Prognostic value of cell-free DNA in patients with oropharyngeal cancers.

Head and Neck Cancer
Session Type and Session Title: 
This abstract will not be presented at the 2016 ASCO Annual Meeting but has been published in conjunction with the meeting.
Abstract Number: 


J Clin Oncol 34, 2016 (suppl; abstr e17511)
Julia Beck, Margret Rave-Fraenk, Kirsten Bornemann-Kolatzki, Markus Schirmer, Michael Oellerich, Howard Urnovitz, Ekkehard Schütz, Martin Canis; Chronix Biomedical, Göttingen, Germany; Department of Radiotherapy and Radiation Oncology, University Medicine Göttingen, Göttingen, Germany; Department of Clinical Pharmacology, University Medicine Göttingen, Göttingen, Germany; Department of Otorhinolaryngology, Head and Neck Surgery, University Medicine Göttingen, Göttingen, Germany

Abstract Disclosures


Background: Tumor derived plasma cell-free DNA (TcfDNA) is described as biomarker to monitor tumor burden in cancer. Large somatic genome aberrations are hallmarks of malignancies and detectable in cfDNA. However, levels of TcfDNA are variable according to cancer types. We obtained copy-number instability (CNI) scores of cfDNA in treatment-naïve head and neck cancers before and after therapy. Methods: In a blinded study, plasma cfDNA from 21 cancer (13 hypopharyngeal, 8 laryngeal) patients were investigated before therapy (3 surgery alone and 18 followed by platinum-based chemoradiotherapy with total radiation doses of 64Gy) together with 4 non-malignant controls. Samples during treatment were obtained for five patients. Plasma cfDNA was subjected to whole genome sequencing (Illumina NextSeq500). Copy-numbers in 5.5Mbp (HG19) windows with coverage of 24,000-fold per bin were transformed into log2 ratios and converted into Z-values vs. a reference population. Values reaching the critical distance to zero were used to calculate a general CNI-score. Results: CNIs above the 97.5thpercentile (Score: 30) of the reference population were detected in 18 (Score: 41-3880) of 21 pre-surgical samples and none in the benign run controls. Significant differences in CNIs were detected, stratified by nodal stages < N2 (mean: 172, SE: 70, n = 7) versus ≥ N2 (mean: 882, SE: 362, n = 12, p = 0.05). In contrast, no significant differences were observed for primary tumor stage ( ≤ T2 vs. > T2). Rapid decreases of CNIs immediately after surgery were seen in 5|5 patients with close follow-up to 10% of baseline values. One patient, not falling below 30 after surgery with subsequently rising CNIs was suffering a relapse thereafter. Complete remission of one tumor (initial CNI = 41) was reflected by a CNI of ≤ 6 after surgery and during 3 months follow-up. Conclusions: Even small sized head and neck tumors are detectable by cfDNA analyses if nodal stages are higher, therefore CNI-scores also might be of prognostic value. If CNIs stay elevated or rise during follow up, a relapse must be suspected. Overall cfDNA shows an unexpectedly high sensitivity in such cancers and might be a valuable addition to patient risk stratification and monitoring.