According to Alberto Sobrero, MD, Head of Medical Oncology at Ospedale San Martino in Genova, Italy, the results of the pivotal IDEA trial, which evaluated the optimal duration of adjuvant chemotherapy for stage III colon cancer, were not clear to clinicians.1 The combined analysis of six individual studies asked whether 3 months of adjuvant oxaliplatin-based therapy was as effective as, and less toxic than, 6 months of treatment.
“The answer to the question was not yes or no. The outcome was much more nuanced. We walked out of the  ASCO Plenary Session without knowing if the patient should get 3 months, or with which regimen, or which patients need to continue with 6 months,” he said. At the European Society for Medical Oncology (ESMO) 2017 Congress, Dr. Sobrero moderated a special session that promised to clarify the results for clinicians.2
The Big IDEA
The IDEA trial enrolled 12,834 patients with stage III colon cancer into 6 studies conducted in 20 countries. Of them, 7,763 received FOLFOX (leucovorin, fluorouracil, oxaliplatin) and 5,071 received CAPOX (capecitabine, oxaliplatin) by physician choice, randomly assigned to receive 3 or 6 months of adjuvant therapy.
Clinically speaking, we have two populations— the fighters and the fatalists—and we always have to adjust our choices to this reality.— Alberto Sobrero, MD
Tweet this quote
IDEA was designed as a noninferiority trial in which 3 months of therapy would not sacrifice more than 12% of the benefit of the standard 6 months. Statistically, the upper 95% confidence interval (CI) of the hazard ratio (HR) could not exceed 1.12, to prove noninferiority in 3-year disease-free survival.
Although the shorter regimen was associated with a dramatic reduction in neurotoxicity, the efficacy endpoint was not met: 3-year disease-free survival was 74.6% with 3 months of chemotherapy and 75.5% with 6 months (HR = 1.07; 95% CI = 1.00–1.15). The investigators’ recommendations were that the duration of adjuvant chemotherapy for stage III patients should be risk-based and that 3 months of treatment is acceptable for low-risk patients.
Nuances by Risk Group
At the ESMO Congress, Axel Grothey, MD, of the Mayo Clinic Cancer Center, Rochester, Minnesota, presented more granular details of the analysis, and Dr. Sobrero led the investigators in a panel discussion of the findings.
Some differences were observed according to the regimen. A greater compromise in efficacy (though still small) was observed in FOLFOX-treated patients receiving 3 months of treatment, whereas noninferiority was proved for 3 months of CAPOX. With FOLFOX, the disease-free hazard ratio for 3 months’ duration was 1.16 (95% CI = 1.06–1.26), whereas with CAPOX, the hazard ratio was 0.95 (95% CI = 0.85–1.06), suggesting the shorter CAPOX regimen may, in fact, be preferred.
Axel Grothey, MD
The level of recurrence risk also mattered and was the topic of discussion among the investigators. “Tumor (T) and nodal (N) stage were grouped together to make a pragmatic choice between low-risk (T1–3 N1) and high-risk (T4 and/or N2) patients. We observed a large difference—about 20% at 3 years—between these cancers,” Dr. Grothey pointed out.
“For the two regimens combined, low-risk patients can consider 3 months of treatment, but this is mainly due to the effect in the CAPOX-treated group. The P value interaction test was positive for the regimen (P = .0061) and not the risk group (P = 0.11),” Dr. Grothey noted. “Overall, for all patients treated with CAPOX, there was noninferiority independent of the risk group. For low- and high-risk patients combined, FOLFOX for 3 months was inferior to 6 months.... The noninferiority margin was clearly crossed.”
Dr. Sobrero suggested that in clinical practice, however, the findings are not so simple to apply. One reason is that “even very small differences count to patients,” he said. Surveys have shown that about one-third of patients would choose a longer, more toxic treatment to avoid sacrificing even a 1% to 2% chance of cure. Others would sacrifice more for shorter, easier treatment.
“Clinically speaking, we have two populations—the fighters and the fatalists—and we always have to adjust our choices to this reality,” Dr. Sobrero commented.
A second sticking point is that, within the stage III population, there are different levels of recurrence risk, and this impacts outcome and efficacy of the regimens. Patients with T4 N2 disease fared worse than patients with T1–3 N1 disease, and “unexpectedly, we found the regimen counts,” he added. With CAPOX, 3 months performed as well as 6 months, but with FOLFOX, 6 months of treatment was necessary.
