Metastatic Cancer of Colon or Rectum
In a phase III trial the addition of bevacizumab to bolus irinotecan/5-5-FU/leucovorin (LV) (IFL) therapy resulted in increased survival, progression-free survival, response rate, and duration of response as compared to bolus IFL alone. Patients (n=923) with untreated, metastatic colorectal cancer were randomized to bolus IFL/bevacizumab (n=402), bolus IFL/placebo (n=411), or 5-FU/LV/bevacizumab (n=110). The IFL regimen consisted of intravenous (IV) irinotecan 125 milligrams/square meter (mg/m2), bolus IV 5-FU 500 mg/m2, and LV 20 mg/m2 once weekly for four weeks, with the cycle repeated every six weeks; the 5-FU/LV regimen consisted of bolus IV 5-FU 500 mg/m2, LV 500 mg/m2 once weekly for six weeks, with the cycle repeated every eight weeks. All doses of bevacizumab were 5-mg/kg IV every two weeks. Results reported for the primary comparison group (bolus IFL/bevacizumab vs. bolus IFL/placebo) were as follows: survival (20.3 vs. 15.6 months, p less than 0.001), progression-free survival (10.6 vs. 6.2 months, p less than 0.001), response rate (44.8% vs. 34.8%; p=0.004), and duration of response (10.4 vs. 7.1 months; p=0.001). Grade III to IV adverse events potentially related to bevacizumab included bleeding (3.1%), hypertension (11%), proteinuria (0.8%), and gastrointestinal perforation (1.5%). Thrombotic events such as deep vein thrombosis and pulmonary embolism occurred in 8.9% and 3.6% of patients respectively.
In a randomized phase II study involving previously untreated patients with metastatic colorectal cancer (n=104), objective responses were seen in 40% (95% confidence interval [CI] 24% to 58%) and 24% (95% CI 12% to 43%) of those treated with IV bevacizumab 5-mg/kg and 10- mg/kg, respectively, every two weeks in combination with the standard chemotherapy regimen of 5-FU/LV (Roswell Park regimen); a response rate of 17% (95% CI 7% to 34%) was observed in patients treated with standard chemotherapy alone. Times to disease progression in these three groups were nine months (95% CI 5.8 to 10.9; p=0.005), 7.2 months (95% CI 3.8 to 9.2; p=0.217), and 5.2 months (95% CI 3.5 to 5.6), respectively. Median survival was 21.5, 16.1, and 13.8 months, respectively; 37% of patients in the 5-mg/kg arm were alive at 18 months. Imbalances in randomization Eastern Cooperative Oncology Group (ECOG) performance status, and baseline serum albumin in this study were a limitation and could be a possible explanation, along with a small sample size and a chance for the more effective outcome were seen with the lower dose of bevacizumab.
In a phase II trial the addition of bevacizumab to full or reduced dose of irinotecan, fluorouracil, and leucovorin (IFL) was safe and efficacious in patients with untreated metastatic colorectal cancer. Patients (n=20) received IV bevacizumab 10-mg/kg every other week, irinotecan 125 mg/m2 IV, bolus IV 5-FU 500 mg/m2, and LV 20 mg/m2 weekly for four of six weeks. The remaining patients enrolled into the study received reduced dose IFL; irinotecan 100 mg/m2, 5-FU 400 mg/m2, and LV 20 mg/m2. Overall response rate (based on Response Evaluation Criteria in Solid Tumors (RECIST) was 45.7% (n=70). The overall response obtained in the full and reduced IFL dose groups was 33.3% and 48.3%, respectively. Time to progression (TTP) was similar regardless of IFL regimen used (full IFL group 12.1 months, reduced IFL group 10.8 months). Toxicity data (n=83) showed no statistical difference between grade III or grade IV toxicities based on the starting dose of IFL (p value not reported). Grade III and IV toxicities included: diarrhea, neutropenia, febrile neutropenia, bleeding events, thromboembolism and hypertensions. Proteinuria was infrequent.
In a randomized, placebo-controlled, phase 3 trial (n=1400), Saltz et al. reported treatment with bevacizumab plus XELOX (oxaliplatin and capecitabine) or FOLFOX-4 (oxaliplatin plus leucovorin followed by a bolus and infusional 5-fluorouracil) therapy (combined results) significantly improved progression-free survival (PFS) compared to placebo plus XELOX or FOLFOX-4 (combined results) as first-line therapy in patients with metastatic colorectal cancer.
Second Line Therapy Metastatic Cancer of Colon or Rectum
A randomized, phase III clinical trial demonstrated that the addition of bevacizumab to standard chemotherapy improves overall survival and progression-free survival (PFS) in patients with previously treated advanced or metastatic colorectal cancer. Eligible patients (n=829) had received prior therapy with a fluoropyrimidine and an irinotecan-based regimen and were both oxaliplatin- and bevacizumab-naive. Patients were randomized to receive either high-dose IV bevacizumab (10-mg/kg) once every two weeks plus the FOLFOX4 regimen (oxaliplatin 85 mg/m2, day one; leucovorin 200 mg/m2 as a two-hour IV infusion, followed by fluorouracil 400 mg/m2 IV bolus followed by fluorouracil 600 mg/m2 continuous IV infusion for 22 hours, days one and two) every two weeks (n=289), FOLFOX4 alone (n=290), or bevacizumab alone (n=243). The bevacizumab alone treatment arm was closed early when interim data showed a trend toward inferior survival compared to the other treatment arms. After a median 28 months of follow-up, median overall survival, the primary endpoint, was 12.9 months, 10.8 months, and 10.2 months for the BEV+FOLFOX4, FOLFOX4 alone, and bevacizumab alone groups, respectively (hazard ratio (HR)=0.76 for BEV+FOLFOX4 versus FOLFOX4 alone, p=0.0018). The study had 80% power to detect a 40% improvement overall survival with the addition of bevacizumab to FOLFOX; it did not meet its primary objective (there was a significant 16% improvement in overall survival. Median PFS was 7.2 months, 4.8 months, and 2.7 months for the BEV+FOLFOX4, FOLFOX4 alone, and bevacizumab alone groups, respectively (HR=0.64 for BEV+FOLFOX4 versus FOLFOX4 alone, p less than 0.0001). Overall response rate was significantly higher in the BEV+FOLFOX4 group (21.8%) compared to the FOLFOX4 group (9.2%; p less than 0.0001) or the bevacizumab alone group (3%). The BEV+FOLFOX4 group had significantly higher rates of several grade III to IV toxicities than the FOLFOX4 alone group (all reported in less than 10% of patients): hypertension (p=0.018), bleeding (p=0.011), and vomiting (p=0.010). Grade III neuropathy occurred in 16% of BEV+FOLFOX4 patients and 9% of FOLFOX4 patients; grade IV neuropathy occurred in less than 1% in both groups (p=0.016 for grades III and IV combined). Bowel perforation was only reported in patients receiving bevacizumab (1% in BEV+FOLFOX4 group, 1.3% in BEV alone group).
