Implications of serum pepsinogen I in early endoscopic diagnosis of gastric cancer and dysplasia. Helsinki Gastritis Study Group.

BACKGROUND AND AIMS: The risk of gastric cancer (GCA) is increased in atrophic gastritis. A low serum pepsinogen group I (SPGI) level is a good serologic indicator of atrophic gastritis of the gastric corpus and fundus, and can be used for diagnosis of subjects with atrophic gastritis and of increased risk for GCA. The present study was undertaken to investigate whether SPGI assay and a diagnostic gastroscopy could enable the diagnosis of GCA at an early stage. MATERIAL AND METHODS: The study was carried out as part of the Alpha-Tocopherol, Beta-Carotene Cancer prevention study (ATBC study) in Finland, in which 22,436 male smokers aged 50-69 years were screened by SPGI. Low SPGI levels (< 25 microg/l) were found in 2196 (9.8%) men. Upper GI endoscopy (gastroscopy) was performed in 1344 men (61%) and 78% of these had moderate or severe atrophic corpus gastritis in endoscopic biopsies. A control series of 136 men from the ATBC study cohort with abdominal symptoms, but with SPGI > or = 50 microg/l were similarly endoscopied, and 2.2% of these had corpus atrophy. RESULTS: Neoplastic alterations were found in 63 (4.7%; 95% CI: 3.6%-5.8%) of the 1344 endoscopied men with low SPGI levels. Of these, 42 were definite dysplasias of low grade, 7 dysplasias of high grade, 11 invasive carcinomas, of which 7 were 'early' cancers, and 3 carcinoid tumors. In the control series, 1 man (0.7%) of the 136 men had a definite low-grade dysplasia. Thus, 18 (1.3%; 95% CI 0.7%-2.0%) cases with 'severe' neoplastic lesions (4 advanced cancers, 7 early cancers and 7 dysplasias of high grade) were found in the low SPGI group, but there were none in the control group. All four patients with advanced cancer died from the malignancy within 5 years (mean survival time 2.5 years), whereas surgical treatment in all those with early cancer or high-grade dysplasia was curative. One of the seven patients with early cancer and two of the seven with high-grade dysplasia died within 5 years, but none died from the gastric cancer. Thus, curative treatment was given to 14 of 18 men in whom a malignant lesion was found in gastroscopy. This is about 15% of all gastric cancer cases (92 cases) which were diagnosed within 5 years after SPGI screening in the 22,436 men. Among the gastric cancer cases of the main ATBC study, the 5-year survival rate was 33% (85% of the non-survivors died from gastric cancer). CONCLUSIONS: We conclude that assay of SPGI followed by endoscopy is an approach which can enable the early diagnosis of gastric cancer at a curable stage.

Treatment of acute otitis media with an antiadhesive oligosaccharide: a randomised, double-blind, placebo-controlled trial.

BACKGROUND: Antiadhesive compounds are promising candidates for prevention or treatment of infections. We have investigated the efficacy of such an agent, 3'-sialyllacto-N-neotetraose (NE-1530), given intranasally for prophylaxis of acute otitis media and for effect on nasopharyngeal carriage of bacteria. METHODS: We did a randomised, double-blind placebo-controlled study at one study site. 507 healthy children were randomly assigned either NE-1530 (n=254) or placebo (253) as intranasal sprays twice daily during 3 months. The children were examined by the study physicians once a month and during illness. Treatment efficacy was estimated from Cox proportional hazards model. A sample of nasopharyngeal secretion was taken at every visit for culture of Streptococcus pneumoniae, Haemophilus influenzae, and Moraxella catarrhalis. Adverse events were recorded in study diaries. FINDINGS: At least one event of acute otitis media was diagnosed in 108 (43%) of 254 children in the NE-1530 group and in 86 (34%) of 253 children in the placebo group. The efficacy of treatment was negative, -27% (95% CI -68 to 5; p=0.10). The nasopharyngeal carriage of S pneumoniae, H. influenzae, and M. catarrhalis was not affected by treatment, and the adverse event profiles were almost identical for NE-1530 and placebo. INTERPRETATION: NE-1530 did not have a beneficial effect on the occurrence of acute otitis media or on the nasopharyngeal carriage of bacteria in children.

Risk evaluation of postvagotomy ulcer recurrence by using endoscopic Congo red test and gastric secretion tests.

