The duration of carcinoma of the stomach is very difficult to determine. It has been shown that the onset of the symptoms rarely, if ever, coincides with the actual commencement of the morbid growth, and that no limit can be assigned to this period of latency. The initial phenomena of the disease are also so variable that the degree of indisposition necessary to attract attention may in one case develop within a few weeks, while in another it is delayed for several months. The personal equation, again, is always a factor of great importance, some individuals being affected by a nervous disposition which causes them to seek advice immediately they suffer from indigestion, while others continue to pursue their usual avocations until compelled to relinquish them by excessive debility. The same influence becomes evident when an attempt is made to elicit the history of the illness, the one class being wont to confuse antecedent and extraneous symptoms with those of malignant disease, while the other tend equally to obscure the issue by denying the existence of ill-health before the occurrence of those symptoms of which they immediately complain. These various sources of error are naturally predominant in statistics based upon hospital practice, where patients seldom apply for treatment until the disease has already made considerable progress,1 and who, from want of education, can rarely give an accurate account of the development and course of their illness. Even the period at which the tumour made its appearance cannot be relied upon, since it is usually discovered by accident after it has already attained a considerable size. Nor is the specialist engaged in private practice able from his own observations to determine the usual duration of the malady ; for although he may be able to obtain a fairly accurate history from his patients, the subsequent course of their illness is always difficult, and often impossible, to ascertain. Finally, the general practitioner, who alone possesses the opportunity of studying the disease from beginning to end, seldom sees a sufficient number of cases to allow him to deduce any definite conclusions from them. It is therefore obvious that all the statistics which we at present possess can only afford a rough idea of the average duration of the complaint.

1 In our own cases the average duration of the disease at the time of the patients' admission into hospital was five months.

From the study of 198 cases Brinton estimated the average duration of cancer of the stomach at twelve and a half months ; Lebert considered it to be fifteen months, with a maximum of four years; while Katzenellenbogen placed the usual duration at eighteen months. In our own series 83 per cent, died within twelve months, 11 per cent, between twelve and eighteen months, and about 6 per cent, between eighteen months and two and a half years. In the following table our results are compared with those of Lebert and Osier and McCrae.

Table 28.-The Duration Of Carcinoma Of The Stomach

Percentages of cases

Time of death

Authors'

Lebert

Osier and McCrae

One to three months .

18.4

4

27.6

Three to six months .

32.8

17

29.3

Six to nine months

18.4

19

25.9

Nine to twelve months

13.6

23

25.9

Twelve to eighteen months .

11.2

20

5.2

Above eighteen months

5.6

17

12

100

100

100

It will be seen that while Lebert's figures indicate a comparatively long duration, our own closely agree with those of Osier and McCrae in the main fact that only about 17 per cent, of all cases survive longer than a year. As a rule carcinoma of the stomach in early life runs a much shorter course than in adults, according to Mathieu the mean duration of cases less than thirty years of age being about three months. Colloid cancer is also relatively frequent in the young, and usually displays a considerable rapidity of development.

Most authorities allude to cases which were supposed to have existed for periods varying from three to ten years (twenty years-Ballou), and it is usually stated that Napoleon the First died from a cancer of the stomach of nine years' duration. But before these statements can be accepted certain sources of error require to be eliminated. It has already been remarked that a neoplasm sometimes attacks a simple ulcer or its scar, the evidences of which it obliterates during the course of its growth. It consequently follows that many patients give a history of pain after food, vomiting, haematemesis, and loss of flesh extending over a considerable number of years prior to their death from carcinoma of the stomach. In other instances some functional disorder of the stomach or colon precedes the development of the fatal malady, but owing to the similarity of the symptoms of the two complaints the entire illness is attributed to the final and fatal disease. This seems to be the explanation of the statement concerning Napoleon's illness, for in the numerous histories of his life no evidence is forthcoming of the anorexia, vomiting, progressive loss of flesh, and failure of strength which invariably accompany even the most chronic form of cancer of the stomach ; while the occasional references to ' mon pylore ' during his attacks of abdominal pain were probably the outcome of the confusion that exists in a non-medical mind between symptoms due to disease of the stomach and those that arise from a disorder of the colon. On the other hand, the increasing tendency to corpulence which marked the last few years of his life, the rapid course pursued by the disease when once it became manifest, and the delirium that preceded death for a fortnight, all point to the presence of a growth of great malignancy, and consequently of short duration.

