T

HE results obtained by various investigators with the fractional test meal in cases of gastric, duodenal and mixed ulcers have varied from anacidity, all the way through to hyperacidity. Definite diagnostic value has been claimed for this test by most of the men who have studied it extensively. These investigators do not agree on any common ground as the basis for the interpretation of the curves. Our study was undertaken with the hope of adding evidence as to the diagnostic value of this test in gastric and duodenal ulcer. This report covers 5 normal cases, 9 cases of gastric ulcer, 10 cases of duodenal ulcer, and 5 cases of mixed ulcer (gastric and duodenal).

Schüle (1), in 1895, used a catheter for withdrawing the digestive fluid at various stages of digestion and was the first to make extensive fractional analysis. Ehrenreich (2), in 1912, also used the catheter to make a large number of fractional studies. It was not, however, until Rehfuss (3) published his article entitled "A New Method of Gastric Testing" that fractional method received any marked attention. He called attention to a new tip designed by Palfski. This is attached to a small bore rubber tube.

ANNALS OF CLINICAL MEDICINE, VOL. III, NO. 5

Rehfuss says "this represents a practical method for following the whole gastric digestion with minimum of discomfort enabling us to construct a curve which graphically represents every phase." Rehfuss, Bergeim and Hawk (4) made an intensive study of gastric secretion and classified the curves obtained in normal individuals under the terms, iso-, hyper-, and hyposecretory types. This classification is based on four considerations: (1) a period of ascension (this usually occupies the first thirty minutes and indicates the rapid responses to a known stimulus); (2) the character and height of the high point, (a) accelerated or (b) retarded and whether (a) abrupt or (b) sustained; (3) period of descent or decline; and (4) the character and modification of food residue.

In 1915, Rehfuss (5) concluded that the results of a fractional test meal are of infinitely more value in furnishing information regarding the secretory and motor functions of the stomach than the single test meal. Later the same author (6) pointed out that the protein curve adds to the diagnostic value of the acid curve.

Talbot (7), in 1916, presented 150 cases studied by the fractional method. He concludes that the method has a great field of usefulness. He found,

however, a great variation in the normal individual response to a test meal. He stated that four factors affect the contour of the normal curve: namely, (1) intensity and duration of appetite secretion as influenced by the test meal; (2) the amount of gastric mucus; (3) the concentration of the food secretion; and (4) the emptying time of the stomach.

Crohn and Reiss (8), in 1917, studied gastric secretion with the fractional method and concluded that there is no curve that is absolutely characteristic for ulcer, but they believe that ulcer cases are associated with a curve of acidity that rises rapidly and is sustained, if not throughout the cycle of digestion, at least to within a short time (fifteen minutes) of its end. In their qualitative tests bile was usually absent, mucin was not associated with ulcer, blood was usually absent. On account of the small tendency to trauma, in the method, the authors attach great significance to blood when found.

They discuss the diagnostic significance of the curves as follows:

"One who expects to find in the curves of the fractional estimation of stomach contents new and infallible diagnostic criteria will be disappointed. One who hopes to label each type of curve as representative of a particular disease will raise false hopes. On the other hand, many of the curves are highly suggestive of certain diseases. There is a type of curve quite constant for ulcer. For the demonstration of digestive hypersecretion the method is unequalled. But the importance and greatest value of the method lies in the fact that it offers knowledge regarding the physiology of the entire cycle of digestion from the beginning of the fasting state through to the very end of the digestive period."

Horner (9) in an extensive study of the fractional test meal on normal and pathological cases, concludes that there are fairly constant characteristics of the curve for certain diseases. Among these pathological conditions, he believes that gastric and duodenal ulcer each have rather a characteristic curve.

In 1918, Rehfuss (10) concludes that there is no one form of curve seen in health. He believes that the curves fall into three classes, depending upon the type of individual. He also believes that no acid figures occur in disease that are not duplicated in health. He found that 45 per cent of healthy individuals showed the socalled hyperacidity and 42 per cent of the ulcer series showed the same. far as duodenal ulcer is concerned he claims that the most characteristic finding is that of positive blood at the phases of tryptic regurgitation. In this connection, he further states that in a large group of duodenal ulcer cases, there is present a late hypersecretion, accompanied by periodic regurgitation of duodenal material giving occult blood reaction.

Crohn and Reiss (11), in 1920, studied 34 cases of gastric ulcer with the following results: 11 cases had iso-, or hyposecretory curves, and 23 cases, hypersecretory curves. Six of the 11 cases with iso-, or hyposecretory, showed a lowering of the acid curves with treatment. Five were unaffected. Of the 23 cases with marked hyperacidity only 7 showed a lowering of the acid curves, the remaining 16 being unaffected by persistent and rigid dieting and resting in bed. The conclusions of these investigators are that clinical improve

ment can take place independently of whether the hyperacidity curve is lowered or not and is apparently not dependent upon the hyperacidity. The authors speak of the necessity for a happy frame of mind in the patient. Regarding the diagnostic value of the fractional test meal the authors make the following statement: "On the other hand innumerable instances are met with in which slighter grades of delayed emptying are evident by this chemical method, which evade entirely the roentgenologist."