Third-line combination therapy with 5-FU/LV has a low response rate in patients with advanced or metastatic colorectal cancer who have progressed on or were intolerant to oxaliplatin- and irinotecan-based chemotherapy regimens. Initially, 100 patients were to be entered into the first stage of the study (the response assessment cohort), and if efficacy was observed with that group of patients, the second stage of the study would be an expanded open access protocol (trial TRC-0301). However, due to rapid accrual, an additional 250 patients were included in the first stage. Patients received bevacizumab (5-mg/kg) every two weeks and either the Roswell Park regimen or the de Gramont regimen. A median 33 weeks after first treatment, median TTP in the response assessment cohort (n=100) was 3.5 months (95% confidence interval two months to 4.8 months). One confirmed response was observed in this cohort. After a median 29.9 weeks after starting therapy, 264 patients had discontinued therapy due to progressive disease (64%), toxicity (14%), consent withdrawn (9%), or death unrelated to progressive disease (n=5). Median TTP for all treated patients (n=337) was 3.7 months. Half of all patients had at least one grade III, IV, or V adverse event; 13% of patients discontinued the study due to adverse effects. Grade III or IV hemorrhage occurred in 4.3% of patients, including 3.5% with gastrointestinal tract bleeding. Other common adverse events included grade III or IV fatigue (10.4%) and diarrhea (9%).
In a randomized, phase II clinical trial (BOND II trial) (n=83), combination therapy with cetuximab/bevacizumab (CB) or cetuximab/bevacizumab/irinotecan (CBI) demonstrated favorable efficacy in patients with irinotecan-refractory metastatic colorectal cancer compared to historical control patients. Patients who had previously failed one or more irinotecan-containing chemotherapy regimens, had never received cetuximab, bevacizumab, or any other epidermal growth factor receptor (EGFR)- or VEGF-directed agents, had a ECOG performance status of one or less, and did not require full-dose anticoagulation were randomized to received either CB (n=40; median age, 56 years [yr], range, 24-80 yr; median number of prior regimens, three, range, 1-8) with cetuximab (400 milligrams [mg]m2) loading dose as two-hour IV infusion on day one, then 250-mg/m2 IV weekly starting on day eight) and bevacizumab (5-mg/kg every two weeks beginning on day two; drugs were administered on the same day for subsequent treatments) or CBI (n=43; median age, 64 years [yr], range, 43-86 yr; median number of prior regimens, three, range, 1-6) with cetuximab plus bevacizumab (same dose and schedule as CB) and irinotecan (starting on day one after cetuximab at the same dose and schedule as was last received prior to study). Response was evaluated every eight weeks using the RECIST. Median TTP was longer in the CB arm compared with cetuximab alone historical controls (5.6 months vs. 1.5 months, respectively) and in the CBI arm compared to cetuximab plus irinotecan historical controls (7.9 months vs. 4.1 months, respectively). Response rates for CB and CBI were 23% and 37%, respectively; additionally, these patients receiving the bevacizumab-containing regimens had higher response rates than cetuximab alone and cetuximab plus irinotecan historical controls (10.8% and 22.9%, respectively). At a median follow-up of 28 months, the overall survival time was 11.4 months in the CB arm and 14.5 months in the CBI arm. When comparing common cetuximab adverse effects between the CB and CBI groups, there was a similar incidence of grade III (20% and 21%) skin rash and paronychial grade III cracking (7% and 5%). Five percent of patients in the CB arm experienced grade III headache compared with no patients in the CBI arm; additionally, one patient with a severe headache required hospitalization and treatment with parenteral opioid analgesics. When comparing common irinotecan adverse effects between the CB and CBI groups, there was an increased incidence of grade III neutropenia (0% vs. 23%, respectively), grade III-IV diarrhea (0% vs. 28%), grade III fatigue (0% vs. 9%), and grade III nausea (0% vs. 2%) in the CBI group. Serious adverse events include perirectal fistula (n=1) reported in CB arm and bowel perforation (n=1), nonperforated duodenal ulcer (n=1), and right frontal lacunar infarct (n=1) reported in the CBI arm. Two deaths (n=1, myocardial infarction followed by death 16 days later; n=1, fatal enterococcal endocarditis) were reported in the CB and CBI arms, respectively.
In a randomized, open-label, multinational trial of patients with metastatic colorectal cancer who progressed following first-line therapy with a bevacizumab-containing regimen, treatment with bevacizumab combined with fluoropyrimidine/irinotecan- or fluoropyrimidine/oxaliplatin-based chemotherapy resulted in a statistically significant improvement in overall survival (11.2 vs 9.8 months) compared with fluoropyrimidine/irinotecan- or fluoropyrimidine/oxaliplatin-based chemotherapy without bevacizumab.
On January 23, 2013, the U. S. Food and Drug Administration approved bevacizumab for use in combination with fluoropyrimidine-irinotecan- or fluoropyrimidine-oxaliplatin-based chemotherapy for the treatment of patients with metastatic colorectal cancer whose disease has progressed on a first-line bevacizumab-containing regimen.
