Gene Expression-Based Prediction of Neoadjuvant Chemotherapy Response in Early Breast Cancer: Results of the Prospective Multicenter EXPRESSION Trial

Patient-Derived Xenografts of Triple-Negative Breast Cancer Enable Deconvolution and Prediction of Chemotherapy Responses

Combination chemotherapy remains essential for clinical management of triple-negative breast cancer (TNBC). Consequently, responses to multiple single agents cannot be delineated at the single patient level, even though some patients might not require all drugs in the combination. Herein, we conduct multi-omic analyses of orthotopic TNBC patient-derived xenografts (PDXs) treated with single agent carboplatin, docetaxel, or the combination. Combination responses were usually no better than the best single agent, with enhanced response in only ~13% of PDX, and apparent antagonism in a comparable percentage. Single-omic comparisons showed largely non-overlapping results between genes associated with single agent and combination treatments that could be validated in independent patient cohorts. Multi-omic analyses of PDXs identified agent-specific biomarkers/biomarker combinations, nominating high Cytokeratin-5 (KRT5) as a general marker of responsiveness. Notably, integrating proteomic with transcriptomic data improved predictive modeling of pathologic complete response to combination chemotherapy. PDXs refractory to all treatments were enriched for signatures of dysregulated mitochondrial function. Targeting this process indirectly in a PDX with HDAC inhibition plus chemotherapy in vivo overcomes chemoresistance. These results suggest possible resistance mechanisms and therapeutic strategies in TNBC to overcome chemoresistance, and potentially allow optimization of chemotherapeutic regimens.

Factors influencing pathological complete response and tumor regression in neoadjuvant radiotherapy and chemotherapy for high-risk breast cancer

BACKGROUND

Pathological complete response (pCR) is a well-established prognostic factor in breast cancer treated with neoadjuvant systemic therapy (naST). The determining factors of pCR are known to be intrinsic subtype, proliferation index, grading, clinical tumor and nodal stage as well as type of systemic therapy. The addition of neoadjuvant radiotherapy (naRT) to this paradigm might improve response, freedom from disease, toxicity and cosmetic outcome compared to adjuvant radiotherapy. The factors for pCR and primary tumor regression when neoadjuvant radiation therapy is added to chemotherapy have not been thoroughly described.

METHODS

We performed a retrospective analysis of 341 patients (cT1-cT4/cN0-N+) treated with naRT and naST between 1990 and 2003. Patients underwent naRT to the breast and mostly to the supra-/infraclavicular lymph nodes combined with an electron or brachytherapy boost. NaST was given either sequentially or simultaneously to naRT using different regimens. We used the univariate and multivariate regression analysis to estimate the effect of different subgroups and treatment modalities on pCR (ypT0/Tis and ypN0) as well as complete primary tumor response (ypT0/Tis; bpCR) in our cohort. Receiver operating characteristic (ROC) analysis was performed to evaluate the interval between radiotherapy (RT) and resection (Rx) as well as radiotherapy dose.

RESULTS

Out of 341 patients, pCR and pbCR were achieved in 31% and 39%, respectively. pCR rate was influenced by resection type, breast cancer subtype, primary tumor stage and interval from radiation to surgery in the multivariate analysis. Univariate analysis of bpCR showed age, resection type, breast cancer subtype, clinical tumor stage and grading as significant factors. Resection type, subtype and clinical tumor stage remained significant in multivariate analysis. Radiation dose to the tumor and interval from radiation to surgery were not significant factors for pCR. However, when treatment factors were added to the model, a longer interval from radiotherapy to resection was a significant predictor for pCR.

CONCLUSIONS

The factors associated with pCR following naST and naRT are similar to known factors after naST alone. Longer interval to surgery might to be associated with higher pCR rates. Dose escalation beyond 60 Gy did not result in higher response rates.

Identification of genes and pathways associated with menopausal status in breast cancer patients using two algorithms

BACKGROUND

Menopausal status has a known relationship with the levels of estrogen, progesterone, and other sex hormones, potentially influencing the activity of ER, PR, and many other signaling pathways involved in the initiation and progression of breast cancer. However, the differences between premenopausal and postmenopausal breast cancer patients at the molecular level are unclear.

METHODS

We retrieved eight datasets from the Gene Expression Omnibus (GEO) database. Differentially expressed genes (DEGs) associated with menopausal status in breast cancer patients were identified using the MAMA and LIMMA methods. Based on these validated DEGs, we performed Gene Ontology (GO) functional enrichment and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses. Protein-protein interaction (PPI) networks were constructed. We used DrugBank data to investigate which of these validated DEGs are targetable. Survival analysis was performed to explore the influence of these genes on breast cancer patient prognosis.

