Tracey L. Evans, MD, Department of Medicine
University of Pennsylvania School of Medicine, Philadelphia, PA
In eligible patients, bevacizumab maximizes overall survival in the first-line treatment of advanced non-small cell lung cancer (NSCLC) when administered in combination with platinum-based chemotherapy. Determining patient eligibility for bevacizumab is critical because of potentially fatal treatment complications unique to the lung cancer population. Patients with either primary lung tumors of squamous histology or hemoptysis should clearly never receive bevacizumab due to the risk of life-threatening hemoptysis. Whether bevacizumab is sufficiently safe in lung cancer patients who are on anticoagulation therapy or have either brain metastasis or low-risk squamous histology tumors is as yet unknown and remains an area of active investigation.
Bevacizumab was initially approved by the US Food and Drug Administration (FDA) for the first-line treatment of advanced colorectal cancer in February 2004. This approval was based on a phase III study demonstrating improved survival in patients treated with bevacizumab (5 mg/kg every 2 weeks) in combination with irinotecan, 5-flurouracil, and leucovorin.34 A subsequent randomized trial in previously treated patients with advanced colorectal cancer demonstrated improved survival in patients treated with bevacizumab (10 mg/kg every 2 weeks) in combination with oxaliplatin,5-fluorouracil, and leucovorin.35
Risk of catastrophic hemoptysis with bevacizumab in lung cancer was first noted in AVF0757g, a randomized phase II study. Patients with squamous histology were permitted study entry, whereas patients with central nervous system (CNS) metastasis and those on aspirin or full dose anticoagulation were excluded. Patients were randomly assigned to either chemotherapy alone (carboplatin and paclitaxel) or chemotherapy in combination with one of two doses of bevacizumab (7.5 or 15mg/kg) administered every 3 weeks. Chemotherapy was administered to a maximum of 6 cycles, whereas bevacizumab was continued until disease progression. Patients in the chemotherapy alone control arm of this study were permitted to cross over to bevacizumab (15 mg/kg) at disease progression. The most eye-catching aspect of the phase II study was the occurrence of nine treatment- related deaths (four in each of the bevacizumab arms and one in the control arm). Four of the deaths in the bevacizumab arms were due to massive hemoptysis. In total, six episodes of life-threatening hemoptysis occurred in bevacizumab-treated patients; four of the episodes occurred in patients with squamous histology. A total of 13 patients with squamous histology received bevacizumab, thus the rate of life-threatening hemoptysis in these patients was 31%. Two episodes of life-threatening hemoptysis occurred in the 53 patients with non- squamous histology treated with bevacizumab (i.e., the life-threatening hemoptysis rate was 4%). These study results led to the design of Eastern Cooperative Oncology Group study 4599 (E4599), the choice of bevacizumab dose (15 mg/kg), and the exclusion of patients with squamous histology.
Bevacizumab in the First-Line Treatment of Non-Small Cell Lung Cancer
E4599 established bevacizumab as the current standard of care in nonsquamous NSCLC. In this study, treatment- naïve patients (n=878) were randomized to either chemotherapy alone (carboplatin and paclitaxel) or chemotherapy in combination with bevacizumab (15 mg/kg) every 3 weeks for a maximum of six cycles of chemotherapy. Unlike AVF0757g, E4599 did not allow crossover to the bevacizumab treatment arm. The trial schema is shown in Figure 4; inclusion and exclusion criteria are shown in Table 2. Improvement in overall survival in the bevacizumab arm was statistically signifi- cant (hazard ratio [HR], 0.79; p = 0.003); median survival was 10.3 and 12.3 months in the chemotherapy alone and bevacizumab arms, respectively (Figure 5). With the addition of bevacizumab, 1-year survival increased from 44% to 51%, and 2-year survival increased from 15% to 23%. This was the first time that median survival has surpassed the 1-year mark in a US Cooperative Group trial of patients with advanced NSCLC. Of particular note, survival in the chemotherapy alone arm was also better than is usually seen. Median survival typically observed in phase III studies with chemotherapy alone in the advanced setting is approximately 8 months.36 Perhaps this resulted from the restric- tive eligibility criteria of the study. Improvement in response was statistically significant in the bevacizumab vs. chemotherapy alone arm (35% and 15%, respectively), as was progression-free survival (median 6.2 and 4.5 months, respectively). Of interest, on subgroup analysis, median survival in women was the same in the bevacizumab and chemotherapy alone arms (13.3 and 13.1 months, respectively), despite the fact that bevacizumab- treated women had statistically improved response and progression-free survival rates compared with those treated with chemotherapy alone.37 Whether this lack of apparent survival benefit in women was due to statistical chance, an imbalance in unmeasured prognostic factors, or a true difference based on sex is unclear.
