Stage III Colorectal Cancer: After IDEA Trial, Which Patients Still Need 6 Months of Adjuvant Chemotherapy?

Alberto Sobrero, MD; Julien Taieb, MD, PhD


November 16, 2017

Alberto Sobrero, MD: Hello. I am Alberto Sobrero. I am a medical oncologist and head of the Medical Oncology Unit at the San Martino Hospital, in Genoa, Italy. In addition, I was the scientific chair of the European Society for Medical Oncology (ESMO) 2017 Congress. Welcome to Medscape Oncology. I am speaking from the 2017 Congress in Madrid. Today, we will discuss an important, clinically relevant issue—the duration of adjuvant chemotherapy for stage III colon cancer.

Joining me in this discussion is Dr Julien Taieb, a gastrointestinal oncologist at the Hôpital Européen Georges-Pompidou in Paris, France. Julien, the IDEA collaboration has been the largest combination of trials that has ever been conducted up to this point in this disease. Can you make a general comment on the trials?

Julien Taieb, MD, PhD: This collaboration was a fantastic clinical research experience for all of us. IDEA began with the idea that reducing the duration of adjuvant therapy from 6 to 3 months may be sufficient for the patient with stage III colon cancer. This is very complex because the noninferiority hypothesis of such a trial would have required at least 10,000 patients, and no country is able to do that alone.

Thus, we set up a wonderful international collaboration that included Italy, the United Kingdom, France, the United States, Japan, and Greece. We succeeded in enrolling more than 12,000 patients.[1,2]

"At ASCO, we walked out quite confused."

Dr Sobrero: The combination of trials essentially led to three findings. First, we tested the noninferiority hypothesis; that is, 3 months of adjuvant treatment may be as effective as 6 months of treatment. We found a small difference. The two additional findings were that the stage within stage III—that is, risk—counts a lot; and the regimen, either FOLFOX [leucovorin, 5-fluorouracil, and oxaliplatin] or CAPOX [capecitabine and oxaliplatin], also counts a lot in this 3 months vs 6 months comparison.

This was a surprise that complicated the interpretation of this trial, so that in Chicago at the American Society of Clinical Oncology (ASCO) meeting, at the end of the plenary session on IDEA, we walked out quite confused. I understand this because the data interpretation is always very difficult, especially when the differences are not so large. What was your impression of this confusion that was generated in Chicago?

Dr Taieb: Sometimes statistics do not match your clinical experience, and this is one of the primary things we have learned with IDEA. It is also important to know that the chemotherapy regimen, CAPOX or FOLFOX, was the "dealer's choice"; treatment was not randomized. We did not expect to find any difference between these regimens. But, in reality, we have had quite surprising findings at the end.

Dr Sobrero: Let's go to the main findings. First, treatment for 3 months is much less toxic than for 6 months. How much? We found two to six times lower neurotoxicity, about 20% to 30% less diarrhea, and two to three times less hand-foot syndrome and mucositis with 3 months vs 6 months of treatment.

Second, at 3 years, there was a difference in disease-free survival (DFS) of 0.9% overall, slightly favoring the 6-month regimen. What is your reaction to these two pieces of information—lower toxicity, minimal difference in DFS?

Dr Taieb: I would say that this would be a strong argument to reduce the duration of therapy. The problem is that, as clinicians, we do not want to harm any patients. When we went into more depth in the subgroups of patients and the different regimens, we found different things.

Dr Sobrero: During the last 5 to 6 months, within the IDEA collaboration, we have exchanged more than 1500 emails in order to interpret these data. Things were being complicated by a typical medical variable, and that is the patient. Sometimes we run into patients who are not willing to sacrifice even a 1% advantage. Neil Love showed this about 10 years ago, that 30% of patients are not inclined to lose even 1%. We can call these the fighters. Then we have a second group of patients, and these are those who may look immediately at how much they pay in terms of toxicity. How much do I gain or lose? We can call these the fatalists. Would you agree, as a clinician, on this gross characterization of attitudes of patients?

Dr Taieb: Yes. It is like that for sure, but the problem, in my opinion, is that when you share the medical decision with your patient, often the last question the patient asks is: What would you do for your father or your mother? You spend a lot of time, and finally the conclusion for the patient is, emotionally and medically, what do you, the doctor, recommend to me? It is true that we can define different patient patterns, but I am not convinced that it will solve all of our daily practice issues.

Dr Sobrero: Right. On one hand, we can interpret the whole data in this way. Very simple. Much less toxicity, 0.9% difference in relapse-free survival at 3 years. Why bother with a longer regimen? Period. Why bother?

Dr Taieb: Because all patients are not the same, and also because we know that although this 0.9% may be nothing, clinically speaking, statistically the study was finally negative because the confidence interval margin was wider than what we expected, meaning that we did not agree to lose more than half of what oxaliplatin brought to our patients.

Technically and statistically, that is something to discuss.

And we also must characterize more pragmatically the subgroups that may still need to continue therapy to 6 months. The question today is not, "Is 6 months still the standard?" The question is, "In which patients do we have to continue for 6 months?"