RESULTS

We identified 762 DEGs associated with menopausal status in breast cancer patients. PPI network analysis indicated that these genes are primarily involved in pathways such as the cell cycle, oocyte meiosis and progesterone-mediated oocyte maturation pathways. Notably, several genes played roles in multiple signaling pathways and were associated with patient survival. These genes were also observed to be targetable according to the DrugBank database.

CONCLUSION

We identified DEGs associated with menopausal status in breast cancer patients. The association of these genes with several key pathways may promote understanding of the complex characterizations of breast cancer. Our findings offer valuable insights for developing new therapeutic strategies tailored to the menopausal status of breast cancer patients.

A prospective cohort study identifying radiologic and tumor related factors of importance for breast conserving surgery after neoadjuvant chemotherapy

INTRODUCTION

Neoadjuvant chemotherapy (NAC) is an established treatment option for early breast cancer, potentially downstaging the tumor and increasing the eligibility for breast-conserving surgery (BCS). The primary aim of this study was to assess the rate of BCS after NAC, and the secondary aim was to identify predictors of application of BCS after NAC.

MATERIALS AND METHODS

This was an observational prospective cohort study of 226 patients in the SCAN-B (Clinical Trials NCT02306096) neoadjuvant cohort during 2014-2019. Eligibility for BCS was assessed at baseline and after NAC. Uni- and multivariable logistic regression analyses were performed using covariates with clinical relevance and/or those associated with outcome (BCS versus mastectomy), including tumor subtype, by gene expression analysis.

RESULTS

The overall BCS rate was 52%, and this rate increased during the study period (from 37% to 52%). Pathological complete response was achieved in 69 patients (30%). Predictors for BCS were smaller tumor size on mammography, visibility on ultrasound, histological subtype other than lobular, benign axillary status, and a diagnosis of triple-negative or HER2-positive subtype, with a similar trend for gene expression subtypes. Mammographic density was negatively related to BCS in a dose-response pattern. In the multivariable logistic regression model, tumor stage at diagnosis and mammographic density showed the strongest association with BCS.

CONCLUSION

The rate of BCS after NAC increased during the study period to 52%. With modern treatment options for NAC the potential for tumor response and BCS eligibility might further increase.

TP53 Signature Can Predict Pathological Response From Neoadjuvant Chemotherapy and Is a Prognostic Factor in Patients With Residual Disease

Background

The TP53 signature that predicts the mutation status of TP53 has been shown to be a prognostic factor and predictor of neoadjuvant chemotherapy (NAC) response.

Objectives

The current study sought to investigate the utility of the TP53 signature for predicting pathological complete response (pCR) and its prognostic significance among patients with residual disease (RD).

Design

The study followed a retrospective cohort study design.

Methods

Patients with T1-3/N0-1 from a cohort of those with HER2-negative breast cancer who received NAC were selected. Ability to predict pCR was evaluated using odds ratio, positive and negative predictive values, sensitivity, and specificity. Prognostic factors in the RD group were explored using the Cox proportional hazards model with distant recurrence-free survival (DRFS). Four independent cohorts were used for validation.

Results

A total of 333 eligible patients were classified into the TP53 mutant signature (n = 154) and wild-type signature (n = 179). Among the molecular and pathological factors, the TP53 signature had the highest predictive power for pCR. In 4 independent cohorts (n = 151, 85, 104, and 67, respectively), pCR rate in TP53 mutant signature group was significantly higher than that in the wild-type group. Univariate and multivariate analyses on DRFS in the RD group identified the TP53 signature and nodal status as independent prognostic factors, with the former having a better hazard ratio than the latter. After comparing DRFS between 3 groups (pCR, RD/TP53 wild-type signature, and RD/TP53 mutant signature groups), the RD/TP53 mutant signature group showed significantly worse prognosis compared with others. The RD/TP53 wild-type signature group did not exhibit inferior DRFS compared with the pCR group.

Conclusion

Our results showed that the TP53 mutant signature can predict pCR and that combining pathological response and TP53 mutant signature allows for the identification of subgroups with truly poor prognosis.

Using biological constraints to improve prediction in precision oncology

Many gene signatures have been developed by applying machine learning (ML) on omics profiles, however, their clinical utility is often hindered by limited interpretability and unstable performance. Here, we show the importance of embedding prior biological knowledge in the decision rules yielded by ML approaches to build robust classifiers. We tested this by applying different ML algorithms on gene expression data to predict three difficult cancer phenotypes: bladder cancer progression to muscle-invasive disease, response to neoadjuvant chemotherapy in triple-negative breast cancer, and prostate cancer metastatic progression. We developed two sets of classifiers: mechanistic, by restricting the training to features capturing specific biological mechanisms; and agnostic, in which the training did not use any a priori biological information. Mechanistic models had a similar or better testing performance than their agnostic counterparts, with enhanced interpretability. Our findings support the use of biological constraints to develop robust gene signatures with high translational potential.

