Richard Earlam

CONTENTS

1. Hiatus hernia, duodenal ulcer and gallstones
2. Add gastrooesophageal reflux and oesophagitis
3. Heartburn and epigastric pain reproduction test
4. Duodenal ulcer pain
5. Gastrointestinal hormones
6. Gastric acid secretion
7. Bile reflux
8. Symptomatic analysis and questionnaire
9. Clinical correlations
10.What is the diagnostic endpoint?
11.References

HIATUS HERNIA   DUODENAL ULCER   GALLSTONES-

The three commonest upper gastrointestinal diseases are hiatus hernia, peptic ulcer which includes stomach or gastric ulcer, and gallbladder diseases of which the commonest are gallstones or cholecystitis. Their diagnosis is used as an endpoint for treatment. This may not be correct because the symptoms associated are not specific, they overlap with these pathological diagnoses and many patients, especially those seen in general practice, have symptoms but negative investigations. However, they need symptomatic treatment by doctors for their patho-physiological abnormalities even if they have no diagnosable pathology.

The original endpoint was the radiological diagnosis of the anatomical defect of hiatus hernia associated with heartburn and acid reflux which allowed a generation of thoracic surgeons to perform life time operation series of 1,000 to 2,000 patients. The best operation choice was controversial and I did my surgical thesis on the repair of hiatus hernia in dogs. This did not convince any of the senior surgeons to change their choice of operative procedure. But it did allow me to clarify the evidence provided by better and more accurate investigations using manometry, pH intraoesophageal recording of acid reflux and analyse the different thresholds for diagnosis of these abnormalities. Later studies investigated the pain of duodenal ulcer (15, 29) and gallstones (53)

The problem remains-which is the endpoint? The important step forward is to move from a purely pathological concept of disease to a pathophysiological approach based on understanding gastrointestinal physiology.

ADD GASTROOESOPHAGEAL REFLUX AND OESOPHAGITIS-

The diagnosis of an anatomical pathological defect of a displaced lower gastrooesophageal sphincter is called a hiatus hernia, which is best diagnosed by radiology. Using manometry the diagnostic threshold is lower and more sensitive so it increases the number diagnosed.

Gastrooesophageal reflux can be diagnosed by radiology and screening of barium reflux into the lower oesophagus. A more sensitive test is to use manometry and pH acid measurement.

The mucosal damage in the lower 10 cm of the oesophagus is called oesophagitis. The word includes both the macroscopic naked eye appearance as well as the microscopic histopathological findings. They are not the same and microscopy is more sensitive, but what does it actual mean?

• Are these three endpoints separate diseases?
• What is the relationship of one to another?
• Can they be correlated with symptoms?
• Can each one exist separately?
• Can symptoms occur without these endpoints?

HEARTBURN AND EPIGASTRIC PAIN REPRODUCTION TESTS -

Bernstein reproduced heartburn by pouring 200-300 ml 0.1N hydrochloric acid through a tube into the lower oesophagus, which confirmed that it arose from the lower oesophagus. Placing acid in the stomach itself was an unreliable and mainly negative way of producing the same heartburn symptoms (37). Bernstein did have a few hiatus hernia patients with heartburn which spread to the upper abdomen – the epigastrium.

The epigastric pain reproduction test (15, 29, 107) was a refined development of the Bernstein test for heartburn. The epigastric pain test dripped 20-30 ml in 4 min of the same strength acid directly and specifically onto the lower 10 cm of the oesophagus under manometric control. It was 100 % accurate if the patient had been awoken by such pain during the night before. Curiously if 200-300 ml of acid were placed directly in the stomach such pain was not reproduced. The author could have become worried because such research was against all textbook teaching, in spite of repeated warnings later under such headings as “Dogma Disputed” in the Lancet (83).

DUODENAL ULCER PAIN -

The pain of duodenal ulcer patients can be divided into those with epigastric pain and those with other sites for their abdominal pain (49). There is a third group without any pain but they have an ulcer in the duodenum which is silent; they often present as emergencies with a perforated ulcer or bleeding, never having had pain. This needs an explanation.

Manometrically the duodenal ulcer patients with epigastric pain have weaker sphincters and more reflux than those with non-epigastric pain; they are more likely to have heartburn (27, 32). When studied months after truncal vagotomy and pyloroplasty some patients had a positive epigastric pain reproduction test even though they had had no pain for months. Did they have a hypersensitive lower oesophageal mucosa which was stimulated at lower levels of acid reflux?

