Head and Neck Cancer

PD 10 - H&N 2 - Head and Neck Poster Discussion

1089 - Molecular Classification of Lymph Node Metastases in HPV-Negative Head and Neck Cancers Identify Distinct Phenotypes That Predict for Survival

Tuesday, October 23
1:48 PM - 1:54 PM
Location: Room 217 A/B

Molecular Classification of Lymph Node Metastases in HPV-Negative Head and Neck Cancers Identify Distinct Phenotypes That Predict for Survival
L. Huang1, O. David2, R. J. Cabay2, K. Valyi-Nagy2, V. Macias2, R. R. Weichselbaum1, and M. T. Spiotto1,2; 1University of Chicago, Chicago, IL, 2University of Illinois Hospital and Health Sciences System, Chicago, IL

Purpose/Objective(s): Approximately half of patients with lymph node metastases (LNMs) in HPV-negative head and neck squamous cell cancers (HNSCCs) achieve cure. Given that LNMs likely represent a biological selection of aggressive tumor variants, we hypothesized that molecular classification of LNMs would identify unique biological subtypes that better predict patient outcomes.

Materials/Methods: We performed mRNAseq and miRNAseq on 73 LNMs and 29 matched primary tumors from 33 patients with HPV-negative HNSCCs (sites: 25 oral cavity, 1 oropharynx, 4 larynx, 1 hypopharynx, 2 other) treated with surgery and post-operative radiation therapy. Unsupervised clustering was performed using iCluster+. Functional gene set enrichment analysis was performed using EGSEAv1.6.1. Survival analyses were performed using Kaplan-Meier analysis and Cox proportional hazard models. Three published prognostic HNSCC gene signatures [173 gene signature, (De Cecco et al. 2014); 10 gene radiosensitivity index (RSI; Eschrich et al. 2009); 13 gene risk score (RS; Lohavanichbutr et al. 2013)] were tested on the transcriptomes of LMNs and primary tumors. A 73 gene LNM-specific classifier was developed using Prediction Analysis of Microarrays and validated on prospective and retrospective datasets encompassing breast cancer (GSE56493), melanoma (GSE65904) and HNSCCs (E-TABM-114).

Results: For the entire cohort, median follow-up, 2y locoregional control (LRC) and overall survival (OS) was 25.8m, 66.7% and 52.7%, respectively. iCluster identified 3 distinct subtypes of LNMs having enrichment in immune (Group 1), invasive (Group 2) or metabolic/proliferative (Group 3) gene expression signatures. Group 2 subtype was associated with significantly worse LRC (1y LRC: Group 2: 0.0% vs. Group 1: 91.7% vs. Group 3: 72.9%; P = .002) and survival (P = .02). On multivariate analyses, Group 2 subtype (ref) remained the only predictor for worse LRC (Group 1 HR: 0.1; 95% CI 0.01-0.71; P = .02; Group 3 HR: 0.2; 95% CI 0.03-0.99; P = .04). These LNM subtypes and associated differences in LRC or OS were not observed in matched primary tumors. Published prognostic HNSCC gene signatures stratified high risk cohorts with worse LRC and/or RFS only when applied to the transcriptomes of LNMs but not to primary tumors (173 gene signature P = .04; 10 gene RSI P = .02; 13 gene RS P = .001). A LNM-specific classifier identified similar Group 2 subtypes in breast cancers, melanomas and HNSCCs. In breast cancers and melanomas, a LNM-specific classifier only predicted for worse disease control and/or survival when applied to LNMs but to other sites of disease.

Conclusion: LNM-specific phenotypes were associated with distinct outcomes in HNSCCs and likely other cancers. Identification of aggressive LNMs subtypes will enable the selection of patients who would benefit from treatment intensification.

Author Disclosure: L. Huang: None. R.J. Cabay: None. K. Valyi-Nagy: None. V. Macias: None. R.R. Weichselbaum: None. M.T. Spiotto: None.

Michael Spiotto, MD, PhD

Disclosure:
Employment
The University of Chicago: Assistant Professor: Employee

Presentation(s):

Send Email for Michael Spiotto


Assets

1089 - Molecular Classification of Lymph Node Metastases in HPV-Negative Head and Neck Cancers Identify Distinct Phenotypes That Predict for Survival



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 Molecular Classification of Lymph Node Metastases in HPV-Negative Head and Neck Cancers Identify Distinct Phenotypes That Predict for Survival