Development and Assessment of a Novel Core Biopsy-Based Prediction Model for Pathological Complete Response to Neoadjuvant Chemotherapy in Women with Breast Cancer

Purpose: Pathological complete response (pCR), the goal of NAC, is considered a surrogate for favorable outcomes in breast cancer (BC) patients administrated neoadjuvant chemotherapy (NAC). This study aimed to develop and assess a novel nomogram model for predicting the probability of pCR based on the core biopsy. Methods: This was a retrospective study involving 920 BC patients administered NAC between January 2012 and December 2018. The patients were divided into a primary cohort (769 patients from January 2012 to December 2017) and a validation cohort (151 patients from January 2017 to December 2018). After converting continuous variables to categorical variables, variables entering the model were sequentially identified via univariate analysis, a multicollinearity test, and binary logistic regression analysis, and then, a nomogram model was developed. The performance of the model was assessed concerning its discrimination, accuracy, and clinical utility. Results: The optimal predictive threshold for estrogen receptor (ER), Ki67, and p53 were 22.5%, 32.5%, and 37.5%, respectively (all p < 0.001). Five variables were selected to develop the model: clinical T staging (cT), clinical nodal (cN) status, ER status, Ki67 status, and p53 status (all p ≤ 0.001). The nomogram showed good discrimination with the area under the curve (AUC) of 0.804 and 0.774 for the primary and validation cohorts, respectively, and good calibration. Decision curve analysis (DCA) showed that the model had practical clinical value. Conclusions: This study constructed a novel nomogram model based on cT, cN, ER status, Ki67 status, and p53 status, which could be applied to personalize the prediction of pCR in BC patients treated with NAC.

Deep learning with biopsy whole slide images for pretreatment prediction of pathological complete response to neoadjuvant chemotherapy in breast cancer：A multicenter study

INTRODUCTION

Predicting pathological complete response (pCR) for patients receiving neoadjuvant chemotherapy (NAC) is crucial in establishing individualized treatment. Whole-slide images (WSIs) of tumor tissues reflect the histopathologic information of the tumor, which is important for therapeutic response effectiveness. In this study, we aimed to investigate whether predictive information for pCR could be detected from WSIs.

MATERIALS AND METHODS

We retrospectively collected data from four cohorts of 874 patients diagnosed with biopsy-proven breast cancer. A deep learning pathological model (DLPM) was constructed to predict pCR using biopsy WSIs in the primary cohort, and it was then validated in three external cohorts. The DLPM could generate a deep learning pathological score (DLPs) for each patient; stromal tumor-infiltrating lymphocytes (TILs) were selected for comparison with DLPs.

RESULTS

The WSI feature-based DLPM showed good predictive performance with the highest area under the curve (AUC) of 0.72 among the cohorts. Alternatively, the combination of the DLPM and clinical characteristics offered a better prediction performance (AUC >0.70) in all cohorts. We also evaluated the performance of DLPM in three different breast subtypes with the best prediction for the triple-negative breast cancer (TNBC) subtype (AUC: 0.73). Moreover, DLPM combined with clinical characteristics and stromal TILs achieved the highest AUC in the primary cohort (AUC: 0.82) and validation cohort 1 (AUC: 0.80).

CONCLUSION

Our study suggested that WSIs integrated with deep learning could potentially predict pCR to NAC in breast cancer. The predictive performance will be improved by combining clinical characteristics. DLPs from DLPM can provide more information compared to stromal TILs for pCR prediction.

Patient-derived primary breast cancer cells and their potential for predicting sensitivity to chemotherapy

Chemotherapy resistance exposes patients to side effects and delays the effect of therapy in patients. So far, there are no predictive tools to predict resistance to chemotherapy and select sensitive chemotherapeutic drugs for the patient. Here, we aim to develop an in-vitro primary cell culture model from breast cancer patients to predict sensitivity to chemotherapy. We created the primary breast cancer cell medium BCMI and culture system with higher efficiency of the model establishment. Immunofluorescence staining of ERa, PR and HER2 were done to identify the primary breast cancer cell from the counterpart breast cancer patient. The killing assay showed that these primary breast cancer cells responded differently to doxorubicin and pirarubicin treatment. These results indicate that our established primary breast cancer cell model holds great promise for predicting breast cancer sensitivity to chemotherapy drugs.