BACKGROUND/AIMS: To evaluate the usefulness of the endoscopic Congo red test (ECRT), and to compare sensitivity and specificity of different tests in the discrimination of cases with high risk for postvagotomy recurrent ulcer (RU). METHODOLOGY: In 271 consecutive postvagotomy duodenal ulcer patients the endoscopic Congo red test (ECRT) was used 5-12 years after vagotomy. Further, 39 patients out of 271 were selected and classified into two groups: A--13 ECRT positive cases with RU, B--26 controls without RU (13 ECRT positive and 13 ECRT negative cases). Basal acid output (BAO), maximal acid output (MAO), and nocturnal acid output (NAO) were determined pre- and postoperatively, the serum pepsinogen I (S-PGI) and insulin test were estimated postoperatively. RESULTS: Positive ECRT had 95% sensitivity and 53% specificity for RU. S-PGI > 150 microg/l had 54% sensitivity and 92% specificity (in ECRT positive cases 100% specificity). The insulin test showed 83% sensitivity and 78% specificity. The respective data for the combination of BAO > 1.5 mmol/h + NAO > 30 mmol/12 h were 80% and 81%. CONCLUSION: ECRT should be a primary step in estimating postvagotomy ulcer risk. In negative ECRT cases, the development of recurrent ulcer is unlikely. Additional gastric secretion studies as S-PGI or BAO+NAO or insulin test are needed only in ECRT positive cases.

Gastric cancer and premalignant lesions in atrophic gastritis: a controlled trial on the effect of supplementation with alpha-tocopherol and beta-carotene. The Helsinki Gastritis Study Group.

BACKGROUND: Vitamin E and beta-carotene are considered to decrease the risk of gastric cancer both in humans and in laboratory animals. We studied the effect of dietary supplementation with alpha-tocopherol and beta-carotene on the end-of-trial prevalence of premalignant and malignant lesions of the stomach in older men with atrophic gastritis. METHODS: The study was carried out within the Alpha-Tocopherol, Beta-Carotene Cancer Prevention Study (ATBC study) in Finland, in which 29,133 male smokers aged 50-69 years were randomly assigned to receive daily 50 mg alpha-tocopherol, 20 mg beta-carotene, both of these agents, or placebo, for 5-8 years. Serum pepsinogen was determined at base line and after 3 years' supplementation to find men with atrophic gastritis. A low serum pepsinogen I level, indicating atrophic gastritis of the corpus area of the stomach, was found in 2132 men. These men were invited to have upper gastrointestinal endoscopy (gastroscopy), which was performed on 1344 subjects after a median supplementation time of 5.1 years. RESULTS: Neoplastic alterations were found in 63 of the men (4.7%): 42 with definite dysplasias of low grade (moderate dysplasia), 7 with definite dysplasias of high grade (severe dysplasia), 11 with carcinomas (of which 7 were 'early' cancers), and 3 with carcinoid tumors. Neither alpha-tocopherol (relative risk, 0.98; 95% confidence interval, 0.57-1.69) nor beta-carotene (relative risk, 1.13; 95% confidence interval, 0.65-1.95) supplementation had any association with end-of-trial prevalence of gastric neoplasias after adjustment for other possible risk factors. The effect was not modified by base-line serum level or dietary intake of vitamins, prevalence of Helicobacter pylori infection, or other covariates. CONCLUSIONS: We thus conclude that supplementation with alpha-tocopherol or beta-carotene for 5 years has no major impact on the occurrence of neoplastic changes of the stomach in older male smokers with atrophic gastritis.

Globus pharyngis, commonly associated with esophageal motility disorders.

OBJECTIVES: To evaluate the role of gastrointestinal and psychiatric etiology in globus sensation. METHODS: The study population consisted of 32 consecutive patients with globus sensation without dysphagia referred to the Department of Otorhinolaryngology in Helsinki University Hospital. Eleven patients were excluded from the study: two because of advanced age, one prisoner, and six patients refused further studies. Only two patients (6%) were found to have abnormal otorhinolaryngological status. These patients were also excluded from the study. Esophagogastroduodenoscopy, 24-h pH recording, esophageal manometry, and Bernstein acid perfusion test were carried out in 21 patients (13 females, eight males, mean age 49 yr). Psychiatric evaluation was done in 20 patients; one patient refused the psychiatric consultation. RESULTS: Abnormal endoscopy was found in 12/21 (57%) of the patients, with antral gastritis and hiatal hernia being the most common findings. Two patients had esophagitis. Sixty-seven percent demonstrated abnormalities in esophageal manometry, the most frequent finding being a nonspecific esophageal motility disorder (29%). pH monitoring was normal in 16/21 of patients (76%), whereas the Bernstein test showed positive results in 13/21 (62%). With DSM IIIR as the diagnostic tool, five of 20 patients (25%) received a psychiatric diagnosis. CONCLUSIONS: Globus sensation has a multiple etiology, and local reasons are rare but should first be ruled out. Abnormalities in esophageal motility are commonly found, and these patients seem to be sensitive to esophageal acidity. Esophageal manometry and ambulatory 24-h pH recording should be included in the evaluation of a globus patient. The number of psychiatric disorders does not differ from that in the general population. Treatment of globus sensation should be directed toward the abnormality found behind the symptom.

