Limits...
A prognostic gene expression profile that predicts circulating tumor cell presence in breast cancer patients.

Molloy TJ, Roepman P, Naume B, van't Veer LJ - PLoS ONE (2012)

Bottom Line: The detection of circulating tumor cells (CTCs) in the peripheral blood and microarray gene expression profiling of the primary tumor are two promising new technologies able to provide valuable prognostic data for patients with breast cancer.In the current study, we aimed to develop a novel profile which provided independent prognostic data by building a signature predictive of CTC status rather than outcome.This profile therefore has the potential to not only add prognostic information to currently-available microarray tests but in some circumstances even replace blood-based prognostic CTC tests at time of diagnosis for those patients already undergoing testing by multigene assays.

View Article: PubMed Central - PubMed

Affiliation: Division of Experimental Therapy, The Netherlands Cancer Institute, Amsterdam, The Netherlands.

ABSTRACT
The detection of circulating tumor cells (CTCs) in the peripheral blood and microarray gene expression profiling of the primary tumor are two promising new technologies able to provide valuable prognostic data for patients with breast cancer. Meta-analyses of several established prognostic breast cancer gene expression profiles in large patient cohorts have demonstrated that despite sharing few genes, their delineation of patients into "good prognosis" or "poor prognosis" are frequently very highly correlated, and combining prognostic profiles does not increase prognostic power. In the current study, we aimed to develop a novel profile which provided independent prognostic data by building a signature predictive of CTC status rather than outcome. Microarray gene expression data from an initial training cohort of 72 breast cancer patients for which CTC status had been determined in a previous study using a multimarker QPCR-based assay was used to develop a CTC-predictive profile. The generated profile was validated in two independent datasets of 49 and 123 patients and confirmed to be both predictive of CTC status, and independently prognostic. Importantly, the "CTC profile" also provided prognostic information independent of the well-established and powerful '70-gene' prognostic breast cancer signature. This profile therefore has the potential to not only add prognostic information to currently-available microarray tests but in some circumstances even replace blood-based prognostic CTC tests at time of diagnosis for those patients already undergoing testing by multigene assays.

Show MeSH

Related in: MedlinePlus

A gene expression profile derived from the primary tumor accurately predicts the presence of CTCs in the peripheral blood in breast cancer patients.(A) The CTC-profile indexes of 72 breast tumor samples are highly correlative with CTC status. Samples are ordered according to CTC-profile index and colored based on CTC-status. The dashed line indicates the classification threshold with optimal sensitivity and specificity. (B) A heatmap shows the level of expression of the 34 CTC profile genes for CTC-negative and CTC-positive patients. (C) The ROC curve of CTC-profile indexes compared to actual CTC status.
© Copyright Policy
Related In: Results  -  Collection


getmorefigures.php?uid=PMC3285692&req=5

pone-0032426-g001: A gene expression profile derived from the primary tumor accurately predicts the presence of CTCs in the peripheral blood in breast cancer patients.(A) The CTC-profile indexes of 72 breast tumor samples are highly correlative with CTC status. Samples are ordered according to CTC-profile index and colored based on CTC-status. The dashed line indicates the classification threshold with optimal sensitivity and specificity. (B) A heatmap shows the level of expression of the 34 CTC profile genes for CTC-negative and CTC-positive patients. (C) The ROC curve of CTC-profile indexes compared to actual CTC status.

Mentions: A multi-marker QPCR-based assay was previously used to determine circulating tumor cell (CTC) status from the peripheral blood samples of 192 breast cancer patients [2], [3]. To develop a gene expression profile that could predict CTC status, we analyzed the primary tumors of 72 of these patients using full-genome Agilent 44 K microarrays. A 4-fold cross validation (CV) procedure was used to determine the optimal set of genes expressed in the primary tumor that were associated with CTC status. We identified a set of 34 genes which formed the CTC-predictive profile (Table S1). The CTC-profile showed a significant leave-one-out CV performance of 82% for the prediction of CTC status (Figure 1) with an area under the Receiver Operating Characteristics (ROC) curve (AUC) of 0.88 and an optimal sensitivity of 74% (95%CI: 62%–82%) and specificity of 88% (95%CI: 79%–94%).


