The authors performed a phase 2 study of bevacizumab plus pemetrexed and carboplatin followed by maintenance bevacizumab in patients with advanced, nonsquamous nonsmall cell lung cancer.
The authors performed a phase 2 study of bevacizumab plus pemetrexed and carboplatin followed by maintenance bevacizumab in patients with advanced, nonsquamous nonsmall cell lung cancer.
Previously untreated patients with advanced, nonsquamous nonsmall cell lung cancer and an Eastern Cooperative Oncology Group performance status of 0 or 1 received bevacizumab 15 mg/kg, pemetrexed 500 mg/m2 and carboplatin at an area under the concentration-time curve of 6 intravenously on day 1 every 21 days. Responding or stable patients who completed 6 cycles then received bevacizumab maintenance every 21 days until disease progression.
In total, 43 patients (40 who were evaluable for response) were entered on the study. Treatment-related grade 3/4 toxicities were low and included febrile neutropenia (2%), neutropenia (28%), anemia (18%), thrombocytopenia (11%), hypertension (7%), epistaxis (5%), venous thrombosis (8%), dyspnea (7%), rectovaginal fistula (2.3%), infusion reaction (2%), and cerebrovascular event (2%). One patient died from complications of venous thromboembolism and cerebrovascular accident after Cycle 2. Minimal clinically significant toxicity occurred during maintenance bevacizumab. Two complete responses (5%) were observed, and 17 patients (42%) had a partial response. Fifteen patients (38%) displayed disease stability. The overall disease control rate was 85%. At a median follow-up of 15.8 months, the median progression-free survival was 7.1 months (95% confidence interval, 5.9-8.3 months), and the median overall survival was 17.1 months (95% confidence interval, 8.8-25.5 months).
Combined bevacizumab, pemetrexed, and carboplatin followed by maintenance bevacizumab was well tolerated and displayed remarkable activity in patients with previously untreated, advanced, nonsquamous nonsmall cell lung cancer. Cancer 2012. © 2012 American Cancer Society.
Lung cancer remains a therapeutic challenge. It is the leading cause of cancer death worldwide, and approximately 85% of patients who are diagnosed with this neoplasm will die from their disease. There are more than 1.6 million new cases of lung cancer worldwide annually, and approximately 75% to 80% are nonsmall cell lung cancer (NSCLC).1-3 The majority of patients present with advanced disease, and 40% to 70% of those with stage I through III NSCLC develop distant metastatic disease after initial therapy.
Treatment options for patients advanced NSCLC previously were limited, because most clinical trials of cytotoxic agents produced <1 year of survival, many at the cost of significant toxicities.4-6 In the past, front-line, systemic therapy typically was discontinued after 4 to 6 cycles in responding patients, because there were no convincing data indicating that patients could benefit from continuation of the same therapy, in part because of cumulative toxicity. Therefore, long periods of disease control were not achievable in most patients, because the disease typically would progress within 3 to 4 months of stopping front-line therapy, and second-line agents would then be options only for patients with reasonable performance status.
To our knowledge, Eastern Cooperative Oncology Group (ECOG) study 4599 was the first to suggest that the continuation of 1 component of initial combination therapy in patients who had a benefit could improve survival.7 In that phase 3 study, >800 previously chemotherapy-naive patients with stage IIIB or IV nonsquamous NSCLC were randomized to receive either bevacizumab plus paclitaxel and carboplatin for 6 cycles followed by maintenance bevacizumab or paclitaxel and carboplatin alone for up to 6 cycles. A median survival of 12.3 months was achieved in the group assigned to chemotherapy plus bevacizumab compared with 10.3 months in the chemotherapy-alone group; and the median progression-free survival (PFS) for the 2 groups was 6.2 months and 4.5 months, respectively, with corresponding response rates of 35% and 15% (P < .001). The results from that trial established a new standard of care in this patient population. The mode of action of bevacizumab, a monoclonal antibody to vascular endothelial growth factor (VEGF), and its lack of cumulative toxicity made it particularly compelling for use as maintenance therapy in patients with nonsquamous NSCLC.