Machine learning models based on immunological genes to predict the response to neoadjuvant therapy in breast cancer patients

Breast cancer (BC) is the most common malignancy worldwide and neoadjuvant therapy (NAT) plays an important role in the treatment of patients with early BC. However, only a subset of BC patients can achieve pathological complete response (pCR) and benefit from NAT. It is therefore necessary to predict the responses to NAT. Although many models to predict the response to NAT based on gene expression determined by the microarray platform have been proposed, their applications in clinical practice are limited due to the data normalization methods during model building and the disadvantages of the microarray platform compared with the RNA-seq platform. In this study, we first reconfirmed the correlation between immune profiles and pCR in an RNA-seq dataset. Then, we employed multiple machine learning algorithms and a model stacking strategy to build an immunological gene based model (Ipredictor model) and an immunological gene and receptor status based model ICpredictor model) in the RNA-seq dataset. The areas under the receiver operator characteristic curves for the Ipredictor model and ICpredictor models were 0.745 and 0.769 in an independent external test set based on the RNA-seq platform, and were 0.716 and 0.752 in another independent external test set based on the microarray platform. Furthermore, we found that the predictive score of the Ipredictor model was correlated with immune microenvironment and genomic aberration markers. These results demonstrated that the models can accurately predict the response to NAT for BC patients and will contribute to individualized therapy.

Immunomodulating Therapies in Breast Cancer-From Prognosis to Clinical Practice

The role of the immune system in breast cancer has been debated for decades. The advent of technologies such as next generation sequencing (NGS) has elucidated the crucial interplay between somatic mutations in tumors leading to neoantigens and immune responses with increased tumor-infiltrating lymphocytes and improved prognosis of breast cancer patients. In particular, triple-negative breast cancer (TNBC) has a higher mutational burden compared to other breast cancer subtypes. In addition, higher levels of tumor-associated antigens suggest that immunotherapies are a promising treatment option, specifically for TNBC. Indeed, higher concentrations of tumor-infiltrating lymphocytes are associated with better prognosis and response to chemotherapy in TNBC. An important target within the cancer immune cell cycle is the "immune checkpoint". Immune checkpoint inhibitors (ICPis) block the interaction of certain cell surface proteins that act as "brakes" on immune responses. Recent studies have shown that ICPis improve survival in both early and advanced TNBC. However, this comes at the price of increased toxicity, particularly immune-mediated toxicity. As an alternative approach, individualized mRNA vaccination strategies against tumor-associated neoantigens represent another promising approach leading to neoantigen-specific immune responses. These novel strategies should help to improve treatment outcomes, especially for patients with triple negative breast cancer.

Lineage-specific mechanisms and drivers of breast cancer chemoresistance revealed by 3D biomimetic culture

To improve the success rate of current preclinical drug trials, there is a growing need for more complex and relevant models that can help predict clinical resistance to anticancer agents. Here, we present a three‐dimensional (3D) technology, based on biomimetic collagen scaffolds, that enables the modeling of the tumor hypoxic state and the prediction of in vivo chemotherapy responses in terms of efficacy, molecular alterations, and emergence of resistance mechanisms. The human breast cancer cell lines MDA‐MB‐231 (triple negative) and MCF‐7 (luminal A) were treated with scaling doses of doxorubicin in monolayer cultures, 3D collagen scaffolds, or orthotopically transplanted murine models. Lineage‐specific resistance mechanisms were revealed by the 3D tumor model. Reduced drug uptake, increased drug efflux, and drug lysosomal confinement were observed in triple‐negative MDA‐MB‐231 cells. In luminal A MCF‐7 cells, the selection of a drug‐resistant subline from parental cells with deregulation of p53 pathways occurred. These cells were demonstrated to be insensitive to DNA damage. Transcriptome analysis was carried out to identify differentially expressed genes (DEGs) in treated cells. DEG evaluation in breast cancer patients demonstrated their potential role as predictive biomarkers. High expression of the transporter associated with antigen processing 1 (TAP1) and the tumor protein p53‐inducible protein 3 (TP53I3) was associated with shorter relapse in patients affected by ER⁺ breast tumor. Likewise, the same clinical outcome was associated with high expression of the lysosomal‐associated membrane protein 1 LAMP1 in triple‐negative breast cancer. Hypoxia inhibition by resveratrol treatment was found to partially re‐sensitize cells to doxorubicin treatment. Our model might improve preclinical in vitro analysis for the translation of anticancer compounds as it provides: (a) more accurate data on drug efficacy and (b) enhanced understanding of resistance mechanisms and molecular drivers.