Revealing hidden imprints: Unravelling gastric and oesophageal cancer with COSMIC v104

Gastric and oesophageal cancers remain among the most difficult cancers to catch early and treat effectively. They often develop quietly, shaped over time by environmental exposures and accumulating genetic changes. That reality sits in the background of much of the research we curate: the need to understand not just what mutations are present, but what has left its imprint on the genome, and when.

In this release, we’ve focused curation on gastric and oesophageal cancers, helping piece together those timelines, particularly in upper gastrointestinal cancers, where infection, tissue context and tumour evolution all intersect in complex ways.

A lasting imprint of infection

One notable study curated as part of this release draws attention to the enduring effects of Helicobacter pylori infection. It’s well known as a major risk factor for gastric cancer, but its effects don’t necessarily stop once the infection is treated. A recent study (PMID: 39827855) looked at tumours from patients diagnosed both before and after H. pylori eradication, comparing genetic changes across different tumour subtypes. Even in this small cohort, the presence of multiple TP53 mutations stood out. 

These mutations are thought to be linked to increased activity of activation-induced cytidine deaminase (AID), triggered by infection. What’s striking is the suggestion that this mutational process may continue, or at least leave a lasting imprint, even after the bacteria are gone. It’s a reminder that in cancer, timing matters: by the point we intervene, some key genomic damage may already have taken place.

Scale reveals what small studies cannot

A similar theme of accumulation and scale emerges in oesophageal squamous cell carcinoma (ESCC). The addition of the most comprehensive whole-genome sequencing study of ESCC to date (PMID: 32398863) brings together data from over 500 patients and millions of mutations. Rather than simply increasing volume, it helps reveal patterns that smaller studies cannot resolve, giving a clearer view of how diverse these tumours can be.

For a disease that is often diagnosed late and carries a poor prognosis, understanding that diversity is essential. It opens the door to asking more precise questions about which mutations drive progression, and which might offer points of intervention.

Rare cancers, real gaps

Rarer tumour types bring their own challenges, often because there simply aren’t enough cases to study in detail. Gastric adenocarcinoma of the fundic gland (GAFG) is one such example.

Figure 1: Nuclei stain showing absence of Epstein Barr virus-encoded RNA status in gastric adenocarcinoma of the fundic-gland. Magnification x40 Via: Liu et al PMID: 37745263

By bringing together multiple recent studies (PMIDs: 37745263, 34268625), we’ve been able to increase both the number of recorded cases and the proportion with identified mutations. That may sound incremental, but for rare cancers, even modest gains in sample size can make a real difference. They help move the field beyond isolated case reports towards something more comparable and interpretable, especially when combined with better representation of patient backgrounds, such as ethnicity.

Shared mutations, unexpected opportunities

Stepping slightly outside the main focus, two additional studies highlight how these same principles apply more broadly. A nationwide analysis of gallbladder cancer (PMID: 36335173) found that around half of tumours carry alterations that could, in principle, be targeted using existing therapies developed for other cancers. This kind of finding is encouraging, but it also depends heavily on careful annotation and comparison, linking a mutation seen in one cancer type to evidence gathered in another.

In a different vein, a case report of caecal adenocarcinoma (PMID: 25065654) illustrates how atypical presentations can complicate diagnosis. Here, a tumour with a BRAF V600E mutation metastasised to the bone marrow and mimicked a lymphoproliferative disorder, showing how genomic and clinical features don’t always align in straightforward ways.

Making sense of mutations in context

Across the almost 200 papers curated for v104, the common thread is not only the identification of mutations, but also the context in which they occur: infection, tissue type, tumour subtype, and patient background. Cancer genomes don’t evolve in isolation, and neither can our understanding of them. Carefully curated data helps bring these pieces together, connecting small studies, rare cases and large sequencing efforts into something more coherent.

In v104, the focus goes further than adding data by improving tumour annotations, clearer links between mutations and their causes, and broader patient representation. This helps researchers trace patterns across studies and supports more informed interpretation of variants, particularly where context such as infection or tumour subtype is relevant.

That coherence matters because it brings us closer to answering practical questions. When does a mutation arise? Does it persist after treatment of an underlying cause? Is it common enough to act on, or rare but informative? For diseases like gastric and oesophageal cancer, where outcomes are often still poor, revealing these hidden imprints and building that clearer picture is an essential step forward, a step that releases like v104 are designed to support.

Don’t forget to follow us on our linkedin or subscribe to our newsletter.