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

B. Györffy, Thomas Karn, Zsófia Sztupinszki, Boglárka Weltz, Volkmar Müller, Lajos Pusztai

Research output: Contribution to journalArticle

7 Citations (Scopus)

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 E256). 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/?q5Retraining. 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.

Original languageEnglish
Pages (from-to)2091-2098
Number of pages8
JournalInternational Journal of Cancer
Volume136
Issue number9
DOIs
Publication statusPublished - May 1 2015

Fingerprint

Transcriptome
Breast Neoplasms
Neoplasms
Gene Expression
Recurrence
Survival
Genes
Breast Cancer 3

Keywords

  • Breast cancer
  • Gene expression
  • Survival

ASJC Scopus subject areas

  • Cancer Research
  • Oncology

Cite this

Dynamic classification using case-specific training cohorts outperforms static gene expression signatures in breast cancer. / Györffy, B.; Karn, Thomas; Sztupinszki, Zsófia; Weltz, Boglárka; Müller, Volkmar; Pusztai, Lajos.

In: International Journal of Cancer, Vol. 136, No. 9, 01.05.2015, p. 2091-2098.

Research output: Contribution to journalArticle

Györffy, B. ; Karn, Thomas ; Sztupinszki, Zsófia ; Weltz, Boglárka ; Müller, Volkmar ; Pusztai, Lajos. / Dynamic classification using case-specific training cohorts outperforms static gene expression signatures in breast cancer. In: International Journal of Cancer. 2015 ; Vol. 136, No. 9. pp. 2091-2098.
@article{3e8150f99743436fab35cdc80dc0c081,
title = "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 E256). 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/?q5Retraining. 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.",
keywords = "Breast cancer, Gene expression, Survival",
author = "B. Gy{\"o}rffy and Thomas Karn and Zs{\'o}fia Sztupinszki and Bogl{\'a}rka Weltz and Volkmar M{\"u}ller and Lajos Pusztai",
year = "2015",
month = "5",
day = "1",
doi = "10.1002/ijc.29247",
language = "English",
volume = "136",
pages = "2091--2098",
journal = "International Journal of Cancer",
issn = "0020-7136",
publisher = "Wiley-Liss Inc.",
number = "9",

}

TY - JOUR

T1 - Dynamic classification using case-specific training cohorts outperforms static gene expression signatures in breast cancer

AU - Györffy, B.

AU - Karn, Thomas

AU - Sztupinszki, Zsófia

AU - Weltz, Boglárka

AU - Müller, Volkmar

AU - Pusztai, Lajos

PY - 2015/5/1

Y1 - 2015/5/1

N2 - 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 E256). 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/?q5Retraining. 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.

AB - 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 E256). 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/?q5Retraining. 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.

KW - Breast cancer

KW - Gene expression

KW - Survival

UR - http://www.scopus.com/inward/record.url?scp=84923166135&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84923166135&partnerID=8YFLogxK

U2 - 10.1002/ijc.29247

DO - 10.1002/ijc.29247

M3 - Article

VL - 136

SP - 2091

EP - 2098

JO - International Journal of Cancer

JF - International Journal of Cancer

SN - 0020-7136

IS - 9

ER -