Serum chromogranin A (CgA) is regarded as a reliable marker of neuroendocrine proliferation. We previously described increased serum CgA levels during short-term profound gastric acid inhibition.
Serum chromogranin A (CgA) is regarded as a reliable marker of neuroendocrine proliferation. We previously described increased serum CgA levels during short-term profound gastric acid inhibition.
To investigate serum gastrin and CgA levels in dyspeptic patients during continuous medium- (6 weeks to 1 year), or long-term (1–8 years) gastric acid suppressive therapy.
114 consecutive dyspeptic patients referred for upper gastrointestinal endoscopy were enrolled in a cross-sectional, case-control study [62 patients on continuous antisecretory therapy, either with proton pump inhibitors (n = 47) or H2-receptor antagonists (H2RA) (n = 15) for gastro-oesophageal reflux disease with or without Barrett’s oesophagus or functional dyspepsia, and 52 age- and sex-matched patients without medical acid inhibition and with normal endoscopic findings (control group)]. Omeprazole doses ranged from 20 mg to 80 mg daily and ranitidine from 150 mg to 450 mg daily. Fasting serum CgA and serum gastrin levels were measured by radioimmunoassay (reference values: serum CgA < 4.0 nmol/L; serum gastrin < 85 ng/L).
Fasting serum CgA levels positively correlated with serum gastrin in the entire study population (r = 0.55, P = 0.0001). Median serum CgA values were higher in patients treated with a proton pump inhibitor than H2RA [2.8 (2.0–5.9) nmol/L vs. 2 (1.9–2.3) nmol/L, P < 0.002] and controls [2.8 (2.0–5.9) nmol/L vs. 1.8 (1.5–2.2) nmol/L, P < 0.0001) and did not differ between patients treated with H2RA or controls. Serum gastrin and CgA levels in patients on proton pump inhibitor therapy positively correlated with the degree and duration of acid inhibition. Patients on long-term proton pump inhibitor therapy had significantly higher fasting serum gastrin and CgA than those on medium-term proton pump inhibitor therapy [127 (73–217) ng/L vs. 49 (29–78) ng/L, P < 0.0001 and 4.8 (2.8–8) ng/L vs. 2.1 (1.9–2.6) ng/L, P < 0.001]. No such relation was found in patients on medium- vs. long-term H2RA. Overall, patients with positive Helicobacter pylori serology had higher serum gastrin and CgA levels than those with negative H. pylori serology [51 (27–119) ng/L vs. 27 (14–79) ng/L, P = 0.01, 2.4 (1.9–3.4) nmol/L vs. 2.0 (1.7–2.5) nmol/L, P = 0.05].
During long-term continuous proton pump inhibitor treatment, serum gastrin and CgA levels are significantly elevated compared to H2RA treatment and nontreated dyspeptic controls. H. pylori infection seems to affect gastric ECL cell secretory function. Increased serum CgA values during long-term profound gastric acid inhibition could reflect either gastric enterochromaffin-like cell hyperfunction or proliferative changes.
