The tissue beneath and at the margins of medullary cancer may be predominantly fibrous in texture and contain comparatively few cancerous alveoli. This scirrhous base is often exposed after the destruction of the greater part of the soft cancer by ulceration and sloughing. It is probable that many of the scirrhous cancers are formed in this way secondarily to medullary cancer (Ziegler).

Cylindrical-celled epithelioma presents the same gross appearances and the same tendency to ulceration and to the formation of metastases which characterize medullary cancer. The consistence of cylindrical epithelioma may, however, be firm like that of scirrhus. Not infrequently the alveoli are distended with mucus secreted by the lining epithelium, and then the tumor presents in whole or in part appearances similar to colloid cancer.

Upon microscopical examination are seen spaces resembling more or less closely sections of tubular glands. These spaces are lined with columnar epithelium. Often in certain parts of the tumor the alveolar spaces are filled with cells, so that the structure is a combination of that of ordinary cancer and of epithelioma. The stroma is generally scanty and rich in cells, but it may be abundant. Cysts may be present in this form of tumor, and in one case I have found such cysts nearly filled with papillary growths covered with cylindrical epithelium, so that the appearance resembled closely that of the so-called proliferous cysto-sarcoma of the breast.

The origin of cylindrical epithelioma from the gastric tubules is generally accepted, and is more readily demonstrable than the similar origin claimed for the other forms of gastric cancer.

Scirrhous cancer assumes often the form of a diffuse thickening and induration of the gastric walls, particularly in the pyloric region, where it causes stenosis of the pyloric orifice. Scirrhus may, however, appear as a circumscribed tumor. Irregular hard nodules frequently project from diffuse scirrhous growths into the interior of the stomach. Scirrhous cancer and medullary cancer are often combined with each other.

The dense consistence of scirrhous cancer is due to the predominance of the fibrous stroma, the cancerous alveoli being relatively small in size and few in number.

Colloid cancer generally appears as a more or less uniform thickening of the gastric walls. All of the coats of the stomach are converted into the colloid growth. Nearly the whole of the stomach may be invaded by the new growth.⁸² The tumor has a tendency to spread to the omenta and to the rest of the peritoneum, where it may form enormous masses, but it rarely gives rise to metastases in the interior of organs. Colloid cancer may, however, form a circumscribed projecting tumor in the stomach, and in rare instances it causes abundant secondary colloid deposits in the liver, the lungs, and other parts.

⁸² In a case reported by Storer the whole stomach, except a little of the left extremity over an extent of about an inch, was converted into a colloid mass in which no trace of the normal coats of the stomach could be made out. The colloid growth replacing the gastric wall measured seven-eighths of an inch in thickness in the pyloric region. Digestion was less disturbed in this case than in most cases of gastric cancer (Boston Med. and Surg. Journ., Oct. 10, 1872). In Amidon's case (reported in the Trans. of the N.Y. Path. Soc., vol. iii. p. 38) there seems to have been an equally extensive colloid metamorphosis of the stomach.

Colloid cancer presents, even to the naked eye, an exquisite alveolar structure, whence the name alveolar cancer as a designation of this tumor. Bands of opaque white or gray connective tissue enclose alveolar meshes which are filled with the gelatinous, pellucid colloid substance. This colloid material is thought to be produced by a colloid transformation of the epithelial cells in the alveoli, but the same transformation seems to occur also in the stroma. Few or no intact epithelial cells may be found in the alveoli. Colloid metamorphosis may take place in all forms of gastric cancer, but it is particularly common in cylindrical epithelioma. Colloid cancer may originate in the peritoneum unconnected with any glandular structures. It occurs often at an earlier age than other forms of cancer. Deep ulceration rarely attacks colloid cancer.

Flat-celled epithelioma is found at the cardiac orifice and as a metastatic growth in other parts of the stomach. Originating in the oesophagus, it may extend downward into the stomach. By noting whether the structure is that of squamous or of cylindrical epithelioma it is often possible to determine whether a tumor at the cardiac orifice originates in the oesophagus or in the stomach.

Secondary cancer of the stomach, although rare, is not such a curiosity as is often represented. Without aiming at completeness, I have been able to collect 37 cases of secondary cancer of the stomach, of which the larger number will stand critical examination.⁸³ Of these cases, 17 were secondary to cancer of the breast, 8 to cancer of the oesophagus, 3 to cancer of the mouth or nose, and the remainder to cancer of other parts of the body. The large number of cases secondary to cancer of the breast is explained by the large statistics relating to mammary cancer which were consulted. Gastric cancer is more frequently secondary to cancer of the oesophagus than to cancer of any other part. In this category of course are not included cases of continuous growth of oesophageal cancer into the stomach, but only metastatic cancers of the stomach. A part at least of the gastric cancers secondary to cancer of the alimentary tract above the stomach I refer, with Klebs, to implantation in the mucous membrane of the stomach of cancerous particles detached from the primary growth in the oesophagus, pharynx, or mouth. This view is supported by the absence in some cases of any involvement of the lymphatic glands. The secondary deposits in the stomach conform in structure to the primary growth. They are usually situated in the submucous coat, where they form one or often several distinctly circumscribed tumors. The secondary tumors may or may not ulcerate. They rarely produce symptoms.

