J Neurooncol DOI 10.1007/s11060-015-2025-5
CLINICAL STUDY
The optimal regimen of bevacizumab for recurrent glioblastoma: does dose matter? Blumenthal DT1 • Mendel L2 • Bokstein F1
Received: 31 July 2015 / Accepted: 22 December 2015 Ó Springer Science+Business Media New York 2015
Abstract The FDA-approved schedule and dose of bevacizumab (BVZ) for recurrent glioblastoma (rGB) (10 mg/kg q 2 weeks) were adopted from systemic cancer protocols. No dose-defining studies have been performed for glioblastoma. We began using BVZ for the treatment of rGB in 2005 at the dose of 5 mg/kg every 2 weeks combined with irinotecan, and later as single agent. Our previous report of 20 patients treated with BVZ 5 mg/kg every 2 weeks showed similar response rates and overall survival (OS) compared to other BVZ treatment protocols, with less adverse effects. In this study we retrospectively reviewed our 7 year experience with BVZ in 162 rGB patients. Treatment outcomes were analyzed from 87 patients who received BVZ at 5 mg/kg and 75 patients at 10 mg/kg. While median age was similar in both groups, the median KPS was significantly higher in the group treated with 10 mg/kg BVZ (85 versus 60). There was no significant difference in OS or progression free survival (PFS) between the groups treated with BVZ 5 versus 10 mg/kg. Overall survival was significantly improved in the
Precis for table of contents Bevacizumab 5 mg/kg as monotherapy for the treatment of recurrent glioblastoma may be equivalent to the dose of 10 mg/kg, with less toxicity. Addition of cytotoxic therapy may benefit a subgroup of these patients. Electronic supplementary material The online version of this article (doi:10.1007/s11060-015-2025-5) contains supplementary material, which is available to authorized users. & Blumenthal DT [email protected] 1
Neuro-Oncology Service, Division of Oncology, Tel-Aviv Sourasky Medical Center, 6 Weizmann St., 64239 Tel-Aviv, Israel
2
Omnistat Statistical Consulting, Tel-Aviv, Israel
subgroup treated with cytotoxic therapy in addition to BVZ 10 mg/kg. There were more adverse events seen with BVZ 10 mg/kg. There is no significant difference in OS for rGB treated with BVZ 5 mg/kg versus 10 mg/kg when given as monotherapy. The smaller dose was slightly less toxic. Addition of cytotoxic therapy resulted in prolongation of OS in a small subgroup of BVZ 10 mg/kg. Keywords Glioblastoma Bevacizumab Recurrent Dose Toxicity Standard Chemotherapy Regimen
Background The FDA-approved schedule and dose of bevacizumab (BVZ) for recurrent glioblastoma (rGB) (10 mg/kg q2 weeks) were adopted in May 2009. The dose regimen was modeled from protocols for colon cancer. The antiangiogenic drug bevacizumab (a humanized monoclonal antibody to VEGF) was first approved in 2002 to treat metastatic cancer of the colon or rectum [1] and has since been approved for several additional types of solid cancers, including non-small cell lung cancer, [2] metastatic renal cell [3] and platinum-resistant ovarian carcinoma. (Accelerated FDA approval was granted for the use of BVZ for metastatic breast cancer, but this approval was later revoked [4].) Side effects related to BVZ include potentially serious complications, such as thrombotic events and hemorrhages. No BVZ-dose defining studies have been performed for glioblastoma. We began using BVZ for treatment of rGB in 2005 at the dose of 5 mg/kg every 2 weeks combined with irinotecan. Our previous report of 20 patients treated with BVZ 5 mg/kg every 2 weeks combined with irinotecan showed similar response rates and overall survival (OS) to
123
J Neurooncol
other BVZ treatment protocols, with less adverse effects [5]. In this paper, we report on a retrospective analysis of 162 consecutively-treated rGBM patients treated over 7 years with 4 different regimens of BVZ: 5 mg/kg with cytotoxic chemotherapy, 5 mg/kg as monotherapy, 10 mg/kg with cytotoxic chemotherapy, and 10 mg/kg as monotherapy.
Patients and methods We began to use bevacizumab for recurrent high grade gliomas in 2005, patterned on the regimen reported by Stark-Vance, BVZ 5 mg/kg plus irinotecan 125 mg/m [2] every 2 weeks [6]. After FDA approval of BVZ for rGB in 2009 as monotherapy at the dose of 10 mg/kg, some patients at our center preferred to receive the treatment at the FDA-approved dose as monotherapy or in combination with chemotherapy. The decision regarding the BVZ dose and possible combination with cytotoxic agent was made in each individual case after thorough discussion of all treatment options with the patient and family. A minority of patients received BVZ at the dose of 10 mg/kg within the framework of a clinical trial for rGB. We retrospectively collected information from our neuro-oncology data base. We included adult patients with rGB who were treated in Tel Aviv Sourasky Medical Center from 2005 through 2012 with bevacizumab for tumor progression. Patients with poor performance status or medical comorbidities rendering bevacizumab use infeasible were excluded from anti-angiogenic treatment and from subsequent review. This review was approved by the institutional ethics review board.
Statistical analysis Cohort data was analyzed for each treatment group separately and compared afterwards. Descriptive statistics were used to characterize the groups and differences between the background data of the groups were evaluated using the v2 test of independence for categorical data, and t test or ANOVA for continuous data, as appropriate. Survival analysis was performed using Kaplan–Meier curves and the log-rank test. Difference in median survival times and confidence interval for that difference between the 5 and 10 mg/kg groups were evaluated using bootstrapping (a simple random resampling method to simulate non-standard distributions). Results with p value \ 0.05 were considered significant. All analyses expect for the bootstrapping were performed using SPSS version 22.0. Bootstrapping simulation was performed with R-studio version 0.98.1049.
123
Patient characteristics Patient characteristics are presented in Table 1. One hundred and sixty-two patients were available for analysis, among them 87 patients who received BVZ at the dose of 5 mg/kg (cohort A), and 75 patients who received the drug at the dose of 10 mg/kg (cohort B). The age and sex distribution was similar in the two groups; the median age was 60 years in group A and 58 years in group B. All of the patients in both groups had histologically-proven (primary) glioblastoma; molecular characteristics were available for a minority of cases. Two patients in group A had information regarding IDH mutational status (both were IDH-negative/ wild-type). In group B, IDH mutation stains were performed in 28 cases; 27 were negative and one was positive for IDH mutation. In cohort A (5 mg/kg) MGMT-methylation status was known for 25/87 patients (28.7 %), 23 with a non-methylated promoter and only two methylated. In the B group (10 mg/kg) MGMT-methylation status was known for 11/75 patients (14.7 %); seven were non-methylated and three methylated. All patients had received standard external beam radiation treatment. In cohort A, 79 patients had been treated with radiation treatment (RT) combined with temozolomide (TMZ) according to the standard of care from the EORTC protocol [7] and only 8 patients had received radiation treatment alone. In cohort B all 75 patients had been treated with combined chemo-radiation. Bevacizumab was initiated after failure of at least one previous chemotherapy line. In cohort A, 55 patients (63.2 %) had received only one prior treatment line (TMZ), 27 patients (31.0 %) had been treated with 2 chemotherapy regimens, and 5 patients (5.8 %) had received 3 or more treatment lines. Patients in cohort B were somewhat less pretreated (both in number of prior chemotherapy lines and in number of recurrent surgeries). There were no significant demographic differences seen between the chemotherapy subgroups (cohorts A and B ? chemo) versus (respectively) the wholes (all of cohorts A and B)*. The median time from glioblastoma diagnosis to initiation of BVZ was 9.6 months (2.8–58.7) in cohort A and 10.3 months (2.1–64.1) in cohort B. Prior to BVZ initiation, 55/87 patients (63.2 %) in cohort A had progression of disease while on temozolomide (TMZ), compared to 57/75 patients (76 %) in cohort B. Twenty-three of the 87 patients (26.4 %) patients in cohort A started treatment with BVZ [ 6 months from TMZ, while in cohort B there were 18/75 patients (24 %) who began BVZ [ 6 months from the time of TMZ therapy.
J Neurooncol Table 1 patient characteristics BVZ dose
Cohort A (BVZ 5 mg/kg)
Cohort B (BVZ 10 mg/kg)
Whole group (with and without chemo)
Pts receiving BVZ ? chemo
Whole group (with and without chemo)
Pts receiving BVZ ? chemo
N
87
57/87 (65.5 %)
75
15/75 (20 %)
Sex (f/m)
31/56
20/37
30/45
8/7
Age
60 (28–84)
56 (28–84)
58 (21–84)
58 (40–74)
1st treatment line
RT - 8
P value (comparing A to B)
\0.001 (A higher proportion of patients received chemo in cohort A) NS
RT ? Stupp - all
RT ? TZM (Stupp) - 79 No of Tx lines 1 Line 2 Lines
55 (63.2 %) 27 (31.0 %)
36 (63.1 %) 16 (28.1 %)
56 (74.7 %) 17 (22.7 %)
8 (53.3 %) 7 (46.7 %)
5 (5.7 %)
58 (8.8 %)
2 (2.7 %)
–
NS
3 and more lines NS
44 (50.1 %)
34 (59.6 %)
55 (73.3 %)
11 (73.3 %)
1 Surgery
34 (39.1 %)
16 (28.1 %)
20 (26.7 %)
4 (26.7 %)
2 Surgeries
9 (10.3 %)
7 (12.3 %)
–
–
Time (range) from Dx to BVZ Tx (months)
9.6 (2.8–58.7)
10.9 (3.2–54.3)
10.3 (2.1–64.1)
10.6 (4.6–42.6)
NS
KPS (median, range)
60 (40–100)
70 (50–100)
85 (50–100)
90 (60–100)
\0.001 (comparison of cohort A vs B)*
No of surgeries
3 and more surgeries
* Full statistical details for subgroups available on Supplementary Table
Regarding proximity of time from radiation and beginning BVZ treatment, in cohort A (5 mg/kg) there were 23/87 patients (26.4 %) who started BVZ [ 2 years after RT; and in cohort B, 13/75 patients (17.3 %) started BVZ [ 2 years after RT. *Complete demographic statistical data can be found on the supplementary table on the journal’s website. The median functional score of the patients treated with BVZ at the dose of 5 mg/was substantially lower (inferior) than that of patients in group B who were treated with 10 mg/kg (median KPS 60 (40–100) and 85 (50–100), respectively, p \ 0.001); additionally, there was a nonsignificant trend for patients in cohort A to be more heavily-treated before starting BVZ. In cohort A 57/87 patients (65 %) received BVZ in combination with a cytotoxic drug (2 patients received TMZ, 2 patients carboplatin, and 53 patients irinotecan); in cohort B only 15/75 patients (20 %) received combined treatment (9 patients received TMZ and 6 patients were treated with irinotecan). The difference in concurrent chemotherapy use between the two groups was statistically significant (p \ 0.001), probably reflecting the tendency to prescribe BVZ 10 mg/kg as monotherapy following FDA approval in 2009. None of the patients in either dose cohort
during the time of the study period (2005-2012) received CCNU or other nitrosoureas in combination with bevacizumab. Bevacizumab (with or without cytotoxic agent) was administered as an intravenous infusion in the chemotherapy day unit of the hospital every 2 weeks. Assessment of side effects was performed at every treatment visit. Radiological response was evaluated by gadolinium-enhanced MRI performed every four treatments (2 months). Clinical evaluation was performed every 2 months after MRI studies or as necessary. In cases of prolonged ([9 months) clinical and radiological response we typically increased the intervals between the BVZ treatments to 3 weeks, and in some cases to 4 weeks.
