Surgery-Only Cohort Exposes Genomic Signals Behind Oesophageal Adenocarcinoma Survival
In oesophageal adenocarcinoma, conventional staging leaves too much room for uncertainty. Two patients with similar TNM categories can follow very different postoperative courses, and this preprint asks whether tumour genomics can help explain why. Investigators examined a 451-gene cancer panel in 65 patients who underwent surgical resection without perioperative treatment, creating a cohort in which tumour biology was not blurred by neoadjuvant chemotherapy or radiotherapy.
The design gives the analysis unusual clarity. By restricting the dataset to surgery alone, the study isolates the cancer’s intrinsic behaviour rather than the treatment response layered on top of it.
Median age at operation was 72 years, 88% of patients were male, and the median number of lymph nodes removed was 32. Follow-up extended beyond 11 years, long enough to capture survival patterns that shorter studies often miss. Pathologic staging remained broad: 14% of patients had stage 1 disease, 14% had stage 2, 45% had stage 3, and 6% had stage 4 disease defined by non-locoregional nodal metastases. Copy number changes were common, and the somatic mutation profile was consistent with earlier reports in this disease.
CCND1 and EGFR Amplification Carried the Strongest Adverse Prognostic Signal
Several genomic alterations tracked with worse survival, but two amplification events stood out. CCND1 amplification was associated with a hazard ratio of 3.08 and a p value of 0.02, while EGFR amplification carried a hazard ratio of 2.98 with a p value of 0.01. In practical terms, both point toward aggressive proliferative signaling. CCND1 drives cell-cycle progression through cyclin D biology, while EGFR amplification can intensify growth-factor signaling, invasion, and resistance to cellular stress.
A single genomic copy-number event can mark a tumour that has already escaped normal growth control. In a disease where local extension and occult dissemination often determine outcome, those signals are biologically plausible markers of the cancers most likely to recur after surgery alone.
Loss of ARID2 also associated with inferior survival, with a hazard ratio of 2.12 and a p value of 0.05. KMT2D loss showed a similar direction of effect, with a hazard ratio of 1.99 and a p value of 0.07. Both genes sit in chromatin regulation pathways, where disruption can reshape transcriptional programs, DNA repair, and cell identity. In oesophageal adenocarcinoma, epigenetic instability may be one route to a more adaptable and therapy-resistant phenotype, even when no systemic therapy has been given.
One adverse signal emerged from a missense mutation in RYR1. The association with worse survival was statistically significant, with a hazard ratio of 2.68 and a p value of 0.01. RYR1 is not a canonical oesophageal cancer driver, so this finding invites caution and further validation. Even so, recurrent survival associations in genes outside the usual oncogenic shortlist can reveal pathways that standard panels overlook, especially when the cohort is biologically enriched by long-term follow-up and uniform treatment.
Why a Surgery-Only Design Matters for Prognostic Genomics
Most contemporary oesophageal cancer datasets mix treatment intensities. That makes it difficult to know whether a genomic alteration predicts natural tumour aggressiveness or simply reflects sensitivity to chemotherapy or radiation. This cohort avoids that problem by focusing on patients treated with surgery alone, reflecting the institutional practice at the time and allowing a cleaner read on tumour biology.
The benefit of that approach is especially important in a cancer where perioperative therapy can change both survival and recurrence patterns. When treatment effects are stripped away, amplification, loss, and point mutation become easier to interpret as intrinsic prognostic markers. The result is not a clinical decision tool ready for routine use, but a more credible map of which tumours behave badly despite apparently comparable stage.
The data also fit with the broader problem of TNM staging in oesophageal adenocarcinoma. Stage captures anatomy, but not all the molecular machinery that drives invasion, metastatic spread, and postoperative relapse. A patient with stage-matched disease may still face very different biology underneath the microscope. Genomic profiling offers a route to refine that gap, especially if adverse alterations can be grouped into pathways with targetable vulnerabilities.
Several of the alterations identified here are potentially actionable in principle. EGFR amplification, for example, has obvious therapeutic relevance in other settings, although no treatment claim can be made from this cohort alone. The same applies to cell-cycle and chromatin-remodelling abnormalities. What this study does provide is a prioritised list of candidate biomarkers worth testing in larger series, ideally with independent validation and modern treatment-era cohorts.
The limitations are equally clear. The sample size was small, the study was observational, and the preprint status means the findings have not yet undergone peer review. The cohort came from a single treatment approach, which strengthens internal biological interpretation but limits generalisability. Even with long follow-up, the number of events available for genomic association testing remains constrained, so each statistically significant result should be viewed as hypothesis-generating rather than practice changing.
Still, the central message is hard to ignore. Surgery alone has exposed a set of genomic lesions linked to survival in oesophageal adenocarcinoma, and those lesions cluster around cell-cycle control, growth-factor signalling, and chromatin regulation. If validated, this kind of profiling could help separate patients with anatomically similar disease into biologically distinct risk groups, which is exactly where prognostic pathology has struggled for years.