form the plexus of the submucosa. From this point on they appear as branched strings underneath the movable mucosa, into which they do not penetrate, strange as it may seem. This must be due to the fact that the veins are obliterated at this point. In these cases the primary, usually ulcerated, tumors often contain veins from which plugs of cancercells can be expressed. When, on the other hand, the carcinomatous cells penetrate the walls of the coronary vein from secondary growths in the lymph-glands or the lesser omentum, then the branches of the coronary veins, as well as the veins of the tumor itself, are usually free from malignant thrombi. But the part of the vein nearest to the portal, as well as the portal vein itself, is plugged with tumor-cells, and the peripheral parts of the coronary veins are destroyed by the tumor mass which has broken into and passed through them. From the often dilated coronary vein the tumor-cells invade the portal. The growth first appears as a club-shaped nodule, free in the blood-stream, coming from the stomach and carrying masses of young cancer-cells and pieces of the tumor up into the liver. The external part of the tumor is soft, and has a scant stroma, with thin-walled, wide vessels, and the spaces between these contain masses of loosely adherent young cancer-cells. These are easily recognized from their large nuclei, often surrounded by a narrow zone of protoplasm. The portal blood contains in places groups of young cancer-cells or individual cells, and in other places clumps of the original tumor. These masses accumulate in the smaller branches of the portal vein, where, with the coagulated blood, they form malignant thrombi, which at times appear as round, slightly protruding plugs, visible to the naked eye, and which can be expressed from the branches. Less frequently larger masses of the original tumor become adherent to the walls of the larger branches, for example, at the bifurcation of the stem, and without any direct connection with the cancerous thrombus. This may extend gradually up the vein, the club-shaped mass growing bigger, filling the lumen of the portal vein, which is often markedly dilated. The tumor most frequently extends along through the stem and into the branches to the parenchyma of the liver, but it may also travel down to the roots of the portal vein, especially to the splenic vein, and may even extend into one of the pancreatic veins. The liver is always extensively involved in these cases, and the characteristic fact about the secondary growth in this organ is that small, pea-sized tumors appear in groups. The individual tumors are rarely spheric in shape, but irregular and stellate in outline. In a few cases they are seen to be a part of an injected vascular reticulum formed by the smaller branches of the portal vein. From these the carcinomatous tissue extends into the liver parenchyma. The small groups of tumors rapidly fuse into larger ones, and there is nothing characteristic about appearance and structure of these larger masses. Cases of this type, in which the cancer-cells break into the veins, are of unusual interest both by the light they throw on carcinoma of the stomach and because of the support they give to the mechanical theory, according to which metastatic growths originate as emboli. I shall return to these conditions later. There are comparatively few cases of this type in the literature, and one gets the impression that it is very unusual for carcinoma of the stomach to disseminate in this way. But, according to my own observation, extension by veins is by no means rare, since there occurred 5 cases of venous metastasis among my few cases. I shall first refer to observations by earlier writers, and then give my own cases. In the early part of the nineteenth century physicians began to notice that the veins close to a malignant tumor at times contained masses of cells similar in structure to those of the main tumor. In 1817 Langstaff* observed a mass of this type in the veins of "fungus hæmatodes,” and seven years later Velpeau† reported a case of cancer of the kidney in which the vena cava was filled. These observations were accompanied by others, and soon cancerous cells had been seen in the various veins of the body. About the same time malignant tissue masses were also observed in the portal system. Andral was the first to find cancer of the portal vein associated with carcinoma of the stomach. Carswel§ reports a case in which the neighboring veins were filled with cancerous tissue in a case of a malignant ulcer of the pylorus. Cruveilhier || had not seen cancer of the portal vein associated with carcinoma of the stomach, but he knew of cases of carcinoma of the liver in which the branches of the portal vein contained carcinomatous thrombi. He places special stress on these findings, and bases his theory of the origin of cancer on them, for he holds that the malignant growths originate in the venous capillaries. Cruveilhier does not differentiate clearly between the primary and the secondary carcinomata. He did not know that the multiple growths occurring in the liver are secondary. It is, therefore, possible that many of his cases of cancer of the liver and the portal vein might have been metastatic in nature. This is strongly suggested by his drawings, which show disseminated tumors in the liver. None of the older cases reported was controlled by microscopic examinations. It is, therefore, possible that some of the cases reported as cancerous might have been benign discolored thrombi. Such mistakes, however, could not have been the rule, since the malignant tissue has a different color from that of a simple thrombus and the malignant thrombi usually dilate the vessels, while the ordinary benign forms lead to retraction of the vessel by the organization of the thrombus. A few instances of cancer of the portal vein have been reported since microscopic examination came into use, but only a few of these are given * Med.