Results The median number of BVZ treatments before progression was eight both in group A (range: 1–48) and group B (range 1–58). Bevacizumab was discontinued due to progressive disease in 154 patients (95 %), and due to adverse events in eight patients (5 %) (Table 2).
123
J Neurooncol
Treatment response Response rates were similar in both groups (A and B) (See Table 2).
Survival analysis For the entire group of patients, median PFS was 3.6 months, and median OS 7.3 months. There were no statistically significant differences in PFS and OS between cohorts A and B (Table 2, Figs. 1, 2). For the entire study population (cohorts A and B), median OS of the patients who had progressed while on TMZ was 7.5 months, and of the patients who progressed after treatment with TMZ, median OS was 6.4 months (p = 0.72); hence, the time of first progression did not significantly influence OS. Similarly, median OS of the patients from both (A and B) cohorts who started treatment with BVZ [ 6 months after TMZ, and of the patients who started BVZ \ 6 months after TMZ was equal– 7.3 months (p = 0.33). Hence, the length of time between first line chemotherapy (TMZ) and subsequent BVZ treatment was not found to be a prognostic factor. Median OS for the entire study population (cohorts A and B) was 8.1 months for patients who started BVZ [ 2 years after RT and 7.3 months for patients who started BVZ \ 2 years after RT, a non-significant difference (p = 0.77). Median PFS was 3.4 months for cohort A and 3.7 months for cohort B; median OS for the two cohorts were 7.1 months and 7.8 months, respectively. Thus, comparison of the two cohorts, receiving BVZ at the dose
Fig. 1 PFS of groups A (BVZ 5 mg/kg) and B (BVZ 10 mg/kg)
of 5 mg/kg and 10 mg/kg, did not reveal any statistically significant differences, neither in response rate nor in PFS and OS. In both groups there was a significant proportion of long-term survivors: 26.4 % of patients treated with 5 mg/ kg of BVZ survived more than 1 year, and 5.7 % of patients survived more than 2 years. In cohort B, 1- and 2-year survivals were 25.3 % and 9.3 %, accordingly. Three patients in cohort B are still alive at time of this writing, one more than 6 years (more than three years off-
Table 2 Treatment results
Group A (5 mg) N = 87
Group B (10 mg) N = 75
P value
Median No of BVZ treatments before PD Response
8 (1–48) 4 (4.6 %)
9 (1–58) 2 (2.7 %)
NS NS
CR
36 (41.4 %)
39 (52 %)
NS
PR
13 (14.9 %)
9 (12 %)
SD
34 (39.1 %)
25 (33.3 %)
PD
69 %
66.7 %
PFS
3.4 months
3.7 months
NS
PFS for monotherapy pts
3.1 months
3.6 months
NS
PFS for pts on BVZ ? chemo
3.4 months
7.9 months
NS
OS
7.1 months
7.8 months
NS
OS for monotherapy pts
5.9 months
7.2 months
NS
OS for pts on BVZ ? chemo
7.2 months
14.5 months
p=0.007
[1 year
23 pts (26.4 %)
19 pts (25.3 %)
NS
[2years
5 pts (5.7 %)
7 pts (9.3 %)
NS
Treatment benefit *(CR, PR, SD)
Long term survivors
123
J Neurooncol
Fig. 2 Overall survival of groups A (BVZ 5 mg/kg) and B (BVZ 10 mg/kg)
Fig. 3 Overall survival of subgroups receiving BVZ with or without chemotherapy
treatment), another patient 2 years and 9 months; the third patient is alive at more than 4 years and 6 months. These three longest surviving patients in the 10 mg/kg group had favorable clinical factors of younger age and positive IDH mutation in one of the three. The patient with IDH mutation had a histologically oligodendroglial component to the GBM (1p19q molecular status not known). Subsequently, we analyzed the influence of concomitant chemotherapy with BVZ on patients’ survival. Median PFS and OS were similar for patients treated with BVZ at the dose of 5 mg/kg with or without chemotherapy and for patients who received monotherapy with BVZ at the dose of 10 mg (Table 2, Fig. 3). However, the small subgroup of 15 patients treated with BVZ at the dose of 10 mg/kg combined with chemotherapy (versus the substantially larger subgroup (57 patients) treated with BVZ at 5 mg/kg with a cytotoxic agent) showed a trend for increased median PFS (7.9 versus 3.4 months) and a statistically significant prolongation of OS (with a median of 14.5 versus 7.2 months, p = 0.007 in the log-rank test).
compared to patients on BVZ monotherapy, who had no hematologic toxicity reported at all. Although the frequency of all adverse effects was substantially higher in cohort B, grade 3 and 4 reactions occurred in the two groups relatively equally (18 and 17 cases, accordingly). Toxicities (albeit mostly grades 1 and 2) likely to be directly related to BVZ (in bold-face print in the table below), namely proteinuria and hypertension, were substantially more frequent in the two higher dose (10 mg/kg) cohorts, 83.4 versus 8.6 % in cohort A. (Table 3).
Toxicity We observed 43 cases of adverse effects in cohort A and 59 cases in cohort B. Expectedly, patients in both dose cohorts receiving BVZ in combination with a cytotoxic agent developed significantly more hematologic adverse reactions (40.1 % combined rate, including grades 3 and 4) as
Discussion The commonly-used bevacizumab (BVZ) regimen for the treatment of recurrent glioblastoma, 10 mg/kg administered intravenously every 2 weeks, was adopted without clinical data comparing different dose schedules. The European neuro-oncology community has questioned this dosing regimen, which has not been studied in prospective trials regarding its efficacy, toxicity, or economic implications. The authors note that the vast majority of patients with rGB were treated off-label based on preliminary data, prior to FDA approval [8]. As such, the opportunity to effectively determine the optimal BVZ dose regimen for rGB may have passed. Pharmacokinetic studies of BVZ show the estimated half-life at steady-state treatment to be approximately 20 days (range 11–50 days), hence an every 3 week schedule
123
123 1 (3.3 %)
Hepatitis C
1 (3.3 %)
Fever
Tremor
Skin eruption
Dehydration
1 (3.3 %)
Pruritus
reactivation
1 (3.3 %)
Visual loss
Nausea
1 (3.3 %) 2 (6.6 %)
1 (3.3 %)
Diarrhea
1 (3.3 %)
1 (3.3 %)
1 (3.3 %)
1 (3.3 %)
3
Psychosis
1 (3.3 %) 4 (13.3 %)
Abscess
Other Fatigue
Wound infection
CSF leak
Stroke
9(30 %)
2 (6.6 %)
Intratumoral bleed
Rectal bleeding
1 (3.3 %)
1–2
1 (3.3 %)
1 (1.8 %)
4
BVZ ? chemo
Epistaxis
1 (1.8 %)
3
PE
1 (1.8 %)
3 (5.3 %)
1–2
BVZ
DVT
Proteinuria
Hypertension
Probably BVZ-related
Leukopenia
Thrombocytopenia
Hematologic
Grade
Table 3 Adverse events
1 (3.3 %)
2 (6.6 %)
1 (3.3 %)
3 (10 %)
4
1 (1.7 %)
1 (1.7 %)
5 (8.3 %)
1 (1.7 %)
1 (1.7 %)
4 (6.7 %)
9 (15 %)
1–2
BVZ 10
1 (1.7 %)
1 (1.7 %)
1 (1.7 %)
1 (1.7 %)
1 (1.7 %)
3
1 (1.7 %)
4
1 (6.7 %)
1 (6.7 %)
1 (6.7 %)
1 (6.7 %)
5 (33.3 %)
4 (26.7 %)
5 (33.3 %)
1 (6.7 %)
1 (6.7 %)
1–2
1 (6.7 %)
1 (6.7 %)
1 (6.7 %)
1 (6.7 %)
1 (6.7 %)
1 (6.7 %)
1 (6.7 %)
1 (6.7 %)
3
BVZ 10 ? chemo
1 (6.7 %)
1 (6.7 %)
1 (6.7 %)
4
J Neurooncol
J Neurooncol
should be therapeutic for the majority of patients [9]. Results obtained from 2 phase II studies of a variety of solid tumors including 69 patients suggest that minimum steady state concentrations (determined at 3 months) are similar when bevacizumab is given at 5 mg/kg every 2 weeks or 7.5 mg/kg every 3 weeks [10]. Early trials using BVZ for colorectal cancer compared 5 versus 10 mg/kg doses. A phase II study of approximately 100 patients with metastatic colon cancer randomized to one of three treatment groups: (1) fluorouracil and leucovorin (5-FU/LV); (2) 5FU/LV with BVZ at 5 mg/kg; or (3) bio-chemotherapy with BVZ at 10 mg/kg dose [11]. Response rates were significantly superior for the biochemotherapy arm with BVZ 5 mg/kg compared to the chemotherapy alone arm, with only a trend for improvement with larger dose bio-chemotherapy (17 % [control], 40 % [BVZ 5mg/kg, p=0.029] and 24 % [BVZ 10mg/kg, p=0.434]). Similar responses were seen for median time-toprogression, respectively. Grade 3/4 hypertension was reported in 8.5 % of patients in the BVZ lower dose group compared with 25 % of patients in the BVZ 10mg/kg group, and in none of the control group patients treated with chemotherapy alone. Based on these results, the preferred bevacizumab dose for ongoing trials was 5mg/kg every 2 weeks [12]. The smaller dose correlated with better response and outcome measures, with less toxicity. Standardly used bevacizumab regimens for metastatic colorectal cancer today include bevacizumab 7.5 mg/kg every 3 weeks in combination with XELOX (Oxaliplatin and oral capecitabine); BVZ 5 mg/kg every 2 weeks when given with 5-FU (fluorouracil) and Oxaliplatin (FOLFOX4);13 or BVZ 5 mg/kg every 2 or even 4 weeks, with biweekly irinotecan, leucovorin, and 5-FU (FOLFIRI) [14, 15]. The FDA approved a labeling extension for bevacizumab in combination with carboplatin and paclitaxel, for first-line treatment of patients with unresectable, locally advanced, recurrent, or metastatic, non-squamous, nonsmall cell lung cancer (NSCLC), in October 2006. This recommendation was based on a controlled, open label, multi-center clinical trial which demonstrated statistically significant improvement in overall survival (OS) in patients receiving BVZ (7.5 mg or 15 mg/kg every 3 weeks) with carboplatin and paclitaxel (434 patients) compared to those receiving carboplatin and paclitaxel alone (444). The generally accepted dosing used for BVZ (with cytotoxic agents) in the setting of metastatic NSCLC is 15 mg/kg every 3 weeks [16]. Bevacizumab was approved in 2009 for metastatic renal cell carcinoma (RCC) in combination with alfa interferon, based on the results of a multi-national randomized, double-blind, placebo-controlled trial which showed a fivemonth improvement in median progression-free survival
for the arm receiving BVZ (versus placebo) in addition to interferon; (an overall survival advantage was not shown). The bevacizumab dose used was 10 mg/kg every 2 weeks [17]. An earlier study showed no difference in dose response between BVZ doses of 3 mg/kg and 10 mg/kg in patients with metastatic RCC [18]. More recently (November 2014), BVZ was approved by the FDA in combination with paclitaxel, pegylated liposomal doxorubicin, or topotecan, for the treatment of patients with platinum-resistant recurrent epithelial ovarian, fallopian tube, or primary peritoneal cancer. Approval was based on results of the AURELIA trial which showed significant improvement in PFS and response rate (with a trend for improved survival). Two different regimens of BVZ were used in this phase III randomized trial, 10 mg/kg every 2 weeks or 15 mg/kg every 3 weeks [19].