⁸³ These cases are from Dittrich, 2 (the remainder of his cases I rejected); Cohnheim, 1; Petri, 2; Klebs, 3; Lücke, 1; Weigert, 1; Coupland, 1; Cruse, 1; Hausmann, 1; Bartholow, 1; Oldekop, 5; Edes, 1; V. Török and V. Wittelshöfer, 8; Grawitz, 4; Haren Noman, 5. So-called melanotic cancers, cancers involving only the serous coat of the stomach, and those extending by continuous growth into the stomach, are not included in this list.

Primary cancers may be present at the same time in different organs of the body; for instance, in the uterus and in the stomach.⁸⁴ The possibility of multiple primary cancers is to be borne in mind in considering some of the apparently secondary cancers of the stomach, as well as in determining whether certain cancers are secondary to gastric cancer or not. Here the microscopical examination is often decisive.⁸⁵

⁸⁴ Case of A. Clark's (Trans. N.Y. Path. Soc., vol. i. p. 260), and a similar one reported by J. B. S. Jackson in Extr. from Records of the Boston Soc. for Med. Improvement, vol. i. p. 335.

⁸⁵ The subject of multiple primary cancers is considered by Kauffmann (Virchow's Arch., Bd. 75, p. 317), and by Beck (Prager med. Wochenschr., 1883, Nos. 18 and 19). V. Winiwarter reports a cancer of the stomach in a patient who died one year seven and a half months after extirpation of a cancer of the nose. He regards the case as one of multiple primary cancer.

Gastric cancer often causes important secondary changes in the coats and the lumen of the stomach. In the neighborhood of the tumor are often found hypertrophy of the muscular coat and fibrous thickening of the submucous coat. Polypoid hypertrophy of the mucous membrane near the cancer is not rare. Not only near the tumor, but over the whole stomach, chronic catarrhal gastritis usually exists.

The most important alterations are those dependent upon obstruction of the orifices of the stomach. This obstruction may be caused either by a tumor encroaching upon the orifice or by an annular thickening of the walls of the orifices. Even without apparent stenosis, destruction of the muscular layer at or near the pylorus may be an obstacle to the propulsion of the gastric contents into the duodenum. As a result of obstruction of the pyloric orifice the stomach becomes dilated, sometimes enormously, so as to occupy most of the abdominal cavity. The walls of the dilated stomach, particularly the muscular coat, are usually thickened, but exceptionally they are thinned. Sometimes with pyloric stenosis the stomach is reduced in size. This occurs particularly when a scirrhous growth extends diffusely from the pyloric region over a considerable part of the stomach. Obstruction of the cardiac orifice or in the oesophagus leads to atrophy of the stomach, although here also there are exceptions. Above the obstruction the oesophagus is often dilated. An existing obstruction may be reduced or removed by ulceration or sloughing of the tumor.

Both dilatation and contraction of the stomach may attend gastric cancer without any involvement of the orifices of the stomach in the cancerous growth. The cavity of the stomach may be so shrunken by scirrhous thickening and contraction of the gastric walls that it will hardly contain a hen's egg. Irregular deformities in the shape of the stomach, such as an hour-glass shape and diverticular recesses, may be caused by gastric cancer.

Changes in the shape of the stomach and the weight of the tumor may cause displacements of pyloric cancers, so that these tumors have been found in nearly all regions of the abdomen, and even in the true pelvis.⁸⁶ Such displaced cancers usually contract adhesions with surrounding parts.

⁸⁶ Lebert, op. cit., p. 420.

It is not necessary to dwell upon the formation of adhesions which may bind the stomach to nearly all of the abdominal organs, most frequently to the liver, the pancreas, the intestine, and the anterior abdominal wall. Adhesions of pyloric cancers are found in at least two-thirds of the cases, and probably oftener.⁸⁷

⁸⁷ Gussenbauer and V. Winiwarter found adhesions recorded in 370 out of 542 pyloric cancers. In considering the propriety of resection of gastric cancers it has become a matter of importance to know in what proportion of cases adhesions are present. I agree with Ledderhose and with Rydygier in believing that adhesions are present oftener than appears from Gussenbauer and V. Winiwarter's statistics. The fact that adhesions are not noted in post-mortem records of gastric cancer cannot be considered proof of their absence. Little has been done in the study of gastric cancer from a surgical point of view. Metastases and adhesions were absent in only 5 out of 52 cases of pyloric cancer in which either pylorectomy or exploratory laparotomy was performed (Rydygier).

Cancer of the stomach in the majority of cases is accompanied with metastases in other parts of the body. In 1120 cases of gastric cancer secondary cancers were present in 710, or 63.4 per cent., and absent in 410, or 36.6 per cent.⁸⁸ In about two-thirds of the cases, therefore, secondary deposits were present.

