Optimized primer/probe sets for HPV-16 detection in head and neck carcinoma

Presentation: C055
Topic: Pharynx / Larynx Cancer
Type: Poster
Date: Thursday, April 19, 2018
Session: 9:00 AM - 7:00 PM
Authors: Yuki Saito, MD, PhD1, Alexander V Favorov, PhD2, Shuling Ren, MD1, Akihiro Sakai, MD, PhD1, Takahito Fukusumi, MD, PhD1, Mizuo Ando, MD, PhD3, Chao Liu, MD1, Joseph A Califano, MD1
Institution(s): 1Moors Cancer Center, University of California San Diego, 2Division of Oncology Biostatistics and Bioinformatics, Sidney Kimmel Comprehensive Cancer Center, Department of Oncology, Johns Hopkins University, 3Otolaryngology-Head and Neck Surgery, University of Tokyo

Background: Current commercially available primer/probe sets for HPV-16 have been designed to identify high-risk HPV in cervical carcinoma and have not been updated for 20 years. HPV detection in saliva or plasma suggested being a promising biomarker for early detection, recurrence, and prognosis. To detect HPV-16 genome in a fragmented DNA or even low DNA copy number samples, we sought to design rational primer/probe sets for HPV-16 in head and neck carcinoma (HNSCC) using whole genomic sequencing data.

Methods: We analyze the physical coverage of the HPV-16 genome present in HPV-related HNSCC in TCGA cohorts and our 40 cohorts, and realigned all the whole genome sequencing data against HPV-16 virus genome. We identified the regions with highest maximum read density and minimum read density of all tumors. Then we selected the candidate primer/probe sets by employing standard PCR design techniques Primer-BLAST. These primer/probe candidates was refined by testing using deidentified plasma samples from mixed human controls with spiked HPV genome that contain the region of interest, and we performed characterization of candidate primer/probe sets focusing on specificity, efficiency of amplification, and sensitivity in terms of template copy number and compare to E6 and E7 primer/probe sets.

Result: In our cohort (N=40), total read counts were over 451 billion. Thirty-one of 40 tumors (77%) showed the highest read density in the junction between E5 and L2 lesion and the E1 region contained the minimum read density across all samples. In TCGA samples (N=94), total read counts were 166 thousand. This also showed the highest read density in the junction between E5 and L2 lesion but did not show consistent presence in the E1 region. Candidate PCR probe and primer sets were created, 5 of 12 primer/probe sets confirmed reaction with 30 or less HPV copy number samples mixed human controls by real-time PCR. Of 6 sets, 3 candidate primer/probe sets also showed more sensitive to E6/E7 existing primer/probe sets under the other HPV-related HNSCC cohorts (N=9) diluted by normal DNA.

Conclusion: These new optimized primer/probe sets are promising and specific to HNSCC and demonstrated improved performance relative to prior, historical assays.