Glioblastoma The first documented use of bevacizumab in patients with rGB was a small series of 11 patients with recurrent GBM treated by Stark-Vance [6]. She used BVZ 5 mg/kg plus irinotecan 125 mg/m2 every 2 weeks (a regimen borrowed from the colorectal cancer literature), and demonstrated activity in rGB. Response rate in this series was 42 %, with a PFS 6 of 30 %. The FDA approved bevacizumab as a single agent for patients with recurrent or progressive GB after treatment, on May 5, 2009. This accelerated approval was based on the results of 2 phase II clinical trials, both which used BVZ at a dose of 10 mg/kg every 2 weeks (AVF3708 g and NCI 06-C-0064E). AVF3708 g was an open-label, multi-center trial of 167 rGBM patients. BVZ was administered via two regimens: 10 mg/kg alone or in combination with irinotecan 340 mg/m2 or 125 mg/m2, once every 2 weeks. Primary end points were 6-month progression-free survival and objective response rate. Of the 85 patients treated with BVZ as monotherapy, 26 % showed objective tumor response to the drug [20]. In the NCI trial (06-C-0064E) 56 patients on a single arm (single site) study were treated with BVZ alone at the dose of 10 mg/kg once every two weeks, with 20 % of patients responding to the drug [21]. In both studies, the effect lasted for an average of about 4 months. A meta-analysis of 18 publications reported in patients with malignant gliomas treated with BVZ regimens showed no dose-response benefit comparing 10–15 mg/kg dose cohorts (462 patients) with 5 mg/kg dosing (86 patients). There were no significant differences seen in progression-free survival, overall survival, or in disease response [22].
123
J Neurooncol
A recent retrospective review of 181 patients with rGB treated with different AUCBEV doses (90 patients received \3.6 mg-week/kg and 91 received [3.6 mg-week/kg) within one health maintenance organization (HMO) showed that dosing at less than the ‘‘standard’’ dose of BVZ was equivalent to, and possibly superior to the 10 mg/ kg dose [23]. After several years trend of using of BVZ as monotherapy, the neuro-oncologic community may be gravitating again to the use of concurrent biochemotherapy for rGBM, now with lomustine as the preferred cytotoxic agent. A recent randomized, 3 arm, phase II multicenter trial (‘‘BELOB’’) showed improved responses to lomustine and BVZ (10 mg/kg every 2 weeks in all BVZ arms), and slightly less so with lomustine alone, compared to the slightly inferior results of the group receiving BVZ monotherapy (nine-month survival rates of 59, 43, and 38 % respectively) [24]. Most recently, results from the phase III clinical trial EORTC 26101 suggest that lomustine alone may be the preferred therapy for rGBM. New data were presented at the 2015 20th annual Society for Neuro-oncology Conference (SNO) for the phase III EORTC trial, modeled after the BELOB trial outcomes. This trial randomized patients with rGBM to monotherapy with lomustine versus combination with lomustine and BVZ 10 mg/kg. Unfortunately, a separate arm of BVZ monotherapy was not included in the trial regimen; (this decision stemmed from the lack of efficacy of single arm BVZ from the phase II BELOB trial). Although there was no significant advantage in OS for the combined biochemotherapy arm (9.1 versus 8.6 months for lomustine alone), there was a significant advantage in PFS (4.17 months versus 1.54). Of note, more than 35 % of the patients receiving monotherapy with lomustine added BVZ to their regimen at progression, which may have affected the OS data [25]. There does not appear to be additional significant survival benefit of the addition of BVZ to CCNU. However, BVZ continues to be a standard therapeutic option. The question of efficacy of CCNU versus BVZ monotherapy remains controversial. There is no level I evidence that definitively addresses this clinical conundrum. As such, the proper dose of BVZ to be used is still a relevant issue. These data need to be compared to the historical results of BVZ monotherapy and biochemotherapy with irinotecan. Table 4 demonstrates survival data from several prospective and retrospective trials using BVZ in rGB patients at different dosing schedules. With the exception of the combined (BVZ ? CCNU) arm of BELOB trial (interestingly not reproduced in the EORTC 26101 phase III trial), no significant differences in PFS and OS have been reported for different BVZ doses.
123
BVZ toxicities The most common adverse reactions related to BVZ use are epistaxis, headache, hypertension, rhinorrhea, proteinuria, taste alteration, dry skin, rectal bleeding, excessive tearing, and skin peeling. Standard protocol for administering BVZ includes urine stick for presence of proteinuria and blood pressure measurement before administering therapy. A recent study correlated incidence of hypertension-related to BVZ with improved outcome. Thirty of 82 patients (37 %) with rGB treated with bevacizumab developed (BVZ-related) hypertension. The group with BVZ-induced hypertension had median OS 11.7 months (95 % CI, 9.0–20.5 months), versus 4.9 months (95 % CI, 4.4–6.8 months) in the normotensive group (p\.001) [26]. A potentially catastrophic neurologic complication, posterior reversible encephalopathy syndrome (PRES) has been associated with BVZ use and can usually be managed by withdrawal of the therapy and treatment of hypertension [27].
Impressions Our report attempts to dissect the issue of BVZ dose regimens for rGBM. There are several inherent limitations to our study: first, the study is retrospective in nature, which lends bias to treatment decisions made regarding doses and schedules of the bio-chemotherapy. Our treatment methods were influenced by reigning standards of care at given times, protocol requirements, and patient preferences. Additionally, we acknowledge the heterogeneity in dose exposures of patients for whom the treatment intervals were increased during the course of their therapy. The BVZ dosing interval was routinely increased from 2 to 3 weeks in most patients after proven response and stability on BVZ for several months. Lastly, we are lacking information regarding biologic characteristics of the tumors which may have influenced outcome data. Molecular marker (IDH and MGMT methylation) status was not available in the majority of our cases for the years reported; immuno-histochemistry for IDH became routine at our center in 2010. The small number of patients with data for MGMT promoter methylation and IDH mutational status did not permit statistical analysis of molecular impact upon survival. Our cohort consisted of primary GBM for whom IDH is overwhelmingly of wild-type state [28]. As such, we would not expect the IDH mutational status to impact survival or to be related to a differential response to treatment between the different groups. In our cohort, the least toxic regimen was that of monotherapy with BVZ at 5 mg/kg. We did not observe a significant difference in PFS or OS between the two dose cohorts, despite lower mean KPS and a trend for more
J Neurooncol Table 4 Publications of different dose schedules of BVZ in rGBM Author Ali et al.
No of pts 9
Median age 53 (22–76)
4
Dose of BVZ (mg/kg) 5
Combined with
BVZ treatment cycle (wks)
PFS (months)
OS (months)
Chemotherapy
2
5.5
6.8
10
Ananthnarayan et al.
15
61 (39–90)
5
None, chemotherapy
2
2.4
NA
Bokstein et al.
17
56 (38–74)
5
Chemotherapy
2
4.7
7.0
Nghiemphu et al.
44
55 (26–90)
5
Chemotherapy
2
11
9.0
Vredenburg et al.
23
48 (18–66)
10
Irinotecan
2
6.0
10.5
Raizer et al.
12 52
52 (21–78)
15 15
None
3 3
6.0 10.8
10.5 6.4
14
46
5
None
3
3.6
6.4
NA
NA
NA
4.0
Kaloshi et al. Levin et al.
