Beatrice Wiafe-Addai

Supplementary Table S3: Beta coefficients (β) and 95% confidence intervals (CI) for the associations of breast cancer risk factors and clinicopathological characteristics with tumor grade among breast cancer patients participating in the Ghana Breast Health Study

Supplementary Table S1: Optimized machine learning parameters for cell detection on digitized hematoxylin and eosin-stained breast cancer sections from participants in the Ghana Breast Health Study

Supplementary Figure S1: Correlation between standard Ta-SCD and percent Ta-SCD. Standard Ta-SCD was calculated by dividing the total number of cells in the stroma by the total stromal area (mm2). Standard Ta-SCD was converted to a percentage by multiplying the total number of nucleated cells by the average area (mm2) of a single nucleus to generate total nuclei area (mm2), dividing this by the total stromal (cellular + non-cellular) area (mm2), and multiplying by 100.

Supplementary Table S2: Beta coefficients (β) and 95% confidence intervals (CI) for the associations of breast cancer risk factors and clinicopathological characteristics with tumor-associated stromal cellular density among breast cancer patients participating in the Ghana Breast Health Study

Background:The stromal microenvironment (SME) is integral to breast cancer biology, impacting metastatic proclivity and treatment response. Emerging data indicate that host factors may impact the SME, but the relationship between prediagnostic host factors and SME phenotype remains poorly characterized, particularly among women of African ancestry.Methods:We conducted a case-only analysis involving 792 patients with breast cancer (17–84 years) from the Ghana Breast Health Study. High-accuracy machine-learning algorithms were applied to standard H&E-stained images to characterize SME phenotypes [including percent tumor-associated connective tissue stroma, Ta-CTS (%); tumor-associated stromal cellular density, Ta-SCD (%)]. Associations between prediagnostic host factors and SME phenotypes were assessed in multivariable linear regression models.Results:Decreasing Ta-CTS and increasing Ta-SCD were associated with aggressive, mostly high-grade tumors (P-value < 0.001). Several prediagnostic host factors were associated with Ta-SCD independently of tumor characteristics. Compared with nulliparous women, parous women had higher levels of Ta-SCD [mean (standard deviation, SD) = 31.3% (7.6%) vs. 28.9% (7.1%); P-value = 0.01]. Similarly, women with a positive family history of breast cancer had higher levels of Ta-SCD than those without family history [mean (SD) = 33.0% (7.5%)] vs. 30.9% (7.6%); P-value = 0.03]. Conversely, increasing body size was associated with decreasing Ta-SCD [mean (SD) = 31.6% (7.4%), 31.4% (7.3%), and 30.1% (8.0%) for slight, average, and large body sizes, respectively; P-value = 0.005].Conclusions:Epidemiological risk factors were associated with varying degrees of stromal cellularity in tumors, independently of clinicopathological characteristics.Impact:The findings raise the possibility that epidemiological risk factors may partly influence tumor biology via the stromal microenvironment.See related In the Spotlight, p. 459

Supplementary Table S4: Coefficients and 95% confidence intervals (CI) for the associations of breast cancer risk factors and clinicopathological characteristics with parity (vs nulliparous) among breast cancer patients participating in the Ghana Breast Health Study

Supplementary Table S5: Beta coefficients (β) and 95% confidence intervals (CI) for the associations between breast cancer risk factors and Tumor-associated Stromal Cellular Density (Ta-SCD) among patients <50 and ≥50 years of age participating in the Ghana Breast Health Study

Supplementary Table S6: Beta coefficients (β) and 95% confidence intervals (CI) for the associations between breast cancer risk factors and Tumor-associated Stromal Cellular Density among patients with ER+ and ER- breast cancer participating in the Ghana Breast Health Study

Background: The stromal microenvironment (SME) is integral to breast cancer biology, impacting metastatic proclivity and treatment response. Emerging data indicate that host factors may impact the SME, but the relationship between prediagnostic host factors and SME phenotype remains poorly characterized, particularly among women of African ancestry. Methods: We conducted a case-only analysis involving 792 patients with breast cancer (17–84 years) from the Ghana Breast Health Study. High-accuracy machine-learning algorithms were applied to standard H&E-stained images to characterize SME phenotypes [including percent tumor-associated connective tissue stroma, Ta-CTS (%); tumor-associated stromal cellular density, Ta-SCD (%)]. Associations between prediagnostic host factors and SME phenotypes were assessed in multivariable linear regression models. Results: Decreasing Ta-CTS and increasing Ta-SCD were associated with aggressive, mostly high-grade tumors (P-value < 0.001). Several prediagnostic host factors were associated with Ta-SCD independently of tumor characteristics. Compared with nulliparous women, parous women had higher levels of Ta-SCD [mean (standard deviation, SD) = 31.3% (7.6%) vs. 28.9% (7.1%); P-value = 0.01]. Similarly, women with a positive family history of breast cancer had higher levels of Ta-SCD than those without family history [mean (SD) = 33.0% (7.5%)] vs. 30.9% (7.6%); P-value = 0.03]. Conversely, increasing body size was associated with decreasing Ta-SCD [mean (SD) = 31.6% (7.4%), 31.4% (7.3%), and 30.1% (8.0%) for slight, average, and large body sizes, respectively; P-value = 0.005]. Conclusions: Epidemiological risk factors were associated with varying degrees of stromal cellularity in tumors, independently of clinicopathological characteristics. Impact: The findings raise the possibility that epidemiological risk factors may partly influence tumor biology via the stromal microenvironment. See related In the Spotlight, p. 459