Molecular profiling of signet ring cell colorectal cancer provides a strong rationale for genomic targeted and immune checkpoint inhibitor therapies


Year:

Session type:

Theme:

Muhammad Alvi1,Maurice Loughrey2,Philip Dunne3,Stephen McQuaid1,Richard Turkington3,Marc-Aurel Fuchs1,Claire McGready1,Victoria Bingham1,Brendan Pang4,Perry Maxwell1,Jacqueline James1,Graeme Murray5,Richard Wilson3,Manuel Salto-Tellez1
1Northern Ireland Molecular Pathology Laboratory,2Belfast Health and Social Care Trust,3Queen's University Belfast,4National University Health System, Singapore,5University of Aberdeen

Abstract

Background

Colorectal signet ring cell carcinoma (SRCCa) is rare but has a bleak prognosis. There is very limited information on the molecular pathology of this disease. Most studies are case reports and the few cohort studies have only looked at candidate biomarkers. Employing the use of high throughput molecular pathology techniques we have investigated the potential of personalised medicine in this disease.

Method

Using discovery (n=26) and validation (n=18) cohorts, a comprehensive analysis of mutations, DNA methylation and whole transcriptome was carried out. Microsatellite instability (MSI) status was established and immunohistochemistry (IHC) was used to assess adaptive immunity (CD3) and immune checkpoint (PDL1) protein expression.

Results

Unsupervised clustering of DNA methylation data split the cohorts into hypermethylated (n=18, 40% of cases) and hypomethylated groups (n=26, 60% of cases). The hypermethylated group was enriched for BRAF V600E (P<0.01), MSI (p<0.01) and CpG island methylator phenotype (CIMP) (p<0.01). These cases were also found to have a heavy T-cell immune infiltrate (p<0.01) measured by CD3 IHC and were also positive for PDL1 expression. The hypomethylated group did not show any characteristic molecular features and no statistically significant difference in outcome was observed between the two groups. We also observed a targetable KIT mutation (p.M541L) in 11 out of 44 of our cases.

Conclusion

Our data shows that colorectal SRCCa phenotype comprises of two distinct genotypes. Markers previously associated with this phenotype (e.g. BRAF V600E, MSI and CIMP) are only present in the hypermethylated genotype, which also has a high immune infiltrate and expresses PDL1 making it an ideal candidate for immune checkpoint inhibitor therapy. The hypomethylated genotype does not demonstrate these molecular features and the biological mechanism by which these cancers develop the signet ring cell morphology seems different and remains unknown. In addition, one fourth of SRCCa cases can potentially be targeted by KIT inhibitors.