181
60 (90–89)
AUCBEV C 3.6 mg.wk/kg AUCBEV \ 3.6 mg.wk/kg
18.5
Kreisl et al.
48
53 (21–69)
10
None
2
4.0
7.8
Friedman et al.
85
54 (23–78)
10
None
2
4.2
9.2
(‘‘Brain’’ trial)
82
57 (23–79)
10
Irinotecan
2
5.6
8.7
Taal et al.
50
58 (37–77)
10
None
2
3.0
8.0
52
53 (29–62)
10
CCNU
2
12.0
16.0
57.1
10
CCNU
2
4.2
9.1
(BELOB trial) Wick et al. (EORTC 26101 trial)
288
(23–82)
heavily pretreated status of the 5 mg/kg group (A). Addition of chemotherapy increased the incidence of adverse effects as expected, but its toxic influence was much more dramatic for group A. Bevacizumab monotherapy at 10 mg/kg (group B) was significantly more toxic (37 % adverse event rate) than BVZ administered at 5 mg/kg (10 % adverse event rate), in our study population. The primary toxic events documented more frequently in the higher- dose cohort B were hypertension, proteinuria, and fatigue, the first two of which are directly caused by BVZ.
Conclusion Bevacizumab has been established as an important therapy for recurrent glioblastoma. The ‘‘standard’’ dosing regimen of BVZ 10 mg/kg every 2 weeks is not based on pharmacokinetic data or on level I clinical evidence. Additionally, it is not completely clear whether BVZ monotherapy or addition of cytotoxic therapy is most beneficial for patients with rGB. Dosing bevacizumab as monotherapy at 5 mg/kg every 2–3? weeks lead to non-inferior response rates and survival outcomes (despite less favorable patient characteristics in our cohort A at pre-BVZ baseline) with a reduction in treatment-related toxicity, compared to the dose of BVZ 10 mg/kg. However, we found an improved survival response in patients with rGB treated at the dose of BVZ
10 mg/kg with concomitant chemotherapy (irinotecan). One possible explanation regarding this finding is that the percentage of patients with KPS [ 70 in the subgroup treated with BVZ 10 mg/kg ? chemotherapy was significantly larger than in the other groups/subgroups of BVZ 5 mg/kg and BVZ 5 mg/kg ? chemotherapy. Additionally, the median age of this subgroup was almost 3 years younger (p\0.05); these two demographic factors could have favorably influenced this subgroup’s survival. Lastly, the addition of chemotherapy may indeed add a benefit to certain subgroups of patients with rGB treated with bevacizumab [24, 25]. In conclusion, we recommend considering both dosing options; patients with increased medical or neurologic pre-morbidities prior to beginning anti-angiogenic therapy with bevacizumab (and potentially susceptible to more adverse effects) may be better served with the smaller, less toxic dosing regimen (5 mg/kg). Patients who have difficulty controlling drug-induced hypertension on larger doses of BVZ may better tolerate a reduced dose regimen [authors’ experience]. Our data suggest that there may be additional survival benefit to patients treated with the higher dose of BVZ (10 mg/kg) in combination with chemotherapy, suggesting a dose-response for this subgroup. These findings corroborate the significant benefits seen with bio-chemotherapy treatment of CCNU and bevacizumab in the phase II BELOB trial [24] and possibly the improvement in PFS seen in the
123
J Neurooncol
biochemotherapy group in the most recent phase III EORTC 26101 trial [25]. Justification for the currently used ‘‘standard’’ dose of bevacizumab is lacking. Prospective controlled trials are necessary to establish the optimal dose and regimen of BVZ for the treatment of recurrent glioblastoma.
14.
15.
Compliance with ethical standards Conflict of interest the authors.
There are no conflicts of interest from any of
16.
17.
References 1. Cohen MH, Gootenberg J, Keegan P, Pazdur R (2007) FDA drug approval summary: bevacizumab plus FOLFOX4 as second-line treatment of colorectal cancer. Oncologist 12(3):356–361 2. Cohen MH, Gootenberg J, Keegan P, Pazdur R (2007) FDA drug approval summary: bevacizumab (Avastin) plus Carboplatin and Paclitaxel as first-line treatment of advanced/metastatic recurrent nonsquamous non-small cell lung cancer. Oncologist 12(6):713–718 3. Summers J, Cohen MH, Keegan P, Pazdur R (2010) FDA drug approval summary: bevacizumab plus interferon for advanced renal cell carcinoma. Oncologist 15(1):104–111 4. Rose S (2011) FDA pulls approval for Avastin in breast cancer. Cancer Discov 1(7):1–2 5. Bokstein F, Shpigel S, Blumenthal DT (2008) Treatment with bevacizumab and irinotecan for recurrent high-grade glial tumors. Cancer 112(10):2267–2273 6. Stark-Vance V (2005) Bevacizumab and CPT-11 in the treatment of relapsed malignant glioma. Neuro Oncol. 7:Abstract 369 7. Stupp R, Mason WP, Bent MJ et al (2005) European organisation for Research and Treatment of Cancer Brain Tumor and Radiotherapy Groups; National Cancer Institute of Canada Clinical Trials Group. Radiotherapy plus concomitant and adjuvant temozolomide for glioblastoma. N Engl J Med 352(10):987–996 8. Wick W, Weller M, van den Bent M, Stupp R (2010) Bevacizumab and recurrent malignant gliomas: a European perspective. J Clin Oncol 28(12):e188–e189 9. Lu J-F, Bruno R, Eppler S, Novotny W, Lum B, Gaudreault J (2008) Clinical pharmacokinetics of bevacizumab in patients with solid tumors. Cancer Chemother Pharmacol 62:779–786 10. Gaudreault J, Bruno R, Kabbinavar F, Sing A, Johnson DH, Lu J (2004) Clinical pharmacokinetics of bevacizumab (AvastinÒ) following every 2-or 3-week dosing. J Clin Oncol 23:205 11. Kabbinavar F, Hurwitz HI, Fehrenbacher L et al (2003) Phase II, randomized trial comparing bevacizumab plus fluorouracil (FU)/ leucovorin (LV) with FU/LV alone in patients with metastatic colorectal cancer. J Clin Oncol 21(1):60–65 12. Chong G, Tebbutt NC (2010) Using bevacizumab with different chemotherapeutic regimens in metastatic colorectal cancer: balancing utility with low toxicity. Ther Adv Med Oncol 2(5):309–317 13. Saltz LB, Clarke S, Dı´az-Rubio E et al (2008) Bevacizumab in combination with oxaliplatin-based chemotherapy as first-line
123
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
therapy in metastatic colorectal cancer: a randomized phase III study. J Clin Oncol 26(12):2013–2019 Van Cutsem E, Rivera F, Berry S et al (2009) First BEAT investigators. Safety and efficacy of first-line bevacizumab with FOLFOX, XELOX, FOLFIRI and fluoropyrimidines in metastatic colorectal cancer: the BEAT study. Ann Oncol 20(11):1842–1847 Yildiz R, Benekli M, Ozkan M et al (2012) Bevacizumab every 4 weeks is as effective as every 2 weeks in combination with biweekly FOLFIRI in metastatic colorectal cancer. J Cancer Res Clin Oncol 138(11):1845–1852 Johnson DH, Fehrenbacher L, Novotny WF et al (2004) Randomized phase II trial comparing bevacizumab plus carboplatin and paclitaxel with carboplatin and paclitaxel alone in previously untreated locally advanced or metastatic non-smallcell lung cancer. J Clin Oncol 22(11):2184–2191 Escudier B, Pluzanska A, Koralewski P et al (2007) AVOREN Trial investigators. Bevacizumab plus interferon alfa-2a for treatment of metastatic renal cell carcinoma: a randomised, double-blind phase III trial. Lancet 370(9605):2103–2111 Yang JC, Haworth L, Sherry RM et al (2003) A randomized trial of bevacizumab, an anti-vascular endothelial growth receptor antibody, for metastatic renal cancer. N Engl J Med 349(5):427–434 Pujade-Lauraine E, Hilpert F, Weber B (2014) Bevacizumab combined with chemotherapy for platinum-resistant recurrent ovarian cancer: the AURELIA open-label randomized phase III trial. J Clin Oncol 32(13):1302–1308 Friedman HS, Prados MD, Wen PY et al (2009) Bevacizumab alone and in combination with irinotecan in recurrent glioblastoma. J Clin Oncol 27(28):4733–4740 Kreisl TN, Kim L, Moore K et al (2009) Phase II trial of singleagent bevacizumab followed by bevacizumab plus irinotecan at tumor progression in recurrent glioblastoma. J Clin Oncol 27(5):740–745 Wong ET, Gautam S, Malchow C, Lun M, Pan E, Brem S (2011) Bevacizumab for recurrent glioblastoma multiforme: a metaanalysis. J Natl Compr Canc Netw 9(4):403–407 Levin VA, Mendelssohn ND, Chan J et al (2015) Impact of bevacizumab administered dose on overall survival of patients with progressive glioblastoma. J Neuroncol 122(1):145–150 Taal W, Oosterkamp HM, Walenkamp AM et al (2014) Singleagent bevacizumab or lomustine versus a combination of bevacizumab plus lomustine in patients with recurrent glioblastoma (BELOB trial): a randomised controlled phase 2 trial. Lancet Oncol 15(9):943–953 Wick W, Brandes AA, Gorlia T, et al. Phase III trial exploring the combination of bevacizumab and lomustine in patients with first recurrence of a glioblastoma: the EORTC 26101 trial. Neuro Oncol (2015) 17 (suppl):v1, Abstract LB-05. Abstracts from the 20th Annual Scientific Meeting of the Society for Neuro-Oncology, November 19–22, 2015, San Antonio, Texas Zhong J, Ali AN, Voloschin AD et al (2015) Bevacizumab-induced hypertension is a predictive marker for improved outcomes in patients with recurrent glioblastoma treated with bevacizumab. Cancer 121(9):1456–1462 Seet RC, Rabinstein AA (2012) Clinical features and outcomes of posterior reversible encephalopathy syndrome following bevacizumab treatment. QJM 105(1):69–75 Aldape K, Zadeh G, Mansouri S, Reifenberger G, Deimling A (2015) Glioblastoma: pathology, molecular mechanisms and markers. Acta Neuropathol 129(6):829–848
CLINICAL STUDY
The optimal regimen of bevacizumab for recurrent glioblastoma: does dose matter? Blumenthal DT1 • Mendel L2 • Bokstein F1
Received: 31 July 2015 / Accepted: 22 December 2015 Ó Springer Science+Business Media New York 2015
Abstract The FDA-approved schedule and dose of bevacizumab (BVZ) for recurrent glioblastoma (rGB) (10 mg/kg q 2 weeks) were adopted from systemic cancer protocols. No dose-defining studies have been performed for glioblastoma. We began using BVZ for the treatment of rGB in 2005 at the dose of 5 mg/kg every 2 weeks combined with irinotecan, and later as single agent. Our previous report of 20 patients treated with BVZ 5 mg/kg every 2 weeks showed similar response rates and overall survival (OS) compared to other BVZ treatment protocols, with less adverse effects. In this study we retrospectively reviewed our 7 year experience with BVZ in 162 rGB patients. Treatment outcomes were analyzed from 87 patients who received BVZ at 5 mg/kg and 75 patients at 10 mg/kg. While median age was similar in both groups, the median KPS was significantly higher in the group treated with 10 mg/kg BVZ (85 versus 60). There was no significant difference in OS or progression free survival (PFS) between the groups treated with BVZ 5 versus 10 mg/kg. Overall survival was significantly improved in the
