IRRADIATION AND BEVACIZUMAB IN HIGH-GRADE GLIOMA RETREATMENT SETTINGS

Transcription

1 doi: /j.ijrobp Int. J. Radiation Oncology Biol. Phys., Vol. 82, No. 1, pp , 2012 Copyright Ó 2012 Elsevier Inc. Printed in the USA. All rights reserved /$ - see front matter CLINICAL INVESTIGATION Central Nervous System Tumor IRRADIATION AND BEVACIZUMAB IN HIGH-GRADE GLIOMA RETREATMENT SETTINGS MAXIMILIAN NIYAZI, M.D., M.SC.,* UTE GANSWINDT, M.D.,* SILKE BIRGIT SCHWARZ, M.D.,* FRIEDRICH-WILHELM KRETH, M.D., PH.D., y J org-christian TONN, M.D., PH.D., y JULIA GEISLER, M.D., z CHRISTIAN LA FOUGeRE, M.D., z LORENZ ERTL, M.D., x JENNIFER LINN, M.D., x AXEL SIEFERT, M.D.,* AND CLAUS BELKA, M.D., PH.D.* Departments of *Radiation Oncology, y Neurosurgery, z Nuclear Medicine, and x Neuroradiology, Ludwig-Maximilians-University Munich, Munich, Germany Purpose: Reirradiation is a treatment option for recurrent high-grade glioma with proven but limited effectiveness. Therapies directed against vascular endothelial growth factor have been shown to exert certain efficacy in combination with chemotherapy and have been safely tested in combination with radiotherapy in a small cohort of patients. To study the feasibility of reirradiation combined with bevacizumab treatment, the toxicity and treatment outcomes of this approach were analyzed retrospectively. Patients and Methods: After previous treatment with standard radiotherapy (with or without temozolomide) patients with recurrent malignant glioma received bevacizumab (10 mg/kg intravenous) on Day 1 and Day 15 during radiotherapy. Maintenance therapy was selected based on individual considerations, and mainly bevacizumabcontaining regimens were chosen. Patients received 36 Gy in 18 fractions. Results: The data of the medical charts of the 30 patients were analyzed retrospectively. All were irradiated in a single institution and received either bevacizumab (n = 20), no additional substance (n = 7), or temozolomide (n = 3). Reirradiation was tolerated well, regardless of the added drug. In 1 patient treated with bevacizumab, a wound dehiscence occurred. Overall survival was significantly better in patients receiving bevacizumab (p = 0.03, logrank test). In a multivariate proportional hazards Cox model, bevacizumab, Karnovsky performance status, and World Health Organization grade at relapse turned out to be the most important predictors for overall survival. Conclusion: Reirradiation with bevacizumab is a feasible and effective treatment for patients with recurrent highgrade gliomas. A randomized trial is warranted to finally answer the question whether bevacizumab adds substantial benefit to a radiotherapeutic retreatment setting. Ó 2012 Elsevier Inc. Radiotherapy, Bevacizumab, Malignant glioma, Antiangiogenesis, Glioblastoma. INTRODUCTION The prognosis of patients with high-grade glioma is limited by a high rate of local failures (1 3). The addition of temozolomide (TMZ) increased local control and survival. However, 72.8% of the patients still die within 24 months (4). Thus, there is a currently unmet need for effective retreatment options. Although various approaches have been tested, no standard regimen has been defined yet. In selected patients, a second course of radiotherapy (RT) constitutes a reasonable treatment option (5 7). For years, reirradiation of central nervous system (CNS) tissue was considered to be associated with a high risk of severe side effects (8, 9). However, recent data clearly indicate that reirradiation is feasible, with acceptably low toxicity (5, 10). The widespread availability of modern high-precision RT equipment (10 13) and of improved imaging capabilities, and the fact that animal experiments in primates revealed a substantial recovery of critical CNS structures allowed the reevaluation of this option in clinical practice (5, 14, 15). Therefore, RT has become a realistic treatment option for patients with recurrent malignant glioma (MG). In most retreatment series, time to second progression is about 6 months (16), and time to death is in the range of 10 to 12 months (5, 7). In parallel, other treatment options have become available for patients with recurrent MG (17 20). Inasmuch as most conventional cytotoxic approaches were found to be not adequately effective (21 23), molecularly targeted drugs Reprint requests to: Claus Belka, M.D., Ph.D., Department of Radiation Oncology, Ludwig-Maximilians-University Munich, Marchioninistr. 15, Munich, Germany. Tel: (49) /4521; Fax: (49) ; claus. belka@med.uni-muenchen.de 67 Conflict of interest: none. Acknowledgment The authors thank Dr. B. Suchorska for data acquisition concerning PFS rates and maintenance therapy. Received April 21, 2010, and in revised form Sept 9, Accepted for publication Sept 15, 2010.

