Date of Graduation


Document Type

Thesis (MS)

Program Affiliation

Biochemistry and Molecular Biology

Degree Name

Masters of Science (MS)

Advisor/Committee Chair

Richard Eric Davis, M.D.

Committee Member

Sattva Neelapu, M.D.

Committee Member

Scott Kopetz, M.D., Ph.D.

Committee Member

Jorge Romaguera, M.D.

Committee Member

Rebert Bast, Jr., M.D.


There are important but ill-defined interactions between benign immune cell subsets and neoplastic B cells within follicular lymphoma (FL). Using the novel technique of correlation matrix analysis (CMA) of publicly available FL whole-tumor gene expression profiling (GEP) data, we have identified signatures of immune cell subsets. Overall survival correlated most highly with a model using signatures of macrophages, T cells, and stroma, which was able to add significantly to existing clinical prognostic tools. From our own data of a cohort of 43 FL tumors sorted into B-cell and non-B cell (NB) fractions for GEP, CMA of the tumor infiltrating NB fraction revealed additional immune cell subset signatures, including T follicular helper (TFH) cells. Comparison of gene signatures between FL and tonsils (n=24) suggested that TFH cells and macrophages are qualitatively distinct in FL from normal tissue. “Cross-correlation”, between FL NB fraction signatures and individual B fraction genes, suggests that TFH cells promote proliferation, germinal center stage differentiation, B-cell receptor signaling, and induction of CCL17 and CCL22 by tumor B cells. This novel analytical approach may be broadly applicable to define gene signatures of rare immune cell subsets in the tumor microenvironment, determine their prognostic impact, discover novel therapeutic targets, and identify patients likely to benefit from therapies targeting tumor-stroma interactions.


Follicular lymphoma, lymphoma, correlation matrix, gene signatures, gene expression profiling, immune cells, gene signatures, B-cell lymphoma