Precis for table of contents Bevacizumab 5 mg/kg as monotherapy for the treatment of recurrent glioblastoma may be equivalent to the dose of 10 mg/kg, with less toxicity. Addition of cytotoxic therapy may benefit a subgroup of these patients. Electronic supplementary material The online version of this article (doi:10.1007/s11060-015-2025-5) contains supplementary material, which is available to authorized users. & Blumenthal DT [email protected] 1
Neuro-Oncology Service, Division of Oncology, Tel-Aviv Sourasky Medical Center, 6 Weizmann St., 64239 Tel-Aviv, Israel
2
Omnistat Statistical Consulting, Tel-Aviv, Israel
subgroup treated with cytotoxic therapy in addition to BVZ 10 mg/kg. There were more adverse events seen with BVZ 10 mg/kg. There is no significant difference in OS for rGB treated with BVZ 5 mg/kg versus 10 mg/kg when given as monotherapy. The smaller dose was slightly less toxic. Addition of cytotoxic therapy resulted in prolongation of OS in a small subgroup of BVZ 10 mg/kg. Keywords Glioblastoma Bevacizumab Recurrent Dose Toxicity Standard Chemotherapy Regimen
Background The FDA-approved schedule and dose of bevacizumab (BVZ) for recurrent glioblastoma (rGB) (10 mg/kg q2 weeks) were adopted in May 2009. The dose regimen was modeled from protocols for colon cancer. The antiangiogenic drug bevacizumab (a humanized monoclonal antibody to VEGF) was first approved in 2002 to treat metastatic cancer of the colon or rectum [1] and has since been approved for several additional types of solid cancers, including non-small cell lung cancer, [2] metastatic renal cell [3] and platinum-resistant ovarian carcinoma. (Accelerated FDA approval was granted for the use of BVZ for metastatic breast cancer, but this approval was later revoked [4].) Side effects related to BVZ include potentially serious complications, such as thrombotic events and hemorrhages. No BVZ-dose defining studies have been performed for glioblastoma. We began using BVZ for treatment of rGB in 2005 at the dose of 5 mg/kg every 2 weeks combined with irinotecan. Our previous report of 20 patients treated with BVZ 5 mg/kg every 2 weeks combined with irinotecan showed similar response rates and overall survival (OS) to
123
J Neurooncol
other BVZ treatment protocols, with less adverse effects [5]. In this paper, we report on a retrospective analysis of 162 consecutively-treated rGBM patients treated over 7 years with 4 different regimens of BVZ: 5 mg/kg with cytotoxic chemotherapy, 5 mg/kg as monotherapy, 10 mg/kg with cytotoxic chemotherapy, and 10 mg/kg as monotherapy.
Patients and methods We began to use bevacizumab for recurrent high grade gliomas in 2005, patterned on the regimen reported by Stark-Vance, BVZ 5 mg/kg plus irinotecan 125 mg/m [2] every 2 weeks [6]. After FDA approval of BVZ for rGB in 2009 as monotherapy at the dose of 10 mg/kg, some patients at our center preferred to receive the treatment at the FDA-approved dose as monotherapy or in combination with chemotherapy. The decision regarding the BVZ dose and possible combination with cytotoxic agent was made in each individual case after thorough discussion of all treatment options with the patient and family. A minority of patients received BVZ at the dose of 10 mg/kg within the framework of a clinical trial for rGB. We retrospectively collected information from our neuro-oncology data base. We included adult patients with rGB who were treated in Tel Aviv Sourasky Medical Center from 2005 through 2012 with bevacizumab for tumor progression. Patients with poor performance status or medical comorbidities rendering bevacizumab use infeasible were excluded from anti-angiogenic treatment and from subsequent review. This review was approved by the institutional ethics review board.
Statistical analysis Cohort data was analyzed for each treatment group separately and compared afterwards. Descriptive statistics were used to characterize the groups and differences between the background data of the groups were evaluated using the v2 test of independence for categorical data, and t test or ANOVA for continuous data, as appropriate. Survival analysis was performed using Kaplan–Meier curves and the log-rank test. Difference in median survival times and confidence interval for that difference between the 5 and 10 mg/kg groups were evaluated using bootstrapping (a simple random resampling method to simulate non-standard distributions). Results with p value \ 0.05 were considered significant. All analyses expect for the bootstrapping were performed using SPSS version 22.0. Bootstrapping simulation was performed with R-studio version 0.98.1049.
123
Patient characteristics Patient characteristics are presented in Table 1. One hundred and sixty-two patients were available for analysis, among them 87 patients who received BVZ at the dose of 5 mg/kg (cohort A), and 75 patients who received the drug at the dose of 10 mg/kg (cohort B). The age and sex distribution was similar in the two groups; the median age was 60 years in group A and 58 years in group B. All of the patients in both groups had histologically-proven (primary) glioblastoma; molecular characteristics were available for a minority of cases. Two patients in group A had information regarding IDH mutational status (both were IDH-negative/ wild-type). In group B, IDH mutation stains were performed in 28 cases; 27 were negative and one was positive for IDH mutation. In cohort A (5 mg/kg) MGMT-methylation status was known for 25/87 patients (28.7 %), 23 with a non-methylated promoter and only two methylated. In the B group (10 mg/kg) MGMT-methylation status was known for 11/75 patients (14.7 %); seven were non-methylated and three methylated. All patients had received standard external beam radiation treatment. In cohort A, 79 patients had been treated with radiation treatment (RT) combined with temozolomide (TMZ) according to the standard of care from the EORTC protocol [7] and only 8 patients had received radiation treatment alone. In cohort B all 75 patients had been treated with combined chemo-radiation. Bevacizumab was initiated after failure of at least one previous chemotherapy line. In cohort A, 55 patients (63.2 %) had received only one prior treatment line (TMZ), 27 patients (31.0 %) had been treated with 2 chemotherapy regimens, and 5 patients (5.8 %) had received 3 or more treatment lines. Patients in cohort B were somewhat less pretreated (both in number of prior chemotherapy lines and in number of recurrent surgeries). There were no significant demographic differences seen between the chemotherapy subgroups (cohorts A and B ? chemo) versus (respectively) the wholes (all of cohorts A and B)*. The median time from glioblastoma diagnosis to initiation of BVZ was 9.6 months (2.8–58.7) in cohort A and 10.3 months (2.1–64.1) in cohort B. Prior to BVZ initiation, 55/87 patients (63.2 %) in cohort A had progression of disease while on temozolomide (TMZ), compared to 57/75 patients (76 %) in cohort B. Twenty-three of the 87 patients (26.4 %) patients in cohort A started treatment with BVZ [ 6 months from TMZ, while in cohort B there were 18/75 patients (24 %) who began BVZ [ 6 months from the time of TMZ therapy.
J Neurooncol Table 1 patient characteristics BVZ dose
Cohort A (BVZ 5 mg/kg)
Cohort B (BVZ 10 mg/kg)
Whole group (with and without chemo)
Pts receiving BVZ ? chemo
Whole group (with and without chemo)
Pts receiving BVZ ? chemo
N
87
57/87 (65.5 %)
75
15/75 (20 %)
Sex (f/m)
31/56
20/37
30/45
8/7
Age
60 (28–84)
56 (28–84)
58 (21–84)
58 (40–74)
1st treatment line
RT - 8
P value (comparing A to B)
\0.001 (A higher proportion of patients received chemo in cohort A) NS
RT ? Stupp - all
RT ? TZM (Stupp) - 79 No of Tx lines 1 Line 2 Lines
55 (63.2 %) 27 (31.0 %)
36 (63.1 %) 16 (28.1 %)
56 (74.7 %) 17 (22.7 %)
8 (53.3 %) 7 (46.7 %)
5 (5.7 %)
58 (8.8 %)
2 (2.7 %)
–
NS
3 and more lines NS
44 (50.1 %)
34 (59.6 %)
55 (73.3 %)
11 (73.3 %)
1 Surgery
34 (39.1 %)
16 (28.1 %)
20 (26.7 %)
4 (26.7 %)
2 Surgeries
9 (10.3 %)
7 (12.3 %)
–
–
Time (range) from Dx to BVZ Tx (months)
9.6 (2.8–58.7)
10.9 (3.2–54.3)
10.3 (2.1–64.1)
10.6 (4.6–42.6)
NS
KPS (median, range)
60 (40–100)
70 (50–100)
85 (50–100)
90 (60–100)
\0.001 (comparison of cohort A vs B)*
No of surgeries
3 and more surgeries
* Full statistical details for subgroups available on Supplementary Table
Regarding proximity of time from radiation and beginning BVZ treatment, in cohort A (5 mg/kg) there were 23/87 patients (26.4 %) who started BVZ [ 2 years after RT; and in cohort B, 13/75 patients (17.3 %) started BVZ [ 2 years after RT. *Complete demographic statistical data can be found on the supplementary table on the journal’s website. The median functional score of the patients treated with BVZ at the dose of 5 mg/was substantially lower (inferior) than that of patients in group B who were treated with 10 mg/kg (median KPS 60 (40–100) and 85 (50–100), respectively, p \ 0.001); additionally, there was a nonsignificant trend for patients in cohort A to be more heavily-treated before starting BVZ. In cohort A 57/87 patients (65 %) received BVZ in combination with a cytotoxic drug (2 patients received TMZ, 2 patients carboplatin, and 53 patients irinotecan); in cohort B only 15/75 patients (20 %) received combined treatment (9 patients received TMZ and 6 patients were treated with irinotecan). The difference in concurrent chemotherapy use between the two groups was statistically significant (p \ 0.001), probably reflecting the tendency to prescribe BVZ 10 mg/kg as monotherapy following FDA approval in 2009. None of the patients in either dose cohort
during the time of the study period (2005-2012) received CCNU or other nitrosoureas in combination with bevacizumab. Bevacizumab (with or without cytotoxic agent) was administered as an intravenous infusion in the chemotherapy day unit of the hospital every 2 weeks. Assessment of side effects was performed at every treatment visit. Radiological response was evaluated by gadolinium-enhanced MRI performed every four treatments (2 months). Clinical evaluation was performed every 2 months after MRI studies or as necessary. In cases of prolonged ([9 months) clinical and radiological response we typically increased the intervals between the BVZ treatments to 3 weeks, and in some cases to 4 weeks.