Serum pepsinogen I and serum gastrin in the screening of severe atrophic corpus gastritis.

The possibilities to screen atrophic corpus gastritis with serum pepsinogen I (S-PGI) and serum gastrin (S-gastrin) concentrations have been studied in 774 subjects: 71 index subjects selected from a general population at random, 353 of their first-degree relatives, 276 first-degree relatives of patients with gastric cancer, 53 patients with pernicious anaemia, and 21 of their relatives. Discrimination function analysis was calculated from members of random and gastric carcinoma families. S-PGI less than 30 ng/ml had a high sensitivity for severe diffuse atrophic corpus gastritis (SDAG) alone (89.5%) and SDAG + severe patchy atrophic corpus gastritis (SPAG) (89.1%). Respective figures for specificity were 91.5% and 94.8%. The discriminatory power of S-PGI less than 30 ng/ml and S-PGI less than 25 ng/ml was of the same order. The sensitivity of low S-PGI decreased sharply in detection of slighter forms of atrophic corpus gastritis. The sensitivity of S-gastrin greater than 100 pmol/l to discriminate SDAG was 57.9% and SDAG+SPAG 58.7%. Respective figures for specificity were 90.2% and 92.2%. Diffuse and patchy atrophic changes behaved similarly regarding S-PGI and S-gastrin mean concentrations. Accordingly, the biopsy specimen with the severest atrophic changes indicates the degree of atrophy, which associates closely with the changes in S-PGI and S-gastrin. In conclusion, severe atrophic (diffuse or patchy) corpus gastritis may be screened from a general population with high sensitivity and specificity by low S-PGI less than 30 ng/ml, whereas an increased level of S-gastrin is too insensitive for this.

Serum pepsinogen I and serum gastrin in the screening of atrophic pangastritis with high risk of gastric cancer.

Serum pepsinogen I (S-PGI) and serum gastrin (S-gastrin) were examined in the screening of three types of atrophic gastritis with inherent high risk of gastric cancer: in 102 cases with severe atrophic corpus gastritis (SACG), in 5 cases with severe atrophic antrum gastritis (SAAG), and in 15 cases with severe atrophic pangastritis (SAPG) (atrophy both in corpus and in antrum) found among 916 subjects from three family series (265 from gastric cancer families, 425 from randomly selected control families and 226 from pernicious anaemia families). There is no way to screen directly atrophic gastritis restricted to the antral mucosa. In pangastritis atrophy of antral glands causes a failure of the hypergastrinemic reaction of achlorhydria. The combination of S-PGI less than 25 micrograms/l + S-gastrin less than 200 pmol/l detected 80.0% of our cases with SAPG, and only 17 subjects of 794 (2.1%) were false positives i.e. who had not advanced atrophic gastritis. The risk of gastric cancer may be significantly higher in SAPG than in SACG. The estimated prevalence of SAPG was 3% in random-family members over 60 years. The combination of S-PGI and S-gastrin is recommended when the cost/benefit ratio in the screening program of gastric cancer is considered and people from a general population are selected for endoscopic studies.

Characteristics of gastric mucosa which precede occurrence of gastric malignancy: results of long-term follow-up of three family samples.

