Abstract 2027: Developing an image analysis algorithm to evaluate immunofluorescence intensity for Rb and phospho-Rb expression in circulating tumor cells in HR+/HER2- breast cancer

Abstract Introduction: Our lab is developing a multiplex immunofluorescence (IF) assay that can be used on circulating tumor cells (CTCs) to assess Rb and pRb expression to predict efficacy of CDK4/6 inhibitors. In this study, we aimed to develop an image processing algorithm for analyzing Rb and pRb expression. Methods: Positive control (MCF7 and T47D) and negative control (MDA-MB-468 and T47D palbociclib-resistant) cell lines were stained with anti-Rb-AF594, anti-pRb-AF488, and DAPI. The IF-stained cells were imaged using a Leica SP8 STED confocal microscope and processed with the ImageJ Fiji software. An image analysis pipeline was developed by comparing various parameters for each condition for optimized z project (Max intensity vs. Sum slices), background subtraction (Rolling ball radius vs. Mean), and threshold settings (Default vs. Otsu). Mean fluorescence intensities (MFI) between various parameters were compared using the Mann-Whitney U test. A p-value of <0.05 was considered statistically significant. Signal-to-noise ratio (SNR) was calculated as (mean signal from positive control cells - mean background from negative control cells) / background standard deviation. This pipeline was further validated on mixed cell populations of MCF7 P + leukocytes, which were stained with anti-CK-AF647, anti-CD45-AF555, anti-Rb-AF594, anti-pRb-AF488, and DAPI. At least 3 independent experiments were included. MFI was determined for DAPI+/CK+/CD45- cells. Results: A systematic approach was employed, beginning with standardizing image acquisition parameters, including laser intensity and gain settings, to ensure consistency between samples within an experiment using 2 positive and 2 negative control cell lines. For initial optimization with z-projection settings, significantly higher MFI was obtained with Sum slices compared to Max intensity for both positive and negative control cells. For background subtraction, Rolling ball radius versus Mean background subtraction showed no significant differences between the two settings. For threshold comparisons Otsu method introduced significant background noise in negative controls, compromising intensity accuracy, leading to the selection of Default method. Therefore, proposed pipeline included Sum slices, Rolling ball radius, and Default method optimized using internal controls with an SNR of 16 for Rb and 170 for pRb. Validation with MCF7 P + leukocyte samples produced MFIs (median (range)) of 103, 802 (5, 484 - 627, 148) for Rb and 149, 274 (8, 379 - 791, 202) for pRb for the DAPI+/CK+/CD45- cells. Conclusion: We have established the IF image processing pipeline for Rb and pRb expression assessment, which will be further refined to assess the sensitivity and specificity of the IF assay, laying a foundation for its future application to patients’ isolated CTC samples. Citation Format: Shivaani Suresh Kanna, Mantasha Tabassum, Meghana V. Trivedi. Developing an image analysis algorithm to evaluate immunofluorescence intensity for Rb and phospho-Rb expression in circulating tumor cells in HR+/HER2- breast cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 2027.