Dr Sobrero: That is the point; that is the complicating issue. Regarding the issue of the stage within stage III, we can essentially recognize low-risk stage III—the T1, 2, 3, N1; fewer than three lymph nodes—or high-risk stage III that is either N2 or T4, any N. I believe this distinction was a byproduct of IDEA because, up until the IDEA publication and presentation, this differentiation was not made. Prognostically, can you recall the percentages?

Dr Taieb: This is very important. Globally, patients with T4 or N2, in the last trials in IDEA, are 35% to 48%, [similar to] PETACC-8, N0147,[3] and other trials. The prognosis is totally different. With T4 or N2, DFS is 65%. The other subgroups are closer to 80% or 90%. They are completely different groups. It makes sense because we are used to having high risk and low risk in stage II. And the benefit of the addition of oxaliplatin has been described. Even in PETACC-8, we do see that some of this T4 or N2 population was completely different from the others and may benefit from more aggressive chemotherapy.

Dr Sobrero: Perhaps for the audience, a rule of thumb could be this, in terms of prognosis: Moving from T1 to T2, you lose 10% of 3-year DFS; going from T2 to T3, you lose 10% again, so 20% loss from T1 to T3. If you go from N1 to N2, you lose 20%. If you go to T4, no matter what, you lose an additional 20%-25% from T3. T4 is the most negative prognostic factor. Then comes N2, and then within the T category, you have 10% loss between T1 and T2, and T2 and T3.

This is important because this may highlight the first condition that we have to consider in terms of 6 months or 3 months of adjuvant treatment within these risk groups. What is your interpretation?

"The results clearly show that in the low-risk stage III groups, 3 months of adjunctive therapy is enough."

Dr Taieb: For me, the results clearly show that in T1, T2, T3, and N1—the low-risk stage III groups—3 months of adjuvant therapy is enough. The results are very robust for CAPOX, but I believe that the curve—the shape of the curves following one to each other, not separating, over time—is also good for FOLFOX. I would not be afraid to stop at 3 months for all of these patients clinically, tomorrow, in my daily practice.

Dr Sobrero: Let me interrupt you at this point. We had a special session at ESMO with 11 discussants of these conditions. As chair of that session, I took a vote of presenters about these conditions, and every single discussant—from Japan, the Americas, and Europe—voted the same, exactly what you said. This is the first, very strong, message that is coming out of this meeting. Let's move to the high-risk population.

Dr Taieb: For the high-risk patient, there are more complexities than for the low-risk patient because if you choose FOLFOX, clearly, all of these studies go the same way. High-risk patients who take FOLFOX for 6 months do better than if they took FOLFOX for 3 months. The problem is that because the treatment was the dealer's choice during the IDEA consortium, some countries used CAPOX for 10% of patients, some countries used it for 90%, and the same for FOLFOX. Thus, there is an imbalance between countries in the proportion of patients receiving one or the other regimen.

Once again, these regimens were not randomized, so theoretically they cannot really be compared. But for CAPOX in high-risk colorectal cancer, it seems that 3 months is almost as good as 6 months, although the statistical result is not as robust as for low-risk patients. In my practice, I would probably push to 6 months to be sure I wasn't removing any choice for my patient. But I would stop oxaliplatin very quickly [if needed] to avoid inducing long-lasting neuropathy in these patients.

Dr Sobrero: Right. This is getting complicated. Now, of course, as we move to subgroup analysis, we lose power, and so again, we are shifting away from the "liturgy" of statistics and jumping into an extrapolation for the clinic. If you look at the DFS curves for 3 vs 6 months within the high-risk patient population, you look at the high-risk curves for CAPOX in patients with N2, they are absolutely the same. If you look at the Kaplan-Meier curves for 3 vs 6 months in the T4 population, at some point they tend to diverge. But you made this point in one of these 1500 emails that we exchanged: The data are still a bit too early to commit ourselves to saying that for patients with N2, duration makes no difference at all; and for T4, maybe it makes a difference later on.

Dr Taieb: With CAPOX, I believe that a longer follow-up will be a major help in the future, if we want to recommend 3 months for everyone. For today, we have to stick with the 3 months for low-risk patients and 3-6 months for high-risk, depending on the regimen and the fighter profile of your patient. In the future, we will be able to make a final conclusion when we have all these data that are still missing.

Dr Sobrero: Somehow we have returned to the beginning and the fact that in front of you when you practice, you have different people who may judge the benefit and the cost in a totally different way. So take this conclusion with a grain of salt.

If the patient is a fatalist, meaning that he doesn't look at the 1%-2% potential loss, then no matter what, 3 months is okay. This applies even in the higher-risk group; 3 months of CAPOX, not FOLFOX, because 3 months of FOLFOX was clearly inferior to 6 months in the French study.[4] So again, to simplify things, CAPOX 3 months for all of the fatalists.

If the patient is a fighter and doesn't want to lose even 1%-2%, then if the patient is low risk, 3 months of CAPOX. If he's high risk, 6 months of CAPOX or FOLFOX. That is a fairly clear message that I believe we agree on at this meeting.

Thanks very much, Julien, for discussing this controversy and contributing to the discussion.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: