Advancements in monitoring and predicting of patient-specific response of triple negative breast cancer (TNBC) to immunotherapy (IMT) with and without chemotherapy are needed. Using granzyme B-specific positron emission tomography (GZP-PET) imaging, we aimed to monitor changes in effector cell activation in response to IMT with chemotherapy in TNBC. TNBC mouse models received the paclitaxel (PTX) ± immune checkpoint inhibitors anti-programmed death 1 (anti-PD1) and anti-cytotoxic T-lymphocyte 4 (anti-CTLA4). GZP-PET imaging was performed on treatment days 0, 3, and 6. Mean standard uptake value (SUVmean), effector cell fractions, and SUV histograms were compared. Mice were sacrificed at early imaging timepoints for cytokine and histological analyses. GZP-PET imaging data revealed differences prior to tumor volume changes. By day six, responders had SUVmean ≥ 2.2-fold higher (p < 0.0037) and effector cell fractions ≥ 1.9-fold higher (p = 0.03) compared to non-responders. IMT/PTX resulted in a significantly different SUV distribution compared to control, indicating broader distribution of activated intratumoral T-cells. IMT/PTX resulted in significantly more necrotic tumor tissue and increased levels of IL-2, 4, and 12 compared to control. Results implicate immunogenic cell death through upregulation of key Th1/Th2 cytokines by IMT/PTX. Noninvasive PET imaging can provide data on the TNBC tumor microenvironment, specifically intratumoral effector cell activation, predicting response to IMT plus chemotherapy.

Background: Although a portion of triple negative breast cancer (TNBC) is sensitive to chemotherapeutic treatment with agents such as paclitaxel (PTX), patients have a high risk of recurrence and short overall and progression-free survival. Clinical trials for TNBC using immune checkpoint inhibitors such as anti-programmed cell death 1 (PD1) have met some success; however, only a small percentage of patients have a positive response to immunotherapy and there is risk for severe side effects. Therefore, there is a need for improvements in monitoring and predicting patient-specific response. The goal of this study is to use non-invasive visualization of the tumor microenvironment to collect longitudinal information that can be used to study tumor response to immunotherapy. By applying positron emission tomography (PET) imaging techniques to investigate changes in the tumor microenvironment (hypoxia and immune cell activation), we seek to predict early response to immunotherapy and better identify which treatments will be effective for individual tumors in TNBC. Methods: A TNBC mouse mammary carcinoma cell line, 4T1, was transduced with CMV-luciferase and 2x105 cells were injected into the third mammary fat pad of 5-6 week old female Balb/c mice (N=28). Mice in cohort 1 received: PTX (10 mg/kg), anti-PD1 (200 µg), both, or vehicle control (saline) (n=4/group). Mice in cohort 2 received: anti-PD1 (n=5), combination PTX/anti-PD1 (n=4), or vehicle control (saline; n=3). Treatments were administered intraperitoneally on days 0, 2, and 5 for cohort 1 (n=16) who underwent granzyme B-PET imaging ([68Ga]-NOTA-GZP-PET) and on days 0, 2, 5, and 8 for cohort 2 (n=12) who underwent hypoxia imaging with [18F]-fluoromisonidazole (FMISO)-PET imaging. Bioluminescence (BLI) imaging and caliper measurements were performed to track tumor size changes at multiple timepoints and tumors were collected for histological validation on day 20. Mean standard uptake value (SUVmean) was calculated as percent of day 0, and statistical analyses were performed with unpaired t-tests and Wilcoxon-rank sum tests. Results: Voxel analysis of GZP-PET images revealed an 42.9% increase in T cell activation of TNBC tumors treated with single-agent PTX compared to anti-PD1 alone on day 3 (p=0.08). FMISO-PET revealed that tumors treated with anti-PD1 alone had lower hypoxic fraction compared to control group tumors (p=0.10) and tumors treated with combination PTX/anti-PD1 (p=0.17) on day 3. BLI data showed that treatment with PTX and anti-PD1 significantly decreased viability signal between days 3 and 6 for cohort 1 (p=0.04). Non-responders to treatment had a significantly higher tumor volume compared to responders starting on day 6 (p&lt;0.05). SUVmean, indicating T cell activation, was significantly higher for responders compared to non-responders on days 3 and 6 (p&lt;0.05). There was a signifciant decrease in hypoxia, as measured thorugh FMISO-PET SUVmean on day 6 for responders compared to non-responders (p=0.04). Conclusion: It is known that the tumor microenvironment can impact neovascularization, tumor cell invasion, tumor oxygenation and growth. Results from this study show that noninvasive PET imaging can provide data on changes in T cell activation and hypoxia in response to chemotherapeutic treatment and immunotherapy in TNBC. Utilizing clinically-translatable advanced imaging strategies to understand the biologically distinct features of the TNBC tumor microenvironment can aid in personalizing anti-cancer therapies. Acknowledgements: We thank the American Cancer Society for RSG-18-006-01-CCE, NIH NCI R01 CA240589, and NIH P30CA013148. Citation Format: Tiara S. Napier, Chanelle L. Hunter, Patrick N. Song, Ben M. Larimer, Anna G. Sorace. Molecular imaging of hypoxia and granzyme B alterations during combination treatment with immunotherapy in triple negative breast cancer [abstract]. In: Proceedings of the 2020 San Antonio Breast Cancer Virtual Symposium; 2020 Dec 8-11; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2021;81(4 Suppl):Abstract nr PS3-07.

