We infused Escherichia coli endotoxin, 0.07-1.33 microgram/kg, intravenously into chronically instrumented unanesthetized sheep and measured pulmonary arterial and left atrial pressures, lung lymph flow, lymph and blood plasma protein concentrations, and arterial blood gases. Endotoxin caused a biphasic reaction: an early phase of pulmonary hypertension and a long late phase of steady state increased pulmonary vascular permeability during which pulmonary arterial and left atrial pressures were not increased significantly and lung lymph flow was 5 times the baseline value. Lymph: plasma total protein concentration ratio during the late phase (0.76 +/- 0.04) was significantly (P less than 0.05) higher than during baseline (0.66 +/- 0.03). The lymph response was reproducible. Lung lymph clearance of endogenous proteins with molecular radii (r) 35.5 to 96 A was increased during the steady state late phase of the reaction, but, as during baseline, clearance decreased as r increased. The endotoxin reaction was similar to the reaction to infusing whole Pseudomonas bacteria, except that endotoxin had less effect on pressures during the steady state response and caused a relatively larger increase in lymph clearance of large proteins. We conclude that E. coli endotoxin in sheep causes a long period of increased lung vascular permeability and may have a greater effect on large solute pathways across microvessels than do Pseudomonas bacteria.

A B S T R A C T To see whether methylprednisolone would affect the pulmonary vascular response to endotoxemia, we studied responses to endotoxemia in the presence and absence of methylprednisolone in the same chronically instrumented, unanesthetized sheep. Infusion of Escherichia coli endotoxin (0.70-1.33,g/ kg) caused an initial period of marked pulmonary hypertension followed several hours later by a long period of increased vascular permeability when pulmonary vascutlar presstures were niear base linie (baseline pulmonary artery pressure (PPa) = 21+1 cm H2O SE, left atrial pressure (Pla) = 1+3; experimental PPa = 20±3, Pla = 3±4; P = NS), lung lymph flow (Qlym) was high (base-line Qlym = 7.2±0.2 ml/h; experimental Olym = 23.2+1.0; P < 0.05) and lymph/ plasma protein conceintration (L/P) was high (base-line L/P = 0.65±0.04; experimental L/P = 0.79±0.05; P < 0.05). When methylprednisolone (1.0 g + 0.5 g/h i.v.) was begtuni 30 min before the same dose of endotoxin was infused, the initial ptilmonary hypertension was less and the late phase increase in lutng vascutlar permeability was prevented (experimental PPa = 24± 1, Pla = 1± 1, Olym = 10.0 ± 0.4; L/P = 0.56 ± 0.03). Olym and L/P were significantly (P < 0.05) lower than with endotoxin alone. Methylprednisolone began duiring the initial putlmonary hypertensive response to endotoxin also prevented the late phase increase inl lutiig vasecular permeability, but the druLg had no effect once vascutlar permeability was increased. We conclude that large doses of methylprednisolone given before or sooIn after endotoxemia prevent the increase in luing vasctular permeability that endotoxin cauises, but do not reverse the abnormality once it occurs.

A B S T R A C T Pulmonary diffusing capacity and arterial blood Po2 decrease in humans when 10% fat emulsion is infused. To study its effects on the pulmonary circulation and lung fluid balance, we infused 0.25 g/kg x h of a 10% fat emulsion (Intralipid, Cutter Laboratories, Inc., Berkeley, Calif.) into an awake sheep lung lymph preparation. The emulsion caused a sustained increase in pulmonary artery pressure to approximately twice base line with little change in left atrial pressure. Pa02 decreased an average 13 torr and lung lymph flow increased two-to threefold. Lymph/ plasma total protein concentration fell as lymph flow increased; the magnitude of the lymph/plasma protein decrease was similar to that reported previously when lung vascular pressures were mechanically elevated. Heparin infusion (loading dose = 4,000 U, maintenance dose = 2,000 U/h) cleared the serum of triglycerides but did not alter the response to fat emulsion. Indomethacin infusion (loading dose = 5 mg/kg, maintenance dose = 3 mg/kg x h) blocked the rise in pulmonary artery pressure, the increase in lung lymph flow, and the fall in Pao0. Neither extravascular lung water nor [14C]urea lung vascular permeability surface area products were altered by fat emulsion infusion. We conclude that fat emulsion infusion in sheep increases lung microvascular filtration by increasing vascular pressures, but has no effect on vascular permeability. Since the effects are blocked by indomethacin, they may be prostaglandin mediated.