Results The median number of BVZ treatments before progression was eight both in group A (range: 1–48) and group B (range 1–58). Bevacizumab was discontinued due to progressive disease in 154 patients (95 %), and due to adverse events in eight patients (5 %) (Table 2).
123
J Neurooncol
Treatment response Response rates were similar in both groups (A and B) (See Table 2).
Survival analysis For the entire group of patients, median PFS was 3.6 months, and median OS 7.3 months. There were no statistically significant differences in PFS and OS between cohorts A and B (Table 2, Figs. 1, 2). For the entire study population (cohorts A and B), median OS of the patients who had progressed while on TMZ was 7.5 months, and of the patients who progressed after treatment with TMZ, median OS was 6.4 months (p = 0.72); hence, the time of first progression did not significantly influence OS. Similarly, median OS of the patients from both (A and B) cohorts who started treatment with BVZ [ 6 months after TMZ, and of the patients who started BVZ \ 6 months after TMZ was equal– 7.3 months (p = 0.33). Hence, the length of time between first line chemotherapy (TMZ) and subsequent BVZ treatment was not found to be a prognostic factor. Median OS for the entire study population (cohorts A and B) was 8.1 months for patients who started BVZ [ 2 years after RT and 7.3 months for patients who started BVZ \ 2 years after RT, a non-significant difference (p = 0.77). Median PFS was 3.4 months for cohort A and 3.7 months for cohort B; median OS for the two cohorts were 7.1 months and 7.8 months, respectively. Thus, comparison of the two cohorts, receiving BVZ at the dose
Fig. 1 PFS of groups A (BVZ 5 mg/kg) and B (BVZ 10 mg/kg)
of 5 mg/kg and 10 mg/kg, did not reveal any statistically significant differences, neither in response rate nor in PFS and OS. In both groups there was a significant proportion of long-term survivors: 26.4 % of patients treated with 5 mg/ kg of BVZ survived more than 1 year, and 5.7 % of patients survived more than 2 years. In cohort B, 1- and 2-year survivals were 25.3 % and 9.3 %, accordingly. Three patients in cohort B are still alive at time of this writing, one more than 6 years (more than three years off-
Table 2 Treatment results
Group A (5 mg) N = 87
Group B (10 mg) N = 75
P value
Median No of BVZ treatments before PD Response
8 (1–48) 4 (4.6 %)
9 (1–58) 2 (2.7 %)
NS NS
CR
36 (41.4 %)
39 (52 %)
NS
PR
13 (14.9 %)
9 (12 %)
SD
34 (39.1 %)
25 (33.3 %)
PD
69 %
66.7 %
PFS
3.4 months
3.7 months
NS
PFS for monotherapy pts
3.1 months
3.6 months
NS
PFS for pts on BVZ ? chemo
3.4 months
7.9 months
NS
OS
7.1 months
7.8 months
NS
OS for monotherapy pts
5.9 months
7.2 months
NS
OS for pts on BVZ ? chemo
7.2 months
14.5 months
p=0.007
[1 year
23 pts (26.4 %)
19 pts (25.3 %)
NS
[2years
5 pts (5.7 %)
7 pts (9.3 %)
NS
Treatment benefit *(CR, PR, SD)
Long term survivors
123
J Neurooncol
Fig. 2 Overall survival of groups A (BVZ 5 mg/kg) and B (BVZ 10 mg/kg)
Fig. 3 Overall survival of subgroups receiving BVZ with or without chemotherapy
treatment), another patient 2 years and 9 months; the third patient is alive at more than 4 years and 6 months. These three longest surviving patients in the 10 mg/kg group had favorable clinical factors of younger age and positive IDH mutation in one of the three. The patient with IDH mutation had a histologically oligodendroglial component to the GBM (1p19q molecular status not known). Subsequently, we analyzed the influence of concomitant chemotherapy with BVZ on patients’ survival. Median PFS and OS were similar for patients treated with BVZ at the dose of 5 mg/kg with or without chemotherapy and for patients who received monotherapy with BVZ at the dose of 10 mg (Table 2, Fig. 3). However, the small subgroup of 15 patients treated with BVZ at the dose of 10 mg/kg combined with chemotherapy (versus the substantially larger subgroup (57 patients) treated with BVZ at 5 mg/kg with a cytotoxic agent) showed a trend for increased median PFS (7.9 versus 3.4 months) and a statistically significant prolongation of OS (with a median of 14.5 versus 7.2 months, p = 0.007 in the log-rank test).
compared to patients on BVZ monotherapy, who had no hematologic toxicity reported at all. Although the frequency of all adverse effects was substantially higher in cohort B, grade 3 and 4 reactions occurred in the two groups relatively equally (18 and 17 cases, accordingly). Toxicities (albeit mostly grades 1 and 2) likely to be directly related to BVZ (in bold-face print in the table below), namely proteinuria and hypertension, were substantially more frequent in the two higher dose (10 mg/kg) cohorts, 83.4 versus 8.6 % in cohort A. (Table 3).
Toxicity We observed 43 cases of adverse effects in cohort A and 59 cases in cohort B. Expectedly, patients in both dose cohorts receiving BVZ in combination with a cytotoxic agent developed significantly more hematologic adverse reactions (40.1 % combined rate, including grades 3 and 4) as
Discussion The commonly-used bevacizumab (BVZ) regimen for the treatment of recurrent glioblastoma, 10 mg/kg administered intravenously every 2 weeks, was adopted without clinical data comparing different dose schedules. The European neuro-oncology community has questioned this dosing regimen, which has not been studied in prospective trials regarding its efficacy, toxicity, or economic implications. The authors note that the vast majority of patients with rGB were treated off-label based on preliminary data, prior to FDA approval [8]. As such, the opportunity to effectively determine the optimal BVZ dose regimen for rGB may have passed. Pharmacokinetic studies of BVZ show the estimated half-life at steady-state treatment to be approximately 20 days (range 11–50 days), hence an every 3 week schedule
123
123 1 (3.3 %)
Hepatitis C
1 (3.3 %)
Fever
Tremor
Skin eruption
Dehydration
1 (3.3 %)
Pruritus
reactivation
1 (3.3 %)
Visual loss
Nausea
1 (3.3 %) 2 (6.6 %)
1 (3.3 %)
Diarrhea
1 (3.3 %)
1 (3.3 %)
1 (3.3 %)
1 (3.3 %)
3
Psychosis
1 (3.3 %) 4 (13.3 %)
Abscess
Other Fatigue
Wound infection
CSF leak
Stroke
9(30 %)
2 (6.6 %)
Intratumoral bleed
Rectal bleeding
1 (3.3 %)
1–2
1 (3.3 %)
1 (1.8 %)
4
BVZ ? chemo
Epistaxis
1 (1.8 %)
3
PE
1 (1.8 %)
3 (5.3 %)
1–2
BVZ
DVT
Proteinuria
Hypertension
Probably BVZ-related
Leukopenia
Thrombocytopenia
Hematologic
Grade
Table 3 Adverse events
1 (3.3 %)
2 (6.6 %)
1 (3.3 %)
3 (10 %)
4
1 (1.7 %)
1 (1.7 %)
5 (8.3 %)
1 (1.7 %)
1 (1.7 %)
4 (6.7 %)
9 (15 %)
1–2
BVZ 10
1 (1.7 %)
1 (1.7 %)
1 (1.7 %)
1 (1.7 %)
1 (1.7 %)
3
1 (1.7 %)
4
1 (6.7 %)
1 (6.7 %)
1 (6.7 %)
1 (6.7 %)
5 (33.3 %)
4 (26.7 %)
5 (33.3 %)
1 (6.7 %)
1 (6.7 %)
1–2
1 (6.7 %)
1 (6.7 %)
1 (6.7 %)
1 (6.7 %)
1 (6.7 %)
1 (6.7 %)
1 (6.7 %)
1 (6.7 %)
3
BVZ 10 ? chemo
1 (6.7 %)
1 (6.7 %)
1 (6.7 %)
4
J Neurooncol
J Neurooncol
should be therapeutic for the majority of patients [9]. Results obtained from 2 phase II studies of a variety of solid tumors including 69 patients suggest that minimum steady state concentrations (determined at 3 months) are similar when bevacizumab is given at 5 mg/kg every 2 weeks or 7.5 mg/kg every 3 weeks [10]. Early trials using BVZ for colorectal cancer compared 5 versus 10 mg/kg doses. A phase II study of approximately 100 patients with metastatic colon cancer randomized to one of three treatment groups: (1) fluorouracil and leucovorin (5-FU/LV); (2) 5FU/LV with BVZ at 5 mg/kg; or (3) bio-chemotherapy with BVZ at 10 mg/kg dose [11]. Response rates were significantly superior for the biochemotherapy arm with BVZ 5 mg/kg compared to the chemotherapy alone arm, with only a trend for improvement with larger dose bio-chemotherapy (17 % [control], 40 % [BVZ 5mg/kg, p=0.029] and 24 % [BVZ 10mg/kg, p=0.434]). Similar responses were seen for median time-toprogression, respectively. Grade 3/4 hypertension was reported in 8.5 % of patients in the BVZ lower dose group compared with 25 % of patients in the BVZ 10mg/kg group, and in none of the control group patients treated with chemotherapy alone. Based on these results, the preferred bevacizumab dose for ongoing trials was 5mg/kg every 2 weeks [12]. The smaller dose correlated with better response and outcome measures, with less toxicity. Standardly used bevacizumab regimens for metastatic colorectal cancer today include bevacizumab 7.5 mg/kg every 3 weeks in combination with XELOX (Oxaliplatin and oral capecitabine); BVZ 5 mg/kg every 2 weeks when given with 5-FU (fluorouracil) and Oxaliplatin (FOLFOX4);13 or BVZ 5 mg/kg every 2 or even 4 weeks, with biweekly irinotecan, leucovorin, and 5-FU (FOLFIRI) [14, 15]. The FDA approved a labeling extension for bevacizumab in combination with carboplatin and paclitaxel, for first-line treatment of patients with unresectable, locally advanced, recurrent, or metastatic, non-squamous, nonsmall cell lung cancer (NSCLC), in October 2006. This recommendation was based on a controlled, open label, multi-center clinical trial which demonstrated statistically significant improvement in overall survival (OS) in patients receiving BVZ (7.5 mg or 15 mg/kg every 3 weeks) with carboplatin and paclitaxel (434 patients) compared to those receiving carboplatin and paclitaxel alone (444). The generally accepted dosing used for BVZ (with cytotoxic agents) in the setting of metastatic NSCLC is 15 mg/kg every 3 weeks [16]. Bevacizumab was approved in 2009 for metastatic renal cell carcinoma (RCC) in combination with alfa interferon, based on the results of a multi-national randomized, double-blind, placebo-controlled trial which showed a fivemonth improvement in median progression-free survival
for the arm receiving BVZ (versus placebo) in addition to interferon; (an overall survival advantage was not shown). The bevacizumab dose used was 10 mg/kg every 2 weeks [17]. An earlier study showed no difference in dose response between BVZ doses of 3 mg/kg and 10 mg/kg in patients with metastatic RCC [18]. More recently (November 2014), BVZ was approved by the FDA in combination with paclitaxel, pegylated liposomal doxorubicin, or topotecan, for the treatment of patients with platinum-resistant recurrent epithelial ovarian, fallopian tube, or primary peritoneal cancer. Approval was based on results of the AURELIA trial which showed significant improvement in PFS and response rate (with a trend for improved survival). Two different regimens of BVZ were used in this phase III randomized trial, 10 mg/kg every 2 weeks or 15 mg/kg every 3 weeks [19].