Shortly after the publication of ECOG 4599, a randomized trial of the novel antifolate pemetrexed proved that it was not inferior to docetaxel as second-line treatment for NSCLC.8 A subsequent study revealed that combination pemetrexed and cisplatin was superior to gemcitabine and cisplatin in the front-line setting for patients with nonsquamous histology,9 and pemetrexed is now used routinely in combination with a platinum agent for previously untreated patients with nonsquamous NSCLC.
The finding of superior activity of pemetrexed in patients with nonsquamous NSCLC and its overall favorable toxicity profile suggest that it potentially may build on the results of ECOG 4599 when combined with bevacizumab and carboplatin in a similar fashion. Here, we present the results from our phase 2 study of this combination followed by maintenance bevacizumab in patients with advanced, nonsquamous NSCLC.
Eligible patients were aged ≥18 years and had cytologically or histologically confirmed, nonsquamous NSCLC that was stage IIIB (malignant pleural effusion) or IV and had received no prior chemotherapy, immunotherapy, biologic therapy, or antiangiogenic drugs. Patients also were required to have measurable disease; an ECOG performance status of 0 or 1; and adequate hematologic, hepatic, and renal function as defined by the following required laboratory values: serum creatinine <1.5 mg/dL, absolute neutrophil count ≥1500/mm3, platelet count ≥100,000/mm3, total bilirubin <2 times the upper limit of normal, and aspartate and alanine aminotransferase levels <2.5 times the institutional upper limit of normal (for patients with hepatic involvement of their cancer, aspartate and alanine aminotransferase levels <5 times the upper limit of normal were required). Patients with a history of hypertension were required to be well controlled (blood pressure <150/100) on a stable regimen of antihypertensive therapy. Patients also were required to have completed any radiation therapy >3 weeks before registration. Patients with squamous cell histology, or with a lung carcinoma of any histology in close proximity to a major vessel, or with cavitation or evidence of bleeding diathesis or coagulopathy were excluded as were patients who were receiving anticoagulation or who had a history of gross hemoptysis (defined as bright red blood of a half teaspoon or more). This study was approved by the institutional review board of each participating institution.
At the initiation of the protocol, patients were required to undergo central nervous system imaging as part of study screening and were not eligible if brain metastases were present. Patients with treated brain metastases also were not eligible. After safety data were provided by the sponsor, the protocol was amended in January 2008 to allow patients with treated brain metastases onto the study if there was no evidence of progression or hemorrhage after treatment, as ascertained by clinical examination and brain imaging studies (magnetic resonance imaging or computed tomography) during the screening period, and if they had no ongoing requirement for dexamethasone. Patients were not eligible if they had central nervous system metastases treated by resection or brain biopsy; significant vascular disease within 6 months; a history of abdominal fistula, perforation, or intra-abdominal abscess within 6 months; uncontrolled illness, including prior hypertensive crisis or hypertensive encephalopathy, ongoing or active infection, symptomatic congestive heart failure (New York Heart Association grade II or greater heart failure), and active cerebrovascular or cardiovascular disease within the prior 6 months; a major surgical procedure within 28 days before entry; or minor surgery or needle biopsies within 7 days before entry. Women of childbearing potential and men partnered with women of childbearing potential were required to use an effective method of contraception beginning before study entry, for the duration of study participation, and for a minimum of 3 months after the last dose of chemotherapy. Women of childbearing potential were required to have a baseline negative pregnancy test within 7 days of receiving the first dose of study medication. Patients who were receiving chronic daily treatment with aspirin (>325 mg daily) or other agents known to inhibit platelet function were not eligible.
This was a 2-site, single-arm, phase 2, open-label study. Treatment was administered on an outpatient basis. Vitamin B12 (1000 mg) was administered every 9 weeks starting 1 week before the first cycle until 2 months after the last dose of pemetrexed. Patients received folic acid 1 mg orally every day 1 week before the first cycle and continued until 2 months after the last dose of pemetrexed. Bevacizumab, pemetrexed, and carboplatin were administered on day 1 every 3 weeks. Patients received commercially available pemetrexed 500 mg/m2 as a 10-minute intravenous infusion on day 1, carboplatin (commercially available) at an area under the concentration-time curve of 6 as a 30-minute intravenous infusion on day 1 immediately after pemetrexed, and bevacizumab (supplied by Genentech Inc., South San Francisco, Calif) 15 mg/kg as a 30-minute to 90-minute intravenous infusion on day 1 immediately after carboplatin. Patients who responded or who had stable disease received a maximum of 6 cycles of bevacizumab, pemetrexed, and carboplatin. After the initial 6 cycles, patients were maintained on bevacizumab 15 mg/kg alone every 3 weeks as a 30-minute infusion until they had evidence of disease progression or unacceptable toxicity.
