Why gastric cancer is studied by clustering
Colorectal cancer follows a predictable, logical progression and accumulation of mutations and genetic changes. That logical step-wise pattern cannot be found for gastric cancers, so clustering genetic studies have been performed by different groups instead, grouping tumours by their dominant molecular abnormality rather than by a sequence.
Molecular subtypes
Genomic analysis of gastric adenocarcinoma by several groups has led to a genetic classification into subtypes. The Cancer Genome Atlas (TCGA) divides them into four:
- Epstein-Barr virus (EBV)-positive tumours, with a high Epstein-Barr virus burden
- microsatellite instability-high (MSI-H) with hypermutation, or MSI, which are cancers associated with a defect in the DNA mismatch repair (MMR) system. This most commonly happens because of a mutation in the MLH1 gene, which leads to accumulation of repeated nucleotides in the genome, increasing the chance of mutation and genomic instability. MSI is associated with well-to-moderately differentiated histology, less frequent TP53 mutation, and a distal location of the cancer.
- genomically stable (GS) tumours, which mostly comprise diffuse-type gastric adenocarcinoma. The most common mutations found in the GS group are the CDH1 gene, responsible for expression of E-cadherin, which is under-expressed, and RHOA.
- chromosomally unstable (CIN) tumours, which mostly represent intestinal-type gastric adenocarcinoma and are mostly located in the proximal parts of the stomach.
The Asian Cancer Research Group (ACRG) proposes a different grouping: mesenchymal (MSS/EMT); microsatellite-unstable; microsatellite-stable with active TP53; and microsatellite-stable with inactive TP53.
Recurrent mutations and pathways
After the introduction of clustering genetic studies, several facts about gastric cancer genetics stand out:
- the most common genetic mutation is TP53, in 60 to 70% of gastric cancers
- deletion or suppression of the FHIT gene on chromosome 3p, in 60% of gastric cancers
- diminished expression of genes that inhibit entry into the cell cycle, such as p16 and p27, in 50%. p16 has no measurable prognostic effect, while decreased p27 expression has a poor prognostic effect.
- overexpression of some growth factor pathways: COX-2, 70%; HGF/SF, 60%; VEGF, 50%; c-met, 50%; AIB-1, 40%; β-catenin, 25%; and simultaneous overexpression of EGF and EGFR for autocrine activity, 15%.
- gastric-specific tumour suppressor genes: TFF1, in more than 50% of gastric cancers, and RUNX3, in more than 82%.
Clinical outcome correlates with the genome subtype: the best outcome is in microsatellite-unstable tumours, and the worst in genomically stable and mesenchymal tumours.
Markers tested in clinical practice
In about 20% of patients with gastric adenocarcinoma the tumour shows overexpression or amplification of HER2, which makes the tumour responsive to systemic trastuzumab therapy. Because this finding changes treatment, the routine diagnostic evaluation of gastric adenocarcinoma should include evaluation of HER2 expression, quantification of PD-L1, and assessment of microsatellite instability by PCR.