Validation of a Molecular Test for Risk-stratification of Patients With High-risk Intestinal Metaplasia (GCEP2 Study)

Objectives:

The main objective of this study is to determine whether a customised molecular test can identify prospectively, within patients with intestinal metaplasia (IM), a subgroup of patients with very high risk of dysplasia or gastric cancer (GC). The secondary objectives of this study are to validate the predictive value of a panel of blood microRNA biomarkers for predicting risk of GC, as well as to determine the accuracy of detection of Helicobacter pylori infection using next-generation sequencing, and compare it with the current gold standard, urea breath test.

Study Design:

An international, multicenter cohort to evaluate the clinical utility of a customized molecular test to identify a subset of IM patients at very high risk of GC. Target recruitment is 500 subjects with IM of OLGIM Stage 2 to 4.

As the Updated Sydney biopsy protocol is not routinely performed for clinical purposes and most subjects will be recruited prior to confirming their OLGIM staging at baseline, some subjects will be withdrawn from the study if detected to have no IM, or IM assessed to be OLGIM Stage 1, at baseline OGD.

Collaborating sites are requested to provide the following specimens with associated demographic and clinic-pathological data (as indicated in Case Report Form): Snap-frozen tissue and FFPE tissue from antrum site for DNA extraction, serum for H. pylori antibody test and miRNA profiling, and buffy coat for DNA extraction.

Enrollment:

Patients seen by investigators during clinics will be considered for enrollment.

To minimize the number of subjects who are withdrawn after baseline endoscopy due to failure to detect at least Stage 2-4 OLGIM at baseline, patients with history of moderate or severe IM at either antrum or body can be prioritized for recruitment.

Baseline:

1. Questionnaire Subjects will be asked to complete a baseline questionnaire, providing information on the demographics, personal medical history, lifestyle, and family history of GC. Clinical data and identifiers will be recorded.
2. Urea Breath Test Subjects will fast for 6 hours or overnight before undergoing the Urea Breath Test (UBT). Breath collection will be performed before ingestion of 13C urea, and at 10min, 20min and 30min after ingestion.
3. OGD and biopsies

Subjects will undergo a baseline OGD under high-definition white-light (1080p) and biopsy to ascertain OLGIM status. A video-recording of the OGD will be carried out during the procedure for validation of diagnosis. Gastric mucosal biopsies will be taken as follows:

• 1 biopsy each from AL, AG, IA, BL, BG according to the Updated Sydney System will be fixed in 4% paraformaldehyde (PFA) and sent for histological examination
• 2 adjacent biopsies at AL, and 2 adjacent biopsies at BL, will be snap-frozen in liquid nitrogen

The locations are defined as follows:

• AL - lesser curvature of the antrum, within 2-3cm of the pylorus
• AG - greater curvature of the antrum, within 2-3cm of the pylorus
• IA - incisura angularis
• BL - lesser curvature of the corpus, 4cm proximal to the angulus
• BG - middle portion of the greater curvature of the corpus, 8cm from the cardia

Should the endoscopist observe suspected areas of IM in the antrum and/or corpus that are within the above-defined Updated Sydney System locations, then biopsies for histological examination, as well as 2 adjacent frozen biopsies, should be taken from these suspected areas of IM instead.

Should the endoscopist observe suspected areas of IM in the antrum and/or corpus that are outside of the above-defined Updated Sydney System locations, then the 2 adjacent frozen biopsies should be taken from these suspected areas of IM instead.

All mucosal lesions identified during OGD will be biopsied separately as a part of normal clinical practice and sent for histopathological assessment.

Frozen biopsy samples will be stored at a high-quality tissue bank.

Investigators will follow the above protocol. However, they may deviate from protocol if the interests of the patient require so.

Blood Collection, Processing and Analyses:

20mls of blood will be drawn from each subject. Serum, plasma and buffy coat will be extracted.

