Aromatase Inhibitor Induced Musculoskeletal Inflammation is Observed Independent of Oophorectomy in a Novel Mouse Model

Abstract

Background: Aromatase Inhibitors (AIs) block physiological estrogen production in peripheral tissues and are used clinically to reduce disease recurrences and improve overall survival rates in hormone receptor-positive breast cancer patients. However, half of patients taking these drugs develop aromatase inhibitor induced arthralgia (AIIA), which is characterized by severe pain and inflammation in various joints and the surrounding musculoskeletal tissue. While the pathophysiology is not currently understood, it has been proposed to be associated with systemic estrogen deficiency resulting from AI treatment. Since AIIA leads to suspension of therapy in 20-30% of patients, reducing AIIA incidence may provide sustained AI treatment and enhance long-term survival. Objective: In order to establish a better understanding of disease pathology and to create a platform that can be used to explore future interventional strategies, our objective in this study was to design a novel animal model of AIIA. Methods: Female BALB/C-Tg(NFkB-RE-luc)-Xen mice, which have a firefly luciferase cDNA reporter transgene under the regulation of NFkB binding sites, were oophorectomized and treated with AI (letrozole) by daily subcutaneous injections for 5 weeks. Control groups included oophorectomized mice receiving vehicle injections and non-oophorectomized mice treated with AI. Knee joints and surrounding muscle tissue were imaged on the BioSpec 94/30 micro-MRI. The primary weight-bearing joint (hind limb) was examined histopathologically and NFkB activity was measured by bioluminescent imaging. Serum was collected for cytokine analysis. Additionally, healthy human PBMCs were treated with letrozole, estrogen, or both, and RNA sequencing was performed at 36 hrs. Results: Bioluminescent imaging showed significantly enhanced NFkB activation with AI treatment in the hind limbs compared to controls receiving vehicle treatment. Moreover, analysis of knee joints and legs by MRI showed enhanced signal detection in the joint space and surrounding tissue following daily AI injections. Surprisingly, the enhanced MRI detection and NFkB activation was observed with AI treatment independent of the oophorectomy procedure. This indicates that the induction of musculoskeletal-directed inflammation by AI is not mediated by changes in physiological estrogen levels, which is contrary to proposed mechanisms of disease pathogenesis. Similarly, histopathological analysis showed tenosynovitis and musculoskeletal infiltrates in all mice receiving AI with or without oophorectomy. IHC analysis of the infiltrates demonstrated a predominantly macrophage-mediated inflammatory response with scattered CD4+ T cells. Additionally, serum cytokine levels of IL-2, IL-4, IL-6, and CXCL1 were significantly elevated in mice with AI treatment. RNA sequencing of human PBMCs after in vitro AI stimulation did not demonstrate an AI-specific gene expression pattern associated with immune system activation directly, suggesting that the pathogenesis of AIIA may be mediated through cells in other tissues in vivo. Conclusions: Collectively, these data establish a novel mouse model of AIIA and identify an estrogen-independent stimulation of disease pathology via AI-mediated induction. This suggests that the pathogenesis of AIIA may not be mediated by estrogen deficiency, as previously hypothesized, and indicates that AI-induced inflammation may not be regulated directly through a pathogenic mechanism initially derived from circulating mononuclear cells. Future studies aim to characterize this inflammatory mechanism in vivo with a focus on other cells, including macrophages, synovial cells and chondrocytes, to provide insight into putative therapeutic strategies directed at mitigating disease pathology.