HEMORRHAGE FROM THE STOMACH.
BY W. H. WELCH, M.D.
Hemorrhage from the stomach is a symptom, and not a disease. It is a result of a great variety of morbid conditions in the description of which it receives more or less consideration. Already the symptomatology and treatment of hemorrhage from the stomach have been considered in connection with its two most important causes—namely, gastric ulcer and gastric cancer. It remains to give a summary of the etiology and diagnosis of gastric hemorrhage.
Hemorrhage from the stomach is also called gastrorrhagia. The term hæmatemesis is not synonymous with gastric hemorrhage, for blood may be vomited which has simply been swallowed or has passed from the intestine into the stomach.
ETIOLOGY.—The causes of gastric hemorrhage are as follows:
1. Ulcer of the Stomach.—Simple gastric ulcer is the most frequent cause of abundant hemorrhage from the stomach. Tuberculous gastric ulcers, typhoid gastric ulcers, and the ulcers of phlegmonous gastritis are extremely rare causes of hemorrhage. Hemorrhagic erosion of the stomach, which by many writers is assigned an important place in the etiology of gastric hemorrhage, is not an independent affection, and in my opinion is without any clinical significance.
2. Cancer of the Stomach.—(Non-cancerous tumors of the stomach hardly deserve mention in this connection, so infrequently are they the cause of gastric hemorrhage.)
3. Traumatism (mechanical, chemical, thermic).—a. Acting from without the stomach: severe injury to the abdomen, as by a blow or a fall; penetrating wounds of the stomach.
b. Acting from within the stomach: foreign bodies, particularly sharp-pointed ones; corrosive poisons, as acids and alkalies; other toxic inflammatory irritants; and very hot substances. Here should also be mentioned injury from an inflexible stomach-tube and aspiration of mucous membrane with the stomach-pump.
4. Diseases of the Gastric Blood-vessels.—a. Aneurism of the arteries of the stomach. Miliary aneurisms have been found by Galliard and others as a cause of profuse and even fatal hemorrhage from the stomach. Especially in obscure cases should careful search be made for miliary aneurisms.
b. Varices of the veins are a not unimportant cause of gastric hemorrhage. They are most frequently associated with chronic passive congestion of the stomach, but they may be found without any apparent disturbance of the circulation.
c. Degenerations of the vessels, particularly fatty and atheromatous degeneration of the arteries. Probably gastric hemorrhage in phosphorus-poisoning is to be attributed to fatty degeneration of the arteries. Amyloid degeneration of the blood-vessels is a doubtful cause of hemorrhage.
5. Active Congestion of the Stomach.—Here is usually placed gastric hemorrhage as a result of severe inflammation of the stomach (as acute catarrhal gastritis), although in these cases the inflammatory alteration of the vascular walls is an equally important factor.
With more probability the so-called vicarious hemorrhages from the stomach are to be assigned to active congestion. Sceptical as one is inclined to be as regards vicarious hemorrhages of the menses, the occurrence of such hemorrhages, although rare, must be admitted. Doubtful, however, are alleged cases of gastric hemorrhage taking the place of suppressed hemorrhoidal bleeding or of epistaxis.
6. Passive Congestion of the Stomach.—This embraces an important group of causes of gastric hemorrhage. This hemorrhage is the result of venous congestion caused by some obstruction to the portal circulation. The obstruction may be—
a. In the portal vein itself or its branches within the liver, as in pylethrombosis, cirrhosis of the liver, tumors, such as cancer or echinococcus cysts, compressing the portal vein, occlusion of capillaries in the liver by pigment-deposits in melanæmia, and dilatation of the bile-ducts in the liver from obstruction to the flow of bile. Next to ulcer and to cancer of the stomach, cirrhosis of the liver is the most frequent and important cause of gastrorrhagia.
b. In the pulmonary blood-vessels, as in pulmonary emphysema, chronic pleurisy, and fibroid induration of the lungs.
c. In the heart in consequence of uncompensated valvular and other diseases of the heart.
For evident reasons, obstruction of the pulmonary or of the cardiac circulation is much less likely to cause gastric hemorrhage than is obstruction in the portal vein or the liver.