Histological Assessment:

All biopsy samples will be assessed for degree of chronic gastritis, atrophic gastritis, presence of H. pylori organisms, IM and dysplasia. These will be scored using the Updated Sydney System. Presence of IM is indicated by presence of globet cells based on hematoxylin and eosin (H\&E) staining or Alcian Blue (AB) positive expression, and will be classified into mild, moderate and marked. Dysplasia will be graded by the revised Vienna classification, and classification of carcinoma will be according to the WHO classification of tumors and AJCC staging system. The percentage of gastric epithelial mucosa showing the features of IM in biopsies will be evaluated to derive OLGIM stage. IM will be further sub-typed as type I, II and III based on histological evaluation, Periodic acid-Schiff/Alcian Blue (PAS/AB) stain and Gomori Aldehyde Fuschin/Alcian Blue (GAF/AB)/ High Iron Diamine-Alcian Blue (HID-AB) special stains. Immunohistochemistry staining will be performed for antibodies commonly used to diagnose gastric cancer and its premalignant lesions including but not limited to TROP2, CEACAM5, MUC1, MUC2, and MUC5AC.

Molecular test:

Three main genomic alterations - increased mutation frequencies, somatic copy number alterations (sCNAs) and telomere shortening - were previously established to show associations with disease progression of IM to dysplasia or GC, or persistence of IM. A significant proportion of sCNAs in IM samples were found to be located at chromosome 8q, where the oncogene Myc resides. A customised molecular test will be developed based on a panel primarily targeting genomic regions with recurrently mutated genes, epigenomic changes and chromosome 8q amplification in IM patients. The customized test aims to further stratify IM patients. Total genomic DNA will be extracted from biopsies (frozen tissue and FFPE) and matched blood samples. A targeted gene panel will be applied so that specific genomic regions of interest are captured, reducing the cost and amount of data analysis significantly. Library preparation will be performed using the target enrichment assays according to manufacturer instructions. Mutation calls and chromosome 8q amplification will be determined using the Genome Analysis Toolkit (GATK) software. Subjects are classified as test-positive if a somatic variant with at least 10 variant supporting reads or a copy number variant with segmented mean coverage of at least 2 standard deviations away from the copy neutral mean is identified.

miRNA profiling: Total RNA from serum is isolated and converted to cDNA, which is then quantified. Target miRNA expression levels after normalization of both technical and biological variations are analysed to identify panels of miRNAs with the highest discriminatory power between healthy and disease states. Each subject will be assigned a score based on miRNA expression profile, indicating the possibility of having GC. The result will be compared with OGD and histology which is the gold standard for diagnosis of GC.

Follow-up:

Subjects will be followed-up at the clinic or via telephone for status update at Years 1 and 3. Clinical data will be collected through a questionnaire. A window period of 6 months before or after the anniversary baseline OGD date is acceptable.

Subjects will undergo a surveillance OGD at Years 2 and 4 to assess whether subject has reached endpoint. Biopsies will be taken following the protocol described. Clinical data will be collected through a questionnaire and the database updated. A window period of 6 months before or after the anniversary baseline OGD date is acceptable.

Subjects will be followed-up yearly at the clinic or via telephone for status update for Years 5-10. Clinical data will be collected through a questionnaire. A window period of 6 months before or after the anniversary baseline OGD date is acceptable. Should the incidence of GC at or after Year 4 among the cohort be sufficient to reject the null hypothesis with the level of significance and power as described, yearly follow-up for Years 5-10 may be discontinued.

Sample size calculation:

The proposed sample size for the prospective cohort study was estimated using the logrank test for the time-to-progression outcome. Assuming a 4-year cumulative incidence of progression of 10% in the test positive group and 3% in the test negative group respectively (i.e. hazard ratio, HR = 3.5), a sample size of 480 will be required based on a level of significance of 5% and a power of 85%. Further assuming an attrition of 5%, the overall sample size will be 500.

Statistical Analysis:

Categorical variables will be analysed using chi-square test or Fisher's exact test. Continuous variables will be analysed using Student's t-test or Mann-Whitney U test. Parameters with P value <0.05 in the univariate analysis will be included in the multivariate analysis. Multivariate analysis will be performed using Cox regression analysis. P <0.05 is considered to be statistically significant.