Harnessing Metabolism with Radiation Therapy to Modulate Host Immunity in the Tumor Microenvironment

Abstract

Host immunity affects response to immunotherapy (IO) and dictates T cell presence in the tumor microenvironment (TME). Analyses of patient IO response have not accounted for host metabolism, which contributes to chronic inflammation, directly affecting host immunity. As the use of combined IO and radiation therapy (RT) increases, targeting host metabolism is key. We have previously shown that caloric restriction (CR) can augment apoptotic effects of RT and set out to determine if CR could better prime the host for IO+RT therapies. Forty 12-week old female Balb/c mice were injected with syngeneic triple negative murine breast cancer (4T1) cells (5 x 104) and upon formation of a palpable tumor, were assigned to one of four treatment groups: ad lib (AL) fed, 8 Gy RT to the primary tumor (RT), 30% reduction in caloric intake (CR) or CR+RT. Primary tumor tissue was isolated and processed for analysis of immune markers. In the clinic, an IRB-approved feasibility trial using diet (25% CR for 10 weeks) concurrently with RT (60 Gy) for early-stage breast cancer patients was implemented. Serum was collected from patients before RT and after completion of RT and diet. Multiplex ELISA quantified cytokines in sera from trial patients and historical controls receiving RT alone. MicroRNA expression was evaluated in sera. Gene ontology analysis was done on microarray data as well as human sera. T-tests were used to compare expression levels and p ≤ 0.05 was significant. Preclinical and human data both implicated inflammation as a top pathway affected by RT and CR. Cytokine and microRNA data from the preclinical model led to further analysis of STAT3/5. RT alone increased STAT3 and STAT5 (2.5-fold; p=0.04 and 2.1-fold; p=0.03) in primary tumors, while CR alone decreased STAT3 and STAT5 (1.5-fold and 2.4-fold). This decrease remained with CR+RT. Co-expression of pSTAT3 and pSTAT5, which suggests Treg phenotype, was decreased in both the CR and CR+RT arms. Given that pSTAT3 targets miR-21, it was measured as a potential biomarker. MiR-21 was increased in the RT arm compared to AL (3.5-fold; p=0.003) and was decreased in both CR (2-fold; p=0.03) and CR+RT arms (2.5-fold; p=0.01). With respect to T cell function, IFNg was decreased in RT alone, but increased in both the CR and CR+RT treatment arms. Conversely, IL-10 expression was increased after RT alone, but decreased with CR and CR+RT. In human sera, analysis revealed increased memory T cell cytokines (IL-7, IL-15) in CR+RT patients as compared to RT alone and decreased Treg cytokines (IL-10, IL-23). We show that RT creates an immunosuppressive TME via upregulation of pSTAT3, pSTAT5 and IL-10, with a decrease in cytotoxic IFN-g, while CR creates a favorable antitumor TME. A similar cytokine signature seen in CR clinical samples corresponded to decreased levels of miR-21 in serum, making it a potential biomarker for host immunity in this setting. Future trials should explore the use of CR to improve efficacy of IO and RT+IO combination.