Next to sepsis we must consider hereditary syphilis as an important cause of gastrointestinal Inemorrhage. According to earlier reports (Behrend and others) and especially according to the thorough investi gations of Mraczek, hereditary syphilis brings about changes in the walls of the small and medium sized vessels and of the capillaries, which lead to thickening of the intima, small-celled infiltration with subse quent connective-tissue formation, to thickening of the vascular wall and to a narrowing and even occlusion of its lumen. Thus venous stasis and haemorrhages ensue. These haemorrhages occur also in the intesti nal canal and Mraczek has reported such cases (Esser also recently re ported one case). The assumption that these were cases of sepsis, which develops especially easily in syphilitic children, is negatived by the fact that, in some of Mraczek's cases, Kolisko and Paltauf were able to ex clude sepsis by means of bacteriological examinations. In the vascular affections of hereditary syphilis, haemorrhages can easily occur because of stasis and disturbances of the circulation, and at times syphilis runs its course under the clinical picture of meltena.
The gastro-intestinal hiemorrliages which Kundrat described occur ring with syphilis of the liver are not to he attributed to a syphilitic affection of the vessels hut rather to abnormal stasis in the portal circu lation. However, not only the constitutional diseases just spoken of but also local diseases of the organs may lead to secondary (sympto matic) mehena. To this category belong first and foremost, congenital anomalies of the heart and large vessels; then diseases of the liver, e.g., syphilis of the liver, which Kundrat advanced as the cause of melama; further, tumors of the abdomen (cysts in Schukowski's case), then local diseases of the intestines and stomach, which lead to secondary gastric haemorrhage. Cases of mehena are observed, however, with compara tive frequency, in otherwise perfectly healthy children (true meliena). In these cases the haemorrhage starts within the first four days of life, usually on the first or second day. In the cases terminating fatally, no constant findings in the gastro-intestinal tract have been shown at necropsy. In a part of the cases nothing which could have explained the haemorrhage was found; in others, hyperaemia and stasis in the gastro-intestinal mucosa; and in still others, multiple haemorrhages or ha.morrhagic erosions were found in the mucosa. In a few cases a small ulcer similar to the round ulcer of the adult, was found in the stomach or the small intestine (duodenum) or very rarely in the esophagus. Multiple ulcers were rarely present.
Mehena is most easily explained in the cases in which there is an ulcer present. The erosion of a blood vessel in consequence of the tissue necrosis, leads to bleeding which is often uncontrollable. For these cases, however, the question as to the origin of the ulcers must be raised; Landau advanced the hypothesis that emboli gave rise to them.
The emboli are supposed to originate in a thrombosis of the umbilical vein or of the ductus arteriosus Botalli; the thrombus being swept. into a branch of the pancreatico-duodenal or the gastric artery, occludes the vessel, thus causing necrosis and ulceration. The gastro-intestinal bleeding then ensues secondarily when the vessel is involved in the tissue-necrosis.
Kundrat brought forward against this theory, the fact that Landau had not demonstrated the embolic origin of the ulcers; that in one of his own cases no embolism existed and that the sweeping of an embolus into the vessels of the intestine or stomach was highly improbable be cause the position, calibre and course of these vessels are very unfavor able for the lodgment of emboli. It is much more probable that the gastro-duodenal ulcers are not embolic in nature; they probably develop through hemorrhages into the mucosa by reason of the softening and digestive action of the gastric juice on these infarcted areas (Kundrat). These ulcers are usually situated in the duodenum, less often in the stomach and rarely above the cardia in the (esophagus. Kundrat's conception is in part supported by a case of Chrzanowski's, in which there was bloody effusion throughout the entire thickness of the (esoph agus above the cardia and only superficial defects were demonstrable in the epithelium. Further, the many cases of melana, in which superfi cial hremorrhagie erosions develop on a mucosa riddled with hemor rhages, serve to substantiate hundrat's position; in these cases the necrotic process has not progressed deeply enough to cause ulcer. That there are many cases in which hemorrhages alone are found in the mucosa argues to this effect. In short, from hemorrhage to erosion and then to ulceration we have only a series of pathological changes, the cause of which must lie in a hyperemia of the gastro-intestinal mucosa. Moreover, in the eases in which no change has been found in the gastro intestinal tract, we must, with Kundrat, hold that such a hyperemia did exist during life which, however, at necropsy has disappeared. The cause of this hyperemia is not clearly established; in explanation many theories have been adduced. A hemorrhage in the gastro-intes tinal tract can take place only under two conditions: (1) with a normal state of circulation and blood pressure, the vessel wall being diseased and thus offering inadequate resistance to the pressure of the blood; (2) with normal vessel walls, the blood pressure either local or general, being abnormally high. Of course in every such instance hamatemesis does not necessarily follow, but on the contrary there is usually only stasis and slight hemorrhage in the substance of the mucosa. Naturally, these two factors may coexist in any one case.