NSABP-C-07: Phase III adjuvant study of 5-FU/leucovorin plus or minus oxaliplatin

Design and efficacy data

In the NSABP-C-07 trial, patients with Stage II and III colon cancer who had undergone resection with curative intent were randomly assigned to receive either the Roswell Park regimen of 5-FU/leucovorin given weekly for six out of eight weeks over approximately a six-month period or the same regimen plus oxaliplatin for three weeks during each eight-week cycle.

Data presented at ASCO 2005 showed that disease-free survival improved with the addition of oxaliplatin. The magnitude of benefit seemed quite comparable to that seen in the MOSAIC study, which led the presenter to conclude that this regimen is an acceptable adjuvant therapy for patients with Stage II or III colon cancer (Wolmark 2005; de Gramont 2005; [2.1]).

Toxicity data

The toxicity data from C-07 suggest that both arms had substantial Grade III and IV toxicities, but how those compare with a FOLFOX regimen or an infusional 5-FU-based regimen is uncertain at this point. In the C-07 trial, they used a lower total dose of oxaliplatin than MOSAIC, which is an important distinction. The long-term neurotoxicity with FOLFOX may be worse than that seen with the FLOX regimen used in C-07.

I believe we need more detailed information about toxicity before adopting the C-07 oxaliplatin regimen into routine clinical practice. The C-07 data do provide an alternative for patients in whom we would like to use an oxaliplatin- containing regimen but for one reason or another are not able to use infusional 5-FU. However, for most other patients, I am more comfortable using FOLFOX because we have longer follow-up and more data on acute toxicity with this regimen.

Rationale for comparing capecitabine to 5-FU

A Phase III study presented at ASCO 2005 compared capecitabine/oxaliplatin to 5-FU/oxaliplatin as first-line therapy in patients with metastatic colorectal cancer, and with early follow-up, it appears overall survival is very similar with these two regimens (Arkenau 2005; [2.2]). For patients who are clearly in a palliative mode, particularly for patients in whom surgical resection is not going to be feasible with curative intent, I believe that study lends some credence to the option of capecitabine/oxaliplatin rather than infusional 5-FU-based therapy.

The data evaluating capecitabine in combination with irinotecan and oxaliplatin suggest that we probably do need to conduct large-scale equivalence studies comparing those regimens to their infusional 5-FU counterparts. I believe that only the large-scale investigations will clearly delineate the toxicity differences between the capecitabine-based regimens and the 5-FU-based regimens, which will become important when selecting treatments for our patients.

Another reason for doing such comparison studies in patients with metastatic disease is to determine whether one of the approaches is superior in terms of major antitumor responses — the types of responses that might lead to resectability of liver metastases, for example, with curative intent. It’s plausible that only a small difference, if any, in overall survival exists between capecitabine and 5-FU-based combination regimens; however, if significant responses can be seen early with one regimen versus another, this might impact a small percentage of patients who ultimately can be cured.

ECOG-E5202: Adjuvant trial assigning treatment based on molecular phenotypes

The Eastern Cooperative Oncology Group is poised to activate a Phase III study, ECOG-E5202, in which patients will be assigned treatment based on molecular phenotype (2.3). All the patients with Stage II disease will have their tumors assessed for microsatellite instability and 18q deletions in real time. The patients whose molecular phenotypes suggest an extremely good prognosis will be followed for relapse and survival but will receive no adjuvant therapy. The patients who fall into the higher-risk group will then be randomly assigned to receive either FOLFOX or FOLFOX plus bevacizumab.

I believe it’s premature to use molecular diagnostics to select adjuvant therapy for colon cancer outside of a clinical trial. In this study, the analyses will be conducted by a CLIA-approved laboratory that is highly skilled in performing these studies and, in fact, was responsible for the initial observations related to microsatellite instability and 18q status as prognostic and predictive markers.

ECOG-E3200: FOLFOX with or without bevacizumab

Design and efficacy data

ECOG-E3200 sought to determine whether bevacizumab, when added to chemotherapy in the second- or third-line setting for metastatic disease, would improve survival. Eligible patients had previously received 5-FU and irinotecan either together or in sequential fashion but had not yet received oxaliplatin.

Initially, patients were randomly assigned to one of three treatment arms, either FOLFOX4 alone or FOLFOX4 plus bevacizumab or bevacizumab alone. However, the Data Safety Monitoring Board of ECOG closed the single-agent bevacizumab arm before completion of accrual because of its inferiority to the other arms, so just the two FOLFOX arms continued until the completion of the study.

The data presented at ASCO 2005 showed that FOLFOX plus bevacizumab improved survival when compared to the other two arms, with a median survival benefit of approximately two months, and the survival benefit was associated with an improvement in response rate (Giantonio 2005; [2.4]).

The FOLFOX plus bevacizumab arm had a response rate of approximately 20 percent, versus approximately 10 percent in the FOLFOX arm. The data also showed bevacizumab to have nominal activity, at least in terms of response rate, in patients with advanced colon cancer who had failed prior chemotherapy.

Clinical implications of ECOG-E3200

ECOG-E3200 does validate the concept that bevacizumab adds to the benefit of an oxaliplatin-containing regimen, just as it did with an irinotecan-containing regimen. For clinicians who may not have been comfortable using FOLFOX plus bevacizumab in the front-line setting because there were no data with that combination, those data now exist, and while they were second-line data, I believe they validate the benefit of adding bevacizumab to FOLFOX. If a physician prefers FOLFOX as their front-line chemotherapy backbone, they can now feel comfortable adding bevacizumab.

BOND-2 trial: Bevacizumab/cetuximab with or without irinotecan

The BOND-2 trial is a first effort in colon cancer to combine anti-VEGF and anti-EGFR agents by combining bevacizumab and cetuximab in patients with metastatic colorectal cancer (2.5). Patients who had previously received irinotecan and had progressive disease were randomly assigned to either receive irinotecan again with both cetuximab and bevacizumab or cetuximab/bevacizumab without further chemotherapy (Saltz 2005). It’s important to note that the patients on this study had not received cetuximab or bevacizumab previously.

The response data in both arms of this trial appear to be better than one might have expected, based on historical controls with cetuximab monotherapy or cetuximab plus irinotecan in this type of pretreated patient population. The BOND-2 study validates this combined approach in terms of moving it into large-scale randomized studies to truly determine whether the combination of bevacizumab plus cetuximab is superior to sequential therapies containing one of these antibodies, followed by another regimen containing the other.

Select publications

Dr Meropol is the Director of the Gastrointestinal Cancer Program and Director of the Gastrointestinal Tumor Risk Assessment Program in the Divisions of Medical Science and Population Science at Fox Chase Cancer Center in Philadelphia, Pennsylvania.