Glioblastoma The first documented use of bevacizumab in patients with rGB was a small series of 11 patients with recurrent GBM treated by Stark-Vance [6]. She used BVZ 5 mg/kg plus irinotecan 125 mg/m2 every 2 weeks (a regimen borrowed from the colorectal cancer literature), and demonstrated activity in rGB. Response rate in this series was 42 %, with a PFS 6 of 30 %. The FDA approved bevacizumab as a single agent for patients with recurrent or progressive GB after treatment, on May 5, 2009. This accelerated approval was based on the results of 2 phase II clinical trials, both which used BVZ at a dose of 10 mg/kg every 2 weeks (AVF3708 g and NCI 06-C-0064E). AVF3708 g was an open-label, multi-center trial of 167 rGBM patients. BVZ was administered via two regimens: 10 mg/kg alone or in combination with irinotecan 340 mg/m2 or 125 mg/m2, once every 2 weeks. Primary end points were 6-month progression-free survival and objective response rate. Of the 85 patients treated with BVZ as monotherapy, 26 % showed objective tumor response to the drug [20]. In the NCI trial (06-C-0064E) 56 patients on a single arm (single site) study were treated with BVZ alone at the dose of 10 mg/kg once every two weeks, with 20 % of patients responding to the drug [21]. In both studies, the effect lasted for an average of about 4 months. A meta-analysis of 18 publications reported in patients with malignant gliomas treated with BVZ regimens showed no dose-response benefit comparing 10–15 mg/kg dose cohorts (462 patients) with 5 mg/kg dosing (86 patients). There were no significant differences seen in progression-free survival, overall survival, or in disease response [22].
123
J Neurooncol
A recent retrospective review of 181 patients with rGB treated with different AUCBEV doses (90 patients received \3.6 mg-week/kg and 91 received [3.6 mg-week/kg) within one health maintenance organization (HMO) showed that dosing at less than the ‘‘standard’’ dose of BVZ was equivalent to, and possibly superior to the 10 mg/ kg dose [23]. After several years trend of using of BVZ as monotherapy, the neuro-oncologic community may be gravitating again to the use of concurrent biochemotherapy for rGBM, now with lomustine as the preferred cytotoxic agent. A recent randomized, 3 arm, phase II multicenter trial (‘‘BELOB’’) showed improved responses to lomustine and BVZ (10 mg/kg every 2 weeks in all BVZ arms), and slightly less so with lomustine alone, compared to the slightly inferior results of the group receiving BVZ monotherapy (nine-month survival rates of 59, 43, and 38 % respectively) [24]. Most recently, results from the phase III clinical trial EORTC 26101 suggest that lomustine alone may be the preferred therapy for rGBM. New data were presented at the 2015 20th annual Society for Neuro-oncology Conference (SNO) for the phase III EORTC trial, modeled after the BELOB trial outcomes. This trial randomized patients with rGBM to monotherapy with lomustine versus combination with lomustine and BVZ 10 mg/kg. Unfortunately, a separate arm of BVZ monotherapy was not included in the trial regimen; (this decision stemmed from the lack of efficacy of single arm BVZ from the phase II BELOB trial). Although there was no significant advantage in OS for the combined biochemotherapy arm (9.1 versus 8.6 months for lomustine alone), there was a significant advantage in PFS (4.17 months versus 1.54). Of note, more than 35 % of the patients receiving monotherapy with lomustine added BVZ to their regimen at progression, which may have affected the OS data [25]. There does not appear to be additional significant survival benefit of the addition of BVZ to CCNU. However, BVZ continues to be a standard therapeutic option. The question of efficacy of CCNU versus BVZ monotherapy remains controversial. There is no level I evidence that definitively addresses this clinical conundrum. As such, the proper dose of BVZ to be used is still a relevant issue. These data need to be compared to the historical results of BVZ monotherapy and biochemotherapy with irinotecan. Table 4 demonstrates survival data from several prospective and retrospective trials using BVZ in rGB patients at different dosing schedules. With the exception of the combined (BVZ ? CCNU) arm of BELOB trial (interestingly not reproduced in the EORTC 26101 phase III trial), no significant differences in PFS and OS have been reported for different BVZ doses.
123
BVZ toxicities The most common adverse reactions related to BVZ use are epistaxis, headache, hypertension, rhinorrhea, proteinuria, taste alteration, dry skin, rectal bleeding, excessive tearing, and skin peeling. Standard protocol for administering BVZ includes urine stick for presence of proteinuria and blood pressure measurement before administering therapy. A recent study correlated incidence of hypertension-related to BVZ with improved outcome. Thirty of 82 patients (37 %) with rGB treated with bevacizumab developed (BVZ-related) hypertension. The group with BVZ-induced hypertension had median OS 11.7 months (95 % CI, 9.0–20.5 months), versus 4.9 months (95 % CI, 4.4–6.8 months) in the normotensive group (p\.001) [26]. A potentially catastrophic neurologic complication, posterior reversible encephalopathy syndrome (PRES) has been associated with BVZ use and can usually be managed by withdrawal of the therapy and treatment of hypertension [27].
Impressions Our report attempts to dissect the issue of BVZ dose regimens for rGBM. There are several inherent limitations to our study: first, the study is retrospective in nature, which lends bias to treatment decisions made regarding doses and schedules of the bio-chemotherapy. Our treatment methods were influenced by reigning standards of care at given times, protocol requirements, and patient preferences. Additionally, we acknowledge the heterogeneity in dose exposures of patients for whom the treatment intervals were increased during the course of their therapy. The BVZ dosing interval was routinely increased from 2 to 3 weeks in most patients after proven response and stability on BVZ for several months. Lastly, we are lacking information regarding biologic characteristics of the tumors which may have influenced outcome data. Molecular marker (IDH and MGMT methylation) status was not available in the majority of our cases for the years reported; immuno-histochemistry for IDH became routine at our center in 2010. The small number of patients with data for MGMT promoter methylation and IDH mutational status did not permit statistical analysis of molecular impact upon survival. Our cohort consisted of primary GBM for whom IDH is overwhelmingly of wild-type state [28]. As such, we would not expect the IDH mutational status to impact survival or to be related to a differential response to treatment between the different groups. In our cohort, the least toxic regimen was that of monotherapy with BVZ at 5 mg/kg. We did not observe a significant difference in PFS or OS between the two dose cohorts, despite lower mean KPS and a trend for more
J Neurooncol Table 4 Publications of different dose schedules of BVZ in rGBM Author Ali et al.
No of pts 9
Median age 53 (22–76)
4
Dose of BVZ (mg/kg) 5
Combined with
BVZ treatment cycle (wks)
PFS (months)
OS (months)
Chemotherapy
2
5.5
6.8
10
Ananthnarayan et al.
15
61 (39–90)
5
None, chemotherapy
2
2.4
NA
Bokstein et al.
17
56 (38–74)
5
Chemotherapy
2
4.7
7.0
Nghiemphu et al.
44
55 (26–90)
5
Chemotherapy
2
11
9.0
Vredenburg et al.
23
48 (18–66)
10
Irinotecan
2
6.0
10.5
Raizer et al.
12 52
52 (21–78)
15 15
None
3 3
6.0 10.8
10.5 6.4
14
46
5
None
3
3.6
6.4
NA
NA
NA
4.0
Kaloshi et al. Levin et al.
181
60 (90–89)
AUCBEV C 3.6 mg.wk/kg AUCBEV \ 3.6 mg.wk/kg
18.5
Kreisl et al.