⁸⁸ These cases are from Habershon, op. cit.; Lebert, op. cit.; Trans. N.Y. Path. Soc., vol. i.; and Gussenbauer and Von Winiwarter, loc. cit.

In order to determine the relative frequency of the secondary deposits in various organs of the body, I have constructed the following table, based upon an analysis of 1574 cases of cancer of the stomach in which the situation of the metastases were given:⁸⁹

⁸⁹ These cases include, in addition to those cited in the preceding foot-note, those of Dittrich (Prager Vierteljahrschr., vol. xvii.), Wrany (ibid., vols. xciv. and xcix.), Katzenellenbogen (op. cit.), and Lange (op. cit.). Metastases in the intestine formed only a small number of those under the heading peritoneum, omentum, and intestine, but as they were all included together in Gussenbauer's large statistics, the intestinal metastases could not well be placed separately. In 673 cases the peritoneum and omentum were cancerous in 21.7 per cent.

Secondary cancerous deposits are probably even more frequent in the lymphatic glands than appears from the table. In 1153 cases of gastric cancer in which the situation of the affected lymphatic glands is specified, the abdominal glands, and chiefly those near the stomach, were the seat of cancer in 32½ per cent. In Lange's 210 cases the cervical glands were affected in 4.3 per cent. In other statistics this percentage is much smaller. In nearly one-third of the cases there are secondary cancers in the liver. These may attain an enormous size in comparison with the tumor of the stomach. Cancer of the peritoneum and of the omentum is found in about one-fifth of the cases of gastric cancer. The spleen is rarely involved, except by continuous growth of a cancer of the fundus or in cases of widespread distribution of cancer through the aortic circulation. Cancer of the liver increases the liability to metastases in the lungs, but the latter may be present without any cancerous deposits in the liver. Secondary cancers may be present in the suprarenal capsules, the kidneys, the ovaries, the heart, the thoracic duct, the bones, the skin, etc. In an interesting case reported by Finlay⁹⁰ the subcutaneous tissue of the trunk was thickly studded with small nodules, of which two were excised during life and found to be cylindrical epitheliomata. This led to the diagnosis of a primary tumor of the same nature in the stomach or in the intestine. At the autopsy was found a cylindrical epithelioma of the stomach which had not given rise to characteristic symptoms. Secondary cancer of the intestine is rare if the deposits in the peritoneal coat be excepted. Several cancerous ulcers or multiple cancerous nodules may be found along the intestinal tract, involving the mucous and the submucous coats.⁹¹ These metastases seem best explained by the theory of implantation of cancerous elements which have been carried from the primary growth in the stomach into the intestine. In some of the cases the idea of multiple primary cancers may also be entertained.

⁹⁰ Trans. Path. Soc. London, vol. xxxiv. p. 102. Unfortunately, in Röseler's case of multiple skin-cancers with an ulcerated cancer of the stomach no microscopical examination of the skin-nodules was made. The interpretation of this case is therefore doubtful (Virchow's Archiv, Bd. 77, p. 372).

⁹¹ Cases in point are recorded by Wrany (loc. cit.), Blix (Virchow u. Hirsch's Jahresbericht, 1876, ii. p. 207), Lange, Katzenellenbogen, and Lebert.

It is not rare for gastric cancer to cause secondary deposits in the stomach itself. Sometimes it is difficult to decide which of two or more cancers in the stomach is the primary growth, as in Ripley's case of ulcerated cancer of the cardiac orifice with a similar growth around the pyloric orifice.⁹² It is probable that in very rare instances multiple primary cancers may develop in the stomach.

⁹² J. H. Ripley, Trans. N.Y. Path. Soc., vol. iv. p. 121. Maurizio has also reported a case of scirrhous cancer of the cardia with scirrhous cancer of the pylorus (Annal. univ. di Medicina, Oct., 1869). A similar case was observed by Barth (Gaz. hebdom., 1856, No. 24, p. 424).

Cancerous metastases are produced by the transportation of cancerous elements by the lymphatic current or by the blood-current. In a number of instances the portal vein or some of the branches which help to form it have been found plugged with a cancerous mass which may or may not be organized.⁹³ The cancer in these cases has burst through the walls of the vessel into the lumen, where it may grow both in the direction and against the direction of the current. On serous surfaces, and probably also, although rarely, on mucous surfaces, secondary cancers may develop from cancerous particles detached from a parent tumor and scattered over the surface as a kind of seminium.

⁹³ Cases of this kind have been reported with especial fulness by Spaeth (Virchow's Archiv, Bd. 35, p. 432), Acker (Deutsches Arch. f. kl. Med., Bd. 11, p. 173), and Audibert (De la Généralisation du Cancer de l'Estomac, Paris, Thesis, 1877).

Mention has already been made of the invasion of parts adjacent to the stomach by the continuous growth of gastric cancer. In this way lymphatic glands, the liver, the pancreas, the omenta, the transverse colon, the spleen, the diaphragm, the anterior abdominal wall, the vertebræ, the spinal cord and membranes, and other parts may be involved in the cancerous growth.

