Advances in computational algorithms and tools have made the prediction of cancer patient outcomes using computational pathology feasible. However, predicting clinical outcomes from pre-treatment histopathologic images remains a challenging task, limited by the poor understanding of tumor immune micro-environments. In this study, an automatic, accurate, comprehensive, interpretable, and reproducible whole slide image (WSI) feature extraction pipeline known as, IMage-based Pathological REgistration and Segmentation Statistics (IMPRESS), is described. We used both H&E and multiplex IHC (PD-L1, CD8+, and CD163+) images, investigated whether artificial intelligence (AI)-based algorithms using automatic feature extraction methods can predict neoadjuvant chemotherapy (NAC) outcomes in HER2-positive (HER2+) and triple-negative breast cancer (TNBC) patients. Features are derived from tumor immune micro-environment and clinical data and used to train machine learning models to accurately predict the response to NAC in breast cancer patients (HER2+ AUC = 0.8975; TNBC AUC = 0.7674). The results demonstrate that this method outperforms the results trained from features that were manually generated by pathologists. The developed image features and algorithms were further externally validated by independent cohorts, yielding encouraging results, especially for the HER2+ subtype.

Purpose: Hormone receptor–positive/HER2-negative (HR⁺/HER2⁻) breast cancer is associated with low levels of stromal tumor-infiltrating lymphocytes (sTIL) and PD-L1, and demonstrates poor responses to checkpoint inhibitor therapy. Evaluating the effect of standard chemotherapy on the immune microenvironment may suggest new opportunities for immunotherapy-based approaches to treating HR⁺/HER2⁻ breast tumors. Experimental Design: HR⁺/HER2⁻ breast tumors were analyzed before and after neoadjuvant chemotherapy. sTIL were assessed histologically; CD8⁺ cells, CD68⁺ cells, and PD-L1 staining were assessed immunohistochemically; whole transcriptome sequencing and panel RNA expression analysis (NanoString) were performed. Results: Ninety-six patients were analyzed from two cohorts (n = 55, Dana-Farber cohort; n = 41, MD Anderson cohort). sTIL, CD8, and PD-L1 on tumor cells were higher in tumors with basal PAM50 intrinsic subtype. Higher levels of tissue-based lymphocyte (sTIL, CD8, PD-L1) and macrophage (CD68) markers, as well as gene expression markers of lymphocyte or macrophage phenotypes (NanoString or CIBERSORT), correlated with favorable response to neoadjuvant chemotherapy, but not with improved distant metastasis-free survival in these cohorts or a large gene expression dataset (N = 302). In paired pre-/postchemotherapy samples, sTIL and CD8⁺ cells were significantly decreased after treatment, whereas expression analyses (NanoString) demonstrated significant increase of multiple myeloid signatures. Single gene expression implicated increased expression of immunosuppressive (M2-like) macrophage-specific genes after chemotherapy. Conclusions: The immune microenvironment of HR⁺/HER2⁻ tumors differs according to tumor biology. This cohort of paired pre-/postchemotherapy samples suggests a critical role for immunosuppressive macrophage expansion in residual disease. The role of macrophages in chemoresistance should be explored, and further evaluation of macrophage-targeting therapy is warranted.

Amplification and/or overexpression of human epidermal growth factor receptor 2 (HER2) in breast cancer is associated with an adverse prognosis. The introduction of anti-HER2 targeted therapy has dramatically improved the clinical outcomes of patients with HER2-positive breast cancer. Unfortunately, a significant number of patients eventually relapse and develop distant metastasis. HER2 intratumoral heterogeneity (ITH) has been reported to be associated with poor prognosis in patients with anti-HER2 targeted therapies and was proposed to be a potential mechanism for anti-HER2 resistance. In this review, we described the current definition, common types of HER2 ITH in breast cancer, the challenge in interpretation of HER2 status in cases showing ITH and the clinical applications of anti-HER2 agents in breast cancer showing heterogeneous HER2 expression. Digital image analysis has emerged as an objective and reproducible scoring method and its role in the assessment of HER2 status with ITH remains to be demonstrated.

Cancer stem cells (CSC) play a central role in cancer metastasis and development of drug resistance. miRNA are important in regulating CSC properties and are considered potential therapeutic targets. Here we report that miR-328–3p (miR-328) is significantly upregulated in ovarian CSC. High expression of miR-328 maintained CSC properties by directly targeting DNA damage binding protein 2, which has been shown previously to inhibit ovarian CSC. Reduced activity of ERK signaling in ovarian CSC, mainly due to a low level of reactive oxygen species, contributed to the enhanced expression of miR-328 and maintenance of CSC. Inhibition of miR-328 in mouse orthotopic ovarian xenografts impeded tumor growth and prevented tumor metastasis. In summary, our findings provide a novel mechanism underlying maintenance of the CSC population in ovarian cancer and suggest that targeted inhibition of miR-328 could be exploited for the eradication of CSC and aversion of tumor metastasis in ovarian cancer.Significance: These findings present inhibition of miR-328 as a novel strategy for efficient elimination of CSC to prevent tumor metastasis and recurrence in patients with epithelial ovarian cancer.