Possibly, gastric hemorrhage which is caused by violent acts of vomiting may be caused by venous congestion of the mucous membrane of the stomach. In support of this view, Rindfleisch advances the idea that the veins in the muscular layers of the stomach, in consequence of the thinness of their coats, are much more likely than the arteries to suffer from the compression of the muscle during its contraction.
The occasional occurrence of gastric hemorrhage during pregnancy has also been attributed to passive congestion of the stomach.
7. Acute Infectious Diseases—namely, yellow fever, acute yellow atrophy of the liver, relapsing fever; less frequently cholera, typhoid fever, typhus fever, diphtheria, erysipelas, and the exanthematous fevers, small-pox, measles, and scarlet fever.
The cause of gastric hemorrhage in these diseases is not understood. The usual explanation attributes the hemorrhage to dissolution of the blood-corpuscles and secondary alteration of the walls of the blood-vessels. Plugging of the vessels with micro-organisms has been found in only a few instances. The gastrorrhagia of acute yellow atrophy of the liver has been attributed to dissolution of the blood, not only by some infectious agent, but also by constituents of the bile, and also to obstruction of the portal circulation by destruction and occlusion of capillaries in the liver.
8. Other Constitutional Affections.—a. Hemorrhagic diatheses—namely, scorbutus, purpura, and hæmatophilia. Strictly speaking, a hemorrhagic diathesis exists in other affections of this class.
b. Malaria. Here we may distinguish, first, periodical malarial hemorrhages from the stomach which are cured by quinia; second, pernicious gastric malarial fever, of very grave prognosis; and third, hemorrhages in malarial cachexia due to extreme anæmia. Cases which have been described as malaria with scorbutic complications belong mostly to the last variety. Mention has already been made of gastric hemorrhages attributed to malarial pigmentation of the liver.
c. Profound Anæmias. The most important affections in this category are progressive pernicious anæmia, leucocythæmia, and pseudo-leucocythæmia, including the so-called splenic anæmia.
d. Cholæmia. The hemorrhage is attributed to dissolution of the blood-corpuscles by the action of the biliary salts.
Gastric hemorrhage is a rare event in Bright's disease, occurring more especially with small kidneys. In one such case I found that the fatal hemorrhage was due to the bursting of a miliary aneurism of a small artery in the submucous coat. Probably in all similar cases the hemorrhage is referable to disease of the vascular walls.
9. Neuropathic Conditions.—Although ecchymoses in the mucous membrane of the stomach can be experimentally produced by injury of various parts of the brain and spinal cord, there is no proof that gastric hemorrhage which is of any clinical importance is referable to structural diseases of the nervous system. The occasional occurrence of gastric hemorrhage in progressive paralysis of the insane, in tuberculous meningitis, in epilepsy, is to be attributed to other causes.
In lack of a better explanation, however, the gastric hemorrhages which have been occasionally observed in hysterical women may be classified here. These constitute not the least important class of gastric hemorrhages. The hemorrhages from the stomach in chlorosis belong partly here and partly to anæmia.
10. Melæna Neonatorum.—Although in some cases ulcers have been found in the stomach or duodenum, and in others a general hemorrhagic diathesis exists, it must be said that the etiology of this grave disease is still very obscure.
11. Bursting of Aneurisms or of Abscesses from without into the Stomach.
12. Idiopathic Causes.—Under this unsatisfactory designation are included cases which are aptly described by Flint1 in the following words: "Hemorrhage sometimes occurs from the stomach, as from the bronchial tubes, the Schneiderian membrane, and in other situations, without any apparent pathological connections, neither following nor preceding any appreciable morbid conditions. It is then to be considered as idiopathic." A person in apparent health has suddenly a hemorrhage, often profuse, from the stomach, which is followed only by symptoms immediately referable to the hemorrhage. The hemorrhage is naturally the source of great anxiety. Ulcer or cancer of the stomach or some other grave disease is usually suspected. But the patient develops no further symptoms, and often never has another hemorrhage. Whatever hypotheses one may construct for these cases of so-called idiopathic hemorrhage, the recognition of the clinical fact of their occurrence is important.
