Effects of hyperoxia on ventilation and pulmonary hemodynamics during immersed prone exercise at 4.7 ATA: possible implications for immersion pulmonary edema.

Immersion pulmonary edema (IPE) can occur in otherwise healthy swimmers and divers, likely because of stress failure of pulmonary capillaries secondary to increased pulmonary vascular pressures. Prior studies have revealed progressive increase in ventilation [minute ventilation (Ve)] during prolonged immersed exercise. We hypothesized that this increase occurs because of development of metabolic acidosis with concomitant rise in mean pulmonary artery pressure (MPAP) and that hyperoxia attenuates this increase. Ten subjects were studied at rest and during 16 min of exercise submersed at 1 atm absolute (ATA) breathing air and at 4.7 ATA in normoxia and hyperoxia [inspired P(O(2)) (Pi(O(2))) 1.75 ATA]. Ve increased from early (E, 6th minute) to late (L, 16th minute) exercise at 1 ATA (64.1 +/- 8.6 to 71.7 +/- 10.9 l/min BTPS; P < 0.001), with no change in arterial pH or Pco(2). MPAP decreased from E to L at 1 ATA (26.7 +/- 5.8 to 22.7 +/- 5.2 mmHg; P = 0.003). Ve and MPAP did not change from E to L at 4.7 ATA. Hyperoxia reduced Ve (62.6 +/- 10.5 to 53.1 +/- 6.1 l/min BTPS; P < 0.0001) and MPAP (29.7 +/- 7.4 to 25.1 +/- 5.7 mmHg, P = 0.002). Variability in MPAP among subjects was wide (range 14.1-42.1 mmHg during surface and depth exercise). Alveolar-arterial Po(2) difference increased from E to L in normoxia, consistent with increased lung water. We conclude that increased Ve at 1 ATA is not due to acidosis and is more consistent with respiratory muscle fatigue and that progressive pulmonary vascular hypertension does not occur during prolonged immersed exercise. Wide variation in MPAP among healthy subjects is consistent with variable individual susceptibility to IPE.

Oxygen seizure latency and peroxynitrite formation in mice lacking neuronal or endothelial nitric oxide synthases.

Nitric oxide (NO) from endothelial or neuronal NO synthases (eNOS or nNOS) may contribute both to the cerebrovascular responses to oxygen and potentially to the peroxynitrite-mediated toxic effects of hyperbaric oxygen (HBO(2)) on the central nervous system (CNS O(2) toxicity). In mice lacking eNOS or nNOS (-/-), regional cerebral blood flow (rCBF) and 3-nitrotyrosine (3-NT), a biochemical marker for peroxynitrite (ONOO(-)) formation, were measured in the brain during HBO(2) exposure. These variables were then correlated with EEG spiking activity related to CNS O(2) toxicity. In wild-type (WT) mice, HBO(2) exposure transiently reduced rCBF, but by 60 min rCBF was restored to baseline levels and above, followed by EEG spikes. Mice lacking nNOS also showed initial depression of rCBF followed by hyperemia but the delay in the onset of EEG discharges was greater. In contrast, in eNOS-deficient mice rCBF did not decrease and hyperemia was less pronounced during HBO(2). EEG spike latency was longer in eNOS(-/-) compared to WT or nNOS(-/-) mice. 3-NT gradually increased in all strains during HBO(2) but accumulation was slower in nNOS(-/-) mice, consistent with less ONOO(-) production. These results indicate that NOS-deficient mice have different cerebrovascular responses and tolerance to HBO(2) depending on which enzyme isoform is affected. The data suggest a key role for eNOS-dependent NO production in cerebral vasoconstriction and in the development of hyperoxic hyperemia preceding O(2) seizures, whereas neuronal NO may mediate toxic effects of HBO(2) mainly by its reaction with superoxide to generate the stronger oxidant, peroxynitrite.