2 68 I. J. Radiation Oncology d Biology d Physics Volume 82, Number 1, 2012 either alone or in combination with cytotoxic agents are currently undergoing clinical testing (24). Inasmuch as MG is characterized by the presence of pronounced hypoxia and by high levels of tumor-driven angiogenesis (25 27), it seems rational to evaluate the efficacy of treatments directed against vascular endothelial growth factor (VEGF) (28). In this regard, several trials have documented the efficacy of anti-vegf either alone or in combination with irinotecan (29 35), etoposide (36, 37), nitrosourea (38), or other agents (39, 40). Several groups have investigated the use of bevacizumab a humanized monoclonal antibody against VEGF with an already established role in metastatic colon, breast, and lung cancer (41) for patients with recurrent MG (42). Bevacizumab was also tested in combination with RT and TMZ as up-front treatment of MG (43). Inasmuch as the efficacy of radiation-based retreatment is limited, it is reasonable to test in how far the addition of any radiation response modulator would affect the efficacy of retreatment. In this regard, Combs et al. tested the current drug of choice (TMZ) in the primary setting in combination with reirradiation and obtained promising results. Actuarial survival rates at 6 and 12 months were 81% and 25%. The median progression-free survival (PFS) was 5 months; actuarial PFS rates at 6 and 12 months were 48% and 16% (44). Alternatively, Gutin et al. determined the safety and activity of RT and bevacizumab. After previous treatment with standard RT, 25 patients with recurrent glioblastoma (GBM) or anaplastic astrocytoma received bevacizumab (10 mg/kg intravenous) every 2 weeks of 28-day cycles until tumor progression. For the GBM cohort, overall response rate was 50%, PFS at 6 months (PFS-6) was 65%; median overall survival (mos) was 12.5 months, and 1-year survival was 54% (45). After the publication of these initial results, we integrated the combination treatment of RT plus bevacizumab into our institutional retreatment options. With all inherent limitations of a retrospective analysis in mind, we now tried to determine the value of this approach by comparing the outcomes in patients having received bevacizumab-based reirradiation with those being retreated without bevacizumab. PATIENTS AND METHODS Patient selection All patients treated with a series of reirradiation for recurrent MG at our institution between February 2005 and September 2009 were identified using the department database. Only patients with recurrent MG, histologically proven and/or [ 18 F]fluoroethyltyrosine positron emission tomography/magnetic resonance imaging (FET- PET/MRI) proven recurrent MG and macroscopic tumor in the brain were reirradiated. Altogether, 30 patients were identified who either received bevacizumab (n = 20), no additional substance (n = 7), or TMZ (n = 3) in combination to RT. Despite the proven efficacy of TMZ in combination with radiation, the drug has only limited value in reirradiation settings because most patients have been massively pretreated with TMZ, resulting in TMZ resistance (44). Thus, in our hands TMZ in combination with radiation in retreatment settings has been used only whenever initial TMZ use was limited (3 patients, 2 without TMZ in the pretreatment history). Historically, few patients did not receive any additional systemic therapy concurrently with the reirradiation course. Treatment schedule and follow-up Baseline evaluation included gadolinium-enhanced