PROJECT SUMMARY/ABSTRACT
Multiplex immunofluorescence (mIF) microscopy and accompanying automated image analysis is a widely
used technique which allows for the assessment and visualization of the tumor immune microenvironment
(TIME). Generally, mIF data has been used to simply examine the presence and abundance of immune cells in
cancer patients; however, this aggregate measure assumes uniform patterns of immune cells and overlooks
spatial heterogeneity. The spatial contexture of the TIME has not been adequately explored, in part due to the
lack of available analytical approaches and tools. Therefore, the goal of this research is to develop novel
statistical methods and software for the downstream analysis of mIF data. We will apply these methods to the
study of epithelial ovarian cancer (EOC), the deadliest gynecologic malignancy in the US, to develop an
immunoscore predictive of survival among EOC patients. Aim 1 is focused on developing spatial statistical
approaches for the analysis of mIF data which leverages the spatial architecture of the TIME. Aim 2 will
develop Bayesian models for the analysis of mIF data which accounts for the zero-inflated and over-dispersed
nature of the immune count data for determination of immune subtypes. Lastly, Aim 3 will characterize the
immune landscape of ovarian tumors and develop an immunoscore (Oimmuno) predictive of survival using
existing mIF data from ~2,500 diverse EOC patients enrolled in established epidemiological studies. We will
validate the Oimmuno by generating targeted mIF data in two independent cohorts, each with more than 1,200
EOC patients. In summary, this proposal will develop and test novel statistical methods for the analysis of mIF
data that incorporates the spatial heterogeneity of the TIME, in addition to abundance measures of immune
cells, producing freely available software that can be used in the study of EOC or adapted for use in other
cancer types. The derivation of methods to quantify the spatial contexture of immune cells has important
applications as such biomarker discovery for predictors of outcomes and therapeutic efficacy in cancer
patients, ultimately reducing cancer mortality.