GASTROINTESTINAL HORMONES -

Gastrin and cholecystokinin are two gastrointestinal hormones which increase the sphincteric pressure of the isolated perfused canine gastrooesophageal sphincter. Both are secretomotor hormones. Gastrin produced from the antrum cause gastric acid secretion from the body of the stomach. Cholecystokinin causes the gallbladder to contract.

Secretin and glucagon are also secretomotor and both reduce gastrooesophageal sphincter pressures as a secondary effect. Secretin’s main effect is to cause pancreatic secretion. Glucagon’s main effect, as its name implies, is to raise blood sugar levels. It counteracts the effect of insulin but is probably the most potent hormone for reducing gastrooesophageal sphincteric pressures.

Work was done at the Royal London Hospital by Paul Thomas on these hormones experimentally in dogs and was thought to be an important promising field for later research (35). Clearly all these hormones had influence on both delayed gastric emptying and gastrooesophageal reflux through a weak sphincter together with bile reflux backwards from the duodenum into the stomach.

GASTRIC ACID SECRETION -

In the 1950s and 60s gastric acid was considered to be the main factor in the causation of peptic ulcers. Pentagastrin induced gastric secretion tests were done routinely in most of the professorial university surgical centres. The accepted expert was Jeremy Hugh Baron who wrote Clinical Tests of Gastric Function in 1978. He pointed out that although as a group acid secretion in duodenal ulcer patients was higher than normal, in fact one-third were normal, one-third subnormal and it was only the final third of high levels which pushed the group as a whole into high secretors. Analysis by Earlam confirmed Baron’s work that for the individual the results of pentagastrin tests did not contribute anything to help decision taking for surgery in duodenal ulcer patients (85).

An attempt was made to quantitate the amount of pain suffered by a duodenal ulcer patient by measuring his antacid relief tablet intake to neutralise his pain (54). This seemed a good quantitative measure of symptoms.

BILE REFLUX -

By itself bile in the duodenum does not damage the mucosa. But if it refluxes into the stomach so that the biliary reflux joining the gastric hydrochloric acid causes damage, it can lead to gastritis and gastric ulcer. When this mixed acid/bile fluid refluxes into the lower oesophagus the epithelium is also damaged and can cause heartburn. Studies were made of the surgical procedures of partial gastrectomy and Roux en Y anastomosis, where the bile fluid is diverted in a Y shaped anastomosis at least 10 inches (25 cm) below the stomach (53, 75).

SYMPTOM ANALYSIS AND QUESTIONNAIRE -

Synchronously in the 1970s, together with these manometric and physiological studies, the importance of symptoms became clear because to the patient they were far more important than the original anatomical/pathological diagnosis of hiatus hernia or peptic ulcer or the subsidiary pathophysiological abnormalities.

A huge data base of 319 patients with a radiologically abnormal duodenum was gradually gathered (47, 55). This was done in the early days of computerisation with IBM punch cards combining a) a long symptom analysis, b) pentagastrin gastric acid secretion test results, c) radiological appearances of stomach, duodenum and oesophagus and d) the decision by several different surgeons as to whether to operate or not.

From this data base it was simple to change the endpoints and obtain a computer print out of all the data available to any one given surgeon or clinician. Examples were:

1. a comparison between epigastric and non-epigastric pain duodenal ulcers in regard to reflux symptoms, etc (47).
2. symptoms correlating with maximum gastric acid output.
3. symptoms correlating with radiological defects of the duodenum (99), eg, ulcer crater or scarring, accompanying gastric ulcer or hiatus hernia.
4. decision to operate or not as an end point decided by several general surgeons (85). Not surprisingly this was random, or in other words no logical thought process was discovered. It was probably the GP who had originally decided because he was fed up with repeated visits to his surgery requesting antacid treatment.

SEARCHING FOR CLINICAL CORRELATIONS -                       Wasted Time and Effort

WHAT IS THE DIAGNOSTIC ENDPOINT-

It became increasingly apparent that the endpoints of hiatus hernia, duodenal ulcer or gallstones were actually confusing because their presence alone did not act as an indication for surgery and inevitably should have been qualified by an analysis of symptoms. This led to the realisation that there were two groups one at each end of the spectrum.

A. Those patients with the old style pathological endpoints who had never ever had any symptoms, eg, asymptomatic gallstones.
B. Those who had typical symptoms of say a hiatus hernia or duodenal ulcer but who on repeated radiological, endoscopic or other investigation had negative investigations. Do you tell the patient that he has the wrong end point or the incorrect symptoms or advise him to go away until he gets the correct one?

