Date of Award

January 2019

Document Type


Degree Name

Medical Doctor (MD)



First Advisor

Wendell G. Yarbrough

Second Advisor

Natalia Issaeva


Epidemiological, molecular, biological and clinical differences have led to recognition of HPV-positive (HPV(+)) and HPV-negative (HPV(-)) head and neck squamous cell carcinomas (HNSCC) as distinct diseases. However, knowledge of mechanisms and treatments for HPV(+) HNSCC is currently lacking in part due to the scarcity of appropriate research models. Compared to HPV(+) HNSCC, prognosis remains poor for HPV(-) HNSCC, underscoring the importance of improving and personalizing therapy though improved understanding of the heterogeneity of tumor biology. To advance our understanding of HNSCC, this work focuses on (1) establishment of a repository of human head and neck cancer biospecimens for laboratory studies; (2) development of a novel HPV(+) research model from banked biospecimens; and (3) investigation of a prognostic biomarker for HPV(-) HNSCC. In Part (1), a diverse group of head and neck cancer blood and tumor specimens are collected from Yale surgeries and clinics, with emphasis on augmenting the repository of both HPV(+) and HPV(-) viably frozen HNSCC. In Part (2), an HPV(+) patient-derived xenograft (PDX) mouse model is developed from viably frozen tumors in the biobank, representing a novel method that no longer relies on fresh tissue for establishing PDXs. We show that an HPV(+) PDX generated from viably frozen tissue retains the histologic characteristics, p16 staining, and expression of HPV genes of its primary tumor. Finally, in Part (3) we identify subgroups of inactivating NSD1 mutations in HPV(+) and HPV(-) HNSCC cohorts and show that these mutations are associated with opposite survival effects in HPV(+) and HPV(-) tumors. NSD1 mutations correlate with worse survival in HPV(+) tumors and improved survival in HPV(-) tumors, with the survival advantage of HPV(-) NSD1 mutants in part due to increased sensitivity to the DNA-crosslinking agents cisplatin and carboplatin. Our study emphasizes the need for novel research models to investigate HPV(+) HNSCC and highlights the advantages of using viably frozen tissue to establish PDXs for translational studies. Furthermore, this work augments our understanding of NSD1 as a prognostic biomarker for HPV(-) HNSCC and explores the therapeutic differences between tumors with and without NSD1 mutations.


This thesis is restricted to Yale network users only. This thesis is permanently embargoed from public release.