It is widely accepted that profound acid inhibition leads to antral gastrin-producing (G) cell hyperplasia and hypergastrinaemia1, 2. Gastrin has a specific effect on the enterochromaffin-like (ECL) cell, regulating its function and promoting its growth.3–5 In rats, gastric carcinoids may occur after long-term acid suppression.6–10 However, in humans, the development of gastric ECL cell carcinoids in response to chronic hypergastrinaemia has been restricted to patients with the Zollinger–Ellison syndrome (ZES) as a part of multiple endocrine neoplasia (MEN)—type I11–13 or with chronic atrophic gastritis type A (A-CAG),14–17 pointing towards the involvement of additional and yet unidentified factors. The genomic events associated with MEN-1-ZES, e.g. the loss of a suppressor oncogene in 11q13 or the intragastric environmental changes due to achlorhydria in type A CAG, might explain gastric endocrine cell proliferation in these particular situations.18–20
Well-documented prospective studies, focused on the safety of long-term profound medical acid inhibition, do not substantiate the risk of carcinoid tumour formation.21–24 Recently, the combined effects of profound gastric acid inhibition and Helicobacter pylori infection on serum gastrin levels have elicited research interest.25–28 In a follow-up study designed to assess serum gastrin levels and gastric endocrine and nonendocrine mucosal changes during a 5-year treatment with lansoprazole, H. pylori infection was regarded as a risk factor for the progression of fundic gastritis and the development of ECL cell hyperplasia.29 Taking into consideration the slow rate of progression from hyperplasia to dysplasia and to carcinoid tumour formation, it seems premature to conclude that ECL cell proliferation due to pharmacologically induced hypergastrinaemia is an innocuous phenomenon and that clinical surveillance is unnecessary. Moderate gastrin elevations exert a constant trophic effect on the ECL cells and the clinical consequences of this over many years warrant further consideration.30–32
The identification of a noninvasive, accurate biochemical marker of ECL cell function and growth is therefore an important issue for clinical practice and research.30–33 Recent studies have described serum chromogranin A (CgA) as a reliable marker of neuroendocrine cell proliferation.34–38 Increased serum CgA concentrations have been reported in patients with gastrinoma,39, 40 as well as in patients with type A autoimmune gastritis.41 Waldum et al. described elevated serum CgA levels after short-term treatment with omeprazole, suggesting its possible use as a test to evaluate ECL cell hyperplasia in patients with hypergastrinaemia secondary to acid inhibition.42 Our group has also shown that after short-term medical acid inhibition with three different drug regimens (ranitidine 300 mg b.d., omeprazole 20 mg o.m., omeprazole 40 mg o.m.) in healthy volunteers, serum gastrin and chromogranin A increase in relation to the degree of acid inhibition and that they are strongly correlated.43
The present study was performed to investigate the relation between serum gastrin and CgA levels during medium- and long-term profound acid inhibition. The influence of Helicobacter pylori infection on serum gastrin and chromogranin A levels, in these circumstances, was also evaluated.
One hundred and fourteen consecutive dyspeptic patients (F = 52, M = 62) referred for open-access upper gastrointestinal endoscopy between January and August 1997 were enrolled in a cross-sectional, case-control study. All subjects had given informed consent prior to inclusion. The mean age was 50 years, with a range of 19–75 years. The study group consisted of 62 patients on continuous medium- (6 weeks to 1 year) or long-term (1 to 8 years) acid suppressive therapy for gastro-oesophageal reflux disease (GERD) with or without Barrett’s oesophagus, or for functional acid dyspepsia. The control group consisted of 52 dyspeptic patients with a similar distribution of age and gender, who did not receive acid suppressive medication or antacids before referral and who had normal endoscopic findings.
Patients with active or previous peptic ulcer disease, those who had received H. pylori eradication therapy or had undergone previous gastric surgery/vagotomy were not included in the study population. Patients with other gastrointestinal or pancreatic diseases, with known malignancies, endocrine or autoimmune diseases, liver or kidney disorders were not eligible.
The study was approved by the medical ethical committee of the University Hospital Maastricht.
After informed consent was given, an extensive medical history was obtained in all patients and their clinical symptoms, type and daily dose of medication, duration of continuous acid inhibition in recent years, the regularity of their drug use and eventual co-morbidity/co-medication or family history of gastric cancer were rigorously recorded.
Endoscopy was performed after an overnight fast using standard gastroscopes (Pentax EG-2901). At the beginning of the endoscopy, a 5–10 mL sample of gastric juice was obtained through the suction channel of the endoscope into a trap placed in the suction line. Fasting pH was immediately measured, using pH-paper strips with grading steps of 0.5, from pH 0 to pH 14 (Schleicher & Schill GmbH; Dassel, Germany).44 A 10 mL venous blood sample was collected from all subjects for the assessment of fasting serum gastrin, serum chromogranin A and H. pylori serology.
Fasting blood samples were immediately centrifuged and the sera were stored at −20 °C until further analysis. The serum gastrin concentration was measured by a radioimmunoassay method using an antibody raised in a rabbit against synthetic unsulphated human gastrin 2-17 covalently coupled to bovine serum albumin as previously described.45 The antibody binds to all known circulating gastrin fragments. The upper limit of the normal range for fasting state was taken as 85 ng/L.