Patients were evaluated for response every other cycle for the first 6 cycles and after every third cycle during the maintenance phase. Response was determined according to Response Evaluation Criteria in Solid Tumors (RECIST) guidelines. All toxicities were graded using National Cancer Institute Common Toxicity Criteria, version 3.0 (available at: ctep.cancer.gov; [accessed April 5, 2012]).
The primary objective of the study at the time of original design was to determine the time to disease progression (TTP) with the combination of bevacizumab, pemetrexed, and carboplatin. Secondary objectives included the assessment of toxicities and the measurement of response rate, PFS, and overall survival (OS). TTP was chosen originally as the primary endpoint for determining the sample size based on results from the randomized phase 2 trial of bevacizumab, paclitaxel, and carboplatin reported by Johnson et al.10 The results from ECOG 4599 had not been published at the time this study was conceived (OS and PFS were reported). We present the results for PFS in the current report, because this endpoint is more relevant since the publication of ECOG 4599 and the subsequent approval of bevacizumab in advanced nonsquamous NSCLC.
Given the TTP of 8 months observed with the combination of 15 mg/kg bevacizumab plus paclitaxel/carboplatin in the randomized phase 2 trial10 and the expected TTP of 4 months from randomized studies of platinum-based combinations in NSCLC,5 a similar TTP of 8 months with the combination would be considered evidence of activity and would be worthy of further study. With a planned sample size of 40 patients, under the assumption of exponential survival, the median of 8 months suggested a 35.4% TTP at 12 months, whereas the median of 4 months suggested a 17.33% TTP at 12 months. Forty-two patients were to be accrued in a 1-stage design with the assumption that, at most, 2 patients would drop out, so that the final sample size would be 40 patients. If the true 12-month TTP rate proved to be ≤17.33%, then this design would have at most a 3.4% probability of declaring the agent worthy of further study. If the true 12-month TTP was ≥35.4%, then this design would have at least an 80% probability of declaring the agent worthy.
PFS was measured from the time of registration to the time of documented disease progression. Disease progression was assumed in patients who died before documentation of progression or in patients who started therapy and were lost to follow-up. OS was calculated from the date of registration to the date of death. Survival data were analyzed using the Kaplan-Meier method.
Between September 2005 and December of 2009, 43 eligible patients were enrolled at 2 institutions: Cooper Cancer Institute, Voorhees, NJ and the University of Pennsylvania, Philadelphia, Pa. Of these, 40 patients were evaluable for response. Three patients discontinued treatment after 1 cycle because of toxicity. All 43 patients were included in the current toxicity and survival analyses. No patients remained on study at the time of this report.
Patient demographics are provided in Table 1. The median age was 65.3 years (range, 34.6-80.4 years). Twenty-three patients (54%) were women, and 20 patients were men (46%). Thirty-three patients had adenocarcinoma histology, 9 patients had NSCLC not otherwise specified, and 1 patient had large cell carcinoma. Eighty-eight percent of patients had stage IV disease, and 84% had an ECOG performance status of zero.
|Characteristic||No. of Patients (%)|
|Age: Median [range], y||65.3 [35-80]|
|African American||7 (16)|
|East Asian||1 (2)|
|Large cell carcinoma||1 (2)|
|NSCLC NOS||9 (21)|
|ECOG performance status|
|IIIB with malignant effusion||6 (12)|
|Site of metastatic involvement|
The median number of cycles delivered was 9 (range, 1-65 cycles). Twenty-eight patients (65%) received the full 6 or more cycles of chemotherapy and bevacizumab, and 25 patients (58%) went on to receive maintenance bevacizumab. Of those who started bevacizumab maintenance, the median number of cycles was 6 (range, 1-59 cycles). Reasons for stopping treatment included disease progression (34 patients), toxicity (7 patients), patient choice (1 patient), and death (1 patient died of respiratory failure after aspiration pneumonia in the setting of venous thromboembolic complications after Cycle 2). Bevacizumab maintenance was discontinued for disease progression in 22 of 25 patients: One patient developed nephrotic syndrome after 21 cycles, 1 patient had a treatment delay because of comorbidities that required study discontinuation after Cycle 22, and 1 patient transferred care to another institution after Cycle 9.
