DNA mismatch repair deficiency (MMRd) produces a hypermutated tumour phenotype through loss of one or more MMR proteins (MLH1, MSH2, MSH6, PMS2), resulting in microsatellite instability and an increased mutational burden. MSI/MMRd defines a biologically and clinically distinct tumour subgroup across several tumour types, including gastrointestinal cancers. In oesophagogastric adenocarcinoma, MSI/MMRd is more frequent in gastric primaries, less common at the oesophagogastric junction and rare in oesophageal tumours. Reported MSI/MMRd frequencies in the upper gastrointestinal tract vary widely (4–24%) owing to heterogeneity in tumour location, stage, patient demographics and assay methodology.
Perioperative or neoadjuvant cytotoxic chemotherapy is a standard component of multimodal management for resectable oesophagogastric adenocarcinoma based on landmark trials demonstrating a survival benefit versus surgery alone. Despite this, treatment selection has limited integration of molecular biomarkers such as MMR status, and the prognostic significance of MMRd in patients treated with perioperative cytotoxic chemotherapy has been uncertain.
The investigators assessed MMR status in tissue from 1,424 patients with oesophagogastric adenocarcinoma enrolled in two large randomized UK trials: MRC OE05 and ST03. These trials investigated perioperative chemotherapy regimens in resectable disease. Associations were evaluated between MMR status and clinicopathological variables, overall survival (OS), progression-free survival (PFS) and post-progression survival (PPS).
The provided source reports the population size, number and proportion of MMRd tumours and the principal survival analyses and multivariable models. Details of the pathological assays, scoring criteria, central review procedures, and the specific chemotherapy regimens used in each trial are not included in the supplied excerpt.
Among 1,424 tumours analysed, 78 (5.5%) were classified as MMRd. The majority of MMRd cases (80.8%) were attributed to MLH1 loss. The source reports that MMR status was correlated with survival outcomes; however, detailed associations with other clinicopathological variables (for example, tumour location, stage distribution, histological subtype or patient demographics) are not provided in the excerpt.
Overall survival differed significantly by MMR status. Patients with MMRd tumours had improved OS compared with those with MMR proficient (MMRp) tumours: median OS was unreached in the MMRd group versus 28.0 months in the MMRp group (p = 0.0009). Five-year OS rates were 50.1% for MMRd tumours versus 34.3% for MMRp tumours.
Progression-free survival was also longer in patients with MMRd tumours: median PFS was 56.7 months for MMRd versus 21.4 months for MMRp (p = 0.0032).
Post-progression survival (PPS) did not differ by MMR status in the analysed cohort according to the reported results.
After adjustment in multivariable models, MMRd remained independently associated with improved survival outcomes. For overall survival the hazard ratio was 0.60 (95% CI 0.42–0.87; p = 0.007), indicating a 40% relative reduction in the hazard of death compared with MMRp tumours. For progression-free survival the hazard ratio was 0.64 (95% CI 0.45–0.90; p = 0.011). These analyses support that MMRd is an independent favourable prognostic factor in this cohort.
This analysis of patients enrolled in the OE05 and ST03 perioperative chemotherapy trials suggests that MMRd retains a favourable prognostic association in resectable oesophagogastric adenocarcinoma treated with perioperative cytotoxic chemotherapy. The results address clinical concerns that cytotoxic chemotherapy might be detrimental in MMRd tumours by showing improved OS and PFS prior to disease progression in MMRd cases. However, the absence of a difference in PPS suggests that MMR status did not influence survival after progression in this dataset.
The findings therefore challenge the assumption that perioperative cytotoxic chemotherapy negates the prognostic advantage of MMRd in the pre-progression setting. Integration of MMR testing into clinical decision-making for resectable oesophagogastric adenocarcinoma may merit further consideration given these outcomes.
The supplied source excerpt does not include several details that would be relevant for interpretation and clinical application: granular clinicopathological correlates by MMR status (such as tumour location breakdown, stage distribution or demographic differences), assay methodology and quality control for MMR assessment, exact chemotherapy regimens and dosing used in each trial arm, and subgroup or sensitivity analyses. Details of subsequent lines of therapy after progression and how PPS was measured, beyond the summary statement that PPS did not differ by MMR status, are also not provided in the text supplied.
In summary, in this pooled analysis of 1,424 patients from the OE05 and ST03 trials, MMRd occurred in 5.5% of tumours (predominantly MLH1 loss) and was independently associated with improved overall and progression-free survival following perioperative cytotoxic chemotherapy, while post-progression survival did not vary by MMR status. The excerpted report does not provide complete methodological or subgroup detail; readers should consult the full article for comprehensive methods and supplementary data.