CHROMOSOMAL ABNORMALITIES AND GENOMIC IMBALANCES IN LUNG CANCER: THE PUZZLING PICTURE

Alterations in the DNA content have been well documented
in lung carcinomas by flow cytometry and static tissue morphometry.
A metaanalysis including data from 4033 NSCLC
patients from 35 published studies has shown that the majority
of NSCLC were aneuploidy and patients with aneuploid
tumors had a significantly shorter survival duration than those
with normal DNA content reflecting both diploid and pseudodiploid
chromosome NSCLC. 17 However, the aneuploid
chromosomal complement in the tumor cells included great
variety of structural and numerical changes, many of which
could be random events.
The search for recurrent abnormalities in lung cancer, the
ones most likely to play specific roles in cancer development,
has started long ago. The first recurrent changes in lung cancer,
the deletions of 3p in SCLC, were identified by classical karyotypic
analysis. 18 However, probably because of the complexity
of the chromosomal alterations and the limitations of the conventional
techniques, few karyotype reports of primary lung
tumors or cell lines were published in the 2 decades following
that seminal publication. 19–26 Loss of large chromosomal segments
in 3p and 8p, gain of whole chromosomal arms, such
as 5p, and amplification by homogeneously staining regions
(HSR) and double minutes (DM) were reported. However, the
conclusions of all those studies were by and large limited, often
yielding incomplete karyotypes.
The advent of the CGH technology brought new momentum
to the cancer field and has also impacted discoveries.
Novel and histological type-specific gains and losses of chromosome
segments in lung cancers were revealed in addition
to those previously reported by conventional cytogenetic approaches.
Chromosomal gains were detected in the long arms
of chromosomes 8, 17, and 19 in NSCLC and chromosome
arms 3q, 8p, and Xq in SCLC. Chromosome losses were frequent
in 1p, 4q, 5q, 6q, 8p, 9p, 13q, and 17p in NSCLC and
in 5q, 13q, and 17p in SCLC. 21,22,27–31
In the end of the 1990s, studies using M-FISH and SKY
were performed in lung carcinoma cell lines that had been
established previously at the National Cancer Institute laboratories
or were newly established by other investigators, and
from resected tumors. Those studies have resulted in the identification
of a greater degree of chromosomal rearrangements
than it been detected by previous G-banding and mCGH
analyses. 32–43 Chromosomal abnormalities were also detected
in nonmalignant bronchial epithelium of heavy smokers. 44 MFISH
and SKY technologies enabled the disclosure of cryptic
translocations, enhanced the ability to delineate chromosomal
breakpoints when integrating information from conventional
banding analysis, and clarified the chromosomal composition
of unrecognized marker chromosomes. Important similarities
were noticed between karyotypic changes in established cell
lines and primary tumors. The vast majority of translocations
were unbalanced but a significant number of balanced translocations
were also detected. These studies have provided a basis
for the search of genes mapped at the breakpoints that were
potentially deregulated and associated with tumorigenesis.
Despite all these efforts, discovery of recurrent gene fusions
generated by structural rearrangements based on cytogenetics
approaches has been quite rare. One such example was
the identification of a translocation between the chromosomes
15 and 19 [t(15;19)(q11;p13)] in an aggressive lung cancer
metastatic to mediastinum and bone arising in a young woman
without a history of smoking or a family history of cancer. 45
The breakpoint on chromosome 19 was mapped to the 5’ region
of the highly overexpressed NOTCH3 gene, which led to
further investigations of the role of this gene in lung cancer.
Notch3 expression was detected in approximately 40% of resected
lung tumors and positively correlated with epidermal
growth factor receptor (EGFR) expression. Notch inhibition
was shown to increase sensitivity to EGFR tyrosine kinase inhibitors
(TKIs) and decrease mitogen-activated protein kinase
(MAPK) phosphorylation, observations that support a role for
NOTCH3 signaling in lung cancer through EGFR-related
pathways. 46 The translocation breakpoints were later refined
to 15q13.2 and 19p13.1 and the cloning of these regions
identified a novel fusion transcript in which the 3' end of the
BRD4 gene on chromosome 19p was fused to the 5' end of the
NUT gene on chromosome 15q. The BRD4–NUT fusion was
demonstrated to alter the cell cycle kinetics, augmenting the
inhibition of the progression G1 to S phase compared with the
wild type BRD4 gene. However, the exact role of the BRD4–
NUT fusion in the pathogenesis of lung cancers remains unclear
and the t(15;19) has not been found in large lung cancer
cohorts tested, which suggest that it is not common in lung
cancer. 47 Gene fusions detected using other strategies are going
to be discussed later. The detection of the intracellular targets
of these fusions is expected to bring new insights into molecular
pathways that trigger tumor development.
A much more detailed picture of genomic copy number
variations has been achieved in the last years with the
array analyses. A summary of detected focal gain and losses
is presented in Table 6.1 for five studies focusing on SCLC
specimens 19,48–51 and in Table 6.2 for 13 studies focusing on
NSCLC specimens. 19,51–62 Although data are available for
over 70 SCLC and close to 800 NSCLC specimens, including
cell lines and primary tumors, it is difficult to compare
those results. Different platforms had different probes, and it
is not always possible to confirm equivalencies. Despite these
limitations, it is evident that there are important recurrent genomic
changes in lung cancer. The most frequently occurring
high-amplitude focal amplicons in lung cancer determined by
at least two studies are listed in Table 6.3. Among those, are
members of the MYC family ( MYCL1 , MYCN , and MYC ),
participants in the EGFR pathways ( EGFR , PIK3CA , KRAS),
and other genes, such as FGFR1 , TP63 , TERT , and the cyclins
CCND1 and CCNE1 . Some are potentially novel oncogenes
in lung ( NKX2-1 , for instance) that cooperate to promote lung
cancer cell proliferation.
The consolidation of the available data contributes to a
growing body of evidence that multiple cooperating oncogenes
participate in these amplification events in an apparently nonrandom
frequency. These findings have important implications
for the design of functional genomic studies projects aimed
at identifying cancer-relevant genes because single-gene assays
will not uncover activities that rely on interaction among multiple
collaborating genes.