brain MRI with gradient echo sequence and perfusion or FET-PET examination, complete physical and neurologic examination, and blood tests before treatment. Whenever bevacizumab was administered, patients received 10 mg/kg on Days 1 and 15 during RT. The TMZ was given daily at a dosage of 75 mg/m 2. Treatment outcome was evaluated on a regular basis by brain MRI or FET-PET according to the Macdonald criteria and also by neurologic status (34). Table 1. Patient characteristics (N = 30) Characteristic Patients (N = 30) Sex M 17 (56.7%) F 13 (43.3%) Median age (y) (range) 51.5 (18 68) Median KPS (range) 80 (50 100) <70 11 (36.7%) $70 19 (63.3%) Median dose of primary 60 Gy radiotherapy Median dose of reirradiation 36 Gy Bevacizumab (WHO Grade 20 (5/15) (66.7%) III/IV) No bevacizumab (WHO 10 (3/7) (33.3%) Grade III/IV) MGMT methylation status Methylated 11 (36.7%) Not methylated 10 (33.3%) Unknown 9 (30%) Initial WHO grade II 3 (10%) III 6 (20%) IV 21 (70%) WHO grade at relapse III 8 (26.7%) IV 22 (73.3%) Resection Yes 24 (80%, one only at relapse) No 6 (20%) Concomitant TMZ treatment during first RT Yes 16 (53.3%) No 14 (46.7%) Adjuvant chemotherapy No adjuvant chemotherapy 17 Bevacizumab + X 7 (X = nil [2], PC [2], TMZ [2], irinotecan [1]) TMZ + X 3 (X = nil [2], sunitinib [1]) TMZ intensified + X 3 (X = nil [2], BCNU [1]) Abbreviations: KPS = Karnovsky performance status; WHO = World Health Organization; MGMT = O-6-methylguanine-DNA methyltransferase; TMZ = temozolomide; RT = radiotherapy; PC = procarbazine; X = combination partner as defined in the table; BCNU = carmustine.

3 Table 2. Treatment course of each individual patient Patient Primary treatment PFS (d) Retreatment 1 PFS (d) Retreatment 2 PFS (d) Retreatment 3 PFS (d) Further retreatment strategies PFS (d) 1 RTCH, adj. TMZ Re-RT + beva 82 2 Surg., RT TMZ 1,431 CEV 516 TMZ, CPT-11 + beva 641 TMZ + cilengitide, Surg., Re-RT +beva 3 Surg., RT 517 PDT, TMZ 418 Re-RT 5 4 RTCH, adj. TMZ 337 Re-RT + beva, adj. 277 TMZ/BCNU 5 RT 256 Re-RT + TMZ Surg. (2), RTCH, adj. 975 RT + TMZ 346 Re-RT + beva 163 (*) TMZ 7 RTCH, adj. TMZ 383 Re-RT + beva 36 (*) 8 Surg., RTCH, adj. TMZ 183 Surg., gliadel, PC 464 Re-RT + beva 38 (*) 9 RTCH, adj. TMZ 632 Re-RT Surg., RTCH, adj. TMZ 620 Re-RT + beva, adj. TMZ Surg. (2), RT 1775 PCV/TMZ 579 PDT 105 Re-RT Surg., RT 1614 TMZ 570 seed, TMZ 100 Re-RT + beva 23 (*) 13 Surg., RTCH, adj. TMZ 426 Re-RT + beva, adj. beva 17 (*) 14 Surg., RT, adj. TMZ 853 Surg., PC + beva 99 Re-RT + beva, adj. PC/beva Surg., RTCH, adj. TMZ 366 Surg. 61 Surg. + TMZ intens. 131 Re-RT + beva, adj. TMZ +beva 16 Surg., RTCH, adj. TMZ 350 Re-RT 132 Re-RT, adj. TMZ sunitinib 17 Surg., RTCH, adj. TMZ 407 Re-RT + beva 106 (*) 18 Surg., RTCH, adj. TMZ 199 beva + irinotecan, beva 245 Re-RT + beva 224 (*) mono 19 RT 147 Surg. 762 Surg. 59 Re-RT + beva, adj. beva 189 (*) 20 Surg., RT 1244 Surg. 61 PDT + PC 91 TMZ intens. 190 (*) Re-RT + beva 190 (*) 21 Surg., RTCH 165 Re-RT Surg., RTCH, adj. TMZ 335 Re-RT + beva 99 (*) 23 Surg., RT 1096 TMZ 232 Re-RT + beva, adj. TMZ Surg., RT 3286 PDT, TMZ 214 PDT, ALA-Surg., gliadel 647 Re-RT + beva 64 (*) 25 Surg., RTCH, adj. 609 Re-RT 189 TMZ, PC 26 Surg., RT, adj. PCV 2741 Re-RT + TMZ RT, adj. TMZ 207 Re-RT + beva Surg., RT, adj. TMZ 258 Re-RT Surg., RTCH, adj. TMZ 116 Surg. 370 Re-RT + beva, bev + irino Surg., RT, adj. TMZ 3473 Re-RT + TMZ 127 (*) 127 (*) Abbreviations: Surg. = surgery; PFS = progression-free survival (calculated from start of the corresponding treatment/chemoradiation; from date of Re-RT); TMZ = temozolomide; RTCH = radiochemotherapy with TMZ; RT = radiotherapy; beva = bevacizumab; PDT = photodynamic therapy; ALA = Aminolevulinic acid; TMZ intens. = TMZ intensified; irino = irinotecan; PC(V) = procarbazine; BCNU = carmustine; CEV = cyclophosphamide, epirubicin, vincristine. (*) = censored (*) Irradiation and bevacizumab in recurrent malignant glioma d M. NIYAZI et al. 69