Metastatic Non-Small-Cell Lung Cancer (NSCLC)
In an open-label, multicenter, randomized, active-controlled trial (n=878) among chemotherapy-naive patients with locally advanced, metastatic or recurrent NSCLC, the addition of intravenous bevacizumab to paclitaxel and carboplatin significantly improved the median overall survival compared to the combination regimen of paclitaxel and carboplatin alone. Patients were randomized in a 1:1 ratio to receive six cycles of paclitaxel 200 mg/m2, carboplatin target area under the concentration-time curve of six milligrams/milliliter per minute (mg/mL/min), plus bevacizumab 15-mg/kg (PCB) on day one (n=434) or a control regimen of the above paclitaxel/carboplatin (PC) combination (n=444) given IV once every three weeks. Bevacizumab as a single agent once every three weeks was continued in 53% patients in the PCB arm after completion or upon discontinuation of six cycles until disease progression or intolerable toxicity. Most patients (57%) were younger than 65 years of age (median age 63 years). There were more male patients in the PC than PCB group (58% vs. 50%, p=0.03). The vast majority (89%) had newly-diagnosed NSCLC, of whom 12% had Stage IIIB with malignant pleural effusion and 76% had Stage IV disease. Based on the intent-to-treat analysis with a median follow-up of 19 months, patients in the PCB arm achieved higher median overall survival (12.3 months) compared with individuals in the PC arm (10.3 months) HR 0.79; 95% CI, 0.67 to 0.92; p=0.003). Survival rates were superior among patients receiving the PC combination regimen with bevacizumab than without bevacizumab at one year (51% vs. 44%) and two years (23% vs. 15%). The addition of bevacizumab to paclitaxel and carboplatin compared with PC therapy alone also led to improvements in median progression-free survival (6.2 months vs. 4.5 months; HR=0.66; 95% CI, 0.57 to 0.77; p less than 0.001), and in overall response rates (35% vs. 15%, p less than 0.001). Subgroup analysis indicated that disease stage, prior radiation therapy, recurrent disease, and VEGF status did not correlate with overall survival. There was also a higher incidence of grade IV neutropenia (25.5% vs. 16.8%) and thrombocytopenia (1.6% vs. 0.2%) reported in the PCB arm than in the PC arm. PCB therapy was also associated with higher incidence of grades III to IV hypertension (7% vs. 0.7%), hyponatremia (3.5% vs. 1.1%), proteinuria (3.1% vs. 0%), headache (3% vs. 0.5%), rash (2.3% vs. 0.5%), and bleeding (4.4% vs. 0.7%) compared with PC alone (p less than 0.05).
The combination of PCB improved overall survival in patients with advanced NSCLC when compared to PC. In this phase II/III trial, patients (n=855 eligible) with chemotherapy-naive stage IIIB or stage IV NSCLC were randomized to either a treatment regimen of paclitaxel 200 mg/m2, carboplatin target area under the curve of six, plus bevacizumab 15-mg/kg on day one IV every three weeks or a control regimen of the above PC. Bevacizumab alone was continued after six cycles until disease progression. Patients had adenocarcinoma with no history or hemoptysis of brain metastasis. Interim results yielded a median overall survival that was significantly greater in the PCB group compared to the PC group (12.5 months (mo) and 10.2 mo, respectively; p=0.0075). Progression-free survival too was greater in the PCB versus PC group (6.4 mo, 4.5 mo; p less than 0.0001). Overall response rate was 27.2% and 10% (p less than 0.0001) in the PCB and PC groups, respectively. In a subgroup analysis, survival benefit was observed across treatment regimens with the exception of gender. Five complete responses were observed in the PCB group. Differences in grade IV toxicities between the two groups were noted for neutropenia (PCB=24%, PC=16.4%; p=0.006) and thrombocytopenia (PCB=1.4%, PC=0%; p=0.01). Differences in grade III or IV non-hematologic toxicities were hemorrhage (PCB=4.5%, PC=0.7%; p less than 0.001) and hypertension (PCB=6%, PC=0.7%; p less than 0.001). While there was no significant difference in treatment-related deaths between the two groups (PCB=9, PC=2), five deaths in the PCB group were attributed to hemoptysis.
In a randomized, open-label phase II study, bevacizumab with carboplatin and paclitaxel lengthened TTP and showed a trend towards improved tumor response compared to PC alone in patients with non-small cell lung cancer. Ninety-nine patients having stage IIIB, stage IV, or recurrent disease not treated within the previous two weeks were evaluated. The control regimen consisted of carboplatin (area under the curve of six) IV and paclitaxel 200 mg/m2 IV each given once every three weeks. Treatment groups received bevacizumab IV either as low-dose 7.5 mg/kg or high-dose 15-mg/kg with PC. Non-progressing patients could continue bevacizumab alone beyond the planned chemotherapy for up to 18 doses. Six cycles with PC were planned for all groups and a median of six were administered. Overall response rates trended in favor of patients in the bevacizumab high dose group (31.5%) and bevacizumab low dose group (28.1%), followed by the carboplatin and paclitaxel group (18.8%). Median TTP was 7.4 months in the bevacizumab high dose group and 4.2 months in the PC group (p=0.023). TTP was 4.3 months in the bevacizumab low dose group. Blinded independent review facility results reflected a trend similar to the investigators' for TTP; however, those results were not statistically significant. Median survival was 17.7, 11.6, and 14.9 months for the bevacizumab high dose, bevacizumab low dose, and carboplatin and paclitaxel groups, respectively. Treatment group results were not significant compared to control. Grade III to IV toxicities (value represents the number of patients) common to CP, CPB low dose, and bevacizumab high dose were leukopenia (7, 10, 13), nausea (1, 1, 2), vomiting (1, 1, 1), and thrombotic events (3, 2, 5), respectively. Those common to the bevacizumab low dose and bevacizumab high dose were diarrhea (3, 1), fever (2, 2), headache (1, 2), and hemoptysis (3, 1), respectively. Six patients with centrally located tumors close to major blood vessels experienced bleeding (hemoptysis or hematemesis); four events were fatal. Grade III to IV hypertension and peripheral neuritis were noted in one each of the CP group and two each of the bevacizumab high dose group. There were nine deaths not directly related to disease progression; four in each of the bevacizumab-treatment groups and one in the control.
