Biostatistics in Clinical Cancer Studies and AIDS Research
Our research involves two main areas: collaboration on cancer and HIV studies and the statistical problems arising from them. Our work is focused primarily on studies of breast cancer (both laboratory and clinical), radiation oncology, immunology, and evaluation of new immunology assays, particularly those involving flow cytometry. Recent research activities in cancer include (1) modeling the effects of screening and therapy on breast cancer mortality in the United States (with the NIH CISNET group), (2) designing xenograft growth experiments to assess the cancer-promotion abilities of various stromal and fibroblast cells, (3) assessing longitudinal circulating tumor markers in breast cancer, (4) designing breast cancer prevention pilot studies, and (5) analyzing long-term follow-up on several randomized breast cancer therapy trials. Recent research activities in HIV disease include (1) modeling changes in the distributions of percents and counts of various lymphocyte subsets in healthy children from birth to age 18 (for use as controls in pediatric HIV and cancer studies), (2) quantitating the co-localization of CD8 cells and JC virus (JCV)-infected glial cells in brain biopsies from patients with progressive multifocal leukoencephalopathy (PML) disease, (3) analyzing the effect of various immune maneuvers on the effectiveness of vaccines against simian-human immunodeficiency viruses (SHIV), and (4) designing and analyzing studies that compare the within-laboratory and between-laboratory variability in CD4 counts using either current methods or several cheaper and simpler methods being developed for use in the third world. Our statistical interests include methods (1) to analyze nonindependent competing risks, (2) to assess the surrogate value of longitudinal markers in cancer and HIV disease (particularly methods that incorporate external information on within-laboratory and between-laboratory variation in assessing these markers), and (3) to assess the heterogeneity of distributions of various types of cells in pathology specimens or high-definition imaging techniques.