Overall grade 3 and 4 hematologic and nonhematologic toxicities were low and reflected known potential side effects of these agents. Grade 1/2 and Grade 3/4 treatment-related adverse events are listed in Tables 2 and 3, respectively.
|Toxicity During Cycles 1-6||No. of Patients (%)|
|AST/ALT elevation||20 (46)|
|No. of Patients (%)|
|Toxicity||Grade 3||Grade 4|
|Neutropenia||6 (14)||6 (14)|
|Anemia||4 (9)||2 (9)|
|Thrombocytopenia||1 (2)||2 (9)|
|Febrile neutropenia||1 (2)||0 (0)|
|Infusion reaction||1 (2)||0 (0)|
|Vomiting||2 (5)||0 (0)|
|Diarrhea||1 (2)||0 (0)|
|Fistula||1 (2)||0 (0)|
|Fatigue||2 (2)||0 (0)|
|Dyspnea||3 (7)||0 (0)|
|Hypertension||3 (7)||0 (0)|
|Epistaxis||2 (4)||0 (0)|
|Venous thrombosis||3 (7)||1 (2)|
|Cerebrovascular||1 (2)||0 (0)|
|Nausea||1 (2)||0 (0)|
|Headache||1 (2)||0 (0)|
|Proteinuria||0 (0)||1 (2)|
|Elevated AST||1 (2)||0 (0)|
|Response||No. of Patients (%)|
|Complete response||2 (5)|
|Partial response||17 (42)|
|Stable disease||15 (38)|
|Progressive disease||6 (12)|
|Survival||Median [95% CI], mo|
Most hematologic toxicities were without clinical sequelae. Grade 1/2 neutropenia or thrombocytopenia as a worst grade during Cycles 1 through 6 was observed in 12 patients (28%) and 22 patients (51%), respectively. One patient developed febrile neutropenia, 6 patients (14%) had grade 4 neutropenia, 2 patients (4%) had grade 4 anemia, and 2 patients (4%) had grade 4 thrombocytopenia. Supportive therapies like pegfilgrastim were not used as primary prophylaxis.
Grade 1/2 nonhematologic toxicities as a worst grade during Cycles 1 through 6 included: nausea (20 patients; 46%), vomiting (11 patients; 26%), fatigue (33 patients; 77%), stomatitis/mucositis (11 patients; 26%), diarrhea (14 patients; 32%), headache (5 patients; 12%), hypertension (6 patients; 14%), proteinuria (15 patients; 35%), epistaxis (20 patients; 46%), and elevated aspartate and alanine aminotransferase levels (20 patients; 46%). More severe toxicities observed during Cycles 1 through 6 were grade 3/4 fatigue (2 patients; 4%), grade 3 infusion reaction (1 patient; 2%), grade 3 rectovaginal fistula in a patient who had a history of diverticulosis (2%), grade 3 dyspnea (3 patients; 7%), grade 3 hypertension (3 patients; 7%), grade 3 epistaxis (2 patients; 4%), grade 3/4 venous thromboembolism (4 patients; 9%), and cerebrovascular event (1 patient; 2%). There were no episodes of pulmonary hemorrhage. With the exception of hypertension, epistaxis, and fatigue, it is not clear whether these adverse events were attributable to study therapy. Toxicities during maintenance bevacizumab were uncommon: One patient each had grade 3 nausea, grade 3 headache, and grade 4 proteinuria (removed from study), and 1 patient had grade 3 aspartate aminotransferase elevation with progressive liver metastases.
Forty patients were assessable for response (Table 4). Two patients (5%) had a complete response, and 17 patients (42%) had a partial response for an overall response rate of 47%. Fifteen patients (38%) had stable disease. The disease control rate was 85%.