Based on 2012 National Comprehensive Cancer Network (NCCN) guidelines on NSCLC, bevacizumab may be used as:
- First-line therapy in combination with chemotherapy for recurrence or metastasis for performance status 0-1 patients with tumors of nonsquamous cell histology and no history of hemoptysis.
- A single-agent maintenance therapy if given first line with chemotherapy for recurrence or metastasis in patients with performance status 0-1, tumors of nonsquamous cell histology, and no history of recent hemoptysis, who achieve tumor response or stable disease following first-line chemotherapy.
- Second-line therapy in combination with a platinum-based doublet for tumors of nonsquamous cell histology in patients with performance status 0-2 and no history of recent hemoptysis, if erlotinib or crizotinib given first line.
High Grade Gliomas of Brain
In a randomized, noncomparative, phase II trial (n=167), presented in abstract form and as an oral presentation, bevacizumab alone or in combination with irinotecan demonstrated favorable six-month progression-free survival (PFS-6), ORR, and overall survival in patients with relapsed, refractory GBM. Patients with recurrent GBM who had previously received temozolomide were eligible for study enrollment. Treatment consisted of IV BEV 10-mg/kg every two weeks alone (n=85; median age, 54 years [yr]; range, 23-78 yr) or BEV plus (+) irinotecan 125 mg/m2 (patients not on enzyme-inducing anti-epileptic drugs [EIAEDs]) or 340 mg/m2 (patients on EIAEDs) IV over 90 minutes (n=82; median age, 57 yr; range, 23-79 yr). Cycles were repeated every six weeks (BEV alone arm: median duration of BEV treatment, 16 weeks; BEV + irinotecan arm: median duration of BEV treatment, 22 weeks; median duration of irinotecan treatment, 21 weeks). Forty-four patients (52%) who were in the BEV alone arm had irinotecan added in an optional post-progressive phase. Tumor response was evaluated every six weeks using the Macdonald criteria 1990). PFS-6 was achieved in 42.6% (97.5% CI, 29.6%-55.5%) and 50.3% (97.5% CI, 36.8%-63.9%) of patients in the BEV alone and BEV + irinotecan arms, respectively. A complete response + partial response of 28.2% (97.5% CI, 18.5%-40.3%; median duration of response, 5.6 months; 95% CI, 3-5.75 months) and 37.8% (97.5% CI, 26.5%-50.8%; median duration of response, 4.3 months; 95% CI, 4.14 months to not yet reached) was observed in the BEV alone and BEV + irinotecan arms, respectively, with CR rates of 1.2% and 2.4%, respectively. The median overall survival at the data cutoff of November 15, 2007 was 9.2 months (95% CI, 8.2-10.7 months) in the BEV alone arm and 8.7 months (95% CI, 7.8-10.9 months) in the BEV + irinotecan arm. Additionally, corticosteroid use decreased from baseline (BEV alone, n=44; mean dose, 7.3 +/- 7.37-mg/kg; BEV + irinotecan, n=46; 8 +/- 4.96-mg/kg) to treatment day 168 (BEV alone, n=12; 1.1 +/- 1.04-mg/kg; BEV + irinotecan, n=15; 4.4 +/- 4.13- mg/kg) in both arms. Serious adverse events were reported in 26.2% of patients in the BEV alone arm (n=84) and 43% of patients in the BEV + irinotecan arm (n=79), with 2.4% and 1.3% of patients, respectively, having an adverse event which resulted in death. In the BEV alone and BEV + irinotecan arms, respectively, grade three or greater toxicities included infection (9.5% and 13.9%), venous thrombotic events (3.6% and 8.9%), and wound healing complications (2.4% and 1.3%). Hemorrhage (grade III or greater) was reported in two patients (2.5%) treated with BEV + irinotecan, with one of these patients experiencing a central nervous system bleed.
In a phase II study (n=35), treatment with BEV + irinotecan exhibited favorable PFS-6, response rates, and overall survival in patients with recurrent GBM. Patients (median age, 48 years [yr]; range, 18-66 yr; dexamethasone use at baseline, 22%) with recurrent GBM (median number of progressions, two; range, 1-5 progressions) who had previously received radiotherapy and concurrent temozolomide were eligible for enrollment. Patients received either IV BEV 10- mg/kg plus irinotecan 125-mg/m2 (patients not EIAEDs) or 340 mg/m2 patients on EIAEDs IV every two weeks (n=23) or bevacizumab 15-mg/kg IV every 21 days plus irinotecan 125 mg/m2 (patients not on EIAEDs or 350-mg/m2 patients on EIAEDs on days 1, 8, 22, and 29 of a 42-day cycle (n=12). Tumor response was evaluated using the Macdonald criteria; additionally, patients needed to be stable or clinically improved and have a stable or decreased corticosteroid dose to meet criteria for achieving a response. Because there was no statistical difference in results between patients treated with either BEV plus irinotecan dosing schedules, efficacy results were combined. PFS-6 was 46% (95% CI, 32%-66%), with a median PFS of 24 weeks (95% CI, 18-36 weeks). Twenty patients (57%; 95% CI, 39%-74%) had a partial tumor response or better; with seven patients having no residual high-grade tumor via positron emission tomography (PET) scan after 54 weeks of therapy. Additionally, the median overall survival was 42 weeks (95% CI, 35-60 weeks) at a median follow-up of 68 weeks and the six-month overall survival rate was 77% (95% CI, 64%-92%). Reasons for discontinuing treatment in this study included thromboembolic complications (n=4), grade II proteinuria (n=2), grade II fatigue (n=4), and central nervous system hemorrhage (n=1).