Under the head of Complications reference has already been made to various lesions which may be associated with gastric cancer. As regards the manifold complications caused by perforation of gastric cancer, in addition to what has already been said the article on gastric ulcer may be consulted. In general, the various fistulous communications caused by gastric cancer are less direct than those produced by gastric ulcer. The wasting of various organs of the body in cases of gastric cancer may be found on post-mortem examination to be extreme. Habershon mentions a case in which the heart of a woman forty years old weighed only 3½ ounces after death from cancer of the pylorus. As in other profoundly anæmic states, the embryonic or lymphoid alteration of the marrow of the bones is often present in gastric cancer.

PATHENOGENESIS.—The problems relating to the ultimate causation and origin of gastric cancer belong to the pathenogenesis of cancer in general. Our knowledge with reference to these points is purely hypothetical. It will suffice in this connection simply to call attention to Virchow's doctrine, that cancer develops most frequently as the result of abnormal or of physiological irritation, hence in the stomach most frequently at the orifices; and to Cohnheim's theory, that cancer as well as other non-infectious tumors originate in abnormalities in development, more specifically in persistent embryonic cells. According to the latter view, gastric cancer develops only in those whose stomachs from the time of birth contain such embryonic remnants. These unused embryonic cells may lie dormant throughout life or they may be incited to cancerous growth by irritation, senile changes, etc. According to Cohnheim's theory, the orifices of the stomach are the most frequent seat of cancer on account of complexity in the development of these parts.

For a full consideration of these theories the reader is referred to the section of this work on General Pathology.

DIAGNOSIS.—The presence of a recognizable tumor in the region of the stomach outweighs in diagnostic value all other symptoms of gastric cancer. The detection of fragments of cancer in the vomit or in washings from the stomach is of equal diagnostic significance, but of rare applicability. The discovery of secondary cancers in the liver, in the peritoneum, or in lymphatic glands may render valuable aid in diagnosis. Of the local gastric symptoms, coffee-ground vomiting is the most important. The relation between the local and the general symptoms may shed much light upon the case. While anorexia, indigestion, vomiting, and epigastric pain and tenderness point to the existence of a gastric affection, the malignant character of the affection may be surmised by the development of anæmia, emaciation, and cachexia more rapid and more profound than can be explained solely by the local gastric symptoms. The value to be attached in the diagnosis of gastric cancer to the absence of free hydrochloric acid from the contents of the stomach must still be left sub judice. The age of the patient, the duration, and the course of the disease are circumstances which are also to be considered in making the diagnosis of gastric cancer. These symptoms of gastric cancer have already been fully considered with reference to their presence and absence and to their diagnostic features.

It remains to call attention to the differential diagnosis between gastric cancer and certain diseases with which it is likely to be confounded. The points of contrast which are to be adduced relate mostly to the intensity and the frequency of certain symptoms. There is not a symptom or any combination of symptoms of gastric cancer which may not occur in other diseases. Hence the diagnosis is reached by a balancing of probabilities, and not by any positive proof. Notwithstanding these difficulties, gastric cancer is diagnosed correctly in the great majority of cases, although often not until a late stage of the disease. Errors in diagnosis, however, are unavoidable, not only in cases in which the symptoms are ambiguous or misleading, but also in cases in which all the symptoms of gastric cancer, including gastric hemorrhage and tumor, are present, and still no gastric cancer exists. Cases of the latter variety are of course rare.

In the absence of tumor the diseases for which gastric cancer is most liable to be mistaken are gastric ulcer and chronic gastric catarrh. In the following table are given the main points of contrast between these three diseases:

The diagnosis between gastric cancer and gastric ulcer is more difficult than that between cancer and gastritis, and sometimes the diagnosis is impossible. The differential points mentioned in the table are of very unequal value. An age under thirty, profuse hemorrhage, and absence of tumor are the most important points in favor of ulcer; tumor, advanced age, and coffee-ground vomiting continued for weeks are the most important points in favor of cancer. As cancer may have been preceded by ulcer or chronic gastritis for years, it is evidently unsafe to trust too much to the duration of the illness. As has already been said, it is best to place no reliance in the differential diagnosis upon the character of the pain. Any peculiarities of the vomiting, the appetite, or the digestion are of little importance in the differential diagnosis. Cachexia is of more importance, but it is to be remembered that ulcer, and even chronic gastritis in rare instances, may be attended by a cachexia indistinguishable from that of cancer. Cases might be cited in which very decided temporary improvement in the symptoms has been brought about in the course of gastric cancer, so that too much stress should not be laid upon this point. Enough has been said under the Symptomatology with reference to the diagnostic bearings of the absence of free hydrochloric acid from the stomach, of the presence of cancerous fragments in fluids from the stomach, and of secondary cancers in different parts of the body.

One must not lose sight of the fact that the whole complex of symptoms, the order of their occurrence, and the general aspect of the case, make an impression which cannot be conveyed in any diagnostic table, but which leads the experienced physician to a correct diagnosis more surely than reliance upon any single symptom.

