572P - Neoantigen heterogeneity among subtypes in colorectal cancer

Background

Colorectal cancer (CRC) is a heterogeneous disease with multiple subtypes that differ in their clinical and molecular features. The presence of neoantigens derived from somatic mutations is a promising avenue for immunotherapy in CRC. However, the heterogeneity of the neoantigen landscape among CRC subtypes remains unclear.

Methods

We analyzed the whole-genomes and transcriptomes of 1063 CRC patients from a Swedish large cohort. HLA alleles were called by at least two HLA callers including HLA-HD, HLA-LA, Kourami, OptiType, PHLAT, Polysolver, xHLA for each patient. Neoantigens were predicted by pVACseq pipeline using the default settings.

Results

In total, 83,630 MHC class I neoantigens derived from 29,106 mutations (511,159 mutations were inputted) were predicted in 979 patients. Average 6.1 neoantigens were from 4,480 frameshift INDELs while only average 2.3 neoantigens were from 24,276 SNVs. Patients older than 80 years carried more neoantigens than those under 65 years, and women tended to have more neoantigens than men (Wilcoxon test, P<0.01). Tumor neoantigen burden was significantly higher in right-colon, stage II, and hypermutated (HM) tumors (P<0.0001). In total, 850 mutations were predicted to generate neoantigens in at least two patients, including KRAS G12D (6.3%), RNF43 G659X (6.1%), KRAS G12V (5.2%), TBC1D23 K647X (2.9%), and MSH3 K381X (2.8%). Neoantigen numbers from PODXL2 and CRAT mutations were significantly higher in consensus molecular subtype (CMS) 3 tumors, whereas neoantigens from JAG2 mutations were more common in CMS2 (chi-square test, P <0.05). Neoantigens derived from DNAH3, KRAS and ERBB2 mutations (37%, 47% and 52% respectively) in non-HM patients were significantly more than those (5%, 14% and 13% respectively) in HM patients. Furthermore, in stage I-III, nHM patients (n=11) with DNAH3 neoantigens have significantly longer overall survival than patients (n=24) with only DNAH3 mutations.

Conclusions

Our study constitutes the largest neoantigen dataset from integrated genome and transcriptome of CRC to date and illustrates the heterogeneity of the neoantigen landscape among CRC subtypes. The identified neoantigens may contribute to future individualized CRC immunotherapies.

Clinical trial identification

Editorial acknowledgement

Legal entity responsible for the study

Uppsala University; Umeå University; BGI-Shenzhen.

Funding

The Swedish Cancer Society; the Uppsala Cancer Foundation; the Guangdong Provincial Key Laboratory of Human Disease Genomics; the Swedish Government (CancerUU); the Erling-Persson Foundation.

Disclosure

All authors have declared no conflicts of interest.