Role of cyclooxygenase inhibition and hyperoxia in regulating pulmonary perfusion in dogs.

To assess the roles of cyclooxygenase inhibition and hyperoxia in regulating pulmonary perfusion, we studied 13 dogs with diffuse granulomatous lung disease (DGLD) and 13 normal dogs. Baseline observations were obtained at fractional inspired O2 (FIO2) 0.21 and 1.0 and repeated after infusion of meclofenamate (Mec; n = 8) or saline (n = 5). Resistance to flow was evaluated from the pulmonary end-diastolic gradient (PDG) and by ohmic pulmonary vascular resistance (PVR). Distribution of blood flow was evaluated with sulfur hexafluoride in DGLD and with multiple inert gas alveolar ventilation-perfusion (VA/Q) plots in normal dogs. Before infusion, there were no differences between the saline and Mec groups at either FIO2. Saline induced no significant changes at either FIO2. After Mec in DGLD, PDG at FIO2 0.21 rose from 4 +/- 2 to 6 +/- 4 mmHg (P < 0.04), PVR increased from 297 +/- 98 to 484 +/- 181 dyn.s.cm-5.m-2 (P < 0.01), whereas shunt flow (Qs/Qt) fell form 13.6 +/- 12.0 to 6.2 +/- 5.3% (P < 0.03). At FIO2 1.0 PDG rose from 3 +/- 2 to 4 +/- 3 mmHg (P < 0.02), PVR increased from 262 +/- 78 to 374 +/- 139 dyn.s.cm-5.m-2 (P < 0.01), whereas Qs/Qt fell from 14.5 +/- 13.3 to 6.4 +/- 5.2% (P < 0.02). After Mec in normal dogs, PDG at FIO2 0.21 rose from 3 +/- 1 to 4 +/- 1 mmHg (P < 0.015) and PVR increased from 256 +/- 92 to 340 +/- 101 dyn.s.cm-5.m-2 (P < 0.05); at FIO2 1.0 PDG and PVR were unchanged from preinfusion levels. In normal dogs, no parameters of VA/Q changed significantly with hyperoxia or Mec. These data suggest that perivascular inflammation enhances perfusion in DGLD by elaboration of vasodilator prostaglandins (PG). By inhibiting PG synthesis, Mec selectively increases resistance in diseased lung at FIO2 0.21 and lowers Qs/Qt. In contrast, there was vasoconstriction without flow redistribution in normal dogs, suggesting that vasodilator PGs contribute to the low tone in the normal pulmonary bed. The vasodilation without flow redistribution in both models during hyperoxia after Mec suggests an effect of O2 that is related neither to PG synthesis nor to hypoxic vasoconstriction.