4 70 I. J. Radiation Oncology d Biology d Physics Volume 82, Number 1, 2012 Radiotherapy Patients underwent MRI with #3 mm slices within 2 weeks of the treatment planning computed tomography (CT) with 3-mm slices. MRI was fused with the treatment planning CT, and a gross tumor volume (GTV) was delineated based on the contrast-enhancing lesion in the T1w + Gd MRI; the maximum GTV diameter did in general not exceed 5 cm. The planning target volume (PTV) was defined typically as the GTV plus a 10-mm margin (maximum value). Additional FET-PET information was used to modify the GTV if necessary. Patients were immobilized with a thermoplastic mask system. Treatment planning was performed by use of the Oncentra treatment planning system (OTP MasterPlan, Nucletron, Solingen, Germany). A single isocenter plan (three-dimensional conformal RT technique) was used for each patient, and a total dose of 36 Gy (2 Gy single dose, 18 fractions) was prescribed to the reference point according to International Commission on Radiation Units and Measurements Report 50 (95% isodose line encompassing the GTV) for all treatment plans. This fractionation scheme was chosen based on the case compilation provided by Combs et al. as a reasonable compromise between efficacy and normal tissue damage (44). Toxicity evaluation Adverse events were evaluated retrospectively, and toxicity was assessed using the National Cancer Institute s Common Toxicity Criteria, version 3.0. Particular attention was given to the (to date) most frequently reported bevacizumab-related adverse events, which include infection, fatigue, headache, hypertension, epistaxis, diarrhea, bleeding/hemorrhage, CNS hemorrhage, venous and arterial thromboembolic events, wound healing complications, proteinuria, gastrointestinal perforation, and reversible posterior leukoencephalopathy (46). Statistics Outcome measures of this study were safety of bevacizumab given in combination with RT for recurrent MG and overall survival in patients treated with or without bevacizumab. Demographic, safety, laboratory data, and treatment response were analyzed using descriptive statistics. Comparisons between groups were carried out using Fisher s exact test/asymptotic chi-square tests or Mann- Whitney U test. Survival analyses were based on Kaplan-Meier estimates, and univariate and multivariate modeling was performed using a Cox proportional hazards analysis. For all patients, overall survival was measured from the first day of reirradiation until death or last follow-up. A two-tailed p value <0.05 was considered significant. RESULTS Patient characteristics From February 2005 to September 2009, 30 patients with MG were treated at our department and included in this retrospective analysis. World Health Organization (WHO) Grade II glioma was initially diagnosed in 10% of these patients, progressing to a secondary high-grade glioma at relapse). The median age was 52 years (range, years), and median Karnovsky performance status (KPS) was 80 (range, ). In 15 patients, the recurrent MG was histologically proven. Further patient characteristics are shown in Table 1. The majority of patients underwent