Multiplex Immunohistochemistry Provides Insight Into Cross‐Talk Between Myoepithelial And Immune Cells In Ductal Carcinoma In Situ (DCIS) Progression

Abstract

Every year in the U.S. ~50,000 women are diagnosed with Ductal Carcinoma in Situ (DCIS). Long term studies of untreated DCIS show ~30% of women progressed to invasive breast cancer (IBC), providing strong rational for treating DCIS similar to early stage IBC. Disease management entails breast‐conserving surgery (BCS) in ~70% of women with or without adjuvant radiation therapy (RT), or unilateral and/or bilateral mastectomy in about 30% of women. While RT and mastectomy reduce risk of IBC or DCIS ipsilateral recurrence, support of potential overtreatment is suggested by long term outcome studies demonstrating low mortality from DCIS (1–2% at 10 years) regardless of treatment received. Thus, prognostic markers that can distinguish DCIS at high versus low risk of progression are needed to personalize treatment decisions. The tumor microenvironment, including myoepithelial and immune cells, has been shown by us and others to play a pivotal role in tumor progression, and we have advanced murine models to study human DCIS progression. In an immune compromised model, the loss of myoepithelial cell calponin, a regulatory component of smooth muscle contraction, specifically corresponded with a ~130‐fold increased probability of adjacent human tumor cells expressing the poor prognostic tumor biomarker p63; implicating calponin as a tumor suppressor protein that acts on adjacent tumor cells. Using immune competent murine models, we find faster progression to invasive tumors, suggesting a supportive role for the immune system in DCIS to IBC progression. We hypothesized that pro‐tumorigenic immune cells will be found near DCIS lesions that have lost myoepithelial differentiation markers and/or tumor cells that have a poor‐prognostic basal phenotype. To test this hypothesis, we have developed novel computer‐assisted computational methods to characterize the integrity of the myoepithelial cell layer in DCIS lesions and to provide near‐neighbor analysis to identify proximity to immune cells. This method involves image segmentation to identify the myoepithelial cell layer using the combined IHC images of SMA, calponin, and CK18, followed by geometric quantification to determine relative expression as well as physical proximity between distinct biomarker (Figure 1). We validated the ability of this computational method to determine loss of myoepithelial cell‐differentiation markers by comparing results to a pathologist's visual annotation, and found a Pearson's‐correlation coefficient of R = 0.959. Utilizing FFPE breast tissues from a cohort of women diagnosed with DCIS (OHSU IRB #15514), we performed multiplex sequential immunohistochemical (IHC) staining for markers of myoepithelial differentiation state (p63, SMA, calponin), immune cells (lymphoid: CD45, CD8, CD3, Foxp3, Granzyme B, PD‐1) and prognostic biomarkers (cyclooxygenase‐2, p16, ki67, ER, PR, Her‐2, tenascin‐c, cytokeratin 5). Our results show that loss of the myoepithelial differentiation marker calponin is associated with a relative decrease in CD4+Foxp3+ T regulatory cells and a concurrent increase in CD8+PD‐1+ T cells. These data suggest that 1) during DCIS progression there may be activation of cytotoxic effectors and concomitant decrease in Tregs to support tumor suppression, and 2) myoepithelial cell integrity may facilitate the activation state of local immune cells. Future studies include assessment for the role of myeloid lineages in myoepithelial cell integrity and DCIS progression.Support or Funding InformationThis research work is supported by OHSU Knight Cancer Institute – Borchard Early Investigator Award to SJ, OHSU OCTRI ‐ NW Kaiser specimen retrieval opportunity to PS and SJ, Susan G. Komen Post‐Doctoral fellowship to EM (PDF17480342).This abstract is from the Experimental Biology 2018 Meeting. There is no full text article associated with this abstract published in The FASEB Journal.