Lung Cancer

1
Lung CancerMolecular Network Disease Cheng
ShujunCancer Institute, Chinese Academy of
Medical Sciences, Peking Union Medical College
2
Fortune Magazine, March 22,2004 The five-year
survival rate did not improve when a cancer has
spread
3

• The challenge we faced in cancer therapy
  may be related to the complexity of gene network
  changes in lung cancer cells, especially at late
  stages.

4
(Li Ding et al. Nature, 2008, Oct. 455
1069-)DNA sequencing of 623 genes in 188 lung
adenocarcinomas. 26 genes are mutated at
significantly high frequencies . Several
important pathway involved in lung adenocarcinoma
5
A small-cell lung cancer genome with complex
signatures of tobacco exposure/nature Published
online 16 December 2009

• .
• They sequenced a small-cell lung cancer
  cell line, NCI-H209, NCI-BL209 (an
  EpsteinBarr-virus-transformed lymphoblastoid
  line has been generated from the patient. ) to
  explore the mutational burden associated with
  tobacco smoking.
• A total of 22,910 somatic substitutions
  (including 134 in coding exons ) were identified
  in a small-cell lung cancer cell line .
• They estimated one mutation for every 15
  cigarettes smoked.

6
(No Transcript)
7

• What we may learn from the recent
  studies
• Pathway rather than individual genes
  appear to govern the course of tumorigenesis.
• The wide variation in tumor behavior
  and responsiveness to therapy may relate to the
  diversity of gene function abnormalities
  (network) in different patients from the same
  type of tumor.
• The acquisition of numerous somatic
  mutations, each with a small fitness advantage,
  may also drive tumourigenesis ?

8

• Previous report indicated that many cancer
  genes play critical roles in cellular development
  and growth
• Cancer might be a molecular network disease
  caused by cellular abnormal growth and
  differentiation, which may be related to
  developmental genome disorder

9
During the past two yeas, we investigated gene
expression profiles in different time of human
lung embryonic development and lung cancer
tissues
10
Mid FL
Early FL
Adjacent lung tissue
Embud
Lung Cancer
AduL
Developmental landscape We projected all the
embryonic tissue samples (Embud, early and middle
fetal lung ( Early FL Mid FL) and the mature
lung samples (AduL) adjacent lung tissues
(Adjacent Lung) and the lung cancer tissues
(Lung Cancer )onto a two dimensional space with
the principle component analysis (PCA) to
construct the developmental landscape. Every spot
represents one sample. The color of the spot
indicates its tissue type. . (cycle direction
wide distribution for cancer(hetrogenecity)
11
Gene-expression in human fetal lung tissues and
lung cancers
P53 signaling pathway
NOSTRIN mediated eNOS trafficking
Metabolism of nitric oxide
Mammalian Wnt signaling pathway
Inhibition of matrix metalloproteinases
TGFBR
Signaling events mediated by HDAC Class III
RNA polymerase I transcription initiation
RNA polymerase I transcription
RNA polymerase I promoter clearance
DNA replication
DNA replication preinitiation
DNA strand elongation
E2F mediated regulation of DNA replication
G1/S transition
G2/M checkpoints
G2/M DNA damage checkpoint
Metaphase/anaphase transition
Mitotic prometaphase
Mitotic prophase
Mitotic spindle checkpoint
Mitotic telophase /cytokinesis
Cheng et al. unpublished data
34
12
The dynamic gene expressing patterns in human
developmental process
We take a bundle of genes (Embryfeature) to test
their clinical significance.
Embryfeature enriched in following GO terms
DNA Replication
DNA Replication Pre-Initiation
DNA strand elongation
E2F mediated regulation of DNA replication
E2F transcriptional targets at G1/S
FOXM1 transcription factor network
FoxO family signaling
G1/S Transition
G2/M Checkpoints
G2/M DNA damage checkpoint
G2/M Transition
M Phase
M/G1 Transition
Mitotic Metaphase/Anaphase Transition
Mitotic Prometaphase
Mitotic Prophase
Mitotic Spindle Checkpoint
Mitotic Telophase /Cytokinesis
13
Clinical Significance of Embryfeature

• The expression level of Embryfeature was
  correlated with the survival time of cancer
  patients.
• Such as
• Lung adenocarcinoma (353 samples)
• 4 independent data sets 49, 117, 125, 62 samples
• Glioma(371 samples)
• 3 independent data sets 100, 191,80 samples
• Breast Cancer(1300 samples)
• 7 independent data sets
• 159, 286, 204, 189, 136, 77, 249 sampels

14
We divided the 49 lung ADC patients into two
groups according to the expression level of
Embryfeature in their cancer tissues. Survival
analysis showed that the prognosis of the
Embryfeature higher patients (H group, red line)
was significantly worse than that of lower ones
(L group, black line).
P 0.041
Overall survival analysis of 49 lung ADC
patients(from our cancer hospital)
15
The same result was confirmed in other three
independent lung adenocarcinoma data sets.
The microarray data and patients clinical
information were downloaded from GEO database of
NCBI.
Relapse-free survival analysis of 62 lung ADC
patients
16
Survival analysis of Glioma patients grouped by
their Embryfeature expression level.
We analyzed 3 independent sets of glioma patients
(371 samples) with the expression level of
Embryfeature in their cancer tissues.
Survival analysis showed that the prognosis of
the Embryfeature higher patients (H group, red
line) was significantly worse than that of lower
ones (L group, black line).
Overall survival analysis of 77 Glioma patients
17
The expression level of Embryfeature was
associated with the relapse-free and overall
survival of the breast cancer patients, which was
confirmed in 7 independent datasets, involving
1,300 samples. Here the survival curves (K-M
curve) of four datasets were shown.
18
The hub genes in the interaction network
constituted a 7-node sub-network shown as below.
Extensive research on the interaction among these
hub genes may provide more hints on understanding
human lung carcinogenesis. Further analysis is
under way.
19

• The embryfeature gene may predict the
  prognosis of several types of tumor (breast
  cancer, glioma, lung adenocarcinoma)located at
  different organs, It may indicate that the
  clinic features of human cancer may not only
  depend on their location, perhaps also on their
  developmental original memory?

20

• The gene network in cancer cells can
  overcome (compensate) the effect of single-agent
  intervention. ( as reported, the amplification of
  Met gene can reactivate PI3K/AKT pathway
  Inhibited by Iressa). The development of drug
  resistance in cancer cells may also relate to
  their gene network response.

21

• Lung cancer is a molecular network disease
  caused by cellular abnormal growth and
  differentiation related to developmental genome.
• It will be difficult to cure cancer at late stage
  with single drug (single gene).
• Multidrug treatments (network drug) are needed
  for cancer therapy in the future

22
(No Transcript)
23

• Key steps for lung cancer research in the future
• To intensify clinical investigation on human lung
  cancer and set up tumor tissue banks.
• To establish high-throughput platforms for fast
  analysis of cancer samples through a synthetic
  approach.
• Systematic analysis of both clinical and basic
  research data with bioinformatics.

24
(No Transcript)
25

• Thank you