Brook, Mark N.

Abstract Background: Prospective validation of breast cancer risk models integrating classical risk factors and genetic variants is required for risk-stratified prevention and screening strategies. The objective of this study was to validate a breast cancer risk model integrating classical risk factors and a 313-variant polygenic risk score (PRS) in multiple prospective cohort studies, and to project five-year risk of breast cancer in six different countries. Methods: The study population included 7,529 cases and 230,103 controls from 14 prospective cohort studies in Australia, Germany, the Netherlands, Sweden, UK, and USA. We used the Individualized Coherent Absolute Risk Estimator (iCARE) tool for risk model building, validation, and risk projection. Expected five-year risk of invasive or in situ breast cancer was compared to observed risk, overall and within deciles of expected risk using goodness of fit statistics. We evaluated calibration of the relative risk through meta-analysis across cohorts, and of the absolute risk within each cohort. Model discrimination was evaluated using the area under the curve (AUC), and percentages of women crossing risk thresholds. Projections of five-year risk distributions were estimated for women of European ancestry aged 50-70 years in the general populations of these six countries. Results: Analysis showed overall good calibration of the integrated iCARE-based model relative risk for both women younger than 50 years (χ2=14.9, P=0.09) and aged 50 years or older (χ2=14.6, P=0.10), with a small overestimation of risk for women in the highest decile of expected risk (RR = 3.5 expected vs 2.3 (95% CI 1.6 to 3.2) observed for women <50 years; and 2.8 expected vs 2.3 (95% CI 2.0 to 2.7) observed for women 50+ years). The age-adjusted AUCs for the integrated model were 63.1 (95% CI 60.9 to 65.3) and 62.9 (95% CI 61.8 to 64.0), for the two age groups respectively. The calibration of absolute risk showed substantial variation across cohorts, particularly for the older group, but had no systematic bias. Model based projections in the general populations showed that compared to the population average, women in the 1st and 99th percentiles of the integrated risk score had relative risks 0.19 and 3.56 respectively. The proportion of women of European ancestry aged 50-70 years with a five-year risk greater than 3% (threshold for consideration of risk-lowering drugs by U.S. Preventive Services Task Force) ranged from 7.1% in Germany to 18.2% in the US, which corresponds to ~5.5 million women in the US. Conclusions: Five-year risk predictions from a model with classical risk factors and PRS are well calibrated and provide substantial risk stratification across multiple cohorts in six different countries. Further studies are needed to evaluate the clinical utility of the validated model for risk stratified screening and prevention of breast cancer. Citation Format: Parichoy Pal Choudhury, Amber Wilcox, Chi Gao, Brian Carter, Anika Husing, Mark Brook, Mikael Eriksson, Kara Martin, Chris Scott, Min Shi, Thomas Ahearn, Michael Jones, Nick Orr, Minouk Schoemaker, Kamila Czene, Jenny Chang-Claude, Jacques Simard, Doug Easton, Marjanka K. Schmidt, Dale Sandler, Clarice R. Weinberg, Celine Vachon, Roger Milne, Per Hall, Anthony Swerdlow, Rudolph Kaaks, Myrto Barrdahl, Mia Gaudet, Antonis Antoniou, Peter Kraft, Montserrat Garcia-Closas, Nilanjan Chatterjee. Validation of breast cancer risk model incorporating classical risk factors and polygenic risk scores in 14 prospective cohort studies in 6 countries [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 962.

AbstractBreast cancer risk factors and clinical outcomes vary by tumour marker expression. However, individual studies often lack the power required to assess these relationships, and large‐scale analyses are limited by the need for high throughput, standardized scoring methods. To address these limitations, we assessed whether automated image analysis of immunohistochemically stained tissue microarrays can permit rapid, standardized scoring of tumour markers from multiple studies. Tissue microarray sections prepared in nine studies containing 20 263 cores from 8267 breast cancers stained for two nuclear (oestrogen receptor, progesterone receptor), two membranous (human epidermal growth factor receptor 2 and epidermal growth factor receptor) and one cytoplasmic (cytokeratin 5/6) marker were scanned as digital images. Automated algorithms were used to score markers in tumour cells using the Ariol system. We compared automated scores against visual reads, and their associations with breast cancer survival. Approximately 65–70% of tissue microarray cores were satisfactory for scoring. Among satisfactory cores, agreement between dichotomous automated and visual scores was highest for oestrogen receptor (Kappa = 0.76), followed by human epidermal growth factor receptor 2 (Kappa = 0.69) and progesterone receptor (Kappa = 0.67). Automated quantitative scores for these markers were associated with hazard ratios for breast cancer mortality in a dose‐response manner. Considering visual scores of epidermal growth factor receptor or cytokeratin 5/6 as the reference, automated scoring achieved excellent negative predictive value (96–98%), but yielded many false positives (positive predictive value = 30–32%). For all markers, we observed substantial heterogeneity in automated scoring performance across tissue microarrays. Automated analysis is a potentially useful tool for large‐scale, quantitative scoring of immunohistochemically stained tissue microarrays available in consortia. However, continued optimization, rigorous marker‐specific quality control measures and standardization of tissue microarray designs, staining and scoring protocols is needed to enhance results.