A Double-Blind, Placebo Controlled, Randomized Phase III trial of Gemcitabine (G) plus Bevacizumab (B) Versus Gemcitabine Plus Placebo (P) in Patients (pts) with Advanced Pancreatic Cancer (PC): A Preliminary Analysis of Cancer and Leukemia Group B (CALG

Reviewer: Eric Shinohara MD, MSCI
Abramson Cancer Center of the University of Pennsylvania
Last Modified: June 3, 2007

Presenter: H. L. Kindler
Presenter's Affiliation: University of Chicago Medical Center
Type of Session: Scientific


  • Advanced pancreatic cancer continues to have an extremely poor prognosis.
  • Gemcitabine is the cornerstone of treatment and objective response rates of 5-10% are seen when gemcitibine is used alone. Treatment with gemcitibine leads to median overall survivals (OS) rates of five to six months.
  • Prior studies have demonstrated overexpression of VEGF in pancreatic cancer. Increased expression of VEGF has been shown to be associated with a worse prognosis in pancreatic cancer with a decrease in overall survival.
  • These findings provided the rational for combining gemcitibine with a VEGF inhibitor, bevacizumab. A phase II study of 52 patients demonstrated a 21% response rate in patients with advanced pancreatic cancer treated with gemcitibine and bevacizumab. Median OS in this study was 8.8 months (Kindler, JCO 2005).
  • These results led to the present Phase III CALGB study comparing gemcitabine and bevacizumab with treatment with gemcitibine with placebo in advance pancreatic cancer patients.

Materials and Methods

  • The present study is a Phase III, randomized, double blinded, placebo controlled study comparing combined treatment with gemcitabine and bevacizumab with treatment with gemcitibine with placebo in patients with advanced pancreatic cancer.
  • The primary endpoint of this study was overall survival (OS) and the secondary endpoints were response rates, progression free survival (PFS), and toxicity.
  • Patient Eligibility:
    • Patients could not have had prior chemotherapy for advanced disease.
    • Patients had to have a performance status (PS) of 0-2.
    • There could not be any tumor invasion of adjacent organs.
    • The patient could not have increased risk for bleeding. Patients on anticoagulants were eligible if the dose of anticoagulant was stable.
  • Patients were stratified by performance status (0/1 or 2), disease extent (locally advanced or metastatic), and prior radiation treatment (yes or no).
  • Chemotherapy Treatment:
    • Gemcitibine was given at a dose of 1,000 mg/m2 over 30 minutes on days 1, 8, and 15 every 28 days.
    • Bevacizumab and placebo were given at a dose of 10 mg/kg on days 1 and 15 every 28 days.
    • Patients underwent a CT scan every two cycles.
  • The study was designed to detect a difference in overall survival of 35% (a change from 6 to 8.1 months) in survival with 90% power.


  • At total of 602 patients were enrolled from June 30, 2004 to April 14, 2006. Interim analysis was performed after 64% of the data on OS had been collected. Interim analysis demonstrated that the futility boundary had been crossed and this information was reported in June of 2006. Patients were informed and unblinded.
  • Patient Characteristics:
    • 302 patients were enrolled in the bevacizumab arm and 300 in the placebo arm.
    • The groups were well balanced.
    • Patient characteristics for the two groups were as follows: male 58% and 51%, median age 63.8 and 65, PS 2 9% and 9%, Stage IV disease 85% and 84%, radiation 11% and 11% for the bevacizumab and placebo groups, respectively
  • Median follow up was 11.3 and 11.7 months. As of 12/22/2006 436 patients, 224 from the bevacizumab arm and 212 from the placebo arm had died (93% of the planned number of events needed at final analysis).
  • Median OS was 5.8 and 6.1 months (p=0.78) and median PFS was 4.9 and 4.7 months (p=0.99) for the bevacizumab and placebo groups, respectively.
  • There was a significant difference in OS when patients were stratified by stage with a median OS of 5.7 months in patients with metastasis and a 9.9 month OS in patients with locally advanced disease.
  • Performance status (PS) was also significantly associated with OS, with patients with a PS of 0 having an OS of 8 months, those with a PS of 1 having an OS of 4.8 months and those with a PS of 2 having an OS of 2.8 months.
  • Toxicity was similar between the two groups aside from hypertension and proteinuria, which were higher in the bevacizumab group. Grade 3 and 4 toxicity was as follows: Neutropenia occurred in 33% and 30%, anemia in 5% and 8%, thrombocytopenia in 12% and 12%, hypertension in 8% and 2%, perforation in 0.4% and 0%, GI bleed in 3% and 2%, cerebral vascular accident in 2% and 2%, proteinuria in 4% and 1%, and venous thrombosis in 9% and 9% of patients in the bevacizumab and placebo groups, respectively.

Author's Conclusions

  • The authors conclude that there is no benefit in OS when bevacizumab is combined with gemcitibine. This was contradictory to earlier Phase II studies which demonstrated an OS benefit with the combination of bevacizumab and gemcitibine (Kindler, JCO 2005).
  • They concluded that this difference may have been due to patient selection. In the prior phase II study there were more patients with a PS of 0. This study demonstrated that PS correlates with OS and hence this difference may have caused the difference seen in these two studies.

Clinical/Scientific Implications

  • Many studies have attempted to combine gemcitibine with a second agent in an attempt to increase its efficacy in pancreatic cancer. Unfortunately, the majority of these studies have been disappointing. Capecitibine in combination with gemcitibine has shown potential, however, only preliminary results are available at present and other studies have contradicted these results.
  • These disappointing results appear to be the case with bevacizumab as well. It is possible that with the use of a molecular marker, patients that would be most likely to benefit from bevacizumab, could be identified. For example, if a patient population with tumors with a high level of VEGF expression could be selected, it may be possible to find a difference in OS with combined treatment with bevacizumab and gemcitibine. However, based on the current data it appears that bevacizumab provides no therapeutic benefit in advance pancreatic cancer. However, question remains if this combination in earlier stage disease and patients with good performance status may be of benefit.