Ryle, in London (12), in 1920, speaks about the increased amount of information obtainable through the fractional test meal as against the single test meal. He advocates the slight change in the composition of the test meal, which was proposed and used by Crohn.

In 1920, Rehfuss and Hawk (13) again discuss the interpretation of the gastric test meal. They state that an Ewald test meal in health is evacuated in from two to two and one-half hours. When a meal is evacuated in one hour and one-half or less, or requires more than three hours, they consider that such findings prove an abnormal stomach. They further state that

"Gastric ulcer can and does occur with a normal secretory and motor mechanism and, so far as we can demonstrate, no functional variation, a point which is simply another link in the chain of evidence to indicate that certainly some forms of ulcer have neither secretory nor motor variations as etiologic factors, but some other cause, probably toxic, to explain their presence."

Lyon, Bartle, and Ellison (14), in 1921, in an article on fractional gastric analysis discuss the increased amount

of information obtainable by this method and also emphasize the value of the information obtainable by the microscopic examination of these fractions and the results of the qualitative tests for bile, blood, mucus, pus, and

enzymes.

Ryle (15), in 1921, concludes that hyperacidity occurs in duodenal ulcer and pyloric ulcer more frequently than in any other condition. Hyperacidity is evidence that secretion is proceeding at a rate too great to be countered by such neutralizing factors as regurgitation from the duodenum and the secretion of mucus.

Nausea and saliva

Ryle (16) in continuation of experiments with the fractional gastric analysis made a large series of personal experiments. He found no difficulty in swallowing the tube after the first trial and no unpleasant symptoms were noted. tion were absent. Talking and even mastication were not interfered with. He found that milk was a marked stimulant for acid secretion, which was not completely neutralized by this food. He suggests that in hypersecretion cream, gruels, and sugar might be more desirable than milk. He finds tea of high value in stimulating acid secretion.

Ryle (17) in association with Bennet, Guy's Hospital, London, in 1921, came to the conclusion that an insufficient number of normals had been studied by means of the fractional test meal. They studied in detail 100 cases using students at Guy's Hospital as the subjects. The meal used was the gruel meal proposed by Crohn and has slightly less stimulating effects than the Ewald test meal. They came to the conclusion

that there was no justification for the classification of normal cases proposed by Rehfuss and Hawk under the names, iso-, hyper-, and hyposecretory types. They published two charts, one showing the different percentage curves, and the other showing what they regard as the normal limits of variation.

Bell (18) studied 425 cases in which gastric analysis was made by the fractional method. According to the Ryle and Bennet standard in 24 cases of proven gastric ulcer, he found 1 achlorhydria, 3 hypochlorhydrias, 4 low normals, 5 normals, 6 high normals, and 5 hyperchlorhydrias. In 34 cases of duodenal ulcer, the classification was as follows: 4 low normals, 3 normals, 9 high normals, and 18

hyperchlorhydrias.

Bell (19) later concludes that gastric ulcer is far more frequently associated with a diminished acidity than is duodenal ulcer, and it is exceptional to find complete achlorhydria with the latter. He says that neither the site or the duration of the ulcer appear to predispose to achlorhydria.

Carlson (20) believes that gastric secretion may vary from hypersecretion through normal down to complete anacidity without producing any disease symptoms. He says that there is no disease known capable of inducing true hyperacidity and that normal tissues are unaffected by excessive acid secretion. Such secretions can only affect tissues that are already pathological.

EXPERIMENTAL

Test meal technique

The patient was instructed to eat a bunch of raisins or grapes with seeds in addition to his regular meal the night before the test meal is given, and to refrain from food and drink the following morning. The Rehfuss tube was passed and the stomach entirely emptied of any residuum. Following the complete emptying of the stomach the Ewald test meal was given. It consisted of 2 medium slices of toast, without butter (35 grams), and a cup of tea (250 cc.), without cream or sugar.

At intervals of exactly fifteen minutes from the time the test meal was eaten, until ten fractions had been taken, 5 to 10 cc. samples of the gastric contents were withdrawn from the stomach. This was accomplished through the Rehfuss tube using a 20 cc. Luer syringe as an aspirator. At the time the tenth fraction was withdrawn an effort was made to completely empty the stomach.

At no time was much force exerted, in passing of the tube of withdrawal of fractions, for fear of slight trauma and possible entrance of blood into the samples.

DISCUSSION OF RESULTS

In a study of the response to the test meal on the part of normal individuals, under our experimental conditions, the results obtained agree in general with those obtained by Rehfuss, Hawk and co-workers (chart 1).

In 9 cases of gastric ulcer, analysis of the curves according to the Rehfuss and Hawk standards, showed 3 of the

hyper- (H+), 1 iso- (I), 1 iso- with