ESMO Congress 2023 | OncologyPRO

Abstract 649P

Background

AA is a rare malignancy with a predilection for peritoneal metastases without hematogenous or lymphatic metastases. Despite this, AA is treated with chemotherapy similar to colorectal cancer. Studies suggest inefficacy of these regimens for many AA patients supporting the likelihood of a unique genomic landscape.

Methods

WES of tumor tissue was performed as part of personalized circulating tumor DNA (ctDNA) testing (Signatera^TM) in 855 patients with AA. Tumor mutational burden (TMB) and microsatellite instability (MSI) were assessed; Kyoto Encyclopedia of Genes and Genomes (KEGG) network analysis was performed. WES data on stage-matched colon (CC, N=833) and rectal cancer tissues (RC, N=841) were compared with AA.

Results

A total of 98,310 variants were identified from 855 AA tumor tissues (median 132.5 variants/tumor). AA had lower TMB (median: 2.14 Muts/Mb) than RC or CC (median: 3.55 and 4.31 Muts/Mb, respectively). Only 1.3% (11/855) of the AA were MSI-high, compared to RC (2.5%) and CC (16.76%). Comparison of frequently mutated genes highlighted that AA has much less frequent APC and TP53 and more frequent GNAS mutations. Despite differences in driver mutations and overall lower TMB, the mutational signatures for AA, CC, and RC were similar with SBS1 (clock-like) and SBS6 (MMR-deficient) being predominant in each tumor type. Co-mutation analysis identified a dense interaction network of 805 co-occurring or mutually exclusive interactions highlighted by connections between Wnt, Ras-MAPK, and DNA repair pathways. Copy number analysis identified that AA has less aneuploidy relative to CC or RC, however, within AA, TP53 mutation was associated with significantly higher aneuploidy relative to wildtype. Interestingly, TP53 mutations were also associated with detectable ctDNA in AA (44.6% vs. 24.9%, p<0.001), but not in CC or RC.

Conclusions

WES data on AA reveals important differences in driver somatic mutations, CNV, and co-mutational networks compared to CC and RC. This is the first large-scale genomic profiling of AA, establishing the genetic landscape of this rare and orphan disease. Future studies should focus on AA-specific treatments rather than the extrapolated CC/RC experience.

Clinical trial identification

Editorial acknowledgement

Legal entity responsible for the study

The authors.

Funding

Has not received any funding.

Disclosure

S. Rivero-Hinojosa: Financial Interests, Personal, Stocks/Shares: Natera; Financial Interests, Personal, Full or part-time Employment: Natera. V. Aushev, A. Jurdi, M.C. Liu: Financial Interests, Personal, Full or part-time Employment: Natera, Inc.; Financial Interests, Personal, Stocks/Shares: Natera, Inc. All other authors have declared no conflicts of interest.