This is an international initiative that coordinated six co-operative groups across twelve countries that in a matter of about eight years has accrued 13,000 stage III colon cancer patients and randomised them to receive either three months or six months of adjuvant CAPOX or FOLFOX. The background lies in the fact that we know that FOLFOX chemo and its variant CAPOX affords about a 20% additional cure rate on top of what the surgery gives. The problem is that the neurologic toxicity of this combination is pretty heavy with about one patient out of two suffering from major, clinically relevant, temporary toxicity that may last, however, even years. There is a 4-10% of irreversible severe neurologic toxicity. So in the background certainly there was our attempt to decrease this neurologic toxicity and leaving the efficacy almost unchanged. That was the reasoning longer than ten years ago when we started this study.

What were your methods?

We generated a hypothesis. We predetermined the non-inferiority margin and we have chosen three year disease free survival as the primary endpoint of the study, knowing perfectly well that we’ll have looked with much interest also to the secondary endpoints, most important of them is the five year overall survival. So the study was already reported two years ago and published in The New England Journal of Medicine two years ago when the primary endpoint was reported, that is three year disease free survival. But that study generated a lot of controversies, debates, because of the discrepancy between the three year disease free survival curves that ran one on top of the other but the statistical analysis saying that non-inferiority could not be shown. So we were puzzled about that. Now it is so important, the latest report at ASCO, showing the secondary endpoint that is so important, the overall survival.

What were the results?

The results were as we expected, that is that six months was not better than three months, as indicated by the two Kaplan-Meier curves of survival being one on top of the other. They were so identical that you cannot even stick the light of a laser pointer in between the lines. So that led us to our immediate interpretation, saying ‘Debate is over. There are no more issues. Three months to everybody.’

Now, things are a bit more complicated than that because we predetermined also the non-inferiority margin for the five year overall survival and we just hit the point of non-inferiority so that the statistical analysis again failed to show non-inferiority although common sense and immediate reaction show identity between the two duration periods of adjuvant therapy. So, again, even the discussant at ASCO agreed with our interpretation, the interpretation of the authors, saying, okay, we are obviously playing the game of clinical research and that has to stick to rules. But, at the same time, when we practice medicine we have to use common sense and we have to counterbalance whatever you can lose, that may be a fraction of one or two percent with a threefold reduction in the occurrence of grade 2, 3 and 4 neuropathy. So the common conclusion is that three months is so much better than six months.

Not only that, there is a chapter two to this story and that is something unexpected. This is the nice aspect of clinical research – you start with the objective to show something and you end up opening up a new avenue. This new avenue was the unexpected interaction between the duration and the regimen type, meaning that these 13,000 patients were treated with FOLFOX but also with CAPOX, the oral version of the fluoropyrimidine. What was found both in the disease free survival and overall survival primary endpoint and secondary endpoint is that if you use CAPOX three months are okay, as good as six months, but if you use FOLFOX that is okay only if the risk within stage III is low. Because if you have a high risk, defined as either T4 or N2, then you have to use FOLFOX for six months if you stick to the fully IV regimen of FOLFOX. So that was the other conclusion of the study.

So, if we want to summarise and make things easy for our clinical practice, we can divide our patients into three groups, and I’m talking of stage III colon cancer. Number one, the low risk. These are 60% of our patients, T3 N1. These patients should get three months of chemo, either CAPOX or FOLFOX, preferably CAPOX. I think that giving six months of chemo is a mistake, that is so strong the conclusion in this first category of patients. The second category is the other extreme, that is the very high risk. Very high risk is qualified as T4 and N2. There is a third category that is in between and that is high risk and that is T4 or N2. But when you get T4 and N2 then the common way of interpreting these results and the subgroup analysis and so on is to give six months of adjuvant chemo, preferably CAPOX. The third category of patients, and that is about 20-30% of patients, so this category is just 10%, the very high risk. The high risk, 20-30% of patients should get either three months of CAPOX or if the patient is unwilling to risk losing even 1-2% then discuss at three months and if it is tolerable proceed and go on to six months.

By distinguishing the patients into the low risk, high risk and very, very high risk, I remind that very, very high risk T4 and N2 has less than 45-50% five year overall survival with chemo. Now, if we grant the chemo a 20-25% chance of long-term overall survival, that means that that category of very high risk patients has 20-30% chances of being cured by surgery alone.