Towards Prognostic Biomarkers for HPV+ OPSCC: Insights into Genomic Heterogeneity, Viral Integration and Tobacco Exposure from Targeted Sequencing.

Presentation: P006
Topic: Cancer Biology
Type: Poster
Authors: Travis P Schrank, MD, PhD
Institution(s): University of North Carolina


Human Papilloma Virus (HPV) associated squamous cell carcinoma of the oropharynx (OPSCC) has become the most common HPV associated malignancy in the United States.  Patients are most often treated with external beam radiation with concomitant cisplatin. High burdens of treatment related morbidity in typically young and otherwise healthy patients have engendered intense interest in improving treatment strategies for this increasingly common disease.  More specifically, efforts to reduce the dose and intensity of chemoradiation are currently being evaluated in clinical trials.  However, no molecular tools are available to help guide which HPV+ OPSCC patients should receive treatment deintensification.  In effort to fill this important gap in translational research, we have analyzed a large single-institution series of targeted next generation DNA sequencing of HPV+ OPSCC tumors. 


Targeted next generation DNA sequencing was performed through the UNCseq program.  Approximately 1,000 cancer related genes, as well as oncogenic viruses were targeted. Clinical data was collected through the IRB approved retrospective chart review.  Copy number analysis was performed using the SynthEx R package. Heterogeneity analysis was performed using PyClone. Somatic mutation calls were performed through the UNCseq program. HPV integration analysis was performed using the ViFi package.  112 cases of p16+ OPSCC were identified, including 99 cases with high numbers of HPV16 viral read counts, thus the reported data represents the largest reported series of HPV+ OPSCC tumors analyzed with NGS DNA sequencing. 


Smoking associated A>C transversion mutation rate was increased with over 20 pack-years of tobacco smoking exposure. This exposure threshold which was also associated with disease-recurrence.  Copy-number alteration burden was found to be highly associated with deep losses of distal chromosome 11q (adjusted p-value < 0.0005).  Risk of recurrence was greatly increased in patients with high burden of CNAs as well as deep loss of 11q (HR = 4.4, p = 0.0004).  43 samples had data sufficient for variant allele frequency heterogeneity (VAF) analysis. High VAF heterogeneity was associated with recurrence (p value = 0.05) as well as distal 11q loss (adjusted p value < 0.005).  Deep coverage (100-10000x) of the HPV16 viral genome has provided a high-resolution picture of HPV16/genomic interactions. 


The report data greatly expand the available NGS data describing HPV+ OPSCC.  Our analyses have revealed potential clinical/prognostic value of CNA burden, A to C transversion mutations, as well as estimated tumor poly-clonality.  Considering the decreasing costs and increasing availability of NGS technology, these findings have immediate translational value. Furthermore, high risk genomic features of CNA burden and VAF heterogeneity were associated with distal 11q loss which harbors the ATM locus. Distal 11q/ATM loss has been previously associated with genomic instability in other solid tumors. Also notably, 11q/ATM loss has been associated with tumor sensitivity to PARP inhibition, suggesting a potential targeted treatment strategy for the identified high-risk tumor groups.