The aim of the study was to evaluate what family characteristics and what morphological, functional and immunological changes of the gastric mucosa precede the development of gastric malignancy in a follow-up of 11-14 years. The material consisted of 301 first-degree relatives of gastric carcinoma patients, 183 relatives of pernicious anaemia patients, and of 358 control relatives of probands computer matched from the general population by age and sex for the carcinoma probands. All subjects were endoscopically examined in 1973-1976 and followed up to the end of 1987. According to cancer registry data, 11 cases of malignant gastric tumours (9 carcinomas, one carcinoid tumour and one anaplastic tumour with suspicion of Hodgkin's disease) had been diagnosed during the follow-up: 6 belonged to gastric carcinoma, 2 to pernicious anaemia and 3 to control families. The occurrence of malignancy was significantly related to the presence of advanced gastritis with atrophy and of intestinal metaplasia before the start of the follow-up. In relatives with achlorhydria and low serum pepsinogen I levels the incidence of malignancy did not significantly differ from that in controls of similar age and sex distribution. The risk of getting malignancy was about four-fold in female members of gastric carcinoma and pernicious anaemia families but was not increased in control families. The risk was increased only in female members and concerned only gastric malignancy being the expected one or even lower than the expected in regard to malignancies of other location.

Chronic gastritis profiles in sibs of probands calculated to carry a highly increased risk of gastric carcinoma.

Of 1161 subjects consisting of 4 family samples (gastric carcinoma patient relatives, pernicious anaemia, duodenal ulcer and control families) 18 subjects representing 17 families had distinct atrophic changes in the gastric mucosa that were considered on the basis of risk calculations to carry an at least 7-fold risk to develop gastric carcinoma. Twelve of these 18 subjects could be used as probands and their 41 sibs were subjected to a closer statistical analysis. Sibs of probands having a very high relative risk of gastric carcinoma (18-fold or more) differed markedly from the general population. All sibs had some form of atrophic gastritis and there was a significantly higher than expected prevalence of subjects with severe corpus mucosal atrophy. It is concluded that sibs of subjects with severe atrophic changes may carry an increased risk of gastric malignancy.

Young onset peptic ulcer disease and non-ulcer dyspepsia are separate entities.

The characteristics of peptic ulcer and non-ulcer dyspepsia in young men were studied in 202 consecutive conscripts who attended Central Military Hospital in Helsinki because of long-standing upper abdominal complaints. Active peptic ulceration (APU) was found in 48 patients, inactive peptic ulcer disease (IPU) was diagnosed in 77 patients, non-ulcer dyspepsia (NUD) was diagnosed in 52 patients. In 25 cases the reason for symptoms was another disease, and these patients were excluded from the study. A control series (CON) consisted of 30 symptomless healthy young male volunteers. The likelihood of discriminating between peptic ulcer disease and non-ulcer dyspepsia in a young male patient with dyspepsia are indicated by odds ratios (OR) and its 95% confidence limits (CL 95). Active peptic ulcer disease differs from NUD, e.g., by 1) presence of antrum gastritis, OR 41.5 (CL 95: 10.1-171), 2) Helicobacter pylori in the gastric mucosa, OR 31.0 (7.4-130), 3) Lewisa+ phenotype, OR 8.9 (1.7-45.4), 4) serum pepsinogen I (S-PGI) greater than 100 micrograms/l, OR 4.6 (1.7-12.4), 5) non-secretor status, OR 4.3 (1.6-11.6), and 6) O-blood group, OR 3.0 (1.2-7.7). In conclusion, the status of gastroduodenal mucosa, gastric secretion pattern and distribution of some genetic markers in patient series indicate that young onset peptic ulcer and non-ulcer dyspepsia are two separate entities. Helicobacter-positive antrum gastritis is the best determinant of ulcer risk, but also high S-PGI, Lewisa+ phenotype, non-secretor status and O-blood group are signs of increased risk of peptic ulcer.

Cumulative 10-year risk of symptomatic duodenal and gastric ulcer in patients with or without chronic gastritis. A clinical follow-up study of 454 outpatients.

The cumulative rate of symptomatic peptic ulcer (PU) was examined in a 10-year clinical follow-up study of 454 consecutive outpatients who had undergone diagnostic gastroscopy, from whom routine biopsy specimens were taken from the antral and corpus mucosa, and who were found to be ulcer-free before and at the time of this initial gastroscopy. During the follow-up period 34 (11%) of 321 patients who showed gastritis in the biopsy specimens at the initial gastroscopy had contracted symptomatic PU (18, 5, 7, and 4 cases of duodenal, pyloric, antral, and angular or corpus ulcer, respectively), which was verified by endoscopy. Only 1 (0.8%) of 133 patients with normal antral and corpus mucosa had contracted PU. It was calculated that the 10-year cumulative probability of PU was 10.6% (95% confidence interval (CI95), 7.2-14.0%) in the patients with gastritis, whereas this probability was only 0.8% (0-2.2%) in the patients who had normal antral and corpus mucosa in the initial specimens. The cumulative probability of PU was found to be highest, 27.3% (1.0-53.6%), in middle-aged men (41-60 years of age) who had chronic antral gastritis or chronic pangastritis (gastritis in both antrum and corpus). It is concluded that chronic gastritis precedes the appearance of PU and that the cumulative 10-year risk of PU is very low when both antral and corpus mucosa are normal but may be high if chronic gastritis is present.

