Evaluation of the biocompatibility of four different types of oxygenator (bubble, membrane, hollow fibre and 'hybrid') was performed on 26 patients undergoing cardiopulmonary bypass during elective coronary surgery. More platelet derangement and an increased degree of hemolysis, revealed by higher plasmatic concentration of beta-thromboglobulin, platelet factor 4 and plasmatic free hemoglobin (p less than 0.05), was seen when using the bubble oxygenator. Damage to blood cells was minimal with the membrane oxygenator while the 'hybrid' and the hollow fibre oxygenators proved to rank at an intermediate level. Complement activation at the beginning of the cardiopulmonary bypass occurred via the alternative pathway as demonstrated by C3ades arg increase (up to nine times) without a concomitant elevation of C4ades arg. Cardiopulmonary bypass complement activation was quantitatively similar with all the oxygenators. A further activation via the classical pathway occurred in all the patients after protamine injection. Consistent differences as far as clinical and biological effects exist among the various commercially available cardiopulmonary bypass apparatus; our study provides guidelines for the evaluation and selection of devices which might reduce postoperative sequelae.

SummaryIsosorbide dinitrate inhibits platelet function in vivo at concentrations about 10 times lower than in vitro (10-7 -10-6 vs. 10-6 -10-5 M). We investigated two possible reasons for this difference. Isosorbide dinitrate and its in vivo longer-lived metabolites, isosorbide-2- and isosorbide-5-mononitrate were incubated for 5 min with human platelet-rich plasma or washed platelets; irreversible aggregation was induced with threshold doses of ADP, adrenaline, collagen, arachidonic acid and thrombin, and thromboxane (TX) B2 production was measured by radioimmunoassay. Moreover, the concentration of exogenous prostacyclin required to inhibit platelet aggregation by 50% (IC50) after preincubation with isosorbide dinitrate or vehicle was determined. At 10-7 M, only isosorbide-2-mononitrate inhibited aggregation (-12%, p >0.05) and TX production (-36%, p >0.01) by ADP. At 10-6 M isosorbide-2-mononitrate inhibited aggregation by adrenaline more than the dinitrate (-41% vs. -25%, p >0.05). In addition, at supra-threshold doses of all the aggregating agents, isosorbide dinitrate decreased IC50 of prostacyclin from 2.7 ± 1.2 to 0.36 ± 0.2 nM. Generation of a platelet-active metabolite and synergism with prostacyclin are new properties of isosorbide dinitrate that may account for antiplatelet effects in vivo.

SummaryTo learn whether glucocorticoids inhibit prostaglandin (PG) production in vascular endothelial cells, we investigated the effects of glucocorticoids on PG synthesis by cultured human umbilical vein endothelial cells (EC). Pretreatment of EC with dexamethasone (DX, 10-9 to 5 x 10-5 M) caused a dose-dependent inhibition of PGI2 production when PG synthesis from endogenous arachidonate was stimulated by human thrombin (0.25-2 U/ml) or ionophore A 23187 (1-5 μM). The inhibition was detectable at 10-7 M DX and maximal at 10-5 M (4.0 ± 0.7 vs. control: 7.7 ± 1.9 ng/ml, mean ± S.D., P <0.01). The production of PGE2 and the release of radiolabelled arachidonate (AA) from prelabelled cells were similarly inhibited. Prolonged incubation of EC with glucocorticoids was required to inhibit PG production or arachidonate release: ranging from 8% inhibition at 5 h to 44% at 38 h. In contrast, prostaglandin formation from exogenous AA was not altered by DX treatment. When thrombin or ionophore-stimulated EC were restimulated with exogenous AA (25 μM), DX-treated cells released more PGI2 than control cells (5.7 ± 0.5 vs. 4.1 ± 0.6 ng/ml, P <0.01). Both the decrease in PGI2 production after thrombin/ionophore and the increase after re-stimulation with AA were blunted in the presence of the protein synthesis inhibitor cycloheximide (0.1-0.2 μg/ml). Thus, incubation of EC with glucocorticoids inhibits PG production at the step of phospholipase activation. The time requirement for these steroid effects and their blunting by cycloheximide are consistent with the induction of regulatory proteins, possibly lipocortins, in endothelial cells.