decreased in hypoxia, but whether this is cause of consequence of decreased VO_2max remains unclear (Wagner, 2000). One of the possible explanations for decreased maximal cardiac output at high altitudes is that the matching of convectional and diffusional O₂ transport mechanisms occurs at a lower cardiac output in hypoxia (Wagner et al, 2000). This would lessen the concern about a hypoxic vasoconstriction as a cause of decreased maximum right ventricular flow output. However, pulmonary vasodilating drugs have been reported to improve hypoxic exercise capacity. The first of these studies were based on the administration of the phosphodiesterase-5 inhibitor sildenafil (Faoro et al, 2007; Ghofrani et al, 2004; Hsu et al, 2006). Yet sildenafil intake was also associated with improvement in arterial oxygenation, so that an alternative explanation for the obtained results could be an improvement in pulmonary gas exchange increasing arterial O₂ content (Faoro et al, 2007). Similar concomitant improvements in arterial oxygenation and decreased pulmonary vascular resistance were reported in HAPE-susceptible subjects exposed to high altitude after preventive administration of dexamethasone (Fischler et al, 2009). The hypothesis that excessive RV afterload might limit exercise capacity could at last be tested by the administration of endothelin receptor blockers, which decreased pulmonary vascular resistance without effect on arterial oxygenation in hypoxic healthy volunteers (Faoro et al, 2009; Naeije et al, 2010). In these studies, isolated decreases in pulmonary vascular resistance by an average of 50% were correlated to an increase in VO_2max by 10%–25%. It is of course understood that correlation is not proof of causality. The notion of hypoxic vasoconstriction to limit right ventricular flow output and thereby limit convectional O₂ transport to exercising muscles needs further experimental support, with direct measurements of maximal cardiac output, however technically difficult this may be, and evaluations of right ventricular function at exercise.

In summary, hypoxic pulmonary vasoconstriction limits exercise capacity by a paradoxical deterioration in pulmonary gas exchange and, possibly, by a limitation of maximal cardiac output. Both effects decrease VO_2max through a decrease in convectional O₂ transport to the exercising muscles.

Almost 2 decades ago, exploring the possibility of noninvasive measurements of hypoxic vasoconstriction to predict tolerance to high altitudes, we were surprised to be unable to detect any changes in pulmonary vascular resistance with low oxygen breathing in athletes who had survived previous climbs at altitudes higher than 8000 m (Vachiery et al, 1993). We now understand that successful high altitude climbs require the lowest levels possible of hypoxic vasoconstriction to preserve gas exchange and cardiac function. Pulmonary vascular reserve is probably essential to extreme athletic achievements.