A sensitive and specific radioimmunoassay method was used for the determination of chromogranin A levels in the serum samples.35, 36 Polyclonal antibodies raised in rabbits against a fragment of chromogranin A (CgA 116–439) purified from urine from a patient with carcinoid tumours were used in the assay. Based on a group of healthy individuals, the upper limit of the normal range for fasting serum chromogranin A levels was < 4 nmol/L.
A commercial Western blotting kit for IgG (Helicoblot 2.0, Imphos BV, Zambon Group, Amersfoort, the Netherlands) was used for the serological assessment of H. pylori. Bands were present for 19.5, 26.5, 30, 35, 89 (vacA), and 116 kDa (cagA) proteins. The patients were considered to be H. pylori-positive if two bands of the 19.5, 26.5, or 30 kDa protein or any of the bands among the 35, 89, or 116 kDa proteins were positive. H. pylori-positive and H. pylori-negative control sera were included at each determination.
Statistical analysis was performed using the SPSS software package. As the data were not normally distributed, values were expressed as medians, with 25–75% quartile ranges, and group comparisons were performed using nonparametric tests. The Kruskal–Wallis test was used for multiple comparisons and the Mann–Whitney U-test for two-group comparisons of continuous variables. Differences in dichotomous variables were evaluated using the χ2-test or the Fisher’s exact test, depending on the number of patients. Correlations were performed with the Spearman rank test. In addition, linear regression analysis was used to assess the relation between serum gastrin, CgA and the duration of proton pump inhibitor therapy. Two-sided P-values < 0.05 were assumed to indicate statistical significance.
Table 1 illustrates the clinical features of the study population, subdivided according to acid inhibition regimen. There was a similar distribution of age and gender in the three subgroups of patients. Omeprazole doses ranged from 20 mg to 80 mg daily and of ranitidine from 150 mg to 450 mg daily. The median length of acid inhibition therapy in the group receiving H2RA (n = 15) was 2.4 years (range 6 weeks–10 years) and in patients on proton pump inhibitors (n = 47) was 1.8 years (range 6 weeks–8 years). In patients on long-term proton pump inhibitors (n = 26) the median duration on therapy was 2.8 years (1–8 years). Eight of these 26 patients (30%) had been treated continuously for 1–2 years, 7 (27%) for 2–3 years, 5 (20%) for 3–4 years and 6 (23%) for more than 4 years. In patients on long-term H2RA therapy (n = 8), the median duration on therapy was 3.2 years (1–10 years). Two (25%) of them had received continuous therapy with H2RAs for 1–2 years, 2 (25%) for 2–3 years, 3 (38%) for 3–4 years and 1 (12%) for about 10 years.
The degree of gastric acid inhibition during endoscopy, as measured by fasting intragastric pH, was higher in patients on proton pump inhibitor therapy compared to those on H2RA therapy or controls, and did not differ between patients treated with H2RAs or the controls (Table 2 ). Moreover, the subset of patients on long-term therapy with proton pump inhibitors (n = 26) showed a stronger suppression of gastric acid secretion at the time of endoscopy than the patients on medium-term proton pump inhibitor therapy (n = 21) (Table 3 ). No such difference was found between patients on medium- or long-term H2RAs.
Basal gastric juice pH values positively correlated with fasting serum gastrin (r = 0.50; P = 0.0001) and chromogranin A levels (r = 0.37; P = 0.0001).
Serum CgA levels correlated positively with serum gastrin in the entire study population (r = 0.55, P = 0.0001). An even stronger correlation (r = 0.76, P = 0.0001) between serum gastrin and CgA was found in the proton pump inhibitor group (Figure 1 1. Correlation between fasting serum gastrin (ng/L) and chromogranin A (nmol/L) in 47 patients on proton pump inhibitor treatment (r = 0.76, P = 0.0001).
Basal serum gastrin and CgA levels were higher in patients taking proton pump inhibitor therapy than in those on H2RAs or controls, but did not differ between patients on H2RA and controls (Table 2 2 ).
