Date of Award

Summer 8-15-2016

Author's School

Graduate School of Arts and Sciences

Author's Department

Mathematics

Additional Affiliations

Statistics

Degree Name

Master of Arts (AM/MA)

Degree Type

Thesis

Abstract

With the fast paced advancement of modern medicine, cancer treatments have improved greatly over the past few decades; however, the overall survival rate has not improved for head neck squamous cell carcinoma (HNSCC). Traditionally, the general affected population of HNSCC was male over 50-60 years of age, whom have had history of alcohol and tobacco use. Conversely, in the recent decades, HNSCC has exhibited significant rise in younger patients, largely due to the increase in human papillomavirus (HPV) infection among young adults.

Generally, HPV as the most prevalent sexually transmitted disease, consisted of strains that do not cause harm to humans. Only handful of strains were found to be carcinogenic, potentially. Furthermore, the carcinogenic property of HPV has been increasing tremendously, and becoming a greater threat to human. For instance, HPV is the leading cause of cervical cancer currently. Recently, HPV related HNSCC has showed significant increase in the last 30 years as well, with oropharyngeal squamous cell carcinoma (OPSCC) as the most prevalent type, and the most increased kind in the HPV related HNSCC groups.

In this study, three methods of survival analysis were used which included non-parametric Kaplan-Meier method, parametric accelerated failure model and Cox proportional hazard method to achieve this data analysis.

First, two best fitted predictive survival models were developed for HNSCC (OPSCC) patients whom have been diagnosed and treated at Barnes Jewish Hospital in St. Louis. The models were initially determined by forward and backward selection of Cox proportional hazard method. The best predictive variables were further identified via forward selection in Kaplan Meier method. As a result, the final model estimates were obtained through accelerated failure time model.

Additionally, using Kaplan Meier method, HPV and HNSCC (OPSCC) relationships were investigated via P16 protein presence, which is an indicator of HPV related OPSCC. Survival rate of P16+ and P16- status were compared and contrasted. Interaction between the presence of P16 protein and other factors such as age groups, tobacco use, loco-regional fail, various stages of cancer defined by tumor differentiation, cancer recurrence, and lymph node found positive for cancer were explored.

Lastly, other factors of interest such as types of treatment, types of chemotherapy, race and anemia were investigated for overall survival rate as well as interactions with presence or absence of P16, also using Kaplan Meier method. Survival graphs were generated for the whole model as well as for the group comparisons.

Language

English (en)

Chair and Committee

Edward Spitznagel

Committee Members

Todd Kuffner, Guoyan Zhan

Comments

Permanent URL: https://doi.org/doi:10.7936/K7FQ9V08

Download

Share

COinS