The effect of vagotomy and antrectomy on serum pepsinogens I and II.

Ninety-seven consecutive patients with gastric surgery for peptic ulcer were studied; 86 had duodenal ulcer (DU), and 11 gastric ulcer (GU). DU patients were surgically treated by proximal vagotomy, proximal vagotomy and pyloroplasty, truncal vagotomy and pyloroplasty, or truncal vagotomy and antrectomy. All GU patients were operated on by the Billroth I method. Serum pepsinogen I(S-PG I), serum pepsinogen II (S-PG II), basal acid output (BAO), and maximal acid output (MAO) were determined before and 3 months and 1 year after the operation. The mean preoperative S-PG I concentration in DU patients (154 +/- 7 micrograms/l; mean +/- SE) was significantly higher than that (97 +/- 9 micrograms/l) in GU patients (p less than 0.001). A significant decrease in the mean S-PG I concentration in DU patients was seen 3 months (92 +/- 6 micrograms/l) and 1 year (66 +/- 4 micrograms/l) after the operation (p less than 0.001). This change did not depend on the type of vagotomy. However, this decrease was not seen in all individual patients as it was in BAO values. Moreover, the mean BAO decrease was much greater at 3 months (7% of the preoperative value) and 1 year (23%) after the operation than the respective decrease in S-PG I concentration. There was also no correlation between S-PG I and acid output (BAO and MAO) before and after the operation. In GU patients the decrease in mean S-PG I value after the Billroth I operation was smaller than in DU patients after vagotomy.(ABSTRACT TRUNCATED AT 250 WORDS)

Fasting levels of serum gastrin in different functional and morphologic states of the antrofundal mucosa. An analysis of 860 subjects.

The relationship of fasting serum gastrin (FSG) levels to the histologic state of antral and body mucosa and to the stimulated acid output (PAO) was examined in 860 subjects. The FSG levels correlated with PAO and atrophy of the body mucosa: the FSG increased linearly with an increase in the grade of body atrophy and increased exponentially when the PAO decreased from 'normal' (greater than 10 meq/h) to zero. In subjects with achlorhydria or marked hypochlorhydria (PAO less than 1.1 meq/h) accompanying moderate or severe atrophy in the gastric body mucosa, FSG decreased linearly with increasing grade of atrophy in the antral mucosa. No such relationship between antral atrophy and FSG was found in subjects who had a PAO above 1.1 meq/h or who had non-atrophic gastric body mucosa. We conclude that the state of the antral mucosa influences the FSG level, but only when the function of antral G cells is maximal--that is, in achlorhydric or nearly achlorhydric conditions in which the inhibitory effect of intragastric acidity on the G cells' secretion of gastrin into the circulation is minimal.

Cyclic AMP and gastric secretion in man. An in vivo study in healthy volunteers, duodenal ulcer patients, and pernicious anaemia patients.

The metabolism of cyclic AMP and HCl secretion has been studied in eight healthy volunteers, in eight duodenal ulcer (DU) patients, and in four pernicious anaemia patients. Pentagastrin showed a tendency to increase adenylate cyclase and cyclic nucleotide phosphodiesterase activities in the fundal mucosa and caused a significant increase in cyclic AMP output into the gastric juice in healthy volunteers and in DU patients. Cimetidine inhibited all these events but had no effect on basal cyclic AMP output. Vagotomy significantly inhibited basal cyclic AMP output. We conclude that cyclic AMP is involved both in basal and in pentagastrin-stimulated gastric secretion in man. Basal secretion is mainly controlled by vagal tone. The main pathway for this stimulus at the parietal cell may be via other than H2-receptors, probably through acetylcholinergic receptors. However, a significant part of the pentagastrin stimulation of the human parietal cell is via H2-receptors.