In the early part of the disease there may be danger of confounding gastric cancer with nervous dyspepsia or with gastralgia, but with the progress of the disease the error usually becomes apparent. What has already been said concerning the symptomatology and the diagnosis of gastric cancer furnishes a sufficient basis for the differential diagnosis between this disease and nervous affections of the stomach.

Chronic interstitial gastritis or fibroid induration of the stomach cannot be distinguished with any certainty from cancer of the stomach. Fibroid induration of the stomach is of longer duration than gastric cancer, and it is less frequently attended by severe pain and hemorrhage. Sometimes a hard, smooth tumor presenting the contours of the stomach can be felt, but this cannot be distinguished from diffuse cancerous infiltration of the stomach.

Non-malignant stenosis of the pylorus is of longer duration than cancer of the pylorus. The symptoms of dilatation of the stomach are common to both diseases. Cicatricial stenosis is the most common form of non-malignant pyloric stenosis. This is usually preceded by symptoms of gastric ulcer which may date back for many years. Non-malignant stenosis more frequently occurs under forty years of age than does cancer. The diagnosis between malignant and non-malignant stenosis of the pylorus is in some cases impossible.

Although the surest ground for the diagnosis of gastric cancer is the appearance of tumor, there are cases in which it is difficult to decide whether the tumor really belongs to the stomach, and even should it be established that the tumor is of the stomach, there may still be doubt whether or not it is cancerous.

The diagnosis between cancerous and non-cancerous tumors of the stomach, such as sarcoma, fibroma, myoma, etc., hardly comes into consideration. The latter group of tumors rarely produces symptoms unless the tumor is so situated as to obstruct one of the orifices of the stomach. Even in this case a positive diagnosis of the nature of the tumor is impossible.

Of greater importance is the distinction between cancerous tumors of the stomach and tumors produced by thickening of the tissues and by adhesions around old ulcers of the stomach. Besides the non-progressive character of the small and usually indistinct tumors occasionally caused by ulcers or their cicatrices, the main points in diagnosis are the age of the patient and the existence, often for years, of symptoms of gastric ulcer antedating the discovery of the tumor. The long duration of symptoms of chronic catarrhal gastritis and of dilatation of the stomach is also the main ground for distinguishing from cancer a tumor produced by hypertrophic stenosis of the pylorus.

Tumors of organs near the stomach are liable to be mistaken for cancer of the stomach. The differential diagnosis between gastric cancer on the one hand, and tumors of the left lobe of the liver and tumors of the pancreas on the other hand, is often one of great difficulty.

Tumors of the liver are generally depressed by inspiration, whereas tumors of the stomach are much less frequently affected by the respiratory movements. The percussion note over tumors of the liver is flat, while a tympanitic quality is usually associated with the dulness over tumors of the stomach. Light percussion will often bring out a zone of tympanitic resonance between the hepatic flatness and the dulness of gastric tumors. Gastric tumors are usually more movable than hepatic tumors. By palpation the lower border of the liver can perhaps be felt and separated from the tumor in case this belongs to the stomach. Most of the points of distinction based upon these physical signs fail in cases in which a gastric cancer becomes firmly adherent to the liver. The basis for a diagnosis must then be sought in the presence or the absence of marked disturbance of the gastric functions, particularly of hæmatemesis, vomiting, and dilatation of the stomach. On the other hand, ascites and persistent jaundice would speak in favor of hepatic cancer. There are cases in which the diagnosis between hepatic cancer and gastric cancer cannot be made. This is especially true of tumors of the left lobe of the liver, which grow down over the stomach and compress it, and which are accompanied by marked derangement of the gastric functions. The frequency with which cancer of the stomach is associated with secondary cancer of the liver should be borne in mind in considering the diagnosis.

There are certain symptoms which in many cases justify a probable diagnosis of cancer of the pancreas, but this disease can rarely be distinguished with any certainty from cancer of the stomach. The situation of the tumor is the same in both diseases. With pancreatic cancer the pain is less influenced by taking food, the vomiting is less prominent as a symptom, and anorexia, hæmatemesis, and dilatation of the stomach are less common than with gastric cancer. Of the positive symptoms in favor of cancer of the pancreas, the most important are jaundice, fatty stools, and sugar in the urine. Of these symptoms jaundice is the most common.

Should there be any suspicion that the tumor is caused by impaction of feces, a positive opinion should be withheld until laxatives have been given.

Mistakes may occur as to the diagnosis between gastric cancer and tumors of the omenta, the mesentery, the transverse colon, the lymphatic glands, and even the spleen or the kidney. Encapsulated peritoneal exudations near the stomach have been mistaken for gastric cancer. Where a mistake is likely to occur each individual case presents its own peculiarities, which it is impossible to deal with in a general way. Of the utmost importance is a careful physical exploration of the characters and relations of the tumor, aided, if necessary, by artificial distension of the stomach or of the colon by gas (see page 549). No less important is the attentive observance of the symptoms of each case. In doubtful cases fluids withdrawn from the stomach by the stomach-tube should be carefully examined for cancerous fragments, and the gastric fluids may be tested for free hydrochloric acid by methods already described.

Pyloric cancers which receive a marked pulsation from the aorta sometimes raise a suspicion of aneurism, but the differential diagnosis is not usually one of great difficulty. Gastric cancer when it presses upon the aorta may simulate aneurism, not only by the presence of pulsation, but also by the existence of a bruit over the tumor. The tumor produced by aneurism is generally smoother and rounder than that caused by cancer. The pulsation of an aneurism is expansile, but the impulse of a tumor resting upon an artery is lifting and generally without lateral expansion. The impulse transmitted to a tumor resting upon the abdominal aorta may be lessened by placing the patient upon his hands and knees. Sometimes the tumor can be moved with the hands off from the artery, so that the pulsation momentarily ceases. A severe boring pain in the back, shooting down into the loins and the lower extremities, and not dependent upon the condition of the stomach, characterizes abdominal aneurism, but is not to be expected in gastric cancer. With aneurism gastric disorders and constitutional disturbance are much less prominent than with cancer of the stomach.⁹⁴

⁹⁴ In a case of pulsating pyloric cancer observed by Bierner the symptoms were much more in favor of aneurism than of cancer. The cancer had extended to the retro-peritoneal glands, which partially surrounded and compressed the aorta. There were marked lateral pulsation of the tumor, distinct systolic bruit, diminution of the femoral pulse, and severe lancinating pain in the back and sacral region. With the exception of vomiting, the gastric symptoms were insignificant. The patient was only thirty-three years old (Ott, Zur Path. des Magencarcinoms, Zurich, 1867, p. 71).

Spasm of the upper part of the rectus abdominis muscle may simulate a tumor in the epigastric region. The diagnosis is made by noting the correspondence in shape and position between the tumor and a division of the rectus muscle, the superficial character of the tumor, the effect of different positions of the body upon the distinctness of the tumor, the tympanitic resonance over the tumor, and, should there still be any doubt, by anæsthetizing the patient, when the phantom tumor will disappear. Spasm of the rectus muscle has been observed in cases of cancer of the stomach.

Attention is also called to the possibility of mistaking in emaciated persons the head of the normal pancreas, or less frequently the mesentery and lymphatic glands, for a tumor.⁹⁵ As emaciation progresses the at first doubtful tumor may even appear to increase in size and distinctness.

⁹⁵ In the case of the late Comte de Chambord the diagnosis of gastric cancer was made upon what appeared to be very good grounds. No cancer, however, existed, and the ill-defined tumor which was felt during life in the epigastric region proved to be the mesentery containing considerable fat (Vulpian, "La dérnière Maladie de M. le Comte de Chambord." Gaz. hebd. de Méd. et de Chir., Sept. 14, 1883).

It is sufficient to call attention to the danger of mistaking, in cases where the gastric symptoms are not prominent and no tumor exists, gastric cancer for pernicious anæmia, senile marasmus, or the chronic phthisis of old age. In some of these cases the diagnosis is impossible, but the physician should bear in mind the possibility of gastric cancer in the class of cases here considered, and should search carefully for a tumor or other symptom which may aid in the diagnosis.

The possibility of mistaking gastric cancer accompanied with peritoneal exudation for cirrhosis of the liver or for tubercular peritonitis is also to be borne in mind.

The diagnosis of the position of the cancer in the stomach can usually be made in cases of cancer of the cardia or of the pylorus. The symptoms diagnostic of cancer of the cardia are dysphagia, regurgitation of food, obstruction in the passage of the oesophageal bougie, and sinking in of the epigastric region in consequence of atrophy of the stomach. It has already been said that catheterization of the oesophagus does not always afford the evidence of obstruction which one would expect. Cancerous stenosis of the cardia is to be distinguished from cicatricial stenosis in this situation. The diagnosis is based upon the history of the case, which is generally decisive, and upon finding fragments of cancer in the tube passed down the oesophagus.

That the cancer is seated at the pylorus is made evident by the situation of the tumor (see p. 561) and by the existence of dilatation of the stomach. There are many more causes of stenosis of the pylorus than of stenosis of the cardia, so that, notwithstanding the absence of tumor, cancer of the cardia is often more readily diagnosticated than cancer of the pylorus.

The greatest difficulty in diagnosis is presented by cancers which do not obstruct the orifices of the stomach. Many of these cancers run an almost latent course so far as the gastric symptoms are concerned, and in case they produce no recognizable tumor and are unattended with hemorrhage, the difficulties in their diagnosis are almost insurmountable.

In general, a diagnosis of the particular form of cancer which is present cannot be made, nor is such a diagnosis of any practical value. In very exceptional cases such a diagnosis might be made by the examination of secondary subcutaneous cancers⁹⁶ or of fragments found in the fluids obtained from the stomach.

⁹⁶ As for example, in Finlay's case, already referred to (p. 567). It is not safe to trust implicitly in this criterion, as the subcutaneous tumors may be of a different nature from the tumor of the stomach, as in an interesting case observed by Leube (op. cit., p. 125).

