Stomach-specific biomarkers in histological staging of gastritis
Research type
Research Study
Full title
A study of the role of stomach-specific biomarkers in the histological staging of gastritis; Is there any relationship of stomach-specific biomarkers with histological diagnosis of different types of gastritis?
IRAS ID
148224
Contact name
Tariq Mahmood
Contact email
Sponsor organisation
United Lincolnshire Hospitals NHS Trust
Duration of Study in the UK
0 years, 6 months, 1 days
Research summary
The Correa cascade describes the prolonged series of precancerous lesions that precede invasive gastric carcinoma [1]. The first recognised histological change of active chronic inflammation can persist as non-atrophic chronic gastritis (no gland loss) or advance to multi-focal atrophic gastritis (loss of normal antral and/or oxyntic glands). Atrophic gastritis can then progress to intestinal metaplasia, dysplasia and eventually invasive carcinoma. Gastritis and atrophy tends to occur first in the antrum and then move upwards with time finally resulting in atrophic gastritis that occupies the whole stomach [2]. One study found the Correa cascade to concern approximately half of gastric cancers [3]
Atrophic gastritis is the single most important precursor condition for gastric cancer [3]. This risk increases with the extent of the disease, with an increase in cancer risk up to 90-fold in atrophic gastritis in both antrum and corpus (i.e. severe pan-atrophy), compared to subjects with healthy stomach mucosa [4]. H pylori infection is implicated as the most important causative agent in chronic active gastritis and therefore subsequent atrophic gastritis [5] [6].
In 2008, gastric cancer was the fourth most common malignancy worldwide (988,000 cases, 7.8% of all cancers), however due to dismal prognosis it was the second leading cause of cancer deaths worldwide (736,000 deaths, 9.7% of total) [7]. A successful strategy for prevention of gastric cancer depends on the diagnosis of pre-cancerous conditions of H pylori gastritis and atrophic gastritis to identify subjects at cancer risk and in whom treatment of H pylori and cancer surveillance are necessary symptoms in order to slow or intercept the Correa cascade [1] [8].
Gastroscopy with microscopic biopsy is the current gold-standard method for diagnosing these pre-cancerous conditions; however several studies have looked at the application of stomach-specific biomarkers in non-invasive diagnosis of atrophic gastritis [8]. In atrophic gastritis the normal functional cells and glands of the gastric mucosa decrease in number and eventually disappear.
Pepsinogen I is secreted solely by chief cells of the corpus mucosa, whereas Pepsinogen II is produced by the chief cells and mucous neck cells of the gastric mucosa, by pyloric glands in the gastric antrum and by Brunner’s glands in the proximal duodenum. Atrophic corpus gastritis will lead to loss of the chief cells of the corpus mucosa leading to a reduction in the secretion of pepsinogen I. The decrease in levels of pepsinogen I and pepsinogen I/II ratio correlate well with the grade and extent of corpus atrophic gastritis [9] [10].
Gastrin-17 is secreted exclusively by gastrin cells (G cells) in the antrum. When dormant, the G cells only secrete small amounts of gastrin-17 with maximal secretion achieved after stimulated conditions e.g. protein stimulation, bobesin (gastrin-releasing peptide) stimulation or when acid secretion in the stomach decreases, is low or absent (e.g. PPI administration). Atrophy of the antral glands leads to loss of antral G cells and both the basal and post-prandial secretion of gastrin decreases [10] [11] [12] [13]. By contrast, in patients with normal stomach mucosa, and a preserved number of antral G cells, the plasma level of gastrin-17 rises markedly after stimulation.
In atrophic gastritis limited to the corpus mucosa, the fasting levels of gastrin 17 are always increased [14]. In subjects with a normal biomarker panel suggesting the absence of H pylori infection or atrophic gastritis, low levels of gastrin-17 suggest high intragastic acidity [15]. Prolonged use of PPIs raises the levels of gastrin-17 which incidentally through a sequence of trophic actions on the corpus mucosa can cause hypertrophy of the oxyntic glands and an increase in plasma pepsinogens [16].
Parallel assays of Pepsinogen I, of the Pepsinogen I/II ratio and of gastrin-17 therefore comprise a exact and validated panel of biomarkers to give a comprehensive diagnosis of the different phenotypes of gastritis including the degree of mucosal inflammation, the extent and grade of AG in the stomach, and the capacity of the existing mucosa to secrete acid and gastrin-17 [9] [11] [12]. In corpus/ fundus atrophic gastritis plasma levels of pepsinogen I and/ or the pepsinogen I/II ratio are low. If atrophic gastritis limited to the corpus/ fundus the fasting level of gastrin-17 will be high, however it will be low or normal if the antrum is also involved. A low fasting level of gastrin -17 is a sign of antral atrophic gastritis or high intra-gastric acidity. These conditions can be differentiated by measuring the biomarker before and after protein stimulation or the administration of a PPI, as Gastrin 17 will rise with the normal mucosal structure in high intragastic acidity. In subjects receiving PPI therapy, atrophic gastritis of the antrum would be suggested by low or normal fasting gastrin-17 levels whereas high levels would indicated normal antral mucosa and an appropriate reduction in acid output to PPI therapy [8].
Several recent studies support the use of stomach specific biomarkers to provide information about the stomach health and function and use as a non-invasive tool for the diagnosis and screening of atrophic gastritis, with a view to eliminate excessive use of endoscopy [8]. However, although less frequent, cancer can still appear in patients with non-atrophic gastritis [17] [18].