-Chir. Transact., 1817, vol. viii, p. 286. Walshe: Loc. cit., p. 45. † Gaz. Méd., 1825, vol. i, p. 357. Walshe: Loc. cit., p. 46. Anat. pathol., Paris, 1829. § Illustrations of the Elementary Forms of Disease, London, 1838. Art. "Carcinoma,” vol. i, Pl. 4, Fig. 4. Puckelt: Das Venensystem in seinen krankhaften Verhältnissen, vol. ii, p. 297. Anat. pathol., Paris, 1829-42. Livr. 12, Pl. 2, 3. as secondary to carcinoma of the stomach. It took a long time before observers became convinced that the primary tumor grew into the veins. Dittrich cites a case in a man, fifty-four years of age, who had a medullary carcinoma of the liver in which the entire portal system, from the smallest capillaries to the veins of the pancreas, was involved by the malignant tumor. Virchow† held that cancer was formed in the thrombi, and he cites 6 cases of cancer in the veins, but does not state which veins. He remarks that cancer of the portal vein often extends so far into the smaller branches that it is impossible to follow the growths to their endpoints, and hence it is difficult to decide just how they originated. Later on he observed a case of cancer of the uterus with metastasis in the lymph-glands along the spinal column and in the liver. Here the portal vein and all its branches were filled with a dirty, reddish-gray fluid, containing white lumps consisting of irregular cells with large nuclei. The malignant tissue extended through the splenic vein into the substance of the organ. He concluded that the cancer-cells, the "geschwanzte Körper" of Müller, were formed in the blood, and hence did not believe they had any connection with the cancer of the uterus. Once § he found cancer of the pylorus in a woman fifty-six years of age. In this case the visceral as well as the parietal peritoneum contained multiple, cauliflower-like tumors, and the primary tumor had extended into the pancreas. In the pancreaticoduodenal vein was found a malignant thrombus which extended into the superior mesenteric vein, and from there, as a club-shaped tumor the size of a nut, into the portal vein. The first accurately described case of involvement of the portal vein is given by Meyer.|| In a man forty years old, dead from ulcerated carcinoma of the pylorus, with multiple metastatic growths in the liver, he found the portal vein filled with a continuous, reddish-gray mass, extending from the bifurcation into both the main branches; this could be traced toward the surface of the liver into the groups of pea-sized tumors present on the surface of the organ. The groups of tumors were about as large as walnuts. The thrombus was adherent to the rough walls of the veins, and contained cells varying in size, and often with more than one nucleus. He does not state whether the tumor-cells of the veins were supported by a stroma. He thought he was dealing with a cancerous phlebitis brought about indirectly by the malignant growth in the stomach (Anregnungsursache), since no direct connection could be found between the tumors of the veins and that of the stomach. This theory he based partly on the fact that phlebitis often occurs some distance away from the primary infection, and partly on the arrangement and appearance of the tumors of the liver, which were very similar to multiple foci found before abscess formation in suppurative hepatitis. It is evident then that Meyer did not consider the cancerous thrombus of the * Prag. Vierteljahresschr., 1846, vol. x, p. 104. † Virchow's Archiv, 1847, vol. i, p. 46. ‡ Ibid., 1849, vol. ii, p. 104. § Gesammelte Abhandlungen, pp. 350, 351, of Traube's Beiträge zur experimentellen Pathologie und Physiologie, Berlin, 1846. "Ueber krebsige Phlebitis," Zeitschr. f. rat. Med., 1853, vol. iii. portal vein to have any direct connection with the malignant ulcer of the pylorus. Bamberger* refers briefly to two cases of cancer of the portal vein. In one of these the entire vein to its finest branches was filled with a coagulum consisting mostly of nucleated cells of different shapes. There was no sign of malignancy anywhere else in the body. He, therefore, holds that the malignant growth originated spontaneously in the portal vein. His second case had ulcerated carcinoma of the pylorus, metastatic tumors in the liver, and cancerous thrombosis of the portal vein. In this case he thinks the malignant growths sprang from an eroded vein in the stomach. Frerichst briefly mentions having seen 5 cases of cancer of the portal vein connected with cancer of the liver. He does not mention the condition of the stomach in his cases. Henoch also cites a case of cancer of the portal vein; but his case is of a doubtful nature. He found a thrombus in the splenic vein extending through the stem of the portal and into its left branch in a patient who had suffered from ascites. The only case of cancer of the portal vein secondary to carcinoma of the stomach which has been thoroughly studied and accurately described is one reported by Spaeth,§ who, in a thirty-one-year-old woman giving symptoms of cancer of the stomach and liver with ascites, found a large ulcerated medullary carcinoma of the pyloric end. The veins of the stomach, both those along the lesser and greater curvature, were dilated