In a phase II study (n=48), treatment with bevacizumab produced a six-month progression-free survival (PFS-6) rate of 29% in patients with recurrent glioblastoma. Patients (median age, 53 years [yr]; range, 21 to 69 yr) with recurrent glioblastoma who had previously received radiotherapy and temozolomide were eligible for enrollment (median number of prior chemotherapy regimens, two; range, 1 to 7 regimens; three or greater regimens, 48%). Patients received intravenous bevacizumab 10 milligrams/kilograms (mg/kg) every 14 days and treatment cycles were repeated every 28 days. Nineteen patients who progressed on single-agent bevacizumab received irinotecan (125 mg/square meter (m2) or 340 mg/m(2) depending on whether patients received enzyme-inducing antiepileptic drugs) in addition to bevacizumab every 28 days as part of a companion trial; however, accrual was stopped before the planned sample size of 20 patients were enrolled due to lack of efficacy. The PFS-6 (primary endpoint) was 29% (95% confidence interval (CI), 18% to 48%), with a median PFS of 16 weeks (95% CI, 12 to 26 weeks). Additionally, the median overall survival (OS) was 31 weeks (95% CI, 21 to 54 weeks) and the six-month OS rate was 57% (95% CI, 44%-75%). The radiographic response rate (using magnetic resonance imaging) was 35% (complete response, n=1; partial response, n=16). Corticosteroid use decreased (average dose reduction from baseline, 59%) in 15 of 26 patients (58%) who were receiving corticosteroids at study enrollment. Thromboembolic events occurred in six patients (12.5%; grade 3, n=2; grade 4, n=4), and included three pulmonary embolisms and one cerebral vascular event. Additionally, one patient experienced a bowel perforation during the study period.
In a single-arm, single institution study (n=56), treatment with bevacizumab produced an objective response rate (based on both World Health Organization (WHO) radiographic criteria and stable or decreasing corticosteroid use) of 19.6% (95% CI, 10.9% to 31.3%) and a median duration of response of 3.9 months (mo) (95% CI, 2.4 to 17.4 mo). In this study, patients (median age 54 years, 54% male) with glioblastoma and documented disease progression after temozolomide and radiation therapy received intravenous bevacizumab 10 milligrams/kilogram (mg/kg) every two weeks until disease progression or unacceptable toxicity.
Metastatic Renal Cell Carcinoma
In a randomized, phase II trial of patients with metastatic renal cancer, bevacizumab prolonged time to progression, yet failed to extend survival. Patients were randomly assigned to bevacizumab (3-mg/kg [low-dose] or 10-mg/kg [high-dose] intravenously every two weeks after a loading dose of 150% of the assigned dose) or placebo after initial stratification for previous interleukin-2 therapy. Plasma levels of VEGF both bound and free were measured. TTP was prolonged by a factor of 1.26 (p=0.053) and 2.55 (p less than 0.001) in the low and high-dose groups, respectively, with the high-dose group reaching statistical significance (4.8 months vs. three months placebo; p less than 0.001). A 10% (95% CI 2.9 to 24.2%), OR (all complete responses) was achieved in the high-dose group. Plasma vascular endothelial growth factor levels were detectable at five and 13 weeks in both the low (median; 155 picogram [pg]/milliliter [mL], 170 pg/mL) and high-dose (196 pg/mL, 246 pg/mL) groups. Of 113 patients, 76 had undetectable pre-treatment levels of VEGF; no association between clinical response and TTP was noted. No grade IV bevacizumab related events were reported. Common adverse events related to bevacizumab were hypertension and proteinuria. Other adverse events that occurred in 10% or more of patients receiving bevacizumab included: epistaxis, fever without infection, malaise, hematuria, hyponatremia, and elevated alanine aminotransferase.
No statistical benefits were noted after twenty-two patients, whose disease progressed while on the placebo arm of the phase II trial, were crossed over to a low-dose arm of bevacizumab (n=10) or a low-dose arm of bevacizumab in combination with thalidomide (n=12). For both arms, a loading dose of bevacizumab 4.5-mg/kg was infused IV over two hours, followed by a treatment dose of 3-mg/kg one week later and then every two weeks after. In the combination arm, thalidomide was given orally the same day as the bevacizumab load at a starting dose of 200 mg then increased by 100-mg every two weeks to a maximum of 800-mg/day. More than 50% of the patients achieved a thalidomide dose of 500-mg/day. The difference in progression-free survival was found to be statistically insignificant between the two groups. In addition, neither treatment arm yielded objective responses. The median time to progression of disease for patients treated with bevacizumab-alone was 2.4 months compared to 3.0 months for subjects given bevacizumab and thalidomide (p=0.63). When compared to bevacizumab-alone, the addition of thalidomide yielded minimal effect on tumor burden. Progression-free survival was found to be similar between this study's two groups of patients and the 37 patients who had been initially randomized to low- dose bevacizumab. Likewise, changes in tumor burden were found to be non-statistically different between the original low-dose bevacizumab group and the two crossover groups described in this study (all p greater than 0.30). In this study, toxicity in the two groups was minimal. Neuropathy occurred in three patients in the combination arm. Only one of ten patients in the bevacizumab-alone arm experienced a grade III to IV toxicity (hand weakness) compared to three of twelve patients in the combination arm; one patient experienced malignant hypertension, one patient experienced abnormal laboratory parameters, and one patient experienced dizziness.
Metastatic Renal Cell Carcinoma, in Combination Therapy
In a multicenter, randomized, double-blind, phase III trial (AVOREN trial) (n=649), bevacizumab (BEV) plus (+) interferon alfa-2a (INF) demonstrated significantly better PFS and ORR in patients with previously untreated metastatic renal cell carcinoma (mRCC) compared to INF plus placebo. After a nephrectomy, patients with greater than 50% clear-cell histology mRCC, a Karnofsky performance status of 70% or greater, and no central nervous system metastases received INF at a target dose of nine million international units subcutaneously three times per week (for up to one year) plus either IV BEV 10-mg/kg every two weeks (BEV+INF, n=327; median age, 61 years [yr]; range, 30-82 yr) or placebo (INF alone, n=322; median age, 60 yr; range, 18-81 yr) until disease progression or unacceptable toxicity. Tumor response was evaluated using the RECIST criteria. At the time of data cutoff, the median duration of BEV therapy was 9.7 (range, 0-24.4 months) and 5.1 (range, 0-24 months) months for the BEV+INF and INF alone arms, respectively. Forty-six percent (n=151) of patients receiving BEV+INF and 70% (n=224) of patients receiving INF alone died or progressed during treatment. At a planned interim analysis (following 251 deaths), overall survival data was not yet mature (median overall survival: BEV+INF, not yet reached; INF alone, 19.8 months ; HR, 0.79; 95% CI, 0.62-1.02; p=0.067); however, because the primary objective of overall survival may have been confounded by patients in the INF alone arm receiving new second-line agents for mRCC, the study was unblinded and patients in the INF alone arm who had not progressed were crossed-over to the BEV+INF arm based on favorable PFS data. The median PFS was significantly longer in the BEV+INF arm (10.2 months vs. 5.4 months; HR, 0.63; 95% CI, 0.52-0.75; p less than 0.0001); additionally, the ORR was significantly higher (p less than 0.0001) in patients treated with BEV+INF (31%; complete response (CR), 1%) compared to patients treated with INF alone (13%; CR, 2%). Death due to causes other than progressive disease occurred in 2% of patients in each study arm, with three of eight deaths attributed to bevacizumab in the BEV+INF arm. Grade III or IV toxicity which occurred more frequently in the BEV+INF arm (n=337) compared to the INF alone arm (n=304) included fatigue (12% vs. 8%), asthenia (10% vs. 7%), proteinuria (7% vs. 0%), hypertension (3% vs. less than 1%), bleeding (3% vs. less than 1%), venous thromboembolism (2% vs. less than 1%), and gastrointestinal perforation (1% vs. 0%).
