Dataset Information

Dynamic classification using case-specific training cohorts outperforms static gene expression signatures in breast cancer.

ABSTRACT: The molecular diversity of breast cancer makes it impossible to identify prognostic markers that are applicable to all breast cancers. To overcome limitations of previous multigene prognostic classifiers, we propose a new dynamic predictor: instead of using a single universal training cohort and an identical list of informative genes to predict the prognosis of new cases, a case-specific predictor is developed for each test case. Gene expression data from 3,534 breast cancers with clinical annotation including relapse-free survival is analyzed. For each test case, we select a case-specific training subset including only molecularly similar cases and a case-specific predictor is generated. This method yields different training sets and different predictors for each new patient. The model performance was assessed in leave-one-out validation and also in 325 independent cases. Prognostic discrimination was high for all cases (n?=?3,534, HR?=?3.68, p?=?1.67 E-56). The dynamic predictor showed higher overall accuracy (0.68) than genomic surrogates for Oncotype DX (0.64), Genomic Grade Index (0.61) or MammaPrint (0.47). The dynamic predictor was also effective in triple-negative cancers (n?=?427, HR?=?3.08, p?=?0.0093) where the above classifiers all failed. Validation in independent patients yielded similar classification power (HR?=?3.57). The dynamic classifier is available online at http://www.recurrenceonline.com/?q=Re_training. In summary, we developed a new method to make personalized prognostic prediction using case-specific training cohorts. The dynamic predictors outperform static models developed from single historical training cohorts and they also predict well in triple-negative cancers.

SUBMITTER: Gyorffy B

PROVIDER: S-EPMC4354298 | biostudies-literature | 2015 May

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Dynamic classification using case-specific training cohorts outperforms static gene expression signatures in breast cancer.

Győrffy Balázs B Karn Thomas T Sztupinszki Zsófia Z Weltz Boglárka B Müller Volkmar V Pusztai Lajos L

International journal of cancer 20141011 9

The molecular diversity of breast cancer makes it impossible to identify prognostic markers that are applicable to all breast cancers. To overcome limitations of previous multigene prognostic classifiers, we propose a new dynamic predictor: instead of using a single universal training cohort and an identical list of informative genes to predict the prognosis of new cases, a case-specific predictor is developed for each test case. Gene expression data from 3,534 breast cancers with clinical annot ...[more]

PMID: 25274406

Similar Datasets

Project description:Background: Epigenetic regulation, including DNA methylation, plays a major role in shaping the identity and function of immune cells. Innate and adaptive immune cells recruited into tumor tissues contribute to the formation of the tumor immune microenvironment (TIME), which is closely involved in tumor progression in breast cancer (BC). However, the specific methylation signatures of immune cells have not been thoroughly investigated yet. Additionally, it remains unknown whether immune cells-specific methylation signatures can identify subgroups and stratify the prognosis of BC patients. Methods: DNA methylation profiles of six immune cell types from eight datasets downloaded from the Gene Expression Omnibus were collected to identify immune cell-specific hypermethylation signatures (IC-SHMSs). Univariate and multivariate cox regression analyses were performed using BC data obtained from The Cancer Genome Atlas to identify the prognostic value of these IC-SHMSs. An unsupervised clustering analysis of the IC-SHMSs with prognostic value was performed to categorize BC patients into subgroups. Multiple Cox proportional hazard models were constructed to explore the role of IC-SHMSs and their relationship to clinical characteristics in the risk stratification of BC patients. Integrated discrimination improvement (IDI) was performed to determine whether the improvement of IC-SHMSs on clinical characteristics in risk stratification was statistically significant. Results: A total of 655 IC-SHMSs of six immune cell types were identified. Thirty of them had prognostic value, and 10 showed independent prognostic value. Four subgroups of BC patients, which showed significant heterogeneity in terms of survival prognosis and immune landscape, were identified. The model incorporating nine IC-SHMSs showed similar survival prediction accuracy as the clinical model incorporating age and TNM stage [3-year area under the curve (AUC): 0.793 vs. 0.785; 5-year AUC: 0.735 vs. 0.761]. Adding the IC-SHMSs to the clinical model significantly improved its prediction accuracy in risk stratification (3-year AUC: 0.897; 5-year AUC: 0.856). The results of IDI validated the statistical significance of the improvement (p < 0.05). Conclusions: Our study suggests that IC-SHMSs may serve as signatures of classification and risk stratification in BC. Our findings provide new insights into epigenetic signatures, which may help improve subgroup identification, risk stratification, and treatment management.

Dataset Information

Dynamic classification using case-specific training cohorts outperforms static gene expression signatures in breast cancer.

Publications

Dynamic classification using case-specific training cohorts outperforms static gene expression signatures in breast cancer.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets