Do microarrays improve breast cancer prognosis? A long story short
Haibe-Kains B
Breast cancer is a global
public health issue. It is the most frequently diagnosed malignancy in
women in the western world and the commonest cause of cancer death in
European and American women. In Europe, one out of eight to ten women,
depending on the country, will develop breast cancer during their
lifetime.
During the last two decades, several clinical and pathological
indicators such as histological grade, tumor size and lymph node
involvement have been used for the survival prediction of breast cancer
patients independently of treatment, also known as prognostication.
Examples of clinical guidelines to the selection of patients who should
receive adjuvant therapy are the St Gallen consensus criteria
(Goldhirsh et al., JCO 2003), the NIH guidelines (Eifel et al., JNCI
2001), the Nottingham prognostic index (Galea et al., BCRT 1992) and
Adjuvant! Online Olivotto et al., JCO 2005). Although BC
prognostication has been the object of intense research, a still open
challenge is how to detect patients who need adjuvant systemic therapy.
The advent of array-based technology and the sequencing of the human
genome brought new insights into breast cancer biology and prognosis.
Several research teams conducted comprehensive genome-wide assessments
of gene expression profiling. Perou et al. showed for the first time
that, not only breast cancer exhibits different clinical outcome, but
these tumors are also heterogeneous at the molecular level (Perou et
al, Nature 2000). Other groups identified prognostic gene expression
signatures. Examples of gene signatures which were obtained by studying
the relationship between gene expression profiles and clinical outcome,
are the 70-gene (van't Veer et al., Nature 2002) and the 76-gene (Wang
et al., Lancet 2005) signatures. Another example of gene signature is
reported in (Sotiriou et al., JNCI 2006). This signature, called GGI,
was defined to characterize at the molecular level the histological
grade, a well-established pathological indicator rooted in the cell
biology of breast cancer. With respect to clinical guidelines, these
signatures were shown to correctly identify a larger group of low-risk
patients not needing treatment. This is particularly relevant for
clinicians since reducing treatments means also reducing potential side
effects and cutting costs.
Other research groups have proposed gene expression signatures that are
predictive of the clinical outcome in breast cancer (Sotiriou et al.,
Nature Cancer Reviews 2007). Although several criticisms arose
concerning the overoptimistic results of early publications (Michiels
et al., Lancet 2005; Ein-Dor et al. Bioinformatics 2005), recent
validation studies (Buyse et al. JNCI 2006; Desmedt et al., CCR 2007;
Haibe-Kains et al. BMC Genomics 2008) supported the good performance of
some prognostic signatures (namely the 70-, 76-gene and GGI). However,
our group recently showed (Desmedt et al. CCR 2008; Wirapati et al.,
BCR 2008; Haibe-Kains et al., Bioinformatics 2008) that proliferation
is the main driving force of these signatures and that their prediction
value is limited to one subtype of tumor (luminal). Current research
focuses on prognostic gene signatures for the other breast cancer
molecular subtypes (basal, her2-positive). It is worth to mention that
the microarray technology is now mature enough to be used in future
clinical applications as showed by the MAQC group (Shen et al., Nature
Biotechnology 2006).
In conclusion, thanks to the validation of the microarray technology
and the recent meta-analyses and reviews which successfully
recapitulated the main discoveries made these late decades, we benefit
from this strong basis to go a step further to improve breast cancer
prognosis using microarrays.