IMPACT: Insights from this project will provide clinical guidance in treatment of immunotherapy in triple negative breast cancer and identify early imaging biomarkers of treatment response. OBJECTIVES/GOALS: Significant research that addresses monitoring and predicting patient response of triple negative breast cancer (TNBC) to immunotherapy is needed. Using positron emission tomography (PET) imaging to probe the tumor microenvironment (hypoxia, T-cell activation), we aim to predict early response to immunotherapy for in mouse models of TNBC tumors. METHODS/STUDY POPULATION: Female Balb/c mice with 4T1-luciferase mammary carcinoma cell tumors were administered paclitaxel (PTX; 10 mg/kg), anti-PD1 (200 µg), both, or vehicle (saline) intraperitoneally. Treatment was given on days 0, 2, and 5 for cohort 1 (n=16) who underwent granzyme B specific (GZP) PET imaging (T-cell activation) and days 0, 2, 5, and 8 for cohort 2 (n=12) who underwent [18F]-fluoromisonidazole (FMISO)-PET imaging (hypoxia). Bioluminescence (BLI) imaging and caliper measurements were performed to track tumor size changes at multiple timepoints and tumors were collected for histological validation on day 20. Mean standard uptake value (SUVmean) was calculated as percent of day 0, and statistical analyses were performed with unpaired t-tests and Wilcoxon-rank sum tests. RESULTS/ANTICIPATED RESULTS: Non-responders to treatment had a significantly higher tumor volume compared to responders starting on day 6 (p<0.05). Although no significant differences in BLI between control and single-agent therapies were found, BLI data revealed that treatment with combination PTX and anti-PD1 significantly decreased viability signal between days 3 and 6 (p=0.04). SUVmean from GZP-PET was over 250% higher in responders compared to non-responders by day 6 (p=0.03). SUVmean from FMISO-PET was 80% less in responders compared to nonresponders, indicating less tumor hypoxia (p=0.04). DISCUSSION/SIGNIFICANCE OF FINDINGS: Non-invasive PET imaging of the tumor microenvironment can provide data on T cell activation and hypoxic response predicting response to combination immunotherapy and chemotherapy. Utilizing advanced imaging to understand biologically distinct features of the TNBC tumor microenvironment can aid in personalizing anti-cancer therapies.

No abstract