A phase III trial comparing mFOLFOX6 to mFOLFOX6 plus bevacizumab in stage II or III carcinoma of the colon: Results of NSABP Protocol C-08

Reviewer: Geoffrey Geiger MD
Abramson Cancer Center of the University of Pennsylvania
Last Modified: June 1, 2009

Presenter: N. Wolmark, G. Yothers, M. J. O'Connell, S. Sharif, J. N. Atkins, T. E. Seay, L. Feherenbacher, S. O'Reilly, C. J. Allegra
Presenter's Affiliation: NSABP Pittsburg, PA
Type of Session: Plenary


  • Bevacizumab is a humanized monoclonal antibody against the VEGF receptor, which inhibits angiogenesis, a rate-limiting step for cancer growth.
  • A study demonstrating that bevacizumab given with standard 5-fluorouracil (5-FU)-based chemotherapy significantly improved overall survival and time to disease progression when used as a first line therapy in metastatic colorectal cancer paved the way for the deployment of the drug for studies for a variety of cancers, including non-small cell lung cancer, renal cell cancer, breast cancer, liver and brain malignancies.
  • Subsequently, the use of this agent was expanded to earlier stage cancers, including stage II/III colon cancer. Initial safety reports indicated that the addition of bevacizumab to modified FOLFOX6 chemotherapy was well tolerated in patients with stage II/III colon cancer and did not appear to appreciably increase the incidence of adverse events that were typically associated with the monoclonal antibody, such as hypertension or thromboembolic events.
  • This two-arm randomized prospective study was to determine whether modified FOLFOX6 (mFOLFOX6) plus bevacizumab (mFF6+B) would prolong disease-free survival (DFS) compared to mFOLFOX6 (mFF6) alone.

Materials and Methods

  • Patients were enrolled between September 2004 and October 2006. A total of 2,672 patients with follow-up were randomized from 292 NSABP centers (1,338 in the mFF6 and 1,334 in the mFF6+B arms) with stage II (24.9%) or III (75.1%) carcinoma of the colon. Patients were randomized to receive either mFF6 (oxaliplatin 85 mg/m2 IV day 1, leucovorin 400 mg/m2 IV day 1, 5-FU 400 mg/m2 IV bolus day 1, and 5-FU 2400 mg/m2 delivered via continuous infusion (CI) over 46 hrs (days 1 and 2) every 14 days for a total of 12 cycles) or mFF6+B (same mFF6 regimen plus bevacizumab 5 mg/kg IV every 2 wks for 1 year).
  • Patients were excluded if they had symptomatic peripheral vascular disease, history of transient ischemic attack or stroke or if they had evidence of an arterial thrombotic episode in the previous year.
  • The primary end point in this trial was disease free survival. Events were defined as first recurrence, second primary cancer, or death.


  • The median follow-up for patients still alive was 36 months.
  • Less than 2% of patients were unanalyzable and patient characteristics were well distributed with 58% of patients under 60 years of age. Overall, 50% were male, 25% were stage II, 45.5% were stage III with 1-3 positive lymph nodes, and 29.5% had 4+ positive nodes.
  • There were no significant differences in grade 3 toxicities between the two groups with respect to venous thrombosis, neutropenias, diarrhea, CNS ischemia or peripheral arterial ischemia. Although wound complications were more frequent with bevacizumab, the difference was not large.
  • Investigators did observe a 6% absolute increase in the rate of grade 2 or higher neuropathy in the bevacizumab arm. The researchers suggested this difference is because 7% more patients in that group received at least 1,000 mg/m2 of oxaliplatin.
  • With respect to the primary endpoint of DFS, there was only a 2% difference at 3 years with a HR of 0.89 and a p-value of 0.15. Examination of the Kaplan-Meier curves, however, reveals a separation at one year before the curves eventually reconvene without crossing over.
  • In the first year, a robust benefit was seen in favor of bevacizumab with a HR of 0.60 (p=0.0004). The size of this benefit is similar to that of trastuzumab seen in NSABP B-31. Magnitude of this benefit became dramatically more attenuated with time: HR was 0.74 at 1.5 years, 0.81 at 2 years, 0.85 at 2.5 years and 0.87 at 3 years. Only at the time of the final analysis at 3 years was the difference found to no longer be statistically significant. The smoothed estimate of the disease free survival HR over time indicated that bevacizumab significantly reduced the risk of a DFS event during the interval spanning 0.5 to 1.0 year.
  • There is no evidence to suggest that patients receiving bevacizumab had a worse DFS as compared to those receiving mFF6 alone following treatment.

Author's Conclusions

  • The addition of bevacizumab to mFF6 did not result in an overall statistically significant prolongation in DFS at 3 years.
  • There was a transient benefit in DFS during the one-year interval that bevacizumab was utilized.
  • As such, consideration may be given to clinical trials assessing a longer duration of bevacizumab administration.
  • Examination of DFS with respect to stage reveals that 75% of the trial population was stage III with a HR of 0.90 and a p value of 0.25 (versus HR of 0.82 and p = 0.035 in stage II), and as such, there was no differential benefit observed when looking at stage groupings.
  • No evidence of a harmful “rebound” effect demonstrated in preclinical murine models in terms of evasive resistance with increased propensity for metastasis. At 36 months of follow-up (2 years after discontinuation of bevacizumab) there were less recurrences and deaths in the bevacizumab group and there was no difference in the risk of second cancers, 2-year rate of recurrence or rate of recurrence at multiple sites.

Clinical/Scientific Implications

  • The addition of bevacizumab did not result in a statistically significant change in overall survival although there was a transient benefit observed in the bevacizumab group with respect to DFS during the one year that bevacizumab was used.
  • This study demonstrates that the use of bevacizumab in the adjuvant setting cannot be routinely recommended at present although more research is clearly needed in stage II and III cancers to examine the effect on overall survival for regimens using bevacizumab for durations longer than the one year used in this study.
  • Under the most favorable circumstances, assuming that the benefit seen during the administration of bevacizumab persisted for a longer duration provided the drug was administered over a three year period, the maximum hypothetical benefit in terms of DFS would be in the range of 3-5%.
  • Authors hypothesize that this potential benefit would be dependent on a longer duration of bevacizumab administration. However, they acknowledge that not only may the optimal duration of administration never be rigorously determined, but also the cost to the patient and society may be prohibitive. Furthermore, several years of continuous every 3 week intravenous therapy is significant from not only time considerations but also in terms of potential adverse effects which may be cumulative.
  • If long-term anti-VEGF therapies are needed, DFS may be a less appropriate endpoint than overall survival and future studies will need to take this into consideration.
  • Authors are currently awaiting the results of AVANT BO17920, a study evaluating bevacizumab in combination with either capecitabine plus oxaliplatin (Xelox) or fluorouracil/leucovorin with oxaliplatin (Folfox-4) on disease-free survival in patients with colon cancer (n=3450). Projected final analysis time of this study is the third quarter of 2010.

See Patient Summary