American Association for Cancer Research
epi-22-1285_supplemental_figure_1_suppsf1.docx (124.81 kB)

Supplemental Figure 1 from Measurement of Ovarian Tumor Immune Profiles by Multiplex Immunohistochemistry: Implications for Epidemiologic Studies

Download (124.81 kB)
journal contribution
posted on 2023-04-12, 16:20 authored by Cassandra A. Hathaway, Jose R. Conejo-Garcia, Brooke L. Fridley, Bernard Rosner, Daryoush Saeed-Vafa, Carlos Moran Segura, Jonathan V. Nguyen, Jonathan L. Hecht, Naoko Sasamoto, Kathryn L. Terry, Shelley S. Tworoger, Mary K. Townsend

Supplemental Figure 1 provides representative images from the T cell panel for high-grade serous, endometrioid, clear cell, and mucinous tumors.



Despite the immunogenic nature of many ovarian tumors, treatment with immune checkpoint therapies has not led to substantial improvements in ovarian cancer survival. To advance population-level research on the ovarian tumor immune microenvironment, it is critical to understand methodologic issues related to measurement of immune cells on tissue microarrays (TMA) using multiplex immunofluorescence (mIF) assays. In two prospective cohorts, we collected formalin-fixed, paraffin-embedded ovarian tumors from 486 cases and created seven TMAs. We measured T cells, including several sub-populations, and immune checkpoint markers on the TMAs using two mIF panels. We used Spearman correlations, Fisher exact tests, and multivariable-adjusted beta-binomial models to evaluate factors related to immune cell measurements in TMA tumor cores. Between-core correlations of intratumoral immune markers ranged from 0.52 to 0.72, with more common markers (e.g., CD3+, CD3+CD8+) having higher correlations. Correlations of immune cell markers between the whole core, tumor area, and stromal area were high (range 0.69–0.97). In multivariable-adjusted models, odds of T-cell positivity were lower in clear cell and mucinous versus type II tumors (ORs, 0.13–0.48) and, for several sub-populations, were lower in older tissue (sample age > 30 versus ≤ 10 years; OR, 0.11–0.32). Overall, high correlations between cores for immune markers measured via mIF support the use of TMAs in studying ovarian tumor immune infiltration, although very old samples may have reduced antigenicity. Future epidemiologic studies should evaluate differences in the tumor immune response by histotype and identify modifiable factors that may alter the tumor immune microenvironment.

Usage metrics

    Cancer Epidemiology, Biomarkers & Prevention