A prognostic gene expression profile that predicts circulating tumor cell presence in breast cancer patients.

Molloy TJ, Roepman P, Naume B, van't Veer LJ - PLoS ONE (2012)

A gene expression profile derived from the primary tumor accurately predicts the presence of CTCs in the peripheral blood in breast cancer patients.(A) The CTC-profile indexes of 72 breast tumor samples are highly correlative with CTC status. Samples are ordered according to CTC-profile index and colored based on CTC-status. The dashed line indicates the classification threshold with optimal sensitivity and specificity. (B) A heatmap shows the level of expression of the 34 CTC profile genes for CTC-negative and CTC-positive patients. (C) The ROC curve of CTC-profile indexes compared to actual CTC status.
© Copyright Policy
Related In: Results  -  Collection

Show All Figures
getmorefigures.php?uid=PMC3285692&req=5

pone-0032426-g001: A gene expression profile derived from the primary tumor accurately predicts the presence of CTCs in the peripheral blood in breast cancer patients.(A) The CTC-profile indexes of 72 breast tumor samples are highly correlative with CTC status. Samples are ordered according to CTC-profile index and colored based on CTC-status. The dashed line indicates the classification threshold with optimal sensitivity and specificity. (B) A heatmap shows the level of expression of the 34 CTC profile genes for CTC-negative and CTC-positive patients. (C) The ROC curve of CTC-profile indexes compared to actual CTC status.
Mentions: A multi-marker QPCR-based assay was previously used to determine circulating tumor cell (CTC) status from the peripheral blood samples of 192 breast cancer patients [2], [3]. To develop a gene expression profile that could predict CTC status, we analyzed the primary tumors of 72 of these patients using full-genome Agilent 44 K microarrays. A 4-fold cross validation (CV) procedure was used to determine the optimal set of genes expressed in the primary tumor that were associated with CTC status. We identified a set of 34 genes which formed the CTC-predictive profile (Table S1). The CTC-profile showed a significant leave-one-out CV performance of 82% for the prediction of CTC status (Figure 1) with an area under the Receiver Operating Characteristics (ROC) curve (AUC) of 0.88 and an optimal sensitivity of 74% (95%CI: 62%–82%) and specificity of 88% (95%CI: 79%–94%).

Bottom Line: The detection of circulating tumor cells (CTCs) in the peripheral blood and microarray gene expression profiling of the primary tumor are two promising new technologies able to provide valuable prognostic data for patients with breast cancer.In the current study, we aimed to develop a novel profile which provided independent prognostic data by building a signature predictive of CTC status rather than outcome.This profile therefore has the potential to not only add prognostic information to currently-available microarray tests but in some circumstances even replace blood-based prognostic CTC tests at time of diagnosis for those patients already undergoing testing by multigene assays.

View Article: PubMed Central - PubMed

Affiliation: Division of Experimental Therapy, The Netherlands Cancer Institute, Amsterdam, The Netherlands.

ABSTRACT
The detection of circulating tumor cells (CTCs) in the peripheral blood and microarray gene expression profiling of the primary tumor are two promising new technologies able to provide valuable prognostic data for patients with breast cancer. Meta-analyses of several established prognostic breast cancer gene expression profiles in large patient cohorts have demonstrated that despite sharing few genes, their delineation of patients into "good prognosis" or "poor prognosis" are frequently very highly correlated, and combining prognostic profiles does not increase prognostic power. In the current study, we aimed to develop a novel profile which provided independent prognostic data by building a signature predictive of CTC status rather than outcome. Microarray gene expression data from an initial training cohort of 72 breast cancer patients for which CTC status had been determined in a previous study using a multimarker QPCR-based assay was used to develop a CTC-predictive profile. The generated profile was validated in two independent datasets of 49 and 123 patients and confirmed to be both predictive of CTC status, and independently prognostic. Importantly, the "CTC profile" also provided prognostic information independent of the well-established and powerful '70-gene' prognostic breast cancer signature. This profile therefore has the potential to not only add prognostic information to currently-available microarray tests but in some circumstances even replace blood-based prognostic CTC tests at time of diagnosis for those patients already undergoing testing by multigene assays.

Show MeSH
Related in: MedlinePlus