World Congress on Gastrointestinal Cancer 14, Barcelona, Spain
Genetic signatures for colon cancer
Professor Pierre Laurent-Puig – University of Paris, France
In fact there are two main questions in colon cancer for the management of the patient. It’s to, let’s say, identify the stage 2 patient that will recur; in fact, 20% of all stage 2 patients have a disease that progresses to a metastatic disease and we have to identify them. The second main question is to identify those with stage 3 that do not recur. Approximately 50% of the patients are cured by surgery alone so we have to identify also this subgroup of patients to avoid inappropriate treatment.
What are some of the genetic differences you can use to identify those patients?
Probably a signature, a gene expression signature, is probably a good way to identify this type of patient.
How does gene expression differ between stage 2 and 3?
We’re not sure because, in fact, you have colon cancer and some stage 2 have recurrence and some stage 3 haven’t recurrence. What it has been already demonstrated by some publications is that some stage 3 have a better prognosis than some stage 2. So there is probably no difference in terms of signature between stage 2 and stage 3 but we have to identify the bad prognostic signature between the two types of cancer.
What are some of the methods currently available?
In fact there are already a lot of, let’s say, validated signatures because they have been identified on a training set and validated apparently on a second or a third set of patients. But at present what is missing is a clinical validation of these signatures because we have to perform a randomised trial using the signature and saying this patient needs chemotherapy or this patient doesn’t need chemotherapy and so see if there is any clinical benefit to having this information before the treatment.
What about the ability to reproduce the results of the current clinical trials?
These signatures seems reproducible, in fact it’s not the reproducibility of the signature which is important, it’s the fact that different signatures give the same information and there is some concern about that. If you compare different signatures already validated for the same patient, sometimes these patients will be classified as high risk of recurrence and sometimes a low risk of recurrence for the same patient with a different signature. So it raises some issues of the clinical use of these signatures. On the population based study, these signatures appear to be validated but at the individual level for a specific patient, probably we have to do some progress about that.
What future research needs to be done?
It’s to set up some clinical validation by randomised trials of these different signatures to be sure that they bring some new information as compared to already existing prognostic factors. The clear demonstration of the clinical value, added value, of these signatures is not so clear, even for breast cancer, so we are approximately on the same point but we need to set up clinical trials to clinically validate the information brought by these signatures.
Does this relate to any of your own research?
In fact we are looking for a signature that predicts specific response to a targeted therapy in colon cancer, especially for anti-GFO therapy. In fact, we have identified a micro-RNA which can predict the response to anti-GFO therapy in colon cancer for KRAS wild-type patients.
What targeted therapy are you using?
Cetuximab or panitumumab.
What is the take home message for these trials?
At the present time the only thing we can say is that this signature is a help in day-to-day decision making. But it’s not clinically validated.
