1. dia - Kaplan

Download Report

Transcript 1. dia - Kaplan

An online tool for the validation of survival-predicting
biomarkers in non small-cell lung cancer using microarray
data of 1,329 1,715 patients
Balázs Győrffy and András Lánczky
Research Laboratory for Pediatrics and Nephrology, Hungarian Academy of
Sciences and Semmelweis University 1st Dept. of Pediatrics, Budapest, Hungary
Background and Objective
Results
1.optimized treatment for non-small cell lung
cancer had lead to improved prognosis, but
the overall survival is still very short
1.DATABASE construction
2.by identifying biomarkers related to survival
we can further understand the molecular
basis of the disease
 Histology (adeno/squamous/large): 50% / 45% / 5%
OBJECTIVE: we present the development
of an online available tool suitable for the
real-time meta-analysis of published lung
cancer microarray datasets to identify
biomarkers related to survival
 n=1,715
 Median OS=40.0 months, age: 64+/-10 yrs
 Stage 1/2/3/4: 63% / 27% / 10% / 1%
2. META-ANALYSIS of biomarker candidates
 Biomarker candidates identified in Pubmed n=22
 For each gene the exact conditions in which it was
identified have been retrieved, and these have been
used as filtering when selecting the patients for the
survival analysis.
 Of the 22, the best performing genes are:
Materials & Methods
Symbol
VEGF
1.DATABASE construction




Repositories: GEO, TCGA, ArrayExpress, caBIG
Platforms: Affymetrix HGU133A, plus2 & A 2.0 arrays
at least 30 patients with survival information
MAS5 normalization + quality control
Cyclin E
Reference
n(1)
Cohort
n(2)
HR
p value
Zhan et al 2009
5386
NSCLC
1404
1.9
3.3e-10
Huang et al 2012
2606
NSCLC
1404
1.59
2.3e-09*
ADE
486
2.44
4.8e-08
CDK1
Zhang et al 2012a
2731
NSCLC
1404
2.56
<1e-16*
CADM1
Botling et al 2012
617
ADE
486
0.38
7e-12*
CDKN2A
Jin et al 2001
106
NSCLC
1404
1.65
1.8e-09
CD24
Lee et al 2010
267
ADE
486
2.45
3.6e-10
Simon et al 2005
51
NSCLC
1404
1.65
1.4e-10
ERCC1
n(1): number of patients in original study, n(2): number of patients in the KM-plotter, HR: hazard ratio,
ADE: adenocarcinoma, NSCLC: all non-small-cell lung cancer patients
2. SURVIVAL analysis
 Kaplan-Meier plot
 „survival” Bioconductor package
 Cox univariate + multivariate analysis
3. Selected KAPLAN-MEIER plots (table: *)
Cyclin E1
CDK1
CADM1
3. ONLINE platform
 Apache web server on Debian Linux
 script developed in PHP
 Open access at: www.kmplot.com/lung
Summary
 we performed a meta-analysis of
survival-associated genes
 an integrated database and an online
tool for future in silico validation of new
candidates has been established
4. META-analysis
 Pubmed search of published biomarkers
 Best cutoff selection: each percentile (of expression)
between the lower and upper quartiles are computed
and the best performing threshold is used as the
final cutoff in the Cox regression analysis.
Web addresses
Online access: http://www.kmplot.com/lung
Group homepage: http://gyer1-6.sote.hu/gyorffy
Contact: [email protected]
Grant support: OTKA PD 83154, PREDICT 259303 (EU Health.2010.2.4.1.-8), KTIA EU_BONUS_12-1-2013-0003,
Alexander von Humboldt-Foundation