STUDIES ON ENDOTHELIAL REACTIONS : VIII. CHANGES IN THE DISTRIBUTION OF TUBERCLE BACILLI AND TUBERCLES IN THE ORGANS OF RABBITS FOLLOWING SPLENECTOMY.

If the spleen be removed from rabbits and tubercle bacilli be injected intravenously, it is found that the lesions produced differ materially from those observed in control animals; the lungs present the most marked contrast, the liver also shows a definite difference in the distribution of the tubercles, and the kidney lesions also differ in the two instances. In discussing these facts we must answer, if possible, the four questions formulated above. It appears that the pulmonary lesions in Group S are small and discrete because of some inhibitory factor that prevents the diffusion of the products of dead tubercle bacilli. The organisms, far from being killed, are more numerous and better preserved in this group, but their destructive action is localized. This seems to be connected with the presence of polymorphonuclear leucocytes, for these cells are more numerous in the lesions and more plentiful in the lumina of the pulmonary capillaries of Group S than they are in the controls. The conclusion to be drawn from this covers the answer to the second question. It is not the tubercle bacillus itself that produces the destructive changes, but the toxins liberated by the breaking down of its substance. The well known experiments of Hodenpyl and Armand-Delille, already referred to in this series of papers, show this to be true; dead tubercle bacilli, or even extracts of these organisms, will produce typical tubercles. Removing the spleen stimulates the production of polymorphonuclear leucocytes, as shown by Johnstone (1922). These cells may prevent the diffusion of split products from the bacilli by removing or neutralizing them. The third question, as to why the liver is more affected after splenectomy than in normal controls, is more readily answered. Probably the spleen acts as a catch-basin for the bacteria; once removed it can no longer withhold them from the portal circulation and the liver receives a larger number than it would were this bacterial filter still operative. On the other hand, there may be an increase in the phagocytic activity of the endothelium of the sinusoids which might take up more bacteria under these changed conditions. Several investigators have claimed, recently, that there is an increased activity of the liver endothelium following splenectomy, their experiments being directed chiefly toward determining the fate of the erythrocytes. Pearce (1918) in reporting the effects of experimental splenectomy in dogs, states that there are definite compensatory changes in the lymph nodes, in the form of an increased proliferation of endothelial phagocytes, and that the stellate cells of the liver sinusoids often show a similar compensatory increase in number. In both cases the cells are, apparently, formed in situ rather than transported to the organs. He says: ‘Such findings suggest the development of a compensatory function on the part of the lymph-nodes and possibly the liver,’ and suggests that, in times of stress ‘the stellate cells of the liver thus assume, in part at least, the function of destroying red blood-corpuscles by phagocytosis.’ Incidentally, he presents an excellent discussion of the history and subject of splenectomy. Motohashi (1922) reports a great increase in the hemophagic power of the hepatic endothelium and an increase in the number of endothelial elements, after some 45 days following splenectomy in rabbits. Nishikawa and Takagi (1922) have observed similar phenomena with white rats, the Kupffer cells taking up erythrocytes in large numbers in splenectomized animals, whereas controls never show similar propensities on the part of these cells. It may be that different substances cause different reactions on the part of the hepatic endothelium. Contributory Experiment.—A side experiment was performed with five rabbits, two splenectomized and three controls, into which uniform doses of pneumococci were injected intravenously. They all died of septicemia after a few days. The results of this experiment strengthen the foregoing conclusions materially. It was found that there were many polymorphonuclear amphophils in the pulmonary capillaries of the splenectomized animals and that there were numerous focal necroses in the livers. The controls showed much fewer polymorphonuclear cells in the lungs and no focal necroses in the livers, while the spleens were actively congested and inflamed. Otherwise the experiment was not of sufficient importance to warrant a separate report. The question as to why the endothelium of the pulmonary capillaries shows no stimulation similar to that observed in the carbon experiment, but rather less activity than that of the controls, must be answered hypothetically for the present. With the carbon, comparatively huge amounts of foreign matter were injected repeatedly; here but one injection of a much smaller amount of suspended tubercle bacilli was administered. The resulting stimulus to the pulmonary endothelium would, therefore, differ materially in the two instances. In one there would be succeeding waves of stimulation following each injection of irritating foreign substance. In the other an entirely different sort of stimulus would result; the bacteria would be withdrawn from the circulation within an hour or two, judging by past experience, and would then multiply, to be cast off into the circulation in driblets, as the lesions containing them broke down. At least it can be said that there is a good theoretical reason for the difference in the endothelial reaction in the lungs of the two groups of animals.