Inflammation, and the organization of collagen in the breast tumor microenvironment, is an important mediator of breast tumor progression. However, a direct link between markers of inflammation, collagen organization, and patient outcome has yet to be established. A tumor microarray of 371 invasive breast carcinoma biopsy specimens was analyzed for expression of inflammatory markers, including cyclooxygenase 2 (COX-2), macrophages, and several collagen features in the tumor nest (TN) or the tumor-associated stroma (TS). The tumor microarray cohort included females, aged 18 to 80 years, with a median follow-up of 8.4 years. High expression of COX-2 (TN), CD68 (TS), and CD163 (TN and TS) predicted worse patient overall survival (OS). This notion was strengthened by the finding from the multivariate analysis that high numbers of CD163 macrophages in the TS is an independent prognostic factor. Overall collagen deposition was associated with high stromal expression of COX-2 and CD163; however, total collagen deposition was not a predictor for OS. Conversely, local collagen density, alignment and perpendicular alignment to the tumor boundary (tumor-associated collagen signature-3) were predictors of OS. These results suggest that in invasive carcinoma, the localization of inflammatory cells and aligned collagen orientation predict poor patient survival. Additional clinical studies may help validate whether therapy with selective COX-2 inhibitors alters expression of CD68 and CD163 inflammatory markers.

Polygenic risk scores (PRSs) have wide applications in human genetics research, but often include tuning parameters which are difficult to optimize in practice due to limited access to individual-level data. Here, we introduce PUMAS, a novel method to fine-tune PRS models using summary statistics from genome-wide association studies (GWASs). Through extensive simulations, external validations, and analysis of 65 traits, we demonstrate that PUMAS can perform various model-tuning procedures using GWAS summary statistics and effectively benchmark and optimize PRS models under diverse genetic architecture. Furthermore, we show that fine-tuned PRSs will significantly improve statistical power in downstream association analysis.

Covariate-adaptive randomization schemes such as minimization and stratified permuted blocks are often applied in clinical trials to balance treatment assignments across prognostic factors. The existing theory for inference after covariate-adaptive randomization is mostly limited to situations where a correct model between the response and covariates can be specified or the randomization method has well-understood properties. Based on stratification with covariate levels utilized in randomization and a further adjustment for covariates not used in randomization, we propose several estimators for model free inference of the average treatment effect, the difference between response means under two treatments. We establish asymptotic normality of the proposed estimators under all popular covariate-adaptive randomization schemes including the minimization method, and we show that the asymptotic distributions are invariant with respect to covariate-adaptive randomization methods. Consistent variance estimators are constructed for asymptotic inference. Asymptotic relative efficiencies and finite sample properties of estimators are also studied. We recommend using one of our proposed estimators for valid and model free inference after covariate-adaptive randomization.

Polygenic risk scores (PRSs) have wide applications in human genetics research. Notably, most PRS models include tuning parameters which improve predictive performance when properly selected. However, existing model-tuning methods require validation data that is independent with both training and testing samples. These data rarely exist in practice, creating a significant gap between PRS methodology and applications. Here, we introduce PUMAS, a novel method to finetune PRS models using summary statistics from genome-wide association studies (GWASs). Through extensive simulations, external validations, and analysis of 65 GWAS traits, we demonstrate that PUMAS can perform a variety of model-tuning procedures (e.g. cross-validation) using GWAS summary statistics and can effectively benchmark and optimize PRS models under diverse genetic architecture. Applied to 211 neuroimaging traits and Alzheimer's disease, we show that fine-tuned PRSs will improve statistical power in association analysis. We believe our method resolves a fundamental problem without a current solution and will greatly benefit genetic prediction applications.

No abstract