48
53 (21–69)
10
None
2
4.0
7.8
Friedman et al.
85
54 (23–78)
10
None
2
4.2
9.2
(‘‘Brain’’ trial)
82
57 (23–79)
10
Irinotecan
2
5.6
8.7
Taal et al.
50
58 (37–77)
10
None
2
3.0
8.0
52
53 (29–62)
10
CCNU
2
12.0
16.0
57.1
10
CCNU
2
4.2
9.1
(BELOB trial) Wick et al. (EORTC 26101 trial)
288
(23–82)
heavily pretreated status of the 5 mg/kg group (A). Addition of chemotherapy increased the incidence of adverse effects as expected, but its toxic influence was much more dramatic for group A. Bevacizumab monotherapy at 10 mg/kg (group B) was significantly more toxic (37 % adverse event rate) than BVZ administered at 5 mg/kg (10 % adverse event rate), in our study population. The primary toxic events documented more frequently in the higher- dose cohort B were hypertension, proteinuria, and fatigue, the first two of which are directly caused by BVZ.
Conclusion Bevacizumab has been established as an important therapy for recurrent glioblastoma. The ‘‘standard’’ dosing regimen of BVZ 10 mg/kg every 2 weeks is not based on pharmacokinetic data or on level I clinical evidence. Additionally, it is not completely clear whether BVZ monotherapy or addition of cytotoxic therapy is most beneficial for patients with rGB. Dosing bevacizumab as monotherapy at 5 mg/kg every 2–3? weeks lead to non-inferior response rates and survival outcomes (despite less favorable patient characteristics in our cohort A at pre-BVZ baseline) with a reduction in treatment-related toxicity, compared to the dose of BVZ 10 mg/kg. However, we found an improved survival response in patients with rGB treated at the dose of BVZ
10 mg/kg with concomitant chemotherapy (irinotecan). One possible explanation regarding this finding is that the percentage of patients with KPS [ 70 in the subgroup treated with BVZ 10 mg/kg ? chemotherapy was significantly larger than in the other groups/subgroups of BVZ 5 mg/kg and BVZ 5 mg/kg ? chemotherapy. Additionally, the median age of this subgroup was almost 3 years younger (p\0.05); these two demographic factors could have favorably influenced this subgroup’s survival. Lastly, the addition of chemotherapy may indeed add a benefit to certain subgroups of patients with rGB treated with bevacizumab [24, 25]. In conclusion, we recommend considering both dosing options; patients with increased medical or neurologic pre-morbidities prior to beginning anti-angiogenic therapy with bevacizumab (and potentially susceptible to more adverse effects) may be better served with the smaller, less toxic dosing regimen (5 mg/kg). Patients who have difficulty controlling drug-induced hypertension on larger doses of BVZ may better tolerate a reduced dose regimen [authors’ experience]. Our data suggest that there may be additional survival benefit to patients treated with the higher dose of BVZ (10 mg/kg) in combination with chemotherapy, suggesting a dose-response for this subgroup. These findings corroborate the significant benefits seen with bio-chemotherapy treatment of CCNU and bevacizumab in the phase II BELOB trial [24] and possibly the improvement in PFS seen in the
123
J Neurooncol
biochemotherapy group in the most recent phase III EORTC 26101 trial [25]. Justification for the currently used ‘‘standard’’ dose of bevacizumab is lacking. Prospective controlled trials are necessary to establish the optimal dose and regimen of BVZ for the treatment of recurrent glioblastoma.
14.
15.
Compliance with ethical standards Conflict of interest the authors.
There are no conflicts of interest from any of
16.
17.
References 1. Cohen MH, Gootenberg J, Keegan P, Pazdur R (2007) FDA drug approval summary: bevacizumab plus FOLFOX4 as second-line treatment of colorectal cancer. Oncologist 12(3):356–361 2. Cohen MH, Gootenberg J, Keegan P, Pazdur R (2007) FDA drug approval summary: bevacizumab (Avastin) plus Carboplatin and Paclitaxel as first-line treatment of advanced/metastatic recurrent nonsquamous non-small cell lung cancer. Oncologist 12(6):713–718 3. Summers J, Cohen MH, Keegan P, Pazdur R (2010) FDA drug approval summary: bevacizumab plus interferon for advanced renal cell carcinoma. Oncologist 15(1):104–111 4. Rose S (2011) FDA pulls approval for Avastin in breast cancer. Cancer Discov 1(7):1–2 5. Bokstein F, Shpigel S, Blumenthal DT (2008) Treatment with bevacizumab and irinotecan for recurrent high-grade glial tumors. Cancer 112(10):2267–2273 6. Stark-Vance V (2005) Bevacizumab and CPT-11 in the treatment of relapsed malignant glioma. Neuro Oncol. 7:Abstract 369 7. Stupp R, Mason WP, Bent MJ et al (2005) European organisation for Research and Treatment of Cancer Brain Tumor and Radiotherapy Groups; National Cancer Institute of Canada Clinical Trials Group. Radiotherapy plus concomitant and adjuvant temozolomide for glioblastoma. N Engl J Med 352(10):987–996 8. Wick W, Weller M, van den Bent M, Stupp R (2010) Bevacizumab and recurrent malignant gliomas: a European perspective. J Clin Oncol 28(12):e188–e189 9. Lu J-F, Bruno R, Eppler S, Novotny W, Lum B, Gaudreault J (2008) Clinical pharmacokinetics of bevacizumab in patients with solid tumors. Cancer Chemother Pharmacol 62:779–786 10. Gaudreault J, Bruno R, Kabbinavar F, Sing A, Johnson DH, Lu J (2004) Clinical pharmacokinetics of bevacizumab (AvastinÒ) following every 2-or 3-week dosing. J Clin Oncol 23:205 11. Kabbinavar F, Hurwitz HI, Fehrenbacher L et al (2003) Phase II, randomized trial comparing bevacizumab plus fluorouracil (FU)/ leucovorin (LV) with FU/LV alone in patients with metastatic colorectal cancer. J Clin Oncol 21(1):60–65 12. Chong G, Tebbutt NC (2010) Using bevacizumab with different chemotherapeutic regimens in metastatic colorectal cancer: balancing utility with low toxicity. Ther Adv Med Oncol 2(5):309–317 13. Saltz LB, Clarke S, Dı´az-Rubio E et al (2008) Bevacizumab in combination with oxaliplatin-based chemotherapy as first-line
123
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
therapy in metastatic colorectal cancer: a randomized phase III study. J Clin Oncol 26(12):2013–2019 Van Cutsem E, Rivera F, Berry S et al (2009) First BEAT investigators. Safety and efficacy of first-line bevacizumab with FOLFOX, XELOX, FOLFIRI and fluoropyrimidines in metastatic colorectal cancer: the BEAT study. Ann Oncol 20(11):1842–1847 Yildiz R, Benekli M, Ozkan M et al (2012) Bevacizumab every 4 weeks is as effective as every 2 weeks in combination with biweekly FOLFIRI in metastatic colorectal cancer. J Cancer Res Clin Oncol 138(11):1845–1852 Johnson DH, Fehrenbacher L, Novotny WF et al (2004) Randomized phase II trial comparing bevacizumab plus carboplatin and paclitaxel with carboplatin and paclitaxel alone in previously untreated locally advanced or metastatic non-smallcell lung cancer. J Clin Oncol 22(11):2184–2191 Escudier B, Pluzanska A, Koralewski P et al (2007) AVOREN Trial investigators. Bevacizumab plus interferon alfa-2a for treatment of metastatic renal cell carcinoma: a randomised, double-blind phase III trial. Lancet 370(9605):2103–2111 Yang JC, Haworth L, Sherry RM et al (2003) A randomized trial of bevacizumab, an anti-vascular endothelial growth receptor antibody, for metastatic renal cancer. N Engl J Med 349(5):427–434 Pujade-Lauraine E, Hilpert F, Weber B (2014) Bevacizumab combined with chemotherapy for platinum-resistant recurrent ovarian cancer: the AURELIA open-label randomized phase III trial. J Clin Oncol 32(13):1302–1308 Friedman HS, Prados MD, Wen PY et al (2009) Bevacizumab alone and in combination with irinotecan in recurrent glioblastoma. J Clin Oncol 27(28):4733–4740 Kreisl TN, Kim L, Moore K et al (2009) Phase II trial of singleagent bevacizumab followed by bevacizumab plus irinotecan at tumor progression in recurrent glioblastoma. J Clin Oncol 27(5):740–745 Wong ET, Gautam S, Malchow C, Lun M, Pan E, Brem S (2011) Bevacizumab for recurrent glioblastoma multiforme: a metaanalysis. J Natl Compr Canc Netw 9(4):403–407 Levin VA, Mendelssohn ND, Chan J et al (2015) Impact of bevacizumab administered dose on overall survival of patients with progressive glioblastoma. J Neuroncol 122(1):145–150 Taal W, Oosterkamp HM, Walenkamp AM et al (2014) Singleagent bevacizumab or lomustine versus a combination of bevacizumab plus lomustine in patients with recurrent glioblastoma (BELOB trial): a randomised controlled phase 2 trial. Lancet Oncol 15(9):943–953 Wick W, Brandes AA, Gorlia T, et al. Phase III trial exploring the combination of bevacizumab and lomustine in patients with first recurrence of a glioblastoma: the EORTC 26101 trial. Neuro Oncol (2015) 17 (suppl):v1, Abstract LB-05. Abstracts from the 20th Annual Scientific Meeting of the Society for Neuro-Oncology, November 19–22, 2015, San Antonio, Texas Zhong J, Ali AN, Voloschin AD et al (2015) Bevacizumab-induced hypertension is a predictive marker for improved outcomes in patients with recurrent glioblastoma treated with bevacizumab. Cancer 121(9):1456–1462 Seet RC, Rabinstein AA (2012) Clinical features and outcomes of posterior reversible encephalopathy syndrome following bevacizumab treatment. QJM 105(1):69–75 Aldape K, Zadeh G, Mansouri S, Reifenberger G, Deimling A (2015) Glioblastoma: pathology, molecular mechanisms and markers. Acta Neuropathol 129(6):829–848