At a median follow-up of 15.8 months (range, 1.1-63 months), the median PFS was 7.1 months (95% CI, 5.9-8.3 months), and the median OS was 17.1 months (95% CI, 8.8-25.5 months). The 1-year and 2-year survival rates were 56% and 21%, respectively. Six patients (14%) were alive >3 years after study entry, and 9 patients remained alive at the date of last follow-up (March 24, 2011). Figures 1 and 2 provide Kaplan-Meier estimates PFS and OS, respectively.
The positive results from this study build on the paradigm shift in the front-line treatment of nonsquamous NSCLC produced by the results from ECOG 4599. Not only did ECOG 4599 demonstrate that an antiangiogenic agent (bevacizumab) added to the efficacy of platinum-based doublet therapy in terms of improved response rate, TTP, and OS, it also established a potential role for maintenance therapy after initial combination therapy. In our study, the substitution of pemetrexed for paclitaxel resulted in better patient tolerance with less hematologic toxicity than what was observed in ECOG 4599 (26% vs 14% grade 4 neutropenia); and it also resulted in less “in-chair” time for patients, because pemetrexed is administered as a 10-minute infusion, whereas paclitaxel is administered as a 3-hour infusion after additional premedication to prevent allergic reaction. Pemetrexed also is associated with less alopecia, musculoskeletal toxicity, and neurologic toxicity than paclitaxel. It is noteworthy that no unexpected toxicities emerged with the addition of pemetrexed to carboplatin and bevacizumab. The efficacy results from this study also compare very favorably to the ECOG 4599 results with an overall response rate of 46% compared with 35% and an overall median survival of 17.1 months versus 12.3 months, recognizing that ours was a phase 2 trial with a higher proportion of patients who had an ECOG performance status of zero (84% vs 40%). Fifty-eight percent of patients were able to receive bevacizumab maintenance on our trial, similar to the rate of 53% in ECOG 4599. Bevacizumab maintenance was discontinued for disease progression in 88% of patients who received it in our study, highlighting our finding that bevacizumab monotherapy was very well tolerated overall.
The only other reported randomized data for front-line bevacizumab-containing combinations versus chemotherapy alone in nonsquamous patients come from the 3-armed Avastin in Lung Cancer (AVAiL) trial of cisplatin, gemcitabine, plus either bevacizumab (7.5 mg/kg or 15 mg/kg) or placebo.11 The authors reported improved PFS in the bevacizuamb-containing arms; however, no differences in OS were noted. Remarkable OS was observed in all 3 arms at >13 months. The investigators believed this was because additional lines of therapy, including pemetrexed, were received by patients in all 3 treatment arms (>60% of patients per arm).
Since the initiation of our study and the reports of ECOG 4599 and AVAiL, we have learned much about the role of pemetrexed in the treatment of NSCLC. We now know that pemetrexed has superior activity in nonsquamous versus squamous NSCLC when received alone or in combination with a platinum agent, and now it is used exclusively in nonsquamous subtypes. Most recently, Ciuleanu et al reported improved PFS (4.5 months vs 2.6 months) and OS (15.5 months vs 10.3 months) in patients with nonsquamous NSCLC when they received pemetrexed as maintenance therapy until progression developed after 4 cycles of a platinum-containing doublet versus placebo plus best supportive care.12 That study led to the approval of “switch” maintenance therapy with pemetrexed. Given the relatively short infusion times and similar schedule of administration, the prospect of the combination of 2 maintenance therapies (bevacizumab and pemetrexed) in nonsquamous NSCLC has become attractive. Patel et al conducted a phase 2 study of bevacizumab, pemetrexed, and carboplatin followed by maintenance bevacizumab and pemetrexed in a total of 51 patients with nonsquamous NSCLC.13 Those authors reported a response rate of 55%, a median PFS of 7.8 months, and a median OS of 14.1 months, not dissimilar to our results. The adverse event profile for the regimen also was fairly similar to that in our study, although grade 3/4 diverticulitis was observed in 8% of their patients, and their study subsequently was amended to exclude patients who had a history of diverticulitis or significant diverticulosis. The development of a rectovaginal fistula in 1 patient on our study who had underlying diverticular disease raises the possibility that this toxicity may be unique to this regimen, although bowel perforation is a known adverse event associated with bevacizumab (with an overall incidence of 2.4% per prescribing information).14 Spigel et al recently reported their results with maintenance bevacizumab alone after 6 cycles of pemetrexed, carboplatin, and bevacizumab in a randomized phase 2 trial comparing that regimen to pemetrexed, gemcitabine, and bevacizumab in elderly patients (aged >70 years) with advanced, nonsquamous NSCLC.15 Improved TTP and OS were observed in the pemetrexed, carboplatin, and bevacizumab arm at 10.2 months and 14.8 months, respectively, and the response rate was 35%. These findings are comparable to our results as well as those of Patel and colleagues with doublet maintenance therapy.
The results of our study demonstrated that combination bevacizumab, pemetrexed, and carboplatin followed by “continuous” maintenance therapy with bevacizumab alone for bevacizumab-eligible patients with nonsquamous NSCLC builds on the current US standard of care with bevacizumab, paclitaxel, and carboplatin in terms of efficacy, tolerability, and ease of administration. The question of maintenance monotherapy versus doublet therapy is important with respect not only to efficacy and toxicity but also to cost, because patients who exhibit a continuing benefit may remain on maintenance therapy for durations ranging from months to years.
Three ongoing randomized trials will help to answer this question. The POINTBREAK trial of bevacizumab, pemetrexed, and carboplatin followed by maintenance bevacizumab plus pemetrexed versus bevacizumab, paclitaxel, and carboplatin followed by bevacizumab monotherapy has completed planned accrual of approximately 900 patients, and results are expected sometime soon.16 It is noteworthy that the maintenance approach in POINTBREAK after bevacizumab, pemetrexed, and carboplatin did not feature bevacizumab monotherapy but, rather, the bevacizumab and pemetrexed combination as piloted by Patel et al.
The AVAPERL1 study (a study of bevacizumab [Avastin] with or without pemetrexed as maintenance therapy after Avastin in first line in patients with nonsquamous NSCLC) comparing maintenance bevacizumab (7.5 mg/kg every 21 days) versus maintenance bevacizumab (7.5 mg/kg) plus pemetrexed after 4 cycles of front-line bevacizumab (7.5 mg/kg), pemetrexed, and cisplatin has completed accrual.17 Final results are pending; however, preliminary results presented at the 2011 European Multidisciplinary Cancer Congress indicated a 50% reduction in the hazard ratio for PFS with doublet maintenance therapy versus monotherapy.18 The results of that study also may raise the question of using cisplatin versus carboplatin in the front-line setting, because the phase 3 results favoring pemetrexed in patients with nonsquamous disease were observed with cisplatin.9 We chose to use carboplatin in our trial to be consistent with ECOG 4599, because the AVAiL results had not been reported yet, and carboplatin is generally the preferred platinum agent for use in advanced NSCLC in the United States. However, it is noteworthy that results from the cisplatin versus carboplatin (CISCA) meta-analysis reported in 2007 (after the initiation of our trial) indicated that cisplatin-based therapies produced superior response rates and survival compared with carboplatin-based regimens, and the survival effect was most notable in patients with nonsquamous histology who were receiving “third-generation” chemotherapy regimens, although no pemetrexed-containing or bevacizumab-containing studies were included.19 Therefore, the AVAPERL1 overall survival results will be of great interest in this regard.
The third trial, ECOG 5508, which is currently in progress, investigates bevacizumab versus pemetrexed given alone versus combination bevacizumab and pemetrexed as maintenance therapy after 4 cycles of bevacizumab, paclitaxel, and carboplatin in patients with advanced, nonsquamous NSCLC.20 The results of these 3 important trials should clarify the preferred maintenance strategy in bevacizumab-eligible patients.
This work was supported by a research grant from Genentech, Inc.
CONFLICT OF INTEREST DISCLOSURES
James P. Stevenson received research funding from Genentech Inc and Eli Lilly and Company. Corey J. Langer is a consultant to and is on the advisory boards of Genentech Inc. and Eli Lilly and Company. Suzanne Walker is on the Speaker's Bureau of Eli Lilly and Company. Beth Eaby-Sandy is on the Speaker's Bureaus of Eli Lilly and Company and Genentech Inc. Alexandre Hageboutros is on the Speaker's Bureau of Eli Lilly and Company.