Reassessing Breast Cancer-Associated Fibroblasts (CAFs) Interactions with Other Stromal Components and Clinico-Pathologic Parameters by Using Immunohistochemistry and Digital Image Analysis (DIA)

SIMPLE SUMMARY: Breast cancer (BC) is primarily classified by immunohistochemistry (IHC) and digital image analysis (DIA) based on distinct immunophenotypes of malignant epithelial cells. When antiangiogenic and immunomodulatory therapies are used for BC, the stromal components such as tumor vessels and immune cells have received more attention. Cancer-associated fibroblasts (CAFs) are extensively investigated, but measuring them in immunostained tissues and determining how they relate to clinical and pathological criteria is still difficult in BC. Our earlier research using CD34 and α-smooth muscle actin (αSMA) immunohistochemistry showed a decrease of CD34-positive CAFs and an increase of αSMA-positive CAFs inside the tumor stroma. This early basic microscopic finding gives us the intention to employ DIA to improve SMA-positive CAF quantification accuracy. Additional DIA processed data correlations to pathologic parameters, survival, invasion, and recurrence indicated substantial differences between BC molecular subtypes, suggesting a unique CD34- and αSMA-positive CAF impact for each subtype. ABSTRACT: Background: Breast cancer (BC) stroma has CD34- and αSMA-positive cancer-associated fibroblasts (CAFs) differently distributed. During malignant transformation, CD34-positive fibroblasts decrease while αSMA-positive CAFs increase. The prevalence of αSMA-positive CAFs in BC stroma makes microscopic examination difficult without digital image analysis processing (DIA). DIA was used to compare CD34- and αSMA-positive CAFs among breast cancer molecular subgroups. DIA-derived data were linked to age, survival, tumor stroma vessels, tertiary lymphoid structures (TLS), invasion, and recurrence. Methods: Double immunostaining for CD34 and αSMA showed different CAF distribution patterns in normal and BC tissues. Single CD34 immunohistochemistry on supplemental slides quantified tumor stroma CD34_CAFs. Digital image analysis (DIA) data on CAF density, intensity, stromal score, and H-score were correlated with clinico-pathologic factors. Results: CD34/αSMA CAF proportion was significantly related to age in Luminal A (LA), Luminal B (LB), and HER2 subtypes. CD34_CAF influence on survival, invasion, and recurrence of LA, LB-HER2, and TNBC subtypes was found to be significant. The CD34/αSMA-expressing CAFs exhibited a heterogeneous impact on stromal vasculature and TLS. Conclusion: BC stromal CD34_CAFs/αSMA_CAFs have an impact on survival, invasion, and recurrence differently between BC molecular subtypes. The tumor stroma DIA assessment may have predictive potential to prognosis and long-term follow-up of patients with breast cancer.