Prostaglandin mediation of unstable hemodynamics during lung perfusion

Positive end-expiratory pressure (PEEP) leads to a fall in both mean arterial blood pressure (MAP) and cardiac output (CO). This study tests the hypothesis that humoral factors, particularly prostanoids, are important mediators of these events. A support dog was used for ex vivo perfusion of an isolated left lung lobe (LLL) at a fixed flow. In group I (n = 10) an isolated, isovolumetrically contracting dog heart was placed in circuit between the support dog and LLL. Indomethacin (5 mg/kg) was used to pretreat the support dogs and LLL donors of group II (n = 9); support dogs of group III (n = 5); and LLL donors of group IV (n = 4). This cyclooxygenase inhibitor was not used for groups I or V (n = 14). The last two groups were similar except that a heart was not included in the circuit of group V. In group V lobe perfusion during simple inspiratory mechanical ventilation led to decreases in support dog MAP from 141 +/- 3 to 118 +/- 15 mm Hg (P less than 0.01) and CO from 4.0 +/- 0.9 to 2.8 +/- 0.8 L/min (P less than 0.01). Application of PEEP further reduced MAP and CO (P less than 0.01). In the perfused isolated hearts of group I, Starling curves were shifted downward during PEEP at five of six left ventricular balloon volumes tested. In groups II and III MAP and CO were unchanged with mechanical ventilation or PEEP and were higher than groups I or V (P less than 0.05). Radioimmunoassay of the stable degradation products of prostacyclin and thromboxane (Tx) A2 demonstrated low concentrations of these prostanoids in support dogs of group III as compared with group V (P less than 0.05). Blocking the LLL donor dog (group IV) produced hemodynamic results and prostanoid concentrations intermediate between groups I and V. The results show that mechanical ventilation and PEEP can cause a circulating agent(s) to be released that results in a decline in MAP and CO. This is prevented with indomethacin.