MUC16 mutations and prognosis in gastric cancer; a little goes a long way

Gastric cancer is a global health problem; although incidence rates are declining, it remains the third most common cause of cancer death worldwide.¹ Gastric cancer patients with advanced disease have limited treatment options available, and most will live for less than two years, therefore exploration of gastric cancer disease biology is warranted to identify new targets.² Recent comprehensive molecular analyses have identified distinct subgroups of gastric cancer which may have therapeutic relevance, however with the exception of microsatellite unstable tumours the potential for genomically guided therapy has not been realised.^3–5

In this edition of JAMA Oncology, Li and colleagues investigate the relationship between MUC16 mutation, tumour mutation load (TML) and survival in patients with gastric adenocarcinoma.⁶ Using data from the TCGA gastric cancer dataset and a second, smaller, Asian validation cohort, the investigators demonstrate that tumours which are MUC16 mutant are more likely to have a higher tumour mutation burden (p<0.001). This association was independent of age, gender, and the presence of mutations in genes affecting genomic stability such as BRCA1/2 and POLE. Independence from microsatellite instability (MSI) status could not be determined as MSI results were not available for every TCGA patient; however, the association between MUC16 mutation and TML also appeared to be independent of derived mismatch repair deficiency signatures. Li et al then proceed to demonstrate that patients who have MUC16 mutant tumours have longer median overall survival than patients with a MUC16 wildtype tumour genotype. In TCGA patients with resected gastric cancer the median overall survival was 26.7 months for MUC16 wildtype versus 46.9 months for MUC16 mutant tumours respectively (p=0.007); similar results were demonstrated in the Asian dataset. This prognostic effect of MUC16 mutation on overall survival also remained statistically significant when adjusted for confounding factors.

The findings presented by Li et al are interesting because if MUC16 mutation is truly predictive of high TML this could have clinical implications. The most readily apparent use of MUC16 mutation as a surrogate for TML would be to identify gastric cancer patients who might benefit from immune checkpoint blockade. As only one in six patients with PD-L1 selected gastroesophageal cancer respond to anti-PD-1 therapy, better enrichment biomarkers are a priority.⁷ Retrospective analysis across multiple cancer types and clinical trials demonstrates a significant correlation between tumour mutational load and response to PD-1 inhibition.⁸ Prospectively in the Checkmate 577 study, non-small cell lung cancer patients with a high tumour mutation load (≥10 mutations per megabase) had significantly superior progression free survival when treated with nivolumab and ipilimumab compared to chemotherapy, independent of PD-L1 expression.⁹ The authors’ transcriptomic results support the suggestion that MUC16 mutation could predict for sensitivity to anti-PD-1 therapy; MUC16 mutant gastric tumours are immunologically “hot”. Most recently, tumour mutation load has been assessed using a commercially available next generation sequencing (NGS) panel.^8,9 However, if TML assessment were to be reduced to MUC16 sequencing only as a surrogate for a larger NGS panel, precise quantification of the relationship between MUC16 mutation and mutation rate per megabase in gastric cancer would be required to inform biomarker development.

Clinical limitations of the current work include the small size, restricted ethnic make-up, and heterogeneous nature of the validation cohort. The first major challenge to the scientific validity of the current work is the risk of a false positive association between MUC16 mutation and high TML because of characteristics associated with MUC16. MUC16 is a very large gene sited in an area with low replication timing. Regions of low replication timing are more likely to accumulate frequent mutations. Considering other relevant mutational co-variates, it is possible that MUC16 has a mutation rate which is in line with what might be expected by chance. Secondly, correlation between MUC16 mutation and high TML does not necessarily imply causation; a functional explanation for the link between MUC16 mutation and high TML is absent. An alternative explanation for the improved long-term survival of MUC16 mutant patients in this study may not relate directly to tumour mutation load. In an assessment of the tumour biology of long term survivors following pancreatic tumour resection, neoantigens (but not non-antigenic mutations) in MUC16 were identified four times more frequently in long term survivors than in patients who had shorter survival.¹⁰ In vitro, MUC16 expression is associated with proliferation and metastatic potential in tumour cells, and MUC16 is also known to inhibit natural killer cell mediated lysis of tumours.^11,12 Thus, elimination of MUC16 neoantigen presenting cancer cells via immune “pruning” could hypothetically leave a residual population of MUC16 non-expressing cells with reduced metastatic and immunosuppressive potential, which could be associated with longer survival; this could present an alternative explanation for the authors’ transcriptomic results on immune infiltrate. Data on MUC16 neoantigen presentation are not provided in the current manuscript; this should be a priority for future research.

There is a paucity of evidence linking MUC16 gene function to cancer, therefore it is unlikely that MUC16 mutations are positively selected oncogenic drivers. This makes the hypothesis that MUC16 mutations act through neoantigen presentation and immunoediting more plausible. Despite the challenges in understanding the relationship between MUC16 mutation and TML, the relationship with prognosis appears to be robust. Although the results presented by Li and colleagues are preliminary and require further validation, they could provide a signpost for clinically relevant research in gastric cancer, or in view of the ubiquity of MUC16 mutations, in other tumour types.