Anna Lokshin

PhD

University of Pittsburgh

Professor

PLCO (Learn more about this study)

PLCO-992

Jun 21, 2022

Analysis of temporal dynamics of pro-inflammatory and anti-inflammatory cytokines during early evolution of ovarian endometrial and lung cancers

Chronic inflammation generated by the tumor microenvironment is known to drive cancer initiation and progression. The tumor microenvironment promotes the secretion of diverse cytokines, in different types and stages of cancers. These cytokines may inhibit tumor development but alternatively may contribute to chronic inflammation that supports tumor growth. Such distinct sets of cytokines from the tumor microenvironment can be detected in the circulation and are
thus potentially useful as biomarkers to detect cancers, predict disease outcomes and manage therapeutic choices. Indeed, analyses of circulating cytokines in combination with cancer-specific biomarkers have been proposed to simplify and improve cancer detection and prognosis. Additionally, the cytokine signaling signatures of the peripheral immune cells, even from patients with localized tumors, are recently found altered in cancer, and may also prove applicable as cancer biomarkers. In this study we propose to use results of analysis of multiple cytokines in PLCO participants who were later diagnosed with ovarian cancer to calculate approximated temporal dynamics in the PLCO study set using generalized additive models (GAM) approach. Such approach does not require serial samples, but provides a reliable estimate of temporal dynamics. However, it may identify cytokine profiles typical for early and late stages of ovarian and pancreatic cancer development. The main objective of the proposed study is to identify a phase in early cancer progression at which growing tumor starts activating the immune system. This approach has not been used by researchers who have previously performed analysis of inflammation related cytokines in pre-diagnostic PLCO samples.
Additionally, if the results for individual samples could be matched with samples that we have obtained from PLCO, we could include cytokines in building diagnostic panels. This approach has not been used by researchers who have previously performed analysis of inflammation related cytokines in pre-diagnostic PLCO samples.

Aim 1. Apply generalized additive models analysis to pre-diagnostic samples to look at approximated cytokine changes and changes in cytokine/cytokine receptor profiles over time. Molecular and mathematical modeling demonstrated that for ovarian cancer, it takes about 7 years from progression from premalignant lesion to advanced cancer. Therefore, we will utilize all data from samples collected 1 to 7 years before diagnosis. From this approach, we expect to obtain preliminary results on temporal changes in inflammatory status during ovarian and pancreatic cancer development and evaluate possible dysregulated cytokine signaling signature.

Aim 2. We will use cytokine quantification to generate diagnostic algorithms as we have done previously with many other proteins. If we obtained data matched to the individual PLCO samples that we have characterized already, we will test whether addition of cytokines is complementary and can improve diagnostic power of our existing multimarker panels.