We don’t need to worry about N2 patients. They have clear-cut noninferiority with 3 months of CAPOX. It’s T4 disease where there’s more uncertainty.— Tim Maughan, MD
Tweet this quote
Although T4 and N2 patients were grouped together in the IDEA analysis, these subsets may represent different tiers of risk, the investigators proposed. In a separate presentation, Tim Maughan, MD, Professor of Clinical Oncology at the University of Oxford in the United Kingdom, suggested that the difference between the two regimens was most likely due to the longer term exposure to fluoropyrimidines with the CAPOX regimen, as patients were on active therapy for 14/21 days per cycles rather than for 2/14 with FOLFOX. He agreed that grouping T4 and N2 patients together “confounded” the analysis, as there was no difference in the effect of duration of therapy between N1 and N2 patients. For N1–2 groups combined, if separated by T stage, the hazard ratio was 1.04 for T1–3 disease (95% CI = 0.96–1.13), meaning noninferiority remained unproven. The difference comes with the T4 patients, where the hazard ratio increased to 1.16 (95% CI = 1.03–1.31), clearly showing inferiority for 3 months of treatment and setting T4 patients apart.
“We don’t need to worry about N2 patients,” Dr. Maughan concluded. “They have clear-cut noninferiority with 3 months of CAPOX. It’s T4 disease where there’s more uncertainty.” Based on hazard ratios for the two subsets, the potential loss of efficacy is much greater over time for T4 patients, potentially up to 5% at 4 years, Dr. Sobrero estimated.
Individual Preferences in High-Risk Patients
John Souglakos, MD, of the University of Crete in Greece, pointed out that some T4 patients probably already have systemic disease and therefore warrant aggressive therapy. Timothy Iveson, MD, of the University Hospital Southampton NHS Foundation Trust in the United Kingdom, concurred: “We know if these patients relapse, they have very poor long-term outcomes. The evidence is not convincing to me that 3 months [of CAPOX] is noninferior to 6 months, and I would be cautious about that.”
Timothy Iveson, MD
A clear preference for 6 months of FOLFOX in high-risk patients was expressed by Julien Taieb, MD, PhD, of -Georges Pompidou European Hospital Paris in France. “With -FOLFOX, we can control what the patient receives, which is not the case with CAPOX. That’s why it’s my preference in high-risk patients,” he said. He treats both T4 and N2 patients -aggressively.
With FOLFOX, we can control what the patient receives, which is not the case with CAPOX. That’s why [FOLFOX is] my preference in high-risk patients.— Julien Taieb, MD, PhD
Tweet this quote
Roberto Labianca, MD, of the Ospedale Papa Giovanni XXIII in Bergamo, Italy, agreed with this choice. “In Italy, we are more confident with FOLFOX, and I like to give my patient a more experienced treatment.” He treats both T4 and N2 patients with 6 months of FOLFOX.
Dr. Grothey called for distinctions between T4a and T4b disease, and said this analysis is forthcoming.
Putting It All Together
With CAPOX, 3 months is as good as 6 months, especially in low-risk patients, and is associated with much less neurotoxicity. With FOLFOX, 6 months adds extra benefit in terms of disease-free survival, especially in high-risk patients. Therefore, the investigators offered the following recommendations, according to Dr. Grothey, with some leeway to accommodate individual tolerances:
As the study’s invited discussant, Dr. Maughan said the findings validate his preferred use of CAPOX, and 3 months of this regimen will become his standard approach. While cautioning that IDEA did not randomize patients to one regimen or the other, he maintained there is “convincing evidence” for FOLFOX prescribers to switch to CAPOX.
Roberto Labianca, MD
For patients not fit for or not tolerant of CAPOX, however, when FOLFOX must be used, then it should be for 6 months’ duration. And for “highly motivated patients looking for every percentage point increase in efficacy,” with T4 disease, he will discuss “the uncertainty of the added benefit” of 3 additional months of treatment, as there remains uncertainty regarding the duration, even with CAPOX in these T4 patients.
“The key benefit of shortened therapy is reduced toxicity for patients, especially the neurotoxicity. There is also a great benefit to cash-strapped health-care systems, as shortening the duration of adjuvant therapy in stage III colon cancer would yield an economic savings in Europe of more than half a billion Euros a year, estimated Dr. Maughan. ■
DISCLOSURE: Drs. Sobrero, Grothey, Maughan, Taieb, Iveson, and Labianca reported no conflicts of interest.
1. Shi Q, Sobrero AF, Shields AF, et al: Prospective pooled analysis of six phase III trials investigating duration of adjuvant oxaliplatin-based therapy (3 vs 6 months) for patients with stage III colon cancer: The IDEA (International Duration Evaluation of Adjuvant Chemotherapy) collaboration. 2017 ASCO Annual Meeting. Abstract LBA1. Presented June 4, 2017.
2. Cervantes A, Sobrero A, co-chairs: When clinical practice demands to go beyond statistics: Adjuvant chemotherapy of colon cancer: The 3 vs 6 month story. ESMO 2017 Congress. Special Session. Presented September 11, 2017.