Although the diagnosis of gastric cancer can generally be made before the death of the patient, unfortunately a positive diagnosis in the early stages of the disease is usually impossible. Should resection of cancer of the stomach become a legitimate operation in surgery, it will be of the utmost importance to make the diagnosis in an early stage of the disease. Only those cases are suitable for resection in which there are no secondary deposits, the general health of the patient is in fair condition, and extensive adhesions have not been formed. It was to be hoped that the ingenious instrument devised by Mikulicz for exploring the interior of the stomach by electrical illumination would prove a valuable aid in diagnosis. The gastroscope in its present construction, however, has proved of little value.⁹⁷ It is, moreover, difficult to manipulate, and is not free from danger to the patient. We may be permitted, however, to hope for improvement in this direction.

⁹⁷ Mikulicz has observed with the gastroscope in a case of pyloric cancer immobility of the pylorus and absence of rugæ in the mucous membrane of the pyloric region (Wiener med. Wochenschr., 1883, No. 24). It does not seem probable that there can be anything peculiar to cancer in these appearances.

In cases in which there is reasonable suspicion of the existence of gastric cancer, and in which there is proper ground to contemplate resection of the tumor, it is justifiable to make an exploratory incision into the abdomen. It can then be decided whether or not cancer exists, and whether the case is suitable for operation. When this incision is made with all of the precautions known to modern surgery, it is attended with little or no danger,⁹⁸ and it should not be made except by surgeons who are practically familiar with these precautions.

⁹⁸ Of 20 exploratory incisions for tumor of the stomach performed by Billroth, not one had ended fatally (Deutsche med. Wochenschrift, 1882, ii.).

PROGNOSIS.—There is no proof that cancer of the stomach has ever ended in recovery. It may be admitted that partial cicatrization of gastric cancer may occur. We have, however, no sufficient reason to believe that cancer of the stomach has ever been completely destroyed by any process of nature or by any medicinal treatment.

A successful resection of a cancer of the pylorus by Billroth in January, 1881, made a great sensation in the medical world. Since that time the operation has been performed successfully ten times, and with fatal issue twenty-seven times. A radical cure has not, however, been effected, although life has been prolonged for a year and a half after the operation.⁹⁹ The possibility of permanent cure of gastric cancer by extirpation must be admitted. Enthusiasm over this possibility, however, is seriously lessened by the fact that a radical cure is not to be expected unless the operation is undertaken when the tumor is of small size, has produced no distant metastases, is free from many adhesions, and the patient is not greatly prostrated. In view of the difficulty of diagnosis in the early stages it is not likely that these favorable conditions can be fulfilled except in the rarest instances. Metastases may already exist when the tumor is small and before it has given rise to any symptoms.¹⁰⁰ Pylorectomy, moreover, will probably be successful in the hands of only comparatively few surgeons. It is therefore but a feeble glimmer of hope which is now admitted to the hitherto relentlessly fatal forecast of this disease.

⁹⁹ Several of the patients are still living (1884), but, so far as I can learn, no patient has survived the operation more than a year and a half.

¹⁰⁰ Birch-Hirschfeld relates a case in which a non-ulcerated cancerous tumor not larger than a silver half-dollar was found in the pyloric region of the stomach of a woman who died from injury. The tumor had given rise to no symptoms. Nevertheless, numerous metastases existed in the lymphatic glands of the omentum and of the lesser curvature (Jahresb. d. gesellschaft f. Natur u. Heilk. im Dresden [1882-83], 1883, p. 37).

TREATMENT.—Even up to the present time various drugs have been vaunted as effecting a radical cure of cancer of the stomach. Some of these, such as mercury, are positively harmful; others, such as conium, belladonna, and condurango, are often palliative; but not one has been proven to be curative. Since its recommendation by Friedreich in 1874, condurango has enjoyed the greatest vogue. The few observations in which, under the use of this agent, tumors, real or apparent, of the stomach have lessened in size or disappeared, admit of other interpretations than as cures of gastric cancer. There is, however, considerable testimony as to the virtues of condurango as a stomachic. In some cases it relieves the pain, vomiting, and indigestion of gastric cancer, but in many cases it is employed without benefit. The drug which passes by the name of condurango in the market is a very variable preparation. According to Friedreich's directions, decoction of condurango is prepared as follows: Macerate oz. ss of cort. condurango for twelve hours with fluidounce xij of water; then boil down to fluidounce vj and strain. The dose is a tablespoonful two or three times daily. The decoction of condurango may be combined with syr. aurantii cort.

While all specific treatment of gastric cancer is to be abandoned, much can be done for the relief and comfort of the patient. The treatment is symptomatic.

In general, the indications are similar to those in gastric ulcer. It is not necessary, however, to restrict the diet to the same extent as in gastric ulcer. The patient's tastes may be consulted to a considerable extent. Still, it will be found, as a rule, that the patient is most comfortable when his diet is confined to easily-digestible substances, such as milk, beef-juice, Leube's beef-solution, rare beefsteak, and other articles mentioned under the treatment of gastric ulcer.

