Carcinoma of the stomach often exhibits an extreme degree of malignancy, and almost always extends beyond the confines of the organ before death. As the result of Brinton's researches it is usually taught that secondary deposits occur in about 50 per cent, of all cases, although Welch concluded from a study of much larger statistics that they are present in at least 63 per cent.1 These and other similar figures, however, only refer to such growths of a metastatic character as were visible to the naked eye, and do not indicate in any way the frequency with which the disease had invaded the lymphatic vessels outside the stomach, and had produced a very real, though perhaps an invisible, infection of the neighbouring viscera.

1 It is noteworthy that modern writers give a much higher percentage of secondary growths, viz. Ewald 75, Perry and Shaw 80, and Osier and McCrae 86.

That extensive dissemination usually occurs even in cases which present no obvious metastases is readily proved by a microscopical examination, for it is found that not only the stomach itself at some distance from the disease, but the perigastric lymphatic glands, the omenta and peritoneal aspect of the diaphragm, present numerous groups of cells which are identical in character with those of the original tumour. It is also the unhappy experience of most surgeons who have been tempted to extirpate an apparently localised and uncomplicated growth of the pylorus, to find that within a comparatively short time a recurrence occurred in one or other of these situations. We are consequently led to infer that almost from its commencement the neoplasm tends to infect the submucous tissue for a considerable distance around its centre of activity, and that within the course of a few weeks some of its epithelial elements gain access to the lymphatic system.

But although a visible extension of the disease to the surrounding viscera constitutes a very crude indication of its distribution, the presence of secondary deposits is of considerable importance from a clinical standpoint, since they not only give rise to important physical signs, but exert a marked influence upon the duration of the complaint. It is therefore convenient to determine as far as possible their relative frequency in different organs, and the means by which the diffusion of the cancer cells is effected.

Out of 131 cases of gastric cancer in which all the organs of the body were carefully examined, we find that the growth had extended beyond the walls of the stomach in 113, or in 86 per cent. The following table shows the percentage frequency with which the various tissues were affected, and also affords a contrast with the conclusions formed by Welch and Lebert.

It will be observed that in several respects our conclusions differ materially from those arrived at by other writers. In the first place, the percentage frequency of glandular infection is twice as great as that usually accepted. This may be explained partly by the difficulty of detecting a few small glands among a mass of adhesions unless a special search is made for them, and partly by the fact that many morbid anatomists allow the existence of disease only when the glands exhibit visible nodules of growth. Since, however, we have found that every swollen gland in the vicinity of a cancerous stomach shows signs of infection when examined by the microscope, we prefer to regard any enlargement which is apparent to the naked eye as indicative of disease.

In the second place, the liver and pancreas appear to be unduly affected in our cases, because wTe have recorded every instance in which they were involved by the growth, instead of drawing a distinction between invasions by contiguity and true metastases. Our reason in so doing is to emphasise the rapid and widespread dissemination of a gastric cancer, the importance of which is apt to be underrated w7hen only metastatic growths are considered worthy of attention. It might also be urged that the pathological distinction between secondary deposits and direct invasions is often more apparent than real. Thus, in many cases where true metastases are scattered through the substance of the liver, the infection of the portal system may be showm to have taken place through the medium of a mass of growth which had spread into it from the adherent stomach ; while in the case of the omentum it is often impossible to determine whether the induration was primarily due to extension by the lymphatics, or to direct invasion from the peritoneal surface of the growth.

1 The exact figures in this case were : multiple discrete tumours = forty-seven, or 35.8 per cent.; direct invasions = eighteen, or 13.7 per cent.

Table 6.-Ax Analysis Of 131 Cases Of Gastric Cancer

Organ

No. of cases

Percentage

Welch

Lebert

Lymphatic glands (en-

larged)

104

79.4

35 %

54.5%

Liver 1 .

65

50

30.2%

40.9%

Peritoneum (and omenta)

46

35

37.5%

Small intestine

2

1.5

22.7%

} 102%

Large intestine

8

6

Pancreas

25

19

7.8%

7 %

Kidney ....

5

3.7

-

} 8 %

Adrenals

3

2.3

-

Spleen ....

5

3.7

1.7%

5.7%

Pleura .... Lungs ....

6 10

4.5 7.6

} 6.2%

8.3%

Heart ....

3

2.3

-

6 % (+ pericardium)

Uterus .... i Ovary ....

2 3

1 5 2.3

} 4.5%

Bladder ....

2

1.5

-

-

Bones ....

2

1.5

-

3.5%

Brain ....

1

.7

.6%

4.5%

Skin ....

3

2.3

-

3 %

Supraclavicular glands .

4

3

-

-

Mediastinal glands

14

10.6

-

Showing the Frequency with which the various other organs were affected by the disease.

The liability of the different organs to infection varies with the situation of the primary disease in the stomach. In some cases this chiefly depends upon the relative proximity of the viscus in question to the growth, as, for example, the invasion of the pancreas by tumours of the posterior surface of the stomach and of the spleen by those of the fundus. In others, however, it is the type of the disease rather than its location which seems to be responsible for its greater infectivity. It will be seen in the following table that cancers of the pylorus which cause constriction of the orifice comparatively rarely affect the liver, for the reason that they are usually scirrhus; while those which attack the body of the organ or the cardia are particularly destructive of other tissues, because they so often possess a medullary or cylindrical-cell structure. Lastly, it will be noticed that the peritoneum is exceptionally prone to suffer when the whole or greater part of the gastric wall is invaded by the disease.

Table 7.-Relative Frequency Of Metastases In Cancer Of Different Regions Of The Stomach

Situation of disease

Pylorus

(with stenosis)

Pylorus (without stenosis)

Walls and curvatures

Cardia

Whole or greater part of organ

Glands

60%

85%

93%

89%

87%

Liver .

24%

64%

70%

84%

37%

Peritoneum .

30%

36%

26%

46%

75%

Pancreas

13%

14%

30%

7%

25%

Lungs .

3.7%

6%

13%

7%

9%

Pleura .

3.7%

6%

4%

7%

25%

Spleen .

3%

4%

14%

6%

Kidneys

1.9%

4%

4%

14%

1%

Carcinoma of the stomach may lead to the invasion of other viscera in several different ways, viz., by direct extension into the surrounding tissues ; by infection of the lymphatic and blood streams ; by the detachment of small particles, and their subsequent transplantation upon serous or mucous surfaces; or by local contact.