We feel endoscopy retains an essential role in the diagnosis of atrophic gastritis and gastric cancer. We seek to conduct a clinical study to correlate stomach-specific plasma biomarkers with the endoscopic and histological diagnosis of these conditions, with a view to the use of stomach-specific biomarkers as an additional complimentary tool to histological diagnosis.
Objective (Research Question, Hypothesis)
The goal of this study is to correlate stomach specific biomarkers with the given histo-pathological diagnosis of gastritis.
We hypothesise that there will be a positive relationship between the stomach-specific biomarkers and the histological diagnosis of gastritis and hope to suggest a rationale for why the use of stomach specific biomarkers as a tool in the endoscopic and histo-pathological staging of gastritis is noteworthy in clinical practice.
Method
Enrolment of the patients in the study will take place at Grantham District Hospital from consecutive potentially eligible patients referred for gastroscopy at the department of endoscopy to a total of 50 patients. Eligible patients are all adult female and males, irrespective of whether symptomatic or asymptomatic from gastritis but have been listed to undergo gastroscopy based upon clinical decision. If gastritis is noted at endoscopy, they will be recruited into the study. Potentially eligible patients will be identified by members of the research team and asked to consent for the study and sign a written consent to participate.
As all patients will have already completed the necessary preparation for gastroscopy, this preparation will already be compliant with the preparatory steps needed for the stomach specific biomarker blood test. Patients should not drink, eat or smoke for at least 4 hours prior to blood sampling. Prescribed regular medication should be taken except ranitidine, famotidine, nizatidine and PPI’s which should be discontinued one week prior to sample collection and mediations neutralising gastric acid secretion and mucosa protection agents which should be discontinued one day prior to sample collection. This is the normal advice given to all patients undergoing endoscopy except in emergency bleeders.
GastroPanel®is an ELISA-based assay designed by Finnish company BioHit HealthCare that will be used to measure the four biomarkers specific for the gastric mucosa 1) Pepsinogen I (P-PGI), 2) Pepsinogen II (P-PGII), 3) Gastrin-17 (P-G-17) and 4) H. pylori antibody (P-HpAb).
In those patient consenting to the study a blood sampling should be done either before gastroscopy or 12-24 hours after gastroscopy (gastroscopy may change the gastrin output and therefore may disturb and bias the G-17 assay if the sampling is done immediately after gastroscopy). Ideally 5ml venous blood is required using an EDTA tube following a minimum four hour fast. Bloods are mixed immediately by inverting tube 5-6 times. The person taking the blood sample is required to complete the test request form (appendix 1).
If the results from the fasting sample indicated the patient is suffering from atrophic gastritis in the antrum, a protein stimulated sample is also recommended to measure maximal gastrin-17 concentrations. Following the fasting blood sample or a minimum of 10 hours fast patient should be given protein drink. Repeat blood sample using EDTA tube is required 20 minutes following protein intake.
The samples require separation by centrifugation for 30 minutes at the laboratory at Grantham hospital within 30 minutes of collection. Samples will require addition of Biohit Gastrin-17 stabiliser to enable storage of sample at room temperature but can be instead refrigerated for 3 days after sample collection to allow for shipment for analysis. If sample cannot be analysed within 4 days the plasma needs to be frozen immediately.
The results of the GastroPanel® examination are evaluated using the GastroSoft® interpretation software. The principles and algorithm used by the GastroSoft® software is based on the Updated Sydney System (USS) for classification of gastritis, as schematically presented in Appendix 2. This also illustrates the most important clinical conditions (disease states) associated with each of the gastritis phenotypes, including the risk of GC. A model report and explanation of gastropanel results is shown in appendix 3.
All patients will proceed to their planned gastroscopy which will be complemented by biopsy sampling from the antrum and corpus as usual. Routine biopsy specimens are taken from the antrum and corpus, at least two biopsies from each. These biopsies are taken from the large and small curvature of the middle antrum (biopsies 1and 4) and from the large curvature of the corpus (biopsies 5 and 6). In addition, two extra biopsies are recommended to be taken from the incisura angularis (biopsies 2 and 3). In endoscopy, all observed abnormal mucosal lesions are noted and photographed, and if necessary (e.g. suspicion of malignancy) subjected to additional biopsy.
Biopsies will be prepared for microscopic examination according to the routine procedures at the pathology laboratory of Grantham hospital. Biopsies are to be examined by expert pathologists at Grantham hospital among the daily routine samples. This is normal histology work generated at endoscopy upon seeing gastritis. Diagnoses are to be reported as per pathologist’s routine procedure.
Statistical analysis
It is expected that the data will be parametric and therefore statistical tests used for analysis of parametric data will be useable otherwise non-parametric tests will be used. All statistical analyses will be performed using the SPSS. The descriptive statistics will be done according to routine procedures. Performance indicators (sensitivity, specificity, positive predictive value(PPV), negative predictive value (NPV) and their 95%CI of individual markers and whole GastroPanel® test will be calculated separately. The area under ROC (Receiver Operating Characteristics) for each biomarker will be calculated. Because GastroPanel® is a quantitative ELISA test, these ROC curves can be used to identify the optimal sensitivity/specificity balance that gives each biomarker an optimal threshold for detection. Significance of the difference between ROC values can be estimated using probably Chi Square test to see of relationship between variables.REC name
London - Hampstead Research Ethics Committee
REC reference
15/LO/0310
Date of REC Opinion
12 Feb 2015
REC opinion
Favourable Opinion