Head and Neck Cancer

SS 06 - H&N 1 - Current Topics in Post-Operative Radiation Therapy

38 - Supplanting the Conventional Risk Groups of Oral Cavity Cancers with Gene Expression-Based Signatures

Sunday, October 21
4:55 PM - 5:05 PM
Location: Room 006

Supplanting the Conventional Risk Groups of Oral Cavity Cancers with Gene Expression-Based Signatures
E. I. Sarihan1, S. Koyfman2, N. M. Woody2, B. Matia1, N. P. Joshi2, J. L. Geiger3, E. Lamarre4, B. Prendes4, J. Ku1, R. R. Lorenz4, J. Scharpf4, B. B. Burkey4, D. J. Adelstein3, and M. Abazeed2; 1Cleveland Clinic, Cleveland, OH, 2Department of Radiation Oncology, Taussig Cancer Institute, Cleveland Clinic, Cleveland, OH, 3Department of Hematology/Oncology, Taussig Cancer Institute, Cleveland Clinic, Cleveland, OH, 4Department of Otolaryngology, Head and Neck Institute, Cleveland Clinic, Cleveland, OH

Purpose/Objective(s): The standard treatment for patients with oral cavity cancer (OCC) after surgery is adjuvant radiotherapy for patients with intermediate risk and chemoradiation for patients with high risk pathologic variables. Despite this, 20-40% of patients experience locoregional failure (LRF) and/or distant failure. We posited that gene expression-derived tumor taxonomies fit to clinical outcomes can predict treatment failures and therefore guide more nuanced clinical decision making. Herein, we report on a score model based on OCC gene expression characteristics that can augment or replace risk stratification and treatment decisions.

Materials/Methods: Formalin-fixed paraffin-embedded (FFPE) tissue samples from patients with OCC were subjected to quantitative nuclease protection and next-generation sequencing. Two gene expression predictors, one for LRF (pLRF) and another for any failure (pAF) were selected on the basis of differential gene expression and predicted biological impact, the latter as measured by gene set enrichment results (GSEA). The Cancer Genome Atlas (TCGA) head and neck squamous cell carcinoma dataset (n=482) was used to validate the prognostic performance of our signature.

Results: Of 123 patients included in the study, 76 and 47 were with intermediate- and high-risk, respectively. All 123 patients received radiation alone, permitting the non-treatment confounded comparison across conventional risk groups. The Kaplan-Meier estimates for locoregional control at 3 years between conventional high and intermediate-risk patients were 0.56 (0.42 to 0.74; 95% CI) and 0.72 (0.62 to 0.84; 95% CI), respectively. There were substantial instances of disagreement between conventional and gene expression predicted risk, suggesting distinct classification scales. GSEA of the gene sets demonstrated a role for apoptosis, p53 pathway, G2-M checkpoint, hypoxia and KRAS signaling. The estimated rate of locoregional control at 3 years for the patients with higher vs lower pLRF scores were 0.59 (0.47 to 0.74; 95% CI) and 0.73 (0.61 to 0.86; 95% CI), respectively (P = 0.067). The estimated rate of progression free survival at 3 years for the patients with higher vs lower pAF scores were 0.39 (0.29 to 0.55; 95% CI) and 0.75 (0.65 to 0.88; 95% CI), respectively (P < 0.0001). On multivariate analysis, the gene expression scores and not the conventional risk groups predicted both LRF and any failure (P = 0.05 and P = 0.04). pLRF also predicted overall survival in the TCGA dataset to validate its prognostic value (P = 0.041).

Conclusion: We developed gene expression classifiers that predict both LRF and any failure in patients with OCC treated with surgery and adjuvant radiotherapy. Retrospective validation on larger datasets and prospective validation are needed. These classifiers can potentially identify patients who have higher risk of failure and hence who should be considered for intensification strategies with the addition of concurrent and/or adjuvant systemic therapy.

Author Disclosure: E. Sarihan: None. S. Koyfman: Research Grant; Merck. B. Matia: None. J.L. Geiger: None. J. Ku: None. D.J. Adelstein: None. M. Abazeed: Research Grant; Siemens, Bayer.

Elif Irem Sarihan, MD

Disclosure:
No relationships to disclose.

Presentation(s):

Send Email for Elif Irem Sarihan


Assets

38 - Supplanting the Conventional Risk Groups of Oral Cavity Cancers with Gene Expression-Based Signatures



Attendees who have favorited this

Please enter your access key

The asset you are trying to access is locked. Please enter your access key to unlock.

Send Email for Supplanting the Conventional Risk Groups of Oral Cavity Cancers with Gene Expression-Based Signatures