Although the question of average duration is endowed with a certain amount of interest, it is of little value to the clinician, who has to offer an opinion as to the probable duration of life in each individual case that comes under his notice. It is therefore necessary to consider the relative influence exerted by certain special conditions upon the progress of the complaint. Of these the most important, in our opinion, is the situation of the growth in the stomach. In the following table we have analysed 125 cases in order to show the proportional death-rate at different periods in carcinoma of various regions of the stomach.

Table 29.-Showing The Duration Of Life In Carcinoma Of Different Regions Of The Stomach

Duration in months j

1-3 months

3-6 months

6-9 months

9-12 months

12-18 months

18-24 months

Cardia (with oesophageal obstruction) Cardia (oesophagus not affected) Walls and curvatures Pylorus (with stricture) . Pylorus (without stricture) General infiltration

30.7%

12. 5% 13%

13. 5% 14% 12.5%

38.5%

50%

30%

27%

28.5%

25%

23.1% 12.5% 26% 27%

11. 4%

12. 5%

7.7% 25% 17.4% 15.4% 22.8%

8.6% 11.5% 14.3% 37.5%

4% 5.6% 9% 12.5%

It will be observed, at once that carcinoma of the cardia is more rapidly fatal than disease of any other part of the organ, and more especially in those cases where the oesophagus is obstructed by the growth. Thus, of the cases in which dysphagia occurred as an early symptom, 92 per cent, died within nine months, while of those where the oesophagus was not involved 75 per cent, succumbed in the same period. In no instance did disease of the cardiac third of the stomach persist for more than a year. This comparatively short duration appears to depend partly upon the abnormal rapidity of growth exhibited by neoplasms in this region, and partly upon the inordinate frequency with which the disease gives rise to dysphagia, inflammation of the left pleura, and secondary growths in the hver and peritoneum.

Next in order of rapid fatality are cases in which the walls and curvatures of the stomach are primarily affected by the neoplasm. Of such, 69 per cent, died within nine months and 86 per cent, within twelve months, while only 4 per cent, survived more than eighteen months. Here again the disease usually exhibited an abnormally rapid growth, and frequently involved the liver and peritoneum.

The duration of life in disease of the pylorus depends to a great extent upon the condition of the orifice. In those cases where there was marked stenosis 67 per cent, died within nine months and 83 per cent, within a year; while of those which presented a patent orifice, or where the obstruction was of secondary importance, only 54 per cent, succumbed within nine months and more than 23 per cent, survived a year.

Contrary, perhaps, to what might have been expected, cases of general infiltration of the stomach often progress comparatively slowly, only 50 per cent, dying within nine months and about the same proportion surviving between one and two years. The fact that in many instances of this kind the symptoms remain slight and ill-defined for a considerable period renders it probable that the average duration of general cancerous infiltration is even longer than that indicated; and from our own experience we should say that, in the absence of pyloric and cardiac obstruction and of secondary growths in the peritoneum, life may sometimes be prolonged for three or even four years.

Early implication of the liver or peritoneum exerts an important influence upon the duration of the disease. Thus, of those cases in our series which died within three months, secondary growths were found in the liver in 60 per cent., but in only 53 per cent, of those who died between the third and sixth months, and in 43 per cent, of those which survived that period. In like manner, the peritoneum was found to be affected in 40 per cent, of those who succumbed within three months, in 24 per cent, of those who died between the third and the sixth months, and only in 18 per cent, of the cases that survived for a longer period.