"Gene Signatures Can Help Doctors Choose Best Treatment For Breast Cancer Patients"
LONDON, U.K. -- November 14, 2007 -- Gene signatures can be used to select the right treatment for patients with breast cancer, with the aim to give patients "tailored therapy" in the future, according to an Article to be published early Online and in the December issue of [The Lancet Oncology.
In the study, about 44% of the patients who received chemotherapy before surgery had a pathological complete response where the tumour completely disappeared. According to the authors if gene signatures had been used to select which type of chemotherapeutic drugs to give to the patients, the response rate would have been about 70%*.
The authors say: "This is such a big improvement that we need to do another study to make sure it is really true, but if it is confirmed it will make a big difference to the way doctors treat breast cancer".**
Systemic chemotherapy for the treatment of breast cancer improves overall survival. Newer chemotherapy regimens containing taxanes further improve survival compared with traditional regimens, but they are expensive, toxic, and might only benefit a small number of people, so decreasing the number of people receiving taxanes without compromising efficacy would be beneficial.
Professor Hervé Bonnefoi (Institut Bergonié, University of Bordeaux, France) and colleagues used tumour samples from patients in a large randomised multicentre trial. Patients were randomly assigned to chemotherapeutic treatments of a non-taxane regimen (fluorouracil, epirubicin, plus cyclophosphamide in the FEC group) or a taxane regimen (docetaxel followed by epirubicin plus docetaxel in the TET group).
The authors calculated how well patients would respond to the two different chemotherapy treatments***.
They found an excellent correlation between regimen-predictive gene signature and pathological complete response (defined as disappearance of the invasive component of the primary tumour, with at most a few scattered tumour cells detected by the pathologist in the resection specimen) in patients treated in the appropriate group.
"The treatment allocation model suggests that selection of the treatment regimen with our genomic signatures has the potential to increase the pathological response rate from 44% to around 70%", say the authors.
Furthermore, the negative predictive value (NPV) for each regimen-predictive gene signature was over 90%, thus identifying patients who are unlikely to achieve a complete disappearance of the tumour after FEC or TET treatment, which include the chemotherapy drugs most commonly used to treat breast cancer.
The authors note: "It would be sensible to offer these patients treatment with new drugs, by using the gene signatures to select patients for inclusion in clinical trials with new drugs or adding new drugs. By doing this, it should be possible to do much smaller trials".
The authors conclude: "If the results are confirmed in the follow-up study, then it would be well worth using our gene signatures to select the treatment for patients with breast cancer".
*The results only apply to a subgroup of patients with tumours that are insensitive to hormonal therapy -- about 30% of patients have this sort of tumour.
** All of the quotes are from the author directly and cannot be found in the text of the article.
*** The mathematical formula used was developed by statisticians in Joe Nevins' group at Duke University in North Carolina, USA. They used a special form of statistics called Bayesian statistics.
SOURCE: The Lancet
|
