Phase III trial of IFL with or without bevacizumab

Trial background

In the initial Phase I studies of bevacizumab, very minimal toxicity and favorable pharmacokinetics were observed — as expected for a monoclonal antibody. The Phase II studies of bevacizumab in conjunction with 5-FU/leucovorin showed a remarkably higher response rate, time to progression and a trend for improvement in survival than 5-FU/leucovorin alone (1.1).

A study reported by Fairooz Kabbinavar in the Journal of Clinical Oncology demonstrated that adding bevacizumab, which had a relatively modest safety profile, to 5-FU/leucovorin accounted for a marked increase in clinical activity.

During the design of the current Phase III study, the standard of care for colorectal cancer in the United States had just evolved once more from 5-FU/ leucovorin to bolus IFL, also known as the Saltz regimen.

That’s why our study initially had three arms: IFL plus placebo, IFL plus bevacizumab, and 5-FU/leucovorin plus bevacizumab. Safety data were not available at that time for the four-drug combination of IFL plus bevacizumab.

We allowed approximately 100 patients to be randomly assigned to each of the three arms and then evaluated the toxicity of each regimen. When the Data Safety Monitoring Committee noted no excess toxicity when bevacizumab was added to IFL, the 5-FU/leucovorin arm was discontinued as prespecified in the protocol. This study was designed essentially to compare a regimen including bevacizumab to the standard of care.

The primary question in our study was: What is the value of adding bevacizumab to IFL? The “clinical trial gods” like simple questions, and our study asked a simple question, was well conducted and had clean results.

Efficacy

The trial involved 923 patients, randomly assigned to IFL or to IFL plus bevacizumab. The patients on the IFL-alone arm had a median survival of 15.6 months, which corresponds to the reports of IFL by Saltz and others. The survival with IFL plus bevacizumab was 20.3 months, which importantly breaks the 20-month barrier (1.2).

Patients’ risk of dying was reduced by approximately one third, which was highly statistically significant. We also saw improvements in time to progression and response rate. The response rate went from approximately 35 percent to 45 percent, which was also highly statistically significant.

The 5-FU/leucovorin plus bevacizumab arm is nicknamed “chemo-light” plus bevacizumab, and it’s particularly intriguing. That regimen has remarkable activity. While the results were not statistically significant, an interesting trend was observed with respect to response rate, time to progression and overall survival when comparing all three arms of the study.

In general, the least efficacious regimen was IFL, followed by 5-FU/leucovorin plus bevacizumab and the four-drug combination of IFL plus bevacizumab. The four-drug combination appears to be best at all points, but it’s remarkable that the 5-FU/leucovorin/bevacizumab — the “chemo-light” plus bevacizumab regimen — benchmarks reasonably well compared to IFL.

Three data sets demonstrate the activity of the 5-FU/leucovorin/bevacizumab regimen: the initial randomized Phase II study reported by Fairooz Kabbinavar, arm 3 of our study with 100 patients and the study reported by Kabbinavar at ASCO this year involving more than 200 patients, again demonstrating that 5-FU/leucovorin plus bevacizumab outperforms 5-FU/leucovorin alone.

Side effects and tolerability

The side-effect profile for bevacizumab was remarkably benign. In general, we saw IFL-related side effects in both arms of the study, as would be expected. Bevacizumab resulted in a slight increase in the total number of Grade III or Grade IV toxicities from approximately 75 to 80 percent. Most of these were Grade III and manageable. Very few additional side effects were noted, and there was no increase in side effects leading to death on study or 60-day all cause mortality (1.3).

Traditional chemotherapy-related side effects, including nausea, vomiting, diarrhea and leukopenia, were essentially the same in both arms of the study. Certain bevacizumab-related side effects were specifically evaluated as part of the study, and we did not see an increase in thrombosis, bleeding or proteinuria. This is probably related to the fact that patients with metastatic colorectal cancer treated with intense chemotherapy are prone to these problems at baseline.

One of the most attractive features of the randomized placebo-controlled design of our study is that we were not only able to have a better evaluation of efficacy, but particularly a better evaluation of the real safety profile for bevacizumab.

We noted mild, clinically insignificant hypertension in about one fourth of patients receiving bevacizumab, and about 11 percent of patients experienced increased blood pressure requiring the addition of an oral antihypertensive. However, no patient experienced a hypertensive crisis, and many antihypertensive agents have been used successfully to manage the hypertension when it occurs.

Gastrointestinal (GI) perforations were an unexpected problem in the bevacizumab arm. These events have a presentation similar to that of severe bowel syndrome (perforation and fistula formation) and have been observed in two to three percent of patients with metastatic colorectal cancer treated with a 5-FUbased regimen.

Patients with GI perforations had variable presentations, sometimes in the setting of tumor response or during a colonoscopy to sort out another complication. GI perforations may have been related to injuries to the bowel caused by chemotherapy side effects. With only six events out of 400 patients (all in the IFL plus bevacizumab arm), it’s very difficult to identify the true predisposing factors.

Of the patients with GI perforations, one died, two had to discontinue treatment permanently, and three restarted therapy after they recovered and experienced no subsequent sequelae. I think the important message is that larger data sets will more clearly identify the predisposing factors, and that studies to address this issue are ongoing. Another important message is that GI perforations are relatively rare, and many patients recover — and an overall survival benefit is associated with bevacizumab.

Select publications

Dr Hurwitz is an Associate Professor of Medicine in the Division of Hematology/Oncology, Clinical Director of the Phase I Program, and Co-leader of the GI Oncology Program at Duke University Medical Center in Durham, North Carolina.