The changes in pulmonary function that occur as a result of experimental pulmonary vascular occlusion have been intensively investigated. The techniques employed for producing pulmonary vascular occlusion have included balloon occlusion (1-3) and the injection of particulate matter (4) into the pulmonary artery. The results of such studies are not directly comparable to clinical pulmonary thromboembolism, where vascular occlusion is produced by autogenous thrombi. It might be anticipated, a priori, that substantial differences would occur because of the different physicochemical nature of the embolic material and the different pattern of embolization. In the present studies the effects of autogenous thromboemboli on ventilation-perfusion relationships, pulmonary gas exchange, pulmonary mechanics, and the ultrastructure of the lung are described.Two questions related to pulmonary thromboembolism are of particular interest. First is the problem of redistribution of ventilation after embolization. Several investigators have described a shift of pulmonary ventilation away from nonperfused lung segments after balloon occlusion of a pulmonary artery (1,5,6). We have developed a technique to evaluate whether redistribution of ventilation occurs after autogenous pulmonary thromboembolism. Second is the evaluation of the mechanisms resulting in arterial hypoxemia after autogenous pulmonary thromboembolism.* Submitted for publication December 22, 1964; accepted July 1, 1965. This work was supported in part by grants HE-05059 and HE-05317 from the National Institutes of Health and in part by a grant from the Tuberculosis League of Pittsburgh.
MethodsSixty-five healthy, mongrel dogs ranging in weight from 9 to 26 kg were studied. The animals were anesthetized with either pentobarbital, thiopental sodium, or chloralose by intravenous administration and intubated with an endotracheal tube. Thirty-four dogs were studied during spontaneous ventilation (spontaneous ventilation dogs, SVD), and in 31 dogs spontaneous ventilation was abolished by either d-tubocurarine or succinylcholine administered intravenously, and ventilation was then maintained with an Etsten bellows pump set to deliver a constant tidal volume at a constant frequency (controlled ventilation dogs, CVD). The total amount of barbiturate administered was such as previously observed to maintain adequate anesthesia in nonparalyzed animals. Autogenous thrombi were produced and released into the pulmonary circulation as previously described (7).The following measurements were made before and within 30 minutes after embolization in the SVD and CVD while they were breathing room air. Minute ventilation (VE) was measured by standard techniques. Femoral arterial blood was analyzed for oxygen content and capacity, and C02 content by the technique of Van Slyke and Neill (8). Arterial pH was measured with a glass electrode at 37°C. Plasma C02 content was derived with the correction factors of Van Slyke and Sendroy from the whole blood CO2 content, the arterial hematocrit,...