Search

Background Triple-negative breast cancer (TNBC) is a heterogeneous disease and we have previously shown that rapid relapse of TNBC is associated with distinct sociodemographic features. We hypothesized that rapid versus late relapse in TNBC is also defined by distinct clinical and genomic features of primary tumors. Methods Using three publicly-available datasets, we identified 453 patients diagnosed with primary TNBC with adequate follow-up to be characterized as ‘rapid relapse’ (rrTNBC; distant relapse or death ≤2 years of diagnosis), ‘late relapse’ (lrTNBC; > 2 years) or ‘no relapse’ (nrTNBC: > 5 years no relapse/death). We explored basic clinical and primary tumor multi-omic data, including whole transcriptome (n = 453), and whole genome copy number and mutation data for 171 cancer-related genes (n = 317). Association of rapid relapse with clinical and genomic features were assessed using Pearson chi-squared tests, t-tests, ANOVA, and Fisher exact tests. We evaluated logistic regression models of clinical features with subtype versus two models that integrated significant genomic features. Results Relative to nrTNBC, both rrTNBC and lrTNBC had significantly lower immune signatures and immune signatures were highly correlated to anti-tumor CD8 T-cell, M1 macrophage, and gamma-delta T-cell CIBERSORT inferred immune subsets. Intriguingly, lrTNBCs were enriched for luminal signatures. There was no difference in tumor mutation burden or percent genome altered across groups. Logistic regression mModels that incorporate genomic features significantly outperformed standard clinical/subtype models in training (n = 63 patients), testing (n = 63) and independent validation (n = 34) cohorts, although performance of all models were overall modest. Conclusions We identify clinical and genomic features associated with rapid relapse TNBC for further study of this aggressive TNBC subset.

Purpose: Cancers with short interval between diagnosis and metastasis are associated with aggressive clinical course. Specifically, metastatic relapse of triple-negative breast cancer (TNBC) within 2 years of initial primary diagnosis is associated with marked chemoresistance, rapid progression, and poor prognosis. We hypothesized that rapid relapse TNBCs (rrTNBC; distant metastatic relapse or death 2 years) or no relapse (nrTNBC; no distant relapse or death with at least 5 years follow-up). Patients and Methods: We identified 453 primary TNBCs from three publicly-available datasets and characterized each as rrTNBC, lrTNBC, or nrTNBC. We compiled primary tumor clinical and multi-omic data, including transcriptome (n=453), copy number alterations (CNAs; n=317), and mutations in 171 cancer-related genes (n=317), then calculated expression and immune signatures. Results: Patients with rrTNBC were higher stage at diagnosis (Chi-square p Citation Format: Yiqing Zhang, William Nock, Meghan Wyse, Zachary Weber, Elizabeth J Adams, Asad Sarah, Sinclair Stockard, David Tallman, Jasneet Singh, Junu Bae, Eric P Winer, Nancy U Lin, Yi-Zhou Jiang, Ding Ma, Peng Wang, Leming Shi, Wei Huang, Zhi-Ming Shao, Claire Verschraegen, Mathew Cherian, Maryam B Lustberg, Bhuvana Ramaswamy, Sagar Sardesai, Jeffrey VanDeusen, Nicole Williams, Wesolowski Robert, Daniel G Stover. Machine learning predicts rapid relapse in triple negative breast cancer [abstract]. In: Proceedings of the 2019 San Antonio Breast Cancer Symposium; 2019 Dec 10-14; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2020;80(4 Suppl):Abstract nr P4-05-02.

PurposeMetastatic relapse of triple-negative breast cancer (TNBC) within 2 years of diagnosis is associated with particularly aggressive disease and a distinct clinical course relative to TNBCs that relapse beyond 2 years. We hypothesized that rapid relapse TNBCs (rrTNBC; metastatic relapse or death 2 years).Patients and MethodsWe identified 453 primary TNBCs from three publicly-available datasets and characterized each as rrTNBc, lrTNBC, or ‘no relapse’ (nrTNBC: no relapse/death with at least 5 years follow-up). We compiled primary tumor clinical and multi-omic data, including transcriptome (n=453), copy number alterations (CNAs; n=317), and mutations in 171 cancer-related genes (n=317), then calculated published gene expression and immune signatures.ResultsPatients with rrTNBC were higher stage at diagnosis (Chi-square pTP53mutations were significantly more frequent in rrTNBC compared to lrTNBC (Fisher exact FDR p=0.009). To develop an optimal classifier, we used 77 significant clinical and ‘omic features to evaluate six modeling approaches encompassing simple, machine learning, and artificial neural network (ANN). Support vector machine outperformed other models with average receiver-operator characteristic area under curve >0.75.ConclusionsWe provide a new approach to define TNBCs based on timing of relapse. We identify distinct clinical and genomic features that can be incorporated into machine learning models to predict rapid relapse of TNBC.

Copy number aberrations (CNAs) are gains and losses of large genomic segments present across most cancer types and are a hallmark of cancer genomic alterations. However, the processes underlying CNAs and characteristic patterns of CNAs are poorly understood. Using single nucleotide variant (SNV) data, bioinformatic advances have identified underlying mutational signatures resulting from distinct mutational processes. Mutational signatures have led to a variety of discoveries, several of which are being investigated in clinical management of cancer. The development of algorithms able to uncover similar signatures for CNAs, rather than SNVs, is still in its infancy. Here we present an analysis package for the R programming language called CNSigs that allows for the robust and reproducible derivation of copy number signatures. Based on a list of extracted copy number features previously verified in ovarian cancer, we utilize mixed model approaches and non-negative matrix factorization to derive CNA signatures across cancer types. The development of a package to derive signatures from copy number data allows further investigation of underlying processes that may be responsible for these CNA fingerprints. To verify the reproducibility of the signatures, we derived signatures from two independent breast cancer datasets that use distinct copy number segmentation approaches. From these independent datasets we were able to recover the same signatures with high accuracy. We identified five signatures that are distinct from known breast cancer receptor-based or expression-based subtypes, yet reveal unique associations with underlying mutations, mutational processes, and transcriptional programs. To validate robustness, we applied the pipeline to 11 different cancer types from the TCGA dataset and showed that we were able to derive signatures from all of these cancer types of varying sample sizes. We identified certain signatures that cut across tumor types, while others are distinct to individual cancers. This work lays the groundwork for further analysis into the underlying molecular processes leading to these copy number signatures seen across cancer types. The CNSigs package also allows researchers to easily analyze their own samples to look for signatures in their copy number profiles and to compare these to signatures previously derived for their cancer type. Citation Format: David Tallman, Sinclair Stockard, Zachary T. Weber, Sarah Asad, Katharine Collier, Elizabeth Adams, Jerome Bey, Daniel G. Stover. CNSigs: An R package for the identification of copy number signatures [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 5471.