Avastin Benefits Metastatic Breast Cancer

SAN ANTONIO—Bevacizumab (Avastin), a monoclonal antibody targeting vascular endothelial growth factor (VEGF), conferred additional benefit to paclitaxel when the combination was used as first-line therapy in locally recurrent and metastatic breast cancer patients in the Eastern Cooperative Oncology Group (ECOG) phase III E2100 study. Kathy D. Miller, MD, of Indiana University Cancer Center, Indianapolis, reported the updated results at the 28th Annual San Antonio Breast Cancer Symposium (abstract 3). "This is a positive study," Dr. Miller said. "The addition of bevacizumab to paclitaxel more than doubled the response rate and significantly prolonged progression-free survival."

Tumor growth is dependent upon angiogenesis, and bevacizumab recognizes all VEGF-A isoforms. It has shown activity in patients with highly refractory metastatic breast cancer, yielding a 9% response rate as monotherapy, Dr. Miller noted. "Since proangiogenic factors become more numerous as breast cancers progress, we expect greater activity of this agent in patients with less heavily pretreated disease," she said. "E2100 was designed to test this hypothesis."

Study Protocol

E2100 randomized 722 breast cancer patients with locally recurrent or metastatic disease to paclitaxel alone (90 mg/m2 on days 1, 8, and 15 every 4 weeks), or the same paclitaxel regimen plus bevacizumab 10 mg/kg on days 1 and 15.

Interim analyses were planned after 270 and 425 progression-free survival events, with the final analysis to occur after 546 events. The current data cut-off occurred September 27, 2005, after 484 events, including 426 cases of disease progression and 58 deaths without documented progression.

Patients receiving the combination of paclitaxel plus bevacizumab demonstrated significant improvements in overall response rate and progression-free survival, Dr. Miller reported.

"The addition of bevacizumab more than doubled the objective response rate, from 13.8% to 29.9%. These results were highly significant (P < .0001)," she said. She noted that the response rates are in keeping with patients who have had significant exposure to prior chemotherapy, adding that "the response rates in E2100 are actually 'artificially low' because we included patients who did not have measurable disease. By definition, these patients' responses could only have been progressive disease, complete response, or stable disease; partial response was not an available category for these patients." The evaluation of only patients with measurable disease at entry showed similar results to that of the overall population.

Progression-Free Survival

The improvement in response rate translated into a dramatic improvement in progression-free survival (11.4 months for the combination vs 6.11 months for paclitaxel alone, P < .0001). "The curves separated early and remained widely separate throughout the follow-up," she noted. "With a hazard ratio of 0.51, these results are both highly statistically and clinically meaningful."

The overall survival data are premature. There was a 3-month improvement in survival that has not reached statistical significance, she said.

The addition of bevacizumab contributed the expected toxicities: 15% of patients developed hypertension requiring treatment. Significant thromboembolic events, serious bleeding, or grade 3-4 proteinuria, however, were extremely rare. Bevacizumab did not enhance the typical toxicity profile of paclitaxel, she said. E2100 also assessed quality of life at 17 weeks and at 33 weeks, and found no differences overall between the arms or between time points.

Circulating Angiogenic Factors

The researchers analyzed circulating angiogenic factors in serum and urine at baseline and after the third cycle. "We focused this analysis on two important questions: Do baseline concentrations predict response or disease-free survival, and does the change in concentration with therapy correlate with outcome?" she said.

Investigators measured serum levels of vascular cell adhesion molecule (VCAM-1)—a cell surface receptor that is expressed on endothelial cells in response to VEGF—and urine VEGF concentrations. Baseline VCAM-1 levels were similar in both groups. There was a suggestion of a lower response rate in patients with the highest VCAM-1 levels at study entry, but this trend did not reach significance. Urine VEGF levels were also not different at baseline between the groups. In both arms there was a suggestion of improved response to bevacizumab in patients with the highest baseline VEGF levels, Dr. Miller reported.

Regarding changes from baseline, increases in VCAM-1 levels were observed after treatment with bevacizumab but not with paclitaxel alone. In both groups, the change did not correlate with the likelihood of overall response. In the paclitaxel-alone group, patients with the greatest increase in VCAM-1 by the third cycle had a significantly lower progression-free survival time, compared with patients with lesser changes.

Urine VEGF levels did not change appreciably in either arm, although there was a suggestion of an improved progression-free survival in paclitaxel/bevacizumab patients with the least degree of change. This represented patients with initially undetectable urine VEGF or those with a significant decrease during therapy.

Dr. Miller concluded, "It is now time to move bevacizumab into the adjuvant setting and explore its role there. We also need to continue to develop methods to identify those breast cancer patients who are most likely to benefit from VEGF-targeted therapies."