NEW YORK (GenomeWeb) – Researchers from Cantonal Hospital of Winterther, University Hospital of Zurich, and elsewhere have identified and validated a biomarker signature for pancreatic ductal adenocarcinoma (PDAC) that they claim could help guide therapy selection in patients with resectable disease.
According to the paper, published today in JAMA Surgery, the researchers used two independent cohorts of PDAC patients to run a genome-wide screen for functional SNPs that affect patient survival rates. They identified two SNPs located in cancer-linked genes that are known to regulate disease progression, invasion, and metastasis and are associated with survival after PDAC resection, according to the paper. The found that patients who carry risk alleles at one of both SNP loci were at higher risk for tumor-associated death compared to patients with protective genotypes.
Specifically, the team analyzed data from a prospective cohort of 195 PDAC cases in Europe who had undergone pancreatic resection looking for high-frequency SNPs associated with allelic differences in tumor-associated survival. They looked for SNPs that either alter protein structure or reside in known regulatory noncoding genomic regions. They then used data from 136 patients from the pancreatic cancer cohort in the Cancer Genome Atlas database to validate the findings.
To come up with an initial list of SNPs, the researchers genotyped over 2 million SNPs from the European cohort using Illumina's high-density Human Omni 2.5-8 SNP array. They pared down the list of variants using allele frequency information, and used the Ensembl Variant Effect Predictor software to find overlaps between SNP loci and known regulatory regions from publicly available experimental genome-wide assays. They also looked for common coding SNPs that are known to have deleterious effects on protein function. The final list contained more than 38,000 SNPs.
The researchers them performed a statistical analysis of this much smaller batch of SNPs to identify those that show allelic differences in tumor-associated death. This generated a set of 73 candidate SNPs that were tested for similar association with PDAC survival using the TCGA pancreatic cancer dataset. Additional testing yielded two variants located in the regulatory regions of the CHI3L2 and CD44 genes that showed "considerable allelic differences in their ability to prognosticate tumor-associated survival in the TCGA cohort." Further genotype testing revealed, among other details, that individuals with homozygous forms of the alleles in either SNP were at increased risk for tumor-associated death compared to heterozygous patients. Homozygous patients were estimated to survive 15 months on average following tumor resection compared to an average of 24 months for heterozygous patients.
PDAC accounts for 85 percent of pancreatic cancers and is one of the most malignant forms of the disease. One of the benefits of this signature is that SNP genotyping can be determined at the time of diagnosis by blood test to identify patients at high risk for rapid tumor progression, the researchers note in the paper. In contrast, other prognostic factors currently in clinical use can only be tested after surgical resection and as such have limited relevance to treatment decisions.
According to the researchers, the signature could be used to identify high-risk PDAC patients with a low probability of survival who might be candidates for clinical trials of new therapeutic strategies including neoadjuvant chemotherapy protocols. They also note that these SNPs could aid in the development of individualized genomic strategies for PDAC therapies.
A limitation of the study highlighted in the paper is that the sample size used did not allow for adjusting the survival analysis to the type of treatment used. The researchers note that future studies will be needed elucidate associations between the SNPs and individual therapeutic protocols.