1 Austin Flint, A Treatise on the Principles and Practice of Medicine, 5th ed., p. 513, Philada., 1881.
Inasmuch as hemorrhage from the stomach is usually made manifest by the vomiting of blood, it is important to bear in mind that gastric hemorrhage is not the only cause of hæmatemesis. Blood may gain access to the stomach by being swallowed in cases of hemorrhages from the nose, mouth, throat, bronchi, lungs, and oesophagus. Blood may also enter the stomach from the duodenum in cases of simple ulcer of the duodenum or of typhoid ulcers situated in the upper part of the intestine.
The SYMPTOMS of hemorrhage from the stomach have already been described in connection with ULCER OF THE STOMACH.
MORBID ANATOMY.—As is evident from the enumeration of the causes of gastric hemorrhage, the lesions found after death are manifold. A description of these lesions, however, does not belong here. It is necessary, however, to say a few words concerning the demonstration of the source of the hemorrhage.
The hemorrhage is arterial, venous, or capillary in origin. Ulcerations from the stomach into the heart, which have been mentioned in connection with gastric ulcer, are too infrequent to come into consideration in this connection. If the bleeding is from a large artery or vein or from a medium-sized aneurism or varix, the demonstration of its source is not difficult. Often, however, in cases of fatal gastric hemorrhage the search for the source of the hemorrhage has proved fruitless. It is supposed that in many such cases the hemorrhage is due to diapedesis, and not to rupture of a blood-vessel (rhexis). Doubtless, small gastric hemorrhages, particularly those into the tissues of the stomach, are often the result of diapedesis, but in cases of profuse hemorrhages from the stomach where the source of the hemorrhage is not demonstrable after death, the convenient assumption of hemorrhage by diapedesis, in my opinion, plays too important a rôle. In most cases of profuse gastrorrhagia the symptoms point to a sudden outpouring of blood into the stomach; and our knowledge of diapedesis does not warrant the belief that the red blood-corpuscles can escape through the unsuffused walls of the vessels with that combined rapidity and abundance which would be necessary to explain the sudden and profuse hemorrhage. In these cases hemorrhage by rhexis is altogether more probable even when ruptured vessels cannot be demonstrated. It often requires a long-continued and careful search to find a small vessel which is ruptured. In the case above mentioned of fatal hæmatemesis from miliary aneurism over an hour of continuous searching was required to find the pinhole perforation in the mucous membrane in the bottom of which lay the small aneurism. Chiari2 has reported a fatal gastric hemorrhage due to rupture of a submucous vein. The erosion of the mucous membrane was not larger than a hempseed. The examination of the mucous membrane is often much impeded by the closely-adherent mucus and blood, which have to be carefully removed. Injection of the vessels of the stomach with fluid may aid in finding a ruptured vessel. It is not intended to assert that in all cases of fatal gastric hemorrhage a painstaking search would reveal the source of the hemorrhage, but it is believed that cases of fatal gastric hemorrhage would less frequently be reported with negative anatomical result as hemorrhages by diapedesis, or by exhalation, or as parenchymatous or capillary hemorrhages, if such a search were made.
2 Prag. med. Wochenschr., 1882, No. 50.
DIAGNOSIS.—Undoubtedly, small hemorrhages into the stomach often occur which are not recognized, and in the absence of vomiting even larger hemorrhages may escape detection unless a careful examination of the stools be made. The patient may die from abundant gastric hemorrhage before any blood has been vomited or has passed by the bowels.
When there is doubt whether the black color of the stools is due to blood or to the administration of iron or of bismuth, it generally suffices to add water to the stools. If blood be present, the water will acquire a reddish color. Should doubt still remain, then the microscope, the spectroscope, or the test for hæmin crystals may be called into requisition (see p. [545]).
Whether black, tarry stools are produced by hemorrhage from the stomach or by hemorrhage in the upper part of the intestine can be decided only by the clinical history. If hæmatemesis be likewise present, the presumption is strongly in favor of gastric hemorrhage. The diagnosis, however, between hemorrhage from duodenal ulcer and that from gastric ulcer is very difficult, and can rarely be positively made.
