Is 7-Gene Signature a New Paradigm for CRC Patients?
Is 7-Gene Signature a New Paradigm for CRC Patients?
Hello. I'm David Kerr. I'm Professor of Cancer Medicine from the University of Oxford and President of the European Society of Medical Oncology (ESMO). I have just come back from the ESMO World Gastrointestinal Cancer Congress in Barcelona, where some really some fascinating data were presented. We have a new paradigm as to how we may select chemotherapy for patients with stage 2 colon cancer.
Let me declare a conflict of interest in that the presentation I'm discussing was made by me. I had an educational study grant from Genomic Health to help me complete the work that I am about to describe.
We worked in perhaps one of the most effective transatlantic cooperations we have had in the field of gastrointestinal cancer, with the National Surgical Adjuvant Breast and Bowel Project, great friends of mine, Mike O'Connell, Norman Wolmark, colleagues from the Cleveland Clinic, and from our QUASAR [Quick and Simple and Reliable] adjuvant trials group in the United Kingdom. We studied 3 and a half thousand patients with early-stage colon cancer. We developed a 7-gene prognostic signature. The technique was recovery of small fragments of RNA from paraffin-embedded colon cancer blocks. Then using polymerase chain reaction (PCR), we identified a multigene RNA signature, which is prognostic, and predicts the likelihood of the patient's cancer relapsing following curative surgery. Therefore, it may allow us to decide which patients would benefit most from chemotherapy.
The largest trial examining the role of chemotherapy vs best supportive care for stage 2 colon cancer was the QUASAR trial, results of which were published Lancet a few years ago. In that study, we were the first group ever to show, in a trial that recruited almost 3500 patients, that chemotherapy conferred a small, but important and statistically significant survival benefit of around 3%-4%. Clearly this means that we have to over-treat to cure that minority of patients. Therefore, if we had a tool that would allow us to select those at high risk for recurrence, then those are the patients in whom we would be more inclined to make a stronger recommendation for 6 months of fluoropyrimidine-based chemotherapy. It does look as though the Genomic Health tool, this multigene RNA signature, Oncotype DxColon, seems to do the trick.
In a validatory QUASAR study in 1500 patients, we showed that there was a very clear statistical link between recurrence score and the chance of the cancer returning. In multivariate analysis, the factors that came out as being prognostically relevant were T4 stage (patients with advanced disease), usually around 10% -15% of stage 2 cancers, and microsatellite mismatch repair (MMR)-deficient tumors, also around the 10%-15% mark. These have a very good prognosis and I would not be inclined to offer those patients chemotherapy because although proportional benefits are seen, the absolute benefits in the group with MMR-deficient disease are so small as to be not worth the cost, toxicity, or societal upset to patients.
The third factor that entered the multivariate model was the recurrence score --the multigene signature. For those 75% of patients with stage 2 and T3 disease, they are MMR proficient, and their recurrence scores define a gradient of risk of around twofold. Therefore, at the top end of the recurrence score, patients have approximately 20%-25% chance of cancer recurrence, compared with around 10% at the other end. Therefore, a recurrence score of 2 will be a useful additional discriminant.
In context, if we use T4 stage, MMR status, and recurrence score, we can define populations of patients with stage 2 colon cancer who do not require chemotherapy and those who are at higher risk, in whom we feel that we would make a stronger recommendation for chemotherapy. This is a good example of personalized medicine, a great example of transatlantic cooperation, and it shows the size of the study you have to do to be able to validate one of these novel molecular tools.
In terms of developing the assay and validating it in QUASAR, a 3500-patient study was required to give sufficient statistical power to validate this tool with a degree of certainty. We are changing the molecular paradigm for treating colon cancer and recurrence score has a useful role to play. Thank you.
Hello. I'm David Kerr. I'm Professor of Cancer Medicine from the University of Oxford and President of the European Society of Medical Oncology (ESMO). I have just come back from the ESMO World Gastrointestinal Cancer Congress in Barcelona, where some really some fascinating data were presented. We have a new paradigm as to how we may select chemotherapy for patients with stage 2 colon cancer.
Let me declare a conflict of interest in that the presentation I'm discussing was made by me. I had an educational study grant from Genomic Health to help me complete the work that I am about to describe.
We worked in perhaps one of the most effective transatlantic cooperations we have had in the field of gastrointestinal cancer, with the National Surgical Adjuvant Breast and Bowel Project, great friends of mine, Mike O'Connell, Norman Wolmark, colleagues from the Cleveland Clinic, and from our QUASAR [Quick and Simple and Reliable] adjuvant trials group in the United Kingdom. We studied 3 and a half thousand patients with early-stage colon cancer. We developed a 7-gene prognostic signature. The technique was recovery of small fragments of RNA from paraffin-embedded colon cancer blocks. Then using polymerase chain reaction (PCR), we identified a multigene RNA signature, which is prognostic, and predicts the likelihood of the patient's cancer relapsing following curative surgery. Therefore, it may allow us to decide which patients would benefit most from chemotherapy.
The largest trial examining the role of chemotherapy vs best supportive care for stage 2 colon cancer was the QUASAR trial, results of which were published Lancet a few years ago. In that study, we were the first group ever to show, in a trial that recruited almost 3500 patients, that chemotherapy conferred a small, but important and statistically significant survival benefit of around 3%-4%. Clearly this means that we have to over-treat to cure that minority of patients. Therefore, if we had a tool that would allow us to select those at high risk for recurrence, then those are the patients in whom we would be more inclined to make a stronger recommendation for 6 months of fluoropyrimidine-based chemotherapy. It does look as though the Genomic Health tool, this multigene RNA signature, Oncotype DxColon, seems to do the trick.
In a validatory QUASAR study in 1500 patients, we showed that there was a very clear statistical link between recurrence score and the chance of the cancer returning. In multivariate analysis, the factors that came out as being prognostically relevant were T4 stage (patients with advanced disease), usually around 10% -15% of stage 2 cancers, and microsatellite mismatch repair (MMR)-deficient tumors, also around the 10%-15% mark. These have a very good prognosis and I would not be inclined to offer those patients chemotherapy because although proportional benefits are seen, the absolute benefits in the group with MMR-deficient disease are so small as to be not worth the cost, toxicity, or societal upset to patients.
The third factor that entered the multivariate model was the recurrence score --the multigene signature. For those 75% of patients with stage 2 and T3 disease, they are MMR proficient, and their recurrence scores define a gradient of risk of around twofold. Therefore, at the top end of the recurrence score, patients have approximately 20%-25% chance of cancer recurrence, compared with around 10% at the other end. Therefore, a recurrence score of 2 will be a useful additional discriminant.
In context, if we use T4 stage, MMR status, and recurrence score, we can define populations of patients with stage 2 colon cancer who do not require chemotherapy and those who are at higher risk, in whom we feel that we would make a stronger recommendation for chemotherapy. This is a good example of personalized medicine, a great example of transatlantic cooperation, and it shows the size of the study you have to do to be able to validate one of these novel molecular tools.
In terms of developing the assay and validating it in QUASAR, a 3500-patient study was required to give sufficient statistical power to validate this tool with a degree of certainty. We are changing the molecular paradigm for treating colon cancer and recurrence score has a useful role to play. Thank you.