The results inevitably lead to questions about the expense and inconvenience spent in trying to find an expected or requisite endpoint radiologically, endoscopically or histologically. But nobody seemed interested in this problem as an intellectual exercise. They may have been far more interested in the money generated by all the unnecessary tests and investigations.

In the middle between the tails of the bell-shaped Gaussian curve lay the majority of hospital patients with a mixture of varying pathological endpoints (HH, duodenal ulcer, gallstones), an assortment of investigations including gastric secretion and emptying, oesophageal motility, pH measurement, histology and finally endoscopy. Did the patients have a disease, a symptom complex, helicobacter pylori or just indigestion or dyspepsia? How could you assess it, measure it, investigate or diagnose? In the end the patient needed treatment while the doctors scratched their heads in doubt.

REFERENCES -

15. Production of epigastric pain in duodenal ulcer by lower oesophageal acid perfusion.[PDF]
Earlam R J
Brit Med J 4:714-6 (1970) PMID 5491256

27. The gastro-oesophageal junction in patients with duodenal ulceration. [PDF]
Earlam R J
Rend Gastroenterol 4:69-72 (1972)

29. Further experience with the epigastric pain reproduction tests in duodenal ulceration. [PDF]
Earlam R J
Brit Med J 2:683-5 (1972) PMID 5031711

32. (a) The gastro-oesophageal junction before and after operations for duodenal ulcer. [PDF]
Thomas P A, Earlam R J
Brit J Surg 60:717-9 (1973) PMID 4741188

(b) The gastro-oesophageal junction in duodenal ulcer after operation.
Thomas P, Earlam R
Br J Surg 59:309 (1972) PMID 5020769

35. (a) The action of gastro-intestinal polypeptide hormones on the isolated perfused
gastro-oesophageal sphincter. [PDF]
Thomas P A, Earlam R J
Brit J Surg 60:306 (1973) PMID 4700236

(b) Proceedings: The effects of gastrin and glucagon on gastro-oesophageal function in the isolated perfused canine stomach and oesophagus.
Thomas P A, Earlam RJ
Clin Sci Mol Med 46:5P (1974) PMID 4811875

37. Pain in gastric ulcer. [PDF]
Earlam R J
Lancet 1:744-5 (1974) PMID 4132476

47. A computerized questionnaire analysis of duodenal ulcer symptoms. [PDF]
Earlam R J
Gastroenterology 71:314-7 (1976) PMID 780183

49. Clinical tests of Oesophageal Function. [PDF]
Earlam R J
Crosby Lockwood Staples St Albans 383 pp 150 illustrations (1976) Book

53. The clinical significance of gallstones and their radiological investigation. [PDF]
Earlam R J, Thomas M
Brit J Surg 65:164-7 (1978) PMID 638425

54. An antacid preparation in the treatment of duodenal ulceration. [PDF]
Chaput de Saintonge D M, Earlam R J, Wright J T, Evans S JW, Hillenbrand P, Lancaster-Smith M J, Balme R H,
Barnardo D E
Practitioner 220: 321-3 (1978)

55. Can duodenal ulcer symptoms be measured? [PDF]
Earlam R J, Chaput de Saintonge D M
Ital J Gastroenterol 10:133-8 (1978)

75. Bile reflux and the Roux en Y anastomosis. [PDF]
Earlam R J
Brit J Surg 70:393-7 (1983) PMID 6347313

81. Histological appearances of oesophagus, antrum and duodenum and their correlation with duodenal ulcer symptoms. [PDF]
Earlam R, Amerigo J, Kakavoulis T, Pollock D.J.
Gut 26:95-100 (1985) PMID 3965370

83. Dogma Disputed: on the origin of duodenal ulcer pain. [PDF]
Earlam R J
Lancet 1: 973-4 (1985) PMID 2859422

85. Decision analysis in elective operation for duodenal ulcer. [PDF]
Earlam R J, Evans S J
Theoretical Surgery 1:89-95 (1986)

99. The clinical significance of radiological changes in duodenal ulcer patients. [PDF]
Earlam R.J, Murray R S, Swann J C, Slater E G, Evans S J
Ital J Gastroenterology 18:88-92 (1986)

107. Esophageal pain. [PDF]
Earlam R J
Gastroenterology 98: 250 (1990) PMID 2293591

115. Endoscopic and histological findings in subjects with dyspepsia. [PDF]
Earlam R J
Brit Med J 302: 1153 (1991) May 11 PMID 1878055