'Gene signature' predicts chemotherapy survival

Scientists have identified an eight-gene 'signature' that can predict patients' response to chemotherapy, according to a study published in journal BMC Medicine.

The researchers, from Academia Sinica and the National Taiwan University College of Medicine, looked at the effect of 99
different chemotherapy drugs on 60 cancer cell lines, representing leukaemia, melanoma, and breast cancer, among
others.

They looked at the pattern of activation of these genes in
each of the cell lines and found that eight genes in particular influenced the
cells' response to the drugs.

Using this 'gene signature', the researchers were able to
predict which people would have a better chance of surviving after chemotherapy
without relapse. The researchers calculated whether patients had an
above-average risk of relapse following chemotherapy by studying which of these
eight genes were expressed. Patients with a low score were found to have longer
relapse-free survival times after chemotherapy.

Professor Ker-Chau Li, who
conducted the research, said: 'Our study found eight genes which were involved
in invasion, and the relative activation of these genes correlated to
chemotherapy outcome'.

The finding hints at the possibility of personalised cancer
treatment, as people who are classified as 'high-risk' can receive more
targeted therapy. Having a better understanding of the genetic makeup of the
tumours means that the most effective chemotherapy drug can be used.

The researchers made their predictions in people with breast
and lung cancer, but they hope their findings can be extended to other forms of
cancer.

Professor Pan-Chyr Yang of
National Taiwan University said: 'The eight-gene signature obtained here may
help choice of treatment as part of individualised cancer therapy and our
method of gene discovery may be applicable in studying other cancers'.

The researchers acknowledged that the cancer cells in the
body may react differently to the cell lines they tested, saying: 'There is room to improve our eight-gene signature
for drug-sensitivity prediction [...] Other genomic components, such as
DNA copy number, single nucleotide polymorphisms, methylation and microRNA,
have not been considered in our study'.