and filled with cancerous masses. Thickened injected branches extended 1 cm. to 3 cm. over the wall of the stomach. The portal vein was dilated and filled with masses of carcinomatous tissue. These were of the medullary type, and had a definite stroma, with distinct vessels. All the portal tributaries, the splenic vein, and the veins of the mesentery and mesocolon were also involved. The cancerous masses in the latter extended to within 1 to 2 cm. from the intestine. In the surface of the liver were seen multiple tumors arranged in groups, and the small branches of the portal vein were injected in the neighborhood of the carcinomatous nodules, so that they appeared as "feine gewunden verlaufende, vielfach sternförmig sich verzweigende Striefen von gelblich-weisser Farbe." The branches of the portal vein were dilated and filled with masses of malignant cells. These could be traced toward the nodules on the surface, which were filled with a whitish-red, soft mass, and could be identified as the sinus-like ends of the portal vein. Spaeth believed the cancer-cells had broken into the gastroepiploic vein; in the region of the pylorus this vein had been surrounded by the tumor mass and its walls destroyed, so that the place where the * "Krankheiten des chylopoetischen Systems," Virchow's Handbuch der speciellen Pathol. u. Therapie, Erlangen, 1855, p. 589. † Klinik der Leberkrankheiten, 1861, vol. ii, p. 278. Klinik der Unterleibskrankheiten, 1863, vol. iii, p. 144. § "Carcinom im Inneren der Venen des Pfortadergebietes," Virchow's Archiv, 1866, vol. xxxv, p. 432. cancer-cells first broke through was difficult to find. Up to the present time Spaeth's case is the only reported instance of cancer of the portal vein in which the malignant thrombosis had a distinct stroma with vessels of its own, and it is also the only case in which the type of carcinoma definitely has been shown to be medullary. After this review of the cases reported in the literature I wish to report the cases personally observed, and to each of them add such explanatory remarks as the case may need: (a) CASES WITH TUMOR MASSES IN THE SMALLER VEINS OF THE STOMACH, WHICH ORIGINATED IN A PRIMARY TUMOR OF THE STOMACH CASE XXIV.—Medullary carcinoma with medium-sized cells. A small, flat ulcer with soft, slightly infiltrated edges on the lesser curvature near the pylorus. Cancerous thrombosis of the veins surrounding the tumor and also of the rest of the veins from the pyloric end as far as to the superior coronary vein, half of which is involved. Carcinoma cells in the blood of the coronary veins. Multiple tumors of the liver. Infiltration of the lymph-glands of the lesser curvature. Lobar pneumonia of the upper lobe of the right lung. Fredrikke Andersen, age fifty-six, had pains in the right hypochondrium for four months, and felt a tumor first in the region of the right lobe of the liver, and later also in the region of the left. Progressive emaciation. Absolutely free from stomach symptoms up to the last day she lived. Then she vomited several times, and gave signs of lobar pneumonia in the upper lobe of the right lung. During her illness diarrhea was occasionally present. She had been in bed nineteen days. The disease lasted five months. Autopsy, twenty-four hours after death. The body is emaciated. Rigor mortis present. The upper lobe of the right lung is in the state of gray hepatization of lobar pneumonia. The lower lobe is edematous. The peritoneal cavity contains a small amount of reddish, serous fluid. The peritoneum is smooth and unchanged. The liver is markedly enlarged, with a round, nodular tumor in the right lobe, and reaches about 13 cm. below the costal arch. The stomach is of normal size and shape, and contains a small amount of a grayish fluid. The mucosa of the entire fundus is softened. On the lesser curvature, 2 cm. from the pylorus, occurs a round, flat ulcer, 3 cm. in diameter. The floor of the ulcer is smooth, reddish gray, finely granulated, and there appear numerous whitish-yellow points, varying in size from a pinpoint to a pinhead. Small drops of a purulent fluid can be expressed from these, and this fluid is seen to consist of a granular detritus containing round- or polymorphous cells with large round nuclei. The bottom of the ulcer is formed by the not perceptibly thickened submucosa, in which occur small, whitish-yellow, soft spots. The submucosa is adherent to the muscularis, which is not thickened (1 mm.). The edge of the ulcer is flat, smooth, and even, but the mucosa covering it is slightly thickened (1 to 11⁄2 mm.), freely movable on the muscular layer, since the submucosa is not perceptibly infiltrated, but in it are found a few small, yellowish-white, twisted strings, which are from 14 to 1⁄2 mm. in diameter. These are the injected veins. On the inside of the posterior wall of the stomach, about 3 cm. from the ulcer, are outstand ing, raised, branched strings, which in cut section are seen to be veins filled with soft, purulent masses. On the posterior surface of the organ, close to the lesser curvature, and corresponding to the place of the ulcer on the inside, is a mass of dendritically branched strings, which extend from the slightly dilated superior coronary vein, 2 to 3 cm. down on the posterior surface of the stomach, where they disappear, to become visible again on the inside of the organ as the branched strings described above. When the coronary vein is slit open, the portal half of it is seen to be filled with liquid blood, and its walls are unchanged Besides the ordinary corpuscles found in blood, this contains medium-sized |