Background In an open-label, randomized, phase 3 trial, we compared the efficacy and safety of paclitaxel with that of paclitaxel plus bevacizumab, a monoclonal antibody against vascular endothelial growth factor, as initial treatment for metastatic breast cancer.
Methods We randomly assigned patients to receive 90 mg of paclitaxel per square meter of body-surface area on days 1, 8, and 15 every 4 weeks, either alone or with 10 mg of bevacizumab per kilogram of body weight on days 1 and 15. The primary end point was progression-free survival; overall survival was a secondary end point.
Results From December 2001 through May 2004, a total of 722 patients were enrolled. Paclitaxel plus bevacizumab significantly prolonged progression-free survival as compared with paclitaxel alone (median, 11.8 vs. 5.9 months; hazard ratio for progression, 0.60; P<0.001) and increased the objective response rate (36.9% vs. 21.2%, P<0.001). The overall survival rate, however, was similar in the two groups (median, 26.7 vs. 25.2 months; hazard ratio, 0.88; P=0.16). Grade 3 or 4 hypertension (14.8% vs. 0.0%, P<0.001), proteinuria (3.6% vs. 0.0%, P<0.001), headache (2.2% vs. 0.0%, P=0.008), and cerebrovascular ischemia (1.9% vs. 0.0%, P=0.02) were more frequent in patients receiving paclitaxel plus bevacizumab. Infection was more common in patients receiving paclitaxel plus bevacizumab (9.3% vs. 2.9%, P<0.001), but febrile neutropenia was uncommon (<1% overall).
Conclusions Initial therapy of metastatic breast cancer with paclitaxel plus bevacizumab prolongs progression-free survival, but not overall survival, as compared with paclitaxel alone. (ClinicalTrials.gov number, NCT00028990 [ClinicalTrials.gov] .)
Discussion
In this phase 3 trial of paclitaxel plus bevacizumab as the initial treatment of metastatic breast cancer, the safety profile of the combination was similar to profiles reported in previous randomized trials.8,16,17,18 Most toxic effects were minimal, rarely limited therapy, and did not have a detrimental effect on overall quality of life.15
We enrolled patients with predominantly HER2-negative breast cancer; no patient received concurrent trastuzumab. Further studies are needed to assess the efficacy of bevacizumab in patients with HER2-positive metastatic breast cancer.19,20 In our trial, bevacizumab was not given to patients who had a tumor with a specific molecular phenotype. Although benefit was seen across a number of clinically important subgroups, our results would be strengthened by the ability to identify patients most likely to benefit from VEGF-directed therapies.
In a previous phase 3 study, the addition of bevacizumab to capecitabine significantly increased the objective response rate but not progression-free survival or overall survival.8 What might account for the different results in these trials? It seems unlikely that chance could account for the improvement in progression-free survival found in our trial. Investigator or patient bias, always a consideration in open-label studies, is unlikely to explain our results. If such biases had a large role, we would have expected to see a greater improvement in patients with nonmeasurable lesions, where disease assessment is necessarily subjective. Actually, the hazard ratio was more favorable in patients with measurable disease than in those with nonmeasurable disease (0.55 vs. 0.68).
Substantial differences between the patient populations of these studies may account for the disparate results. All patients in the earlier study had received previous anthracycline and taxane therapy, and most (more than 85%) had received chemotherapy for metastatic disease.8 In contrast, 35.2% of our patients had not received any previous chemotherapy, and only 13.2% had received both an anthracycline and a taxane as adjuvant therapy.
A recent phase 2 trial found a median time to disease progression of only 5.7 months (95% confidence interval, 4.9 to 8.4) with capecitabine plus bevacizumab as initial chemotherapy.21 Perhaps paclitaxel is uniquely synergistic with bevacizumab. Indeed, the taxanes have distinct antiangiogenic activity.22 In preclinical studies, VEGF protected endothelial cells from the antiangiogenic properties of docetaxel; bevacizumab overcame this protective effect in vitro and in vivo.23
Despite a striking improvement in progression-free survival, the addition of bevacizumab did not prolong overall survival in this study. Patients with metastatic breast cancer frequently receive multiple therapies during the course of their disease. Data on treatment administered after progression were not collected in this trial, precluding an exploratory analysis of the influence of subsequent therapy on overall survival. Though the mechanisms of resistance to bevacizumab are not well defined,24,25 it is possible that resistance to bevacizumab results in relative resistance to subsequent therapies. Alternatively, rebound increases in VEGF on discontinuation of bevacizumab could result in more aggressive disease. Resistance to paclitaxel, whether mediated by increased expression of the multidrug resistance protein26 or by microtubule mutations,27 could also cause resistance to subsequent chemotherapy.
We found that treatment with bevacizumab early in the course of metastatic breast cancer, when angiogenic pathways are less redundant, improved progression-free survival and the objective response rate. Although our patients were receiving their first treatment for metastatic breast cancer, only a third had never received any chemotherapy. More than 80% had overt visceral involvement, presumably with an established vasculature. In short, first-line therapy for metastatic breast cancer is not “early” in the natural history of breast cancer. Recent laboratory studies suggest that the initial events in the development of metastasis are VEGF-dependent.28,29 If this is true, the most successful clinical application of angiogenesis inhibitors is likely to be in patients with micrometastatic disease in the adjuvant setting.
Supported in part by Public Health Service Grants CA23318, CA66636, CA21115, CA49883, and CA16116 from the National Cancer Institute, National Institutes of Health, Department of Health and Human Services, and by Genentech.
Dr. Miller reports receiving lecture fees and consulting fees for service on breast cancer advisory boards from Roche. Dr. Gralow reports receiving lecture fees and consulting fees for service on breast cancer advisory boards from Genentech and Roche. Dr. Dickler and Dr. Cobleigh report receiving consulting fees from Genentech for service on breast cancer advisory boards. Dr. Cobleigh reports receiving lecture fees from Genentech. Dr. Perez reports receiving consulting fees from Genentech, GlaxoSmithKline, Bristol-Myers Squibb, and Sanofi-Aventis for service on breast cancer advisory boards, as well as grant support for clinical translational studies from Genentech. No other potential conflict of interest relevant to this article was reported.
The views expressed are those of the authors and do not necessarily represent the official views of the National Cancer Institute.
We thank Dr. George Sledge and Dr. Robert Gray for helpful discussions and Carol Chami for providing technical assistance in preparing the manuscript.
Source Information
From Indiana University Cancer Center, Indianapolis (K.M.); Dana–Farber Cancer Institute, Boston (M.W.); Puget Sound Oncology Consortium, Seattle (J.G.); Memorial Sloan-Kettering Cancer Center, New York (M.D.); Rush–Presbyterian–St. Luke’s Medical Center, Chicago (M.C.); Mayo Clinic, Jacksonville, FL (E.A.P.); British Columbia Cancer Agency–Vancouver Cancer Center, Vancouver, BC, Canada (T.S.); Evanston Northwestern Healthcare and Robert H. Lurie Comprehensive Cancer Center of Northwestern University — both in Evanston, IL (D.C.); and Johns Hopkins Kimmel Comprehensive Cancer Center, Baltimore (N.E.D.).
Address reprint requests to Dr. Miller at the Indiana Cancer Pavilion, 535 Barnhill Dr., RT473, Indianapolis, IN 46202, or at kathmill@iupui.edu.