Effect of Normobaric Hypoxia on Exercise Performance in Pulmonary Hypertension Randomized Trial

BACKGROUND: Many patients with pulmonary arterial or chronic thromboembolic pulmonary hypertension (PH) wish to travel to altitude or by airplane, but their risk of hypoxia-related adverse health effects is insufficiently explored. RESEARCH QUESTION: How does hypoxia, compared with normoxia, affect constant work-rate exercise test (CWRET) time in patients with PH, and which physiologic mechanisms are involved? STUDY DESIGN AND METHODS: Stable patients with PH with resting PaO2 >= 7.3 kPa underwent symptom-limited cycling CWRET (60% of maximal workload) while breathing normobaric hypoxic air (hypoxia; FIO2, 15%) and ambient air (normoxia; FIO2, 21%) in a randomized cross-over design. Borg dyspnea score, arterial blood gases, tricuspid regurgitation pressure gradient, and mean pulmonary artery pressure/cardiac output ratio (mean PAP/CO) by echocardiography were assessed before and during end-CWRET. RESULTS: Twenty-eight patients (13 women) were included: median (quartiles) age, 66 (54; 74) years; mean pulmonary artery pressure, 41 (29; 49) mm Hg; and pulmonary vascular resistance, 5.4 (4; 8) Wood units. Under normoxia and hypoxia, CWRET times were 16.9 (8.0; 30.0) and 6.7 (5.5; 27.3) min, respectively, with a median difference (95% CI) of -0.7 (-3.1 to 0.0) min corresponding to -7 (-32 to 0.0)% (P = .006). At end-exercise in normoxia and hypoxia, respectively, median values and differences in corresponding variables were as follows: PaO2: 8.0 vs 6.4, -1.7 (-2.7 to -1.1) kPa; arterial oxygen content: 19.2 vs 17.2, -1.7 (-3 to -0.1) mL/dL; PaCO2: 4.7 vs 4.3, -0.3 (-0.5 to -0.1) kPa; lactate: 3.7 vs 3.7, 0.9 (0.1 to 1.6) mM (P <.05 all differences). Values for Borg scale score: 7 vs 6, 0.5 (0 to 1); tricuspid pressure gradient: 89 vs 77, - 3 (-9 to 16) mm Hg; and mean PAP/CO: 4.5 vs 3.3, 0.3 (-0.8 to 1.4) Wood units remained unchanged. In multivariable regression, baseline pulmonary vascular resistance was the sole predictor of hypoxia-induced change in CWRET time. INTERPRETATION: In patients with PH, short-time exposure to hypoxia was well tolerated but reduced CWRET time compared with normoxia in association with hypoxemia, lactacidemia, and hypocapnia. Because pulmonary hemodynamics and dyspnea at end-exercise remained unaltered, the hypoxia-induced exercise limitation may be due to a reduced oxygen delivery causing peripheral tissue hypoxia, augmented lactic acid loading and hyperventilation.