The Measurement of Gastric Acid.

Since the end of the last century the standard method of assessing gastric acid secretion has been by the test meal, which was developed largely by von Leube (1876), Ewald and Boas (1885) and Rehfuss (1927). Its history and development have been reviewed by Hollander and Penner (1939) and Rovelstad (1963). The vast amounts of data collected in these original studies contributed only little to our knowledge of the secretions of the stomach in health and in disease. With the recent development of precise quantitative methods of assessing gastric function, such investigations are now largely of historical interest. The 'fractional test meal', still performed today, differs little from that described by Rehfuss in 1914. Following a fast, a gastric tube is passed and the stomach emptied. A 'test meal' of carbohydrate-bread, cereal, gruel or biscuits, is ingested and 10 ml. of gastric juice aspirated each 15 minutes for 2-3 hours. The acidity in each sample is estimated by titration with 0.1 N NaOH using two indicators Topfer's dimethylaminoazobenzene for 'free', and phenolphthalein for 'total' acid. The acidity curve is drawn on a chart which shows a 'normal range' based on the curves of 80 of the 100 normal students studied by Bennett and Ryle in 1921. It is of interest that the other 20 students showed curves from extreme hypoto hypersecretion. The fractional test meal has little to commend it as a test of gastric function. It is neither precise, quantitative nor reproducible. The coefficient of variation of the concentration of free acid in a series of 20 consecutive meals performed on a healthy adult ranged from 25.8 to 199.6%-quite outside acceptable limits (Bell and McAdam, 1924). It is therefore not surprising that Enticknap and Merivale (1954), in an analysis of 1000 gruel meals, concluded that they were of little diagnostic value. In a comparison with the augmented histamine test in 114 patients, Marks and Shay (1960) found that, although the peak acidity in a fractional test meal was statistically related to the acid output in the augmented histamine test, the results of a test meal could not be used to predict the true acid output in an individual patient. As the 'test meal' was a poor stimulus of gastric secretion, differing little from plain water (Bergeim, Rehfuss and Hawk, 1914), the incidence of achlorhydria was fairly high. More potent stimuli were used-alcohol (Kast, 1906; Ehrmann, 1912) caffeine (Katsch and Kalk, 1924) and histamine, either in single (Bockus and Bank, 1927; Vanzant, Alvarez, Eustermen, Runn and Berkson, 1932), double (Rivers, Osterby and Vanzant, 1936), or triple (Hitchock, Sullivan and Wangensteen, 1955) doses. However, even with a standard dose of histamine (usually 0.01 mg. histamine acid phosphate per kg. body weight) the response had a wide variation in normal subjects and proved of little routine diagnostic value (Polland and Bloomfield, 1931; Polland, 1933).