Substantial toxicity was seen in the bevacizumab arm, with significantly higher rates of hypertension, proteinuria, bleeding, neutropenia, febrile neutropenia, thrombocytopenia, hyponatremia, rash, and headache (Table 3). Fifteen treatment-related deaths occurred in the beva- cizumab arm (3.5% of patients receiving study treatment), compared with two treatment-related deaths in the chemotherapy alone arm (0.5%). Cause of death in the bevacizumab arm was pulmonary hemorrhage (n=5), complications of febrile neutropenia (n=5), cerebrovascular events (n=2), gastrointestinal hemorrhage (n=2), and possible pulmonary embolus (n=1). Most of the deaths occurred during the first two cycles of chemotherapy. The overall rate of grade 3-5 hemoptysis in the bevacizumab arm was 1.9%. Exclusion of patients with hemoptysis greater than 1/2 tsp was an addition to the study following two episodes of fatal hemoptysis in patients with antecedent hemoptysis.
Bevacizumab was granted FDA approval for the initial systemic treatment of patients with unresectable, locally advanced, recurrent or metastatic, nonsquamous NSCLC on October 11, 2006. The FDA stipulated that bevacizumab was approved for use specifically in combination with paclitaxel and carboplatin. CurrentNational Comprehensive Cancer Network (NCCN)Guidelinesrecommend bevacizumab in conjunction with chemotherapy in patients with ECOG performance status 0-2 who meet the following eligibility criteria: nonsquamous cell histology and no hemoptysis, CNS metastasis, and ongoing therapeutic anticoagulation. The guidelines also state that any regimen with a high risk of thrombocytopenia, and therefore risk of bleeding, should be used with caution.38
Unfortunately, it seems that a substantial proportion of patients with advanced NSCLC may not be eligible for bevacizumab under the current eligibility criteria. In a retrospective analysis of 150 patients with advanced NSCLC evaluated at Fox Chase Cancer Center in 2002, only 45.6% met the eligibility criteria for E4599. Reasons for ineligibility were as follows: 25.6% had an ECOG performance status of 2; 24.3% had CNS metastasis; 13.4% had squamous cell histology; 10.9% were on therapeutic anticoagulation; and 25.6% had more than one characteristic causing ineligibility.39
Conclusions
Ongoing trials of bevacizumab are focused on loosening the above restrictions on eligibility. The ATLAStrial, a randomized phase IIIb study with a total expected enrollment of 1150 patients, was designed primarily to determine whether maintenance erlotinib with bevacizumab following four cycles of bevacizumab with platinum-based chemotherapy improves progression- free survival. Treatment-associated adverse events is a secondary endpoint. The eligibility criteria allow enrollment of patients with stable treated brain metastasis, patients with squamous cell carcinoma who do not have intrathoracic disease or in whom the intrathoracic tumors are peripheral, and patients on anticoagulation with low molecular weight heparin. In addition, patients in the trial can be treated with one of three chemotherapeutic agents in combination with carboplatin: paclitaxel, docetaxel, or gemcitabine. The PASSPORT trial is a phase II study of bevacizumab in combination with either first- or second-line therapy in patients with previously treated CNS metastasis. The AVAIL trial, a European randomized comparison of cisplatin and gemcitabine vs. the same chemotherapy with low-dose (7.5 mg/kg) or high- dose (15 mg/kg) bevacizumab, has completed accrual, with results expected shortly. Preliminary results of phase II studies combining bevacizumab with carboplatin and either gemcitabine 38 or pemetrexed 40 have demonstrated no untoward toxicity. Studies attempting to safely treat patients with squamous cell carcinoma—either by radiating the lung tumor first or starting with chemotherapy alone and then adding bevacizumab with later chemotherapy cycles—are also underway.
Discussion
Dr. Lynch: It's worth noting that although NCCN guidelines recommend bevacizumab in conjunction with chemotherapy in patients with ECOG performance status (PS) 0-2, the study on which approval is based, E4599, only included patients with PS 0-1, so that is what we use. Zero to two is not as evidence-based.
Dr. Evans: Yes. The E4599 criteria for patient selection and exclusion are both important.
Dr. Lynch: I read in the Wall Street Journal that the AVAIL trial results comparing low-dose (7.5 mg/kg) or high-dose (15 mg/kg) bevacizumab show that both doses prolong disease-free survival.
Dr. Heist: Why do you think patients with lung cancer experience hemoptysis in response to bevacizumab?
Dr. Socinski: Perhaps toxicity is a marker of activity with these agents. With bevacizumab, a lot of the severe or even fatal hemorrhages have occurred with dramatic responses—cavitating lesions, squamous cell histology, and so on. The same types of toxicity also have been seen with many of the tyrosine kinase inhibitors that target VEGF. An important challenge is identifying patients who are at high risk of side effects in the context of deriving clinical benefit from these agents.
Dr. Lynch: Any thoughts on the toxicity mechanism?
Dr. Socinski: I believe that it comes down to the interaction of the drug, the spectrum of the targets, and interindividual differences in vasculature. In the small percentage of patients who experience hemoptysis, perhaps something very intrinsic about the tumor is responding or reacting to the drug, which is responsiblefor the side effect. But, we don't yet have any specific mechanistic explanations.