In a phase II clinical trial, combination therapy with erlotinib and bevacizumab produced an objective response rate of 25% in patients with metastatic or unresectable, locally recurrent clear cell renal cell carcinoma. Patients (median age, 61 years; range, 42-79 years) had previous nephrectomy (100%), were previously untreated (68%), or had received one prior treatment with interleukin-2 and/or interferon (32%). Patients (n=63) received IV bevacizumab (10-mg/kg) every two weeks and oral erlotinib (150 mg) once daily for twelve months or until disease progression. Response was evaluated using RECIST criteria after eight weeks of treatment (n=59). Median duration of treatment was eight months (range, one to 19+ months). Fourteen partial responses (PR) and one CR were reported for an ORR of 25% (95% CI, 16% to 37%). Thirty-six patients (61%; 95% CI, 48% to 72%) had stable disease or minor objective responses (MR; 22%). The ORR for patients who were previously untreated (n=39) was 31% compared to 15% in patient who received previous treatment with cytokine therapy (n=20). Median PFS was 11 months while median overall survival has not been reached after a median follow-up of 15 months. The 12-month and 18-month survival rates were 78% and 60%, respectively. Previously untreated patients had a longer PFS (12.9 months) compared to patients who had been previously treated (8.9 months; p=0.038). Grade III or IV toxicities included diarrhea (13%), rash (13%), nausea/vomiting (10%), hypertension (8%), bleeding (7%), proteinuria (8%), and pruritus (3%). Two patients discontinued treatment due to skin toxicity, one was considered severe. One case of grade IV gastrointestinal bleeding occurred.
Triple therapy with bevacizumab, erlotinib, and imatinib produced an overall response rate of 9% and a stable disease/minor response rate of 61% in patients with metastatic or unresectable locally recurrent clear cell renal cell carcinoma. Eligible patients for this phase I/II clinical trial had measurable disease, previous nephrectomy (85%), and no more than one prior regimen for metastatic renal cell carcinoma (71% untreated, 29% one prior therapy). For the first phase of the trial, imatinib doses were escalated to find the maximally tolerated dose (MTD) to use in phase II, which was 400-mg oral imatinib daily. For phase II, patients (n=44) received intravenous bevacizumab 10-mg/kg every two weeks, oral erlotinib 150-mg daily, and oral imatinib 400-mg daily until disease progression. Patients were evaluated for response every eight weeks. After median follow-up of five months (range three to ten months), four patients (9%) had partial response, 27 (61%) had stable disease (including six patients (14%) with minor response), and 13 (30%) had disease progression. Nine-month progression-free and overall survival was 66% and 70%, respectively. The most commonly reported grade III to IV toxicities were diarrhea (29%), rash (27%), and nausea/vomiting (13%).
Patients with treatment-naïve mRCC were evaluated in a multicenter, randomized, double-blind, international study comparing Avastin plus interferon alfa 2a (IFN- α2a) versus placebo plus IFN-α2a. A total of 649 patients who had undergone a nephrectomy were randomized (1:1) to 719 receive either Avastin (10 mg/kg IV infusion every 2 weeks; n = 327) or placebo (IV every two weeks; 720 n = in combination with IFN-α2a (9 MIU subcutaneously three times weekly, for a maximum of 52 weeks). Patients were treated until disease progression or unacceptable toxicity. The main outcome measure of the study was investigator-assessed PFS. Secondary outcome measures were ORR and OS.
The median age was 60 years (range 18−82), 96% were white, and 70% were male. The study population was characterized by Motzer scores as follows: 28% favorable (0), 56% intermediate, 8% poor (3-5), and 7% missing.
PFS was statistically significantly prolonged among patients receiving Avastin plus IFN-α2a compared to those receiving IFN-α2a alone; median PFS was 10.2 months vs. 5.4 months [HR 0.60 (95% CI 0.49, 0.72), p-value < 0.0001, stratified log-rank test]. Among the 595 patients with measurable disease, ORR was also significantly higher (30% vs. 12%, p < 0.0001, stratified CMH test). There was no improvement in OS based on the final analysis conducted after 444 deaths, with a median OS of 23 months in the Avastin plus IFN-α2a arm and 21 months in the IFN-α2a plus placebo arm [HR 0.86, (95% CI 0.72, 1.04)] U.S.
