Genomic sequencing helps identify best candidates for adjuvant chemo in colorectal cancer

Share :
Published: 1 Jun 2014
Views: 3865
Dr Ramon Salazar - L'Hospitalet de Llobregat, Barcelona, Spain

Dr Salazar talks to ecancertv at ASCO 2014 about how colorectal cancer patients that respond better to treatment without adjuvant chemotherapy can be better identified.

ecancer's filming at ASCO has been kindly supported by Amgen through the ECMS Foundation. ecancer is editorially independent and there is no influence over content.




PARSC is a prospective validation study of a prognostic signature previously developed called ColoPrint, developed by the company named Agendia from the Netherlands. We had previously bvalidated the power of this signature to classify the risk of relapse in stage 2 colon cancer patients and in this study we are trying to prospectively validate the results that separate the relapse rates in the high signature from the low signature tumours.

How did you compare?

We developed an eighteen gene signature based on the different recurrence status of the patient. So we initially, previously to the PARSC study, developed this signature performing genomic analyses of those patients that relapsed and of those patients that did not relapse. The genes that differentially defined these two groups of patients were validated in a prognostic signature called ColoPrint which is the signature that we are now using in the PARSC study.

What did you find?

There are two groups, high risk and low risk signature. Those patients with a stage 2 with a low risk signature have an excess of 90% non-recurrence rate; those are the low signatures. Those with a high risk signature, they have 70% recurrence rate. So we are not going to consider treatment with adjuvant chemotherapy in the low risk patients but we could consider adjuvant chemotherapy in the high risk patients.

To validate this you compared with the NCCN classification system. How did it compare to that?

First of all I have to say that the concordance is almost nil; it’s 50% concordance so it means that there is no association between the clinical classification and the genetic classification. That means that the genetic classification adds information to the model. And in the previous retrospective series the genetic classification is more accurate to predict recurrence than the old fashioned clinical classification, the NCCN clinical classification.

How do you think this can be applied by cancer doctors everywhere?

Stage 2 colon cancer patients always pose a conflict. We don’t know whether to treat them or not to treat them with adjuvant chemotherapy because historic trials of adjuvant chemotherapy have yielded very marginal benefit to these patients. So if we could split this population into those with very low risk of relapse, we do not consider chemotherapy on those, and on those who have a standard risk of relapse, like 30%, 35%, risk of relapse, we can discuss with the patient the benefit of treating.

What is the take home message for cancer doctors coming out of this study?

If this study validates our previous findings we will have yet another good tool, prognostic gene signature, to help us discern which patients can benefit from adjuvant chemotherapy.

Are you recommending practice change now?

The PARSC study has just finalised and we will have the final results in the next year. So at this point I have to hold the conclusions because I would not recommend using a gene signature based only on retrospective clinical data validation. The beauty of this study is that it is a prospective study and if this result is positive I will recommend this practice changing.