Serum gastrin and CgA levels showed a slight and significant increase with the duration of continuous proton pump inhibitor therapy (Figures 2a, 2b). Likewise, patients on long-term treatment with proton pump inhibitors (1–8 years) had significantly higher serum gastrin and CgA levels compared to those on medium-term (6 weeks to 1 year) proton pump inhibitor therapy (Table 3 ). However, there was no difference with respect to serum gastrin and CgA between patients on medium- vs. long-term therapy with H2RAs [30 (15–56) ng/L vs. 28 (17–35) ng/L, P = 0.95 and 2 (1.9–2.3) nmol/L vs. 2 (1.8–2.3) nmol/L, P = 0.78).
Serum CgA levels higher than the reference range were found in 15 of 26 (58%) patients on long-term proton pump inhibitor treatment, three of 21 (14%) patients on medium-term proton pump inhibitors, in none of the patients on H2RAs and in one of the 52 (1.9%) controls. All patients with CgA concentrations above the reference range also had serum gastrin levels higher than the upper limit of the normal range.
Among the 114 consecutive dyspeptic patients included in the study, 50 (43.8%) were H. pylori positive and 64 (56.2%) were H. pylori negative. Twenty-four (46%) of the controls, six (40%) of the patients on H2RAs, and 20 (43%) of the patients on proton pump inhibitor therapy had positive H. pylori serology.
Patients with positive H. pylori serology had higher median fasting serum gastrin and median chromogranin A levels compared to patients with negative H. pylori serology [51 (27–119) ng/L vs. 27 (14–79) ng/L, P = 0.011; 2.4 (1.9–3.4) nmol/L vs. 2.0 (1.7–2.5) nmol/L, P = 0.05] (Figures 3a, 3b).
Long-term therapy with potent antisecretory drugs results in functional and proliferative changes of gastric enterochromaffin-like cells.46 Although an achlorhydria-carcinoid sequence seems unlikely in humans, the effect of long-term hypergastrinaemia on the endocrine cells of the gastric oxyntic mucosa remains unclear.
The recently described role of ECL cells in the modulation of gastric acid secretion sheds a new light on the pathophysiology of this complex cell population. The ECL cells are thought to increase gastric acid secretion by releasing histamine in response to gastrin stimulation.47–49 Hence, prolonged hypergastrinaemia might play a central role in rebound acid hypersecretion after long-term acid inhibition.41
The study of gastric endocrine cells based on morphological methods provides unique information about the progression of the hyperplastic changes. However, the accurate evaluation of gastric mucosal endocrine proliferation encounters many limitations in daily clinical practice. Mucosal ECL cell hyperplasia is not endoscopically detectable and, thus, may be under-estimated in routine examinations.50 The co-existence of parietal cell hypertrophy/hyperplasia due to hypergastrinaemic states,51 or the presence of H. pylori-related atrophic gastritis with clustering of the endocrine cells are possible confounding factors in the assessment of endocrine cell hyperplasia. Moreover, the use of laborious morphometric methodologies with large variations in the expression of results makes a comparison between different studies difficult.
For these reasons, the search for rapid, reliable and cost-effective strategies to monitor early changes in gastric ECL cell function and growth related to drug-induced hypergastrinaemia is justified.
The present study describes, for the first time, a correlation between serum gastrin and CgA levels during medium- and long-term acid suppressive therapy, suggesting that serum CgA might reflect functional and/or proliferative changes of the gastric ECL cells. This observation is consistent with our previous study involving healthy volunteers and short-term antisecretory therapy with H2RA and proton pump inhibitors, in which the increase in serum gastrin levels following different degrees of pharmacological acid inhibition was accompanied by a similar increase in serum CgA levels.43 These findings confirm the results of Waldum et al.,42 who described augmented serum CgA values after short-term treatment with omeprazole and suggested its use as a test to evaluate ECL cell hyperplasia secondary to hypergastrinaemia during long-term proton pump inhibitor therapy.
Based on a large sample of consecutive dyspeptic patients free of any conditions known to be associated with elevated serum gastrin and CgA levels, we found increased serum gastrin and CgA in patients on proton pump inhibitor therapy, but not in those on H2RAs, or in nontreated dyspeptic controls with a similar distribution of age and gender. More than 38% of the patients receiving continuous acid inhibition with proton pump inhibitors (15 of the 26 patients on long-term proton pump inhibitor and three of the 21 patients on medium-term proton pump inhibitor therapy) had serum CgA values above the normal range, whereas no patients on H2RA and only one control subject showed such increased levels.