A B S T R A C T Although prostaglandins E2 and F2, have been suggested as mediators of the pulmonary hypertension seen after endotoxin infusion or during alveolar hypoxia, their precursors, the endoperoxides (prostaglandins G2 and H2) are much more potent vasoconstrictors in vitro. In this study we compared the effects of prostaglandin (PG)H2, a stable 9-methylene ether analogue of PGH2 (PGH2-A), PGE2, and PGF2a on pulmonary hemodynamics in awake sheep. The animals were prepared to allow for measurement of (a) INTRODUCTIONThe agents mediating the pulmonary hypertension seen with alveolar hypoxia and gram-negative endotoxin infusion have not been identified. Because prostaglandins (PG)l E2 and F2, are released from the lungs during these reactions, they have been implicated as the mediators, but neither is an impressive vasoconstrictor in vitro (1, 2). In contrast, their precursors, the endoperoxides PGG2 and PGH2, are powerful vasoconstrictors, -100 times more potent in constricting isolated smooth muscle (3).Because the endoperoxides are unstable and difficult to synthesize in the large amounts needed for in vivo studies, we first used a stable 9-methylene ether analogue of PGH2 (PGH2-A; [15S] hydroxy-lla, 9a-[epoxymethano] prosta-5Z, 13E-dienoic acid, [4]) to assess the effects of endoperoxides on pulmonary vascular pressures, permeability of the lung microcirculation, and lung fluid balance in awake sheep. We then prepared PGH2, itself, in amounts large enough for brief steady-state infusions, and compared its effects to those of the analogue, PGE2 and PGF2a,.

A B S T R A C T To see whether antihistamines could prevent and reverse histamine-induced pulmonary edema and increased lung vascular permeability, we compared the effects of a 4-h intravenous infusion of 4 /Ag/kg per min histamine phosphate on pulmonary hemodynamics, lung lymph flow, lymph and plasma protein content, arterial blood gases, hematocrit, and lung water with the effects of an identical histamine infusion given during an infusion of diphenhydramine or metiamide on the same variables in unanesthetized sheep. Histamine caused lymph flow to increase from 6.0±0.5 to 27.0±5.5 (SEM) ml/h (P <0.05), lymph: plasma globulin concentration ratio to increase from 0.62±0.01 to 0.67±0.02 (P < 0.05), left atrial pressure to fall from 1+1 to -3±1 cm H20 (P < 0.05), and lung lymph clearance of eight protein fractions ranging from 36 to 96 A molecular radius to increase significantly. Histamine also caused increases in lung water, pulmonary vascular resistance, arterial Pco2, pH, and hematocrit, and decreases in cardiac output and arterial Po2. Diphenhydramine (3 mg/kg before histamine followed by 1.5 mg/ kg per h intravenous infusion) completely prevented the histamine effect on hematocrit, lung lymph flow, lymph protein clearance, and lung water content, and reduced histamine effects on arterial blood gases and pH. 6 mg/kg diphenhydramine given at the peak histamine response caused lymph flow and lymph: plasma protein concentration ratios to fall. Metiamide (10 mg/ Description of the preparation We used an unanesthetized, chronic sheep preparation described previously (1-4). Each animal was prepared by a series of three staged thoracotomies during which nonpulmonary contributions to a large lymph node in the posterior mediastinum (caudal mediastinal node) were resected;