Metastatic Breast Cancer
In an open-label, randomized, phase III study, women (n=722) with metastatic breast cancer treated with bevacizumab plus paclitaxel had significantly improved progression-free survival (PFS) compared to patients who received paclitaxel alone. Patients with metastatic breast cancer, an ECOG performance status of zero or one, no prior therapy for metastatic disease (hormonal therapy or cytotoxic adjuvant therapy was allowed and previous adjuvant taxane therapy was permitted if patients had a disease-free interval of at least 12 months after treatment), and any human epidermal growth factor receptor type-2 (HER2) status (HER2-postitive breast cancer permitted only if patients had previously received trastuzumab) were randomized to receive either IV paclitaxel 90 mg/m2 on days 1, 8, and 15 plus bevacizumab 10-mg/kg IV on days one and 15 of a 28-day cycle (PAC + BEV) (n=368; median duration of paclitaxel therapy, 7.1 months) or paclitaxel alone (n=354; median duration of paclitaxel therapy, 5.1 months) until disease progression or unacceptable toxicity. Disease response was evaluated at baseline and every 12 weeks until progression using the RECIST. Twenty-one and 28 patients in the PAC + BEV and paclitaxel alone treatment groups, respectively, were considered ineligible by the investigators; therefore, 347 patients in the PAC + BEV arm (median age, 56 years [yr]; range, 29-84 yr; median duration of follow-up, 41.6 months) and 326 patients in the paclitaxel alone arm (median age, 55 yr; range, 27-85 yr; median duration of follow-up, 43.5 months) were included in the efficacy analysis. PFS was significantly improved in patients who received PAC + BEV compared to patients who received paclitaxel alone (11.8 months vs. 5.9 months; HR=0.6; 95% CI, 0.51-0.7; p less than 0.001). Objective response rates (ORR) were also significantly improved in the PAC + BEV arm (36.9%) compared to the paclitaxel alone arm (21.2%; p less than 0.001). There was no significant difference in the secondary endpoint of overall median survival between the two treatment arms (26.7 months vs. 25.2 months; HR=0.88; p=0.16); however, the one-year survival rate was higher in the PAC + BEV arm (81.2% vs. 73.4%; p=0.01). Additionally, most patients’ deaths (88.8%) were attributed to progressive disease. No significant quality of life differences were observed between the two treatment arms (evaluating the mean change in scores from baseline) using the Functional Assessment of Cancer Therapy-Breast (FACT-B) scale, FACT-B subscale, or the Trial Outcome Index. Adverse effects reported more frequently in patients treated with PAC + BEV (n=365) compared to patients treated with paclitaxel alone (n=346) included grade III or IV sensory neuropathy (23.5% vs. 17.7%; p=0.05), hypertension (14.8% vs. 0%; p less than 0.001), infection (9.3% vs. 2.9%; p less than 0.001), fatigue (9.1% vs. 4.9%; p=0.04), proteinuria (3.5% vs. 0%; p less than 0.001), headache (grade III only, 2.2% vs. 0%; p=0.008), and cerebrovascular ischemia (1.9% vs. 0%; p=0.02).
Recurrent Epithelial Ovarian Cancer
Guidelines from the National Comprehensive Cancer Network (NCCN) state that bevacizumab is an acceptable alternative chemotherapeutic regimen for recurrent epithelial ovarian cancer for Stage II, III, and IV patients with partial responses to their primary paclitaxel and platinum-based chemotherapeutic regimens. The guidelines note that bevacizumab has been demonstrated to be active in recurrent epithelial ovarian cancer, although it may cause arterial thrombosis and intestinal perforation.
Wright and colleagues reported a small retrospective study of 23 patients with epithelial ovarian carcinoma treated with bevacizumab in combination with other cytotoxic agents, a best overall response rate of progesterone resistance was noted in eight patients (35%). Ten patients had stable disease (44%) and five patients had progressive disease. Patients with progressive resistance had a median TTP of 5.6 months compared with 2.3 months for patients with stable disease. Three participants had a progression free interval greater than six months. Common adverse effects included proteinuria and hypertension. Thromboembolic complications occurred 9% in addition to two patients (4%) bleeding complications. Bowel obstruction occurred in five (22%) patients, with two patients (9%) experienced bowel perforations.
In a retrospective review of 32 assessable patients with refractory ovarian cancer treated with bevacizumab alone or bevacizumab with additional cytotoxic agents, Monk observed a 16% response rate and 62.5% stable disease rate in those patients treated with bevacizumab alone. The author also noted another study of single agent bevacizumab as third-line therapy for ovarian carcinoma was closed early (2005) due to bowel perforation. Thus, further trials investigating the use of bevacizumab for ovarian cancer are recommended.
Second Line Therapy for Metastatic Cancer of the Cervix
Bevacizumab is reported to have shown a partial response and usefulness in second line therapy for metastatic squamous cell carcinoma of the cervix. In their 2010 Guidelines, the NCCN recommends Bevacizumab as a single (Category 2B) agent for treatment of several forms of cervical cancer and in local or regional recurrence in patients who have received prior radiation therapy and patients with distant metastases. Their recommendation is based on the currently ongoing National Institute of Health Phase III trial.
A vascular endothelial growth factor (VEGF) is a key promoter of tumor progression in cervical carcinoma. The Gynecologic Oncology Group (GOG) is currently conducting a phase III trial to assess the efficacy and tolerability of bevacizumab as a VEGF inhibitor in advanced cervical cancer. Monk et al. concluded that Bevacizumab seems to be well tolerated and active in the second- and third-line treatment of patients with recurrent cervical cancer.
In another review of the Phase II trial by Monk et al. it was concluded that “the dominant role of angiogenesis in cervical cancer seems to be directly related to the human papillomavirus (HPV) inhibition of p53 and stabilization of hypoxia-inducible transcription factor 1 alpha (HIF-1 alpha), both of which increase VEGF. Bevacizumab’s binding and subsequent inactivation of VEGF seems to shrink cervical tumors and delay progression without appreciable toxicity, and is therefore being studied in a Gynecologic Oncology Group (GOG) phase III trial.”
Carcinoma of Prostate
Intravenous bevacizumab alone did not produce impressive results in patients with hormone-refractory prostate cancer in a small study (n=15). Patients were treated with 10-mg/kg every two weeks for six infusions (one cycle); repeat cycles were given to selected patients. No patient achieved a 50% reduction in serum prostate-specific antigen (PSA) levels at 70 days. A decrease in PSA of 25 to 50% occurred in one patient, and up to a 25% decrease was seen in three others. The McGill Present Pain Index indicated a trend toward increased pain during the study. Asthenia was the most common adverse event (40%).