This patient was a 63-year-old man, with a 3-year history of persistent epigastric pain and nausea, without weight loss. His past history did not show any relevant events and an abdominal ultrasonography performed prior to endoscopy was normal. Gastroscopy showed no abnormalities. Fasting intragastric pH was 1.5, serum gastrin and CgA were 11 ng/L and 8.2 nmol/L, respectively, and H. pylori serology was negative. Routine biochemistry including liver and renal functional tests were normal. The family history revealed no findings suggestive of MEN I. As increased serum CgA levels are not totally specific to neuroendocrine tumours, a subclinical malignant process with neuroendocrine differentiation could account for the high circulating CgA. However, neither before nor after the endoscopy was such condition detected.
In agreement with previous studies, fasting serum gastrin levels in the 47 patients on continuous proton pump inhibitor treatment were related to the degree and duration of gastric acid inhibition.20, 22, 29, 33, 52, 53 Moreover, we found that serum CgA described similar time-dependent changes. Linear regression analysis showed a slight but significant increase in serum CgA levels with the duration of continuous proton pump inhibitor therapy. In spite of a similar distribution of the demographic data, patients on long-term (1–8 years) treatment with proton pump inhibitors, had significantly higher serum CgA than those who received medium-term (6 weeks to 1 year) proton pump inhibitor therapy.
In a comprehensive study looking at the long-term effects of lansoprazole on serum gastrin, the progression of fundic gastritis and changes in fundic argyrophil cell morphology, Eissele et al. found a significant increase in fasting serum gastrin levels within the first 3 months of treatment, with no further elevation in the following 5 years of therapy.29 Despite the plateau phenomenon described for serum gastrin, argyrophil cell density showed a slight but steady increase during the 5-year of follow-up investigation. In our study, even slightly elevated median serum gastrin levels (127 ng/mL) in patients on long-term proton pump inhibitor treatment corresponded to an increase in serum CgA levels. The duration of continuous stimulation, rather than the magnitude of hypergastrinaemia, seems to be responsible for the elevated serum CgA values in these patients. Whether increased serum CgA reflects only an enhanced secretory function of the gastric ECL cells, as an early consequence of sustained hypergastrinaemia, or a late phenomenon, such as ECL cell proliferation, remains to be established.
The results of the current study support the role of H. pylori infection in the modulation of gastric ECL cell function and/or growth. Patients with positive H. pylori serology showed significantly higher serum gastrin and CgA levels compared to patients with negative H. pylori serology. The forthcoming hypothesis could be that serum CgA might provide additional information with regard to gastric mucosal endocrine and nonendocrine changes in H. pylori infected patients on antisecretory therapy.
We realize that the present study may suffer from several confounding factors. Its cross-sectional design does not allow conclusions to be drawn concerning the precise time-course of serum CgA changes in patients on long-term proton pump inhibitor therapy. However, the significant increase in CgA with the longer period of acid inhibition is a genuine finding. An interesting question arising in this context is whether a persistent increase in serum CgA levels might interfere with gastric acid secretion. The functional role of circulating chromogranins in humans is poorly defined.54 However, in experimental models pancreastatin, the main CgA-derived peptide, has been proposed as a potent inhibitor of gastric parietal cell secretion.55 In view of this, one might speculate that elevated serum CgA, subsequent to drug-induced hypergastrinaemia, may further up-regulate gastrin secretion, thus explaining their parallel changes during long-term acid inhibition. On the other hand, a reduction of parietal cell secretion could further potentiate the pharmacologically induced acid inhibition, in a self-perpetuating process.
In conclusion, we report significantly higher serum gastrin and CgA levels after long-term proton pump inhibitor therapy compared to H2RA therapy and nontreated dyspeptic controls. A combined effect of H. pylori infection and antisecretory therapy on gastric endocrine cells, reflected in increased serum gastrin and CgA levels, is also suggested. A possible relationship between elevated CgA in serum and the presence and severity of gastric mucosal endocrine proliferation remains to be further evaluated.
This study was supported by an Educational grant from Astra Pharmaceutica, B.V., the Netherlands.