NEW YORK (GenomeWeb) – Researchers from Harvard Medical School and elsewhere suggest there is a link between colorectal cancer samples with high numbers of predicted neoantigens and lymphocyte infiltration and survival.
The researchers took advantage of data collected by two long-running studies, the Nurses' Health Study and the Health Professionals Follow-up Study, to combine epidemiological and pathological data with genomic data. After sequencing the whole-exomes of nearly 620 colorectal cancer samples, the researchers not only identified novel recurrently mutated genes, but also gauged the samples' neoantigen load based on their mutational profiles and correlated that with immune cell infiltration, as they reported today in Cell Reports.
"We were looking for genetic features that predict how extensively a tumor is infiltrated by lymphocytes and which types of lymphocytes are present," co-first author Marios Giannakis from the Dana-Farber Gastrointestinal Cancer Treatment Center and the Broad Institute said in a statement. "We found that tumors with a high 'neoantigen load' … tended to be infiltrated by a large number of lymphocytes, including memory T cells. … Patients whose tumors had high numbers of neoantigens also survived longer than those with lower neoantigen loads."
He and his colleagues performed whole-exome sequencing on 619 colorectal cancer samples that arose in the NHS and HPFS cohorts during follow-up, as well as on matched normal samples. They sequenced these FFPE samples to an average 90X depth, and also collected detailed pathological and clinical data on the patients.
Using MutSigCV, the researchers uncovered 90 significantly mutated genes, many of which has been previously found in the Cancer Genome Atlas Network Analysis. However, 73 of these genes hadn't before been designated as driver genes in colorectal cancer. According to the researchers, these genes have roles in processes ranging from kinase activity to DNA and RNA binding.
Pathways like WNT, RAS, and TGF-beta that have key roles in colorectal cancer development were also included among this 'long tail' of mutated genes, the researchers noted.
Tumors can also develop mutations that lead to the generation of novel peptides that the host immune system then recognizes as foreign to provoke an immune response to disease, Giannakis and his colleagues wrote.
To look for mutations that could have such immunogenic potential, the researchers determined the human leukocyte antigen (HLA) class I type of each sample and then used mutations to predict neopeptides that would bind those personal HLA molecules with high affinity. Significantly mutated colorectal cancer genes like RNF43, KRAS, and NRAS, they found, harbor recurrent neopeptides.
This neoantigen load is associated with an increased overall lymphocytic score, which, in turn, has been linked to longer survival, the researchers reported. In particular, they found that neoantigen load was most significantly associated with tumor-infiltrating lymphocytes (TILs) in their cohort. This suggested that neopeptides might be be recognized by TILs.
Neoantigen load also correlated with CD45RO+ T cell density, but not with CD8+, CD3+, or FOXP3+ T cell density, they found, supporting the notion that memory T cells are important for neoantigen-directed tumor recognition.
The researchers also confirmed that neoantigen load is correlated with better colorectal-cancer specific survival.
Giannakis and his colleagues also uncovered an enrichment of mutations in HLA exon 4, which encodes the T cell receptor-binding domain. They further noted samples with higher TIL grading harbored more HLA mutations, suggesting that these HLA gene mutations may be under positive selection in tumors with high immune infiltrates. This supports the idea that HLA mutations could provide tumors a way of escaping immune-based elimination.
These findings, the team concluded, could help inform immunotherapeutic approaches for colorectal cancer.