Campylobacter pylori is associated with chronic gastritis but not with active peptic ulcer disease.

Campylobacter pylori is supposed to be involved in the pathogenesis of gastroduodenal peptic ulcer diseases and chronic gastritis. In order to study whether the Campylobacter pylori in the stomach of peptic ulcer patients is related to ulcer itself or to a co-existing chronic gastritis, we examined the frequency of the bacteria in Giemsa stained histological sections of biopsy specimens from a series of patients with active peptic ulcer and from series of non-ulcer control subjects. We found no difference in the frequency of Campylobacter- positive cases between ulcer patients and non-ulcer controls when the comparison was done within the same category of chronic gastritis; e.g., within the category of chronic superficial gastritis 74% and 78% of cases showed the bacteria in antral biopsies from ulcer patients and from non-ulcer controls, respectively. In both ulcer patients and control subjects, in similar way in both antral and body mucosa, the Campylobacter pylori was strongly associated with chronic superficial gastritis but was more weakly associated with chronic atrophic gastritis, and the bacteria were only occasionally seen in normal mucosa. We conclude that Campylobacter pylori is associated with chronic gastritis in peptic ulcer patients but is not related to active ulcer.

Gastritis--a misused term in clinical gastroenterology.

Gastritis is an inflammation in the gastric mucosa. The definition, classification and diagnosis of gastritis is based on morphological changes. The term 'gastritis' is widely used as a catchbag for upper abdominal complaints. Although acute or specific forms of gastritis may present upper abdominal symptoms, chronic gastritis is asymptomatic, and is not the cause of long standing upper abdominal complaints. Chronic gastritis is a very common condition in the general population. Hence the probability of finding chronic gastritis in endoscopic biopsies of patients with upper abdominal complaints is high for statistical reasons, and this does not prove any causal relation between chronic gastritis and subjective complaints. Campylobacter pylori is associated with chronic gastritis. It may activate the inflammatory process in the gastric mucosa, or it may just be an innocent bystander, which subsists within a diseased mucosa. The role of campylobacter pylory in the etiology of non-ulcer dyspepsia or peptic ulcer is questionable.

Changed profile of peptic ulcer disease in hospital patients during 1969-1984 in Finland.

The annual number of hospital discharges for peptic ulcer decreased in Finland from 1969 to 1974, mainly because of a smaller number of men with uncomplicated peptic ulcer disease. After the mid-1970s the number of hospital admissions remained stable, whereas the number of elderly patients increased. From 1979 to 1984 the number of hospitalizations for perforated ulcers remained stable, whereas those for ulcer haemorrhages increased, the increase being most marked in the older age groups. The age-specific mortality from peptic ulcer remained stable from 1969 to 1984 and rose thereafter among old patients. The risk of death from peptic ulcer was about 10-fold higher in patients over 75 years old than in younger age groups, and about half of the deaths caused by peptic ulcer occurred in patients over 75 years old. The men to women ratio among hospitalized patients decreased from 3.5 in the late 1960s to 2.1 in the 1980s, and the gastric ulcer to duodenal ulcer ratio was about 1.1 throughout the observation period.

Serum group I pepsinogens during insulin and pentagastrin tests in unoperated and vagotomized duodenal ulcer patients.

The level of serum group I pepsinogens (PG I) has been studied during the conventional insulin-pentagastrin test in 29 duodenal ulcer (DU) patients before and 2 months after proximal selective vagotomy (PSV) and in 31 unoperated DU patients. The mean basal serum PG I level was 191.6 +/- 15.4 micrograms/l (mean +/- SEM) before and 143.7 +/- 24.0 micrograms/l after PSV. A significant increase in mean serum PG I above the initial value was found both in unoperated DU patients and in patients after PSV 1 h after insulin injection. In 29 PSV patients the mean serum PG I showed a paradoxical decrease during the 2nd h after insulin injection, and the mean postvagotomy serum PG I 2 h after insulin injection was significantly (p less than 0.01) lower than the respective preoperative value in the same patients, and the value was close to the basal serum PG I. The low level of serum PG I 2 h after insulin injection in vagotomized patients may reflect the deprivation of the reduced store of PG I in the absence of normal vagal tone. Both the basal serum PG I and serum PG I response during insulin-induced hypoglycaemia showed an overlap between unoperated and vagotomized DU patients. Therefore, serum PG I analyses during the insulin test cannot replace acid secretion tests in the assessment of the completeness of vagotomy.