The mere inspection of the vomit is generally sufficient to determine whether it contain blood or not. Color more or less resembling that of altered blood may be produced in the vomit by iron, bismuth, red wine, various fruits, such as cranberries, and by bile. When a careful inspection by the physician leaves doubt as to the presence of blood—which will rarely be the case—then here also recourse may be had to the microscope, spectroscope, and test for hæmin crystals.
Sometimes blood is swallowed and then vomited by hysterical females or by malingering soldiers or prisoners for purpose of deceit. In such cases there are generally no evidences of acute anæmia or of gastric disease. The blood of some animals can be distinguished by the microscope from human blood. When suspicion of deceit exists, there are generally various ways of entrapping the patient.
When blood is vomited by nursing infants the possibility of its coming from the breast of the mother is to be thought of.
Sometimes blood from the nose or throat is swallowed, particularly when the bleeding occurs during the night. This blood may subsequently be vomited. The inspection of the nose or throat will generally reveal the source of the hemorrhage in such cases.
The diagnosis between hemorrhage from the oesophagus and that from the stomach must be based upon the clinical history. The oesophagoscope, however, has been successfully employed for diagnostic purposes. Several cases have been reported of fatal hemorrhage from varices of the oesophagus. Such hemorrhage cannot be distinguished from gastric hemorrhage.
Much more frequently arises the question whether the hemorrhage is from the stomach or from the lungs. Sometimes the decision of this point is very difficult, and it may even be impossible, especially when the physician is obliged to trust only to the statements of the patient or his friends. Difficulty in the diagnosis results mainly from the fact that coughing and vomiting of blood are often associated with each other. With hæmoptysis blood may be swallowed and then vomited, and with hæmatemesis more or less coughing occurs. The diagnosis is to be based upon the points contained in the following table:
| HÆMOPTYSIS. | HÆMATEMESIS. |
| 1. Usually preceded by symptoms of pulmonary or of cardiac disease. Bronchial hemorrhage, however, without evidence of preceding disease, is not rare. | 1. Usually preceded by symptoms of gastric or of hepatic disease, less frequently by other diseases (see [Etiology]). |
| 2. The attack begins with a tickling sensation in the throat or behind the sternum. The blood is raised by coughing. Vomiting, if it occurs at all, follows the act of coughing. | 2. The attack begins with a feeling of fulness in the stomach, followed by nausea. The blood is expelled by vomiting, to which cough, if it occurs, is secondary. |
| 3. The blood is bright red, fluid or but slightly coagulated, alkaline, frothy, and frequently mixed with muco-pus. | 3. The blood is dark, often black and grumous, sometimes acid, and usually mingled with the food and other contents of the stomach. |
| If the blood has remained some time in the bronchi or a cavity, it becomes dark and coagulated. | If the blood is vomited at once after its effusion, it is bright red and alkaline, or it may be alkaline if it is effused into an empty stomach. |
| 4. The attack is usually accompanied and followed by localized moist râles in the chest, and there may be other physical signs of pulmonary or of cardiac disease. | 4. After the attack the physical examination of the lungs is usually negative, but there are generally symptoms and signs of gastric or hepatic disease. |
| Bloody sputum continues for some time, often for days, after the profuse hemorrhage ceases. | Black stools follow profuse hæmatemesis. |
As it is important that the patient should be as quiet as possible during and for some time after the hemorrhage, any physical examination which disturbs the patient, such as percussing the posterior part of the chest or palpating the abdomen, should be avoided.
The diagnosis of the many causes of gastric hemorrhage belongs to the description of the various diseases which have been enumerated under the etiology.
PROGNOSIS.—It is exceptional for gastric hemorrhage to prove immediately fatal. According to Fox, such an occurrence is more frequent with cirrhosis of the liver than with ulcer or cancer of the stomach. The ultimate result of the hemorrhage depends greatly upon the previous condition of the patient. If this condition was good, he often rallies from the most desperate prostration immediately following the hemorrhage. A previously enfeebled patient is of course more likely to yield to the further anæmia and exhaustion caused by profuse hemorrhage. Although the symptoms of gastric ulcer and of cirrhosis of the liver are sometimes improved after hemorrhage from the stomach, nevertheless this hemorrhage can never be regarded as a welcome event.
For the treatment of gastric hemorrhage see ULCER OF THE STOMACH.