Bevacizumab in Advanced Adenocarcinoma of the Pancreas
Because vascular endothelial growth factor (VEGF) appears to play a role in pancreatic cancer, bevacizumab was considered a promising therapy, and the results of two Phase II trials seemed to indicate potential benefit as well. Approximately 89–93% of pancreatic cancer patients have a VEGF mutation, which is associated with early recurrence after surgery, liver metastases, and poor prognosis. Finding VEGF in tumors is also correlated with tumor size.
Five studies were identified for review as part of a 2009 Blue Cross Blue Shield Association (BCBSA) Technology Evaluation Center (TEC) Assessment that looked at the use of bevacizumab in patients with advanced adenocarcinoma. These studies consisted of two Phase III trials, two Phase II trials, and one Phase I trial. In all trials, bevacizumab was added to gemcitabine, the latter of which is considered the current standard of care. Some trials also included other agents as well, including cisplatin and erlotinib.
The two Phase III studies, one by Kindler and colleagues and the second by van Cutsem and colleagues, provided the strongest evidence because of their design. Neither study demonstrated that the addition of bevacizumab resulted in a statistically significant difference in the primary outcome of overall survival (OS). For the secondary outcome of progression-free survival (PFS), the van Cutsem et al. study appeared to show benefit while the Kindler et al. study did not.
Kindler and colleagues randomly assigned 590 patients with advanced cancer (local or metastatic) to gemcitabine with or without bevacizumab. The trial was stopped early when it was determined that the combination of gemcitabine plus bevacizumab could not achieve longer survival than the gemcitabine-alone arm of the trial.
Van Cutsem and colleagues randomly assigned 607 patients with metastatic adenocarcinoma of the pancreas to gemcitabine plus erlotinib, with or without bevacizumab. There was no statistically significant difference between the two groups in the primary outcome of OS. The median OS was 7.1 months for the treatment group and 6.0 months for the control group (hazard ratio [HR]: 0.89; 95% confidence interval [CI]: 0.74–1.07%; p=0.21). The study reported a statistically significant difference in PFS of 4.6 months in the treatment arm and 3.6 months in the control group (HR: 0.73; 95% CI: 0.61–0.86%; p=0.0002). Although this secondary outcome was significant, there were few details given regarding the methods used to assess PFS, which may be subject to greater measurement error than OS.
In 2009, Crane and colleagues assessed the one-year survival in a case series of patients with locally-advanced, unresectable pancreatic cancer. Over expression of VEGF by pancreatic cancer cells may diminish the effectiveness of radiation therapy. The authors suggest that the action of bevacizumab in reducing VEGF expression may therefore improve radiation therapy outcomes. One-year overall survival in patients with locally-advanced disease receiving radiation and paclitaxel without VEGF inhibition was 43% in a previous study To examine the effectiveness of the combination of bevacizumab, capecitabine, and radiation, 94 patients were recruited to detect a 15% improvement in the one-year survival rate over the 43% seen in the previous study. Crane et al. reported a 47% survival rate (95% CI, 36% to 57%, not statistically different compared to the previous study). The authors concluded that the addition of bevacizumab does not contribute to an improved one-year survival rate.
Javle and colleagues reported the results of a single-arm study of 50 patients with metastatic pancreatic cancer treated with capecitabine, gemcitabine, and bevacizumab. The primary endpoint was progression-free survival (PFS) based on radiologic measure or tumor size, and serum CA 19-9 levels, the latter a biomarker associated with pancreatic cancer burden. Secondary endpoints included overall survival, response rate, and toxicity. In this study, discontinuation of therapy was higher than anticipated, with 46 of 50 (92%) subjects not continuing due to death, disease progression, or adverse events. Although some radiologic response and reduction in CA 19-9 levels did occur, PFS at one year was 19% (95% CI: 9.4-31.6%). Overall survival at one year reached 35.5% (95% CI: 21.7-49.5). The authors commented that, based on this study and previous results for bevacizumab, they have chosen not to proceed with Phase III studies.
In 2010, Astsaturov and colleagues reported on a comparative, non-blinded trial for patients with metastatic pancreatic cancer. Patients were randomly assigned to receive bevacizumab alone or in combination with docetaxel cytotoxic therapy. The primary endpoint was PFS. However, at four months, only two and three patients were stable, and the trial was discontinued on the study-defined grounds of futility of less than 25% PFS at that time.
Ko and colleagues reported partial results of an observational study of gemcitabine refractory metastatic pancreatic cancer patients treated with bevacizumab and erlotinib. Recruitment stalled after publication of the relative ineffectiveness of bevacizumab in the Phase III trials. Of the 36 patients followed in the study, eight reached the primary endpoint of six-month survival (22%). This survival rate is inferior to published rates of cytotoxic regimens.
According to the National Comprehensive Cancer Network (NCCN) guidelines, gemcitabine is recommended for pancreatic cancer patients with locally advanced or metastatic disease, and studies have suggested some benefit from adding another chemotherapy agent (such as cisplatin or fluoropyrimidine). The guidelines state that the only new targeted drug (such as bevacizumab) for which there is evidence of a statistically significant increase in survival when combined with gemcitabine is erlotinib. NCCN therefore “recommends gemcitabine-erlotinib combination therapy as an option for patients with locally advanced or metastatic disease and good performance status” and makes no recommendations regarding bevacizumab.
Treatment of advanced adenocarcinoma of the pancreas with bevacizumab is not an FDA-approved indication. The available evidence does not clearly demonstrate that addition of bevacizumab to chemotherapy regimens for advanced adenocarcinoma of the pancreas improves the net health outcome of those patients. Therefore, bevacizumab for patients with advanced adenocarcinoma of the pancreas continues to be considered experimental, investigational and unproven.
Disclaimer for coding information on Medical Policies
Procedure and diagnosis codes on Medical Policy documents are included only as a general reference tool for each policy. They may not be all-inclusive.
The presence or absence of procedure, service, supply, device or diagnosis codes in a Medical Policy document has no relevance for determination of benefit coverage for members or reimbursement for providers. Only the written coverage position in a medical policy should be used for such determinations.
Benefit coverage determinations based on written Medical Policy coverage positions must include review of the member’s benefit contract or Summary Plan Description (SPD) for defined coverage vs. non-coverage, benefit exclusions, and benefit limitations such as dollar or duration caps.