The pain of gastric cancer will usually require the administration of opium in some form. There is manifestly not the same objection to the employment of narcotics in a necessarily fatal disease like cancer as in ulcer of the stomach. Opium may be given in pill form or as the tincture or deodorized tincture, or often most advantageously as hypodermic injections of morphia, to which atropia may be added.

Vomiting is sometimes controlled by regulation of the diet, particularly by iced milk. For this symptom also opium or morphia is often necessary. In addition, the customary remedies for relief of vomiting, such as bits of ice, iced champagne, soda-water, hydrocyanic acid, oxalate of cerium, creasote, may be tried. Cold or hot applications to the abdomen and mild counter-irritants, such as mustard plaster or turpentine stupes, sometimes afford relief. If the vomiting be incoercible, it may be well to administer food for a short time exclusively by the rectum, and in case of stenosing cancer of the cardia this method of administering food may be the only one possible.

Acid eructations and heartburn are often relieved by the antacids, as bicarbonate of sodium, lime-water, or calcined magnesia. Against fermentative processes in the stomach have been recommended salicylate of sodium, creasote, carbolic acid, and the alkaline hyposulphites. Charcoal tablets are as useful as, and less likely to disagree than, other antifermentatives. In view of V. d. Velden's investigations, already mentioned, the administration of dilute hydrochloric acid in an hour after a meal is indicated.

Excellent results have been obtained by regularly washing out the stomach in cases of gastric cancer, particularly in pyloric cancer with dilatation of the stomach. By this procedure pain, indigestion, and vomiting are often greatly relieved, and the patient experiences a renewed sense of well-being. Unfortunately, the benefit is only temporary. The syphon process is most conveniently employed. Contraindications to the use of the stomach-tube are copious gastrorrhagia and great weakness of the patient.

When constipation is not relieved by washing out the stomach, enemata should be employed. Drastic purgatives should not be given.

For diarrhoea opium may be given, particularly in the form of small enemata of starch and laudanum.

Scanty hemorrhage in the form of coffee-grounds vomiting requires no treatment. Copious hæmatemesis is to be treated according to the principles laid down under the treatment of hemorrhage from gastric ulcer.

Discussion of the surgical treatment of gastric cancer of course does not belong to this work. The opinion entertained by the physician as to the propriety of surgical interference in gastric cancer is not, however, a matter of indifference, for cases of gastric cancer come first into the hands of the physician, and generally only by his recommendation into those of the surgeon. So long as the physician stands absolutely powerless before this disease, his general attitude as to the propriety of surgical interference should not be one of hostility. Experience only can determine the justification of surgical operation in cases of gastric cancer. As yet, it is too soon to express a positive opinion as to the value of resection of gastric cancer. Of 37 published resections of cancer of the pylorus, 27 died from the effects of the operation, and of the fatal cases 18 within the first twenty-four hours. These results are certainly not calculated to awaken much enthusiasm for the operation. Still, it would be wrong to draw definite conclusions from the existing statistics of resection of the cancerous pylorus, partly because the number of operations is as yet too small, partly because the operation has been done when it was certainly unwarrantable according to the best judges (Billroth, Czerny), and chiefly because the number of operators in proportion to the number of operations is too great. For the 37 published operations there have been 27 operators. Ovariotomy was not considered a justifiable operation until the excellent results of individual operators were obtained. It is probable that to an even greater extent resection of the pylorus will become the specialty of certain operators. Therefore, before concluding as to the value of resection of cancer of the stomach it is necessary to await the results of individual surgeons in a series of cases.¹⁰¹

¹⁰¹ Already, from this point of view, the operation appears more hopeful. Czerny has performed 6 resections of the stomach with only 2 fatal results; 4 of the operations were pylorectomies for cancer. Billroth has performed the operation 8 times with 3 fatal results (Wiener med. Wochenschrift, 1884, Nos. 17 and 29).

So much, however, is now certain, that with our present means of diagnosis the number of cases suitable for extirpation is very small.¹⁰² A radical cure is to be expected only in the rarest instances, so that the value of the operation will depend chiefly upon the condition of the patient after its performance. As regards this point, the results in the successful cases have been encouraging. In several instances the terrible sufferings of the patient have given place to months of comparative health and comfort.

¹⁰² Billroth at the eleventh session of the Congress of German Surgeons said that he was amazed at the number of resections of the pylorus which had been performed. Out of 50 to 60 cases of gastric cancer, only 1 appeared to him suitable for operation.

In cases of extreme cancerous stenosis of the pylorus which are not suitable for resection Wölfler proposed forming a fistulous communication between the stomach and the small intestine (gastro-enterostomy). The results of the operation have not been encouraging. Out of six cases in which this operation has been performed, only two patients lived after the operation.

For the same condition Schede proposed making a duodenal fistula (duodenostomy), but I am not aware that the operation has been performed.

The results of gastrostomy for relief of cancerous stenosis of the cardia or of the oesophagus have not been encouraging.¹⁰³

¹⁰³ Of 76 cases of gastrostomy for the relief of cancer of the oesophagus or of the cardia, only 14 lived over thirty days (Leisrink and Alsberg, Arch. f. klin. Chir., Bd. 28, p. 760, 1882).