resection before RT (77% total and subtotal before initial RT; 1 patient underwent a resection at first relapse without initial surgery), and 53.5% received TMZ during adjuvant/primary RT. Median RT dose during the first treatment course was 60 Gy; in case of recurrence a median dose of 36 Gy was applied. Inasmuch as a systematic analysis of the MGMT methylation status was not performed at our institution before 2006, the MGMT status was available in only 21 of 30 cases. Bevacizumab was administered in the retreatment setting in 20 patients (5 WHO Grade III, 15 WHO Grade IV), and 10 patients (3 WHO Grade III, 7 WHO Grade IV) were retreated without bevacizumab. Adjuvant chemotherapy was applied in 13 patients; 6 patients received TMZ-containing regimens, and the remaining 7 patients received bevacizumab or a combination. Median follow-up for the whole patient cohort was 137 days (range, days). The treatment course of each individual patient is shown in Table 2. Fig. 1. (A) Kaplan-Meier progression-free survival (PFS) curve for all patients retreated with radiotherapy (N = 30). (B) Kaplan-Meier survival curve (calculated from the beginning of reirradiation) for all patients retreated with radiotherapy (N = 30).

5 Irradiation and bevacizumab in recurrent malignant glioma d M. NIYAZI et al. 71 Fig. 2. Images from a patient positive for O-6-methylguanine-DNA methyltransferase who was treated with radiotherapy (RT) and bevacizumab. This patient s glioblastoma was first detected in the right frontal lobe in April The patient received adjuvant RT after tumor excision. Thereafter, several relapses occurred, and subsequently secondary resections, photodynamic therapy, and a couple of chemotherapy regimens were performed. Left, last massive frontal progress as shown by [ 18 F]fluoroethyltyrosine positron emission tomography after intensified treatment with temozolomide from April 2009). Reirradiation with bevacizumab took place in May 2009; treatment response is shown second from the left. Survival and toxicity data Median progression-free survival (PFS) was 189 days (range, days), and median survival was 311 days (95% CI, ) for the whole patient population (Kaplan-Meier plots for the whole patient cohort, Fig. 1). In a second step, outcome (survival and toxicity) of the bevacizumab-treated population was analyzed. It was not possible to calculate mos because many of these patients were censored and the survival curve did not fall below 50 %. Mean survival for this collective was days, and median PFS was 244 days. The clinical course in a representative patient is shown in Fig. 2. Reirradiation with bevacizumab was generally well tolerated (two Grade 2 toxicities, one Grade 3 and one Grade 4 toxicity) with one exception, which needs to be discussed (Table 3). The only Grade 4 toxicity was due to a wound dehiscence after reresection. Given that the patient was noncompliant and did not consult a specialist after the end of re-rt, this wound dehiscence became superinfected and had to be revised surgically. To address the issue of radiation necrosis/leukoencephalopathy, all MRI images were retrospectively examined by neuroradiologists. In parallel, FET-PET examinations were analyzed by experienced nuclear medicine experts. Regarding toxicity, imaging revealed three putative cases of normal tissue damage. In 1 patient, contrast enhancement on MRI compatible with radiation necrosis was described; however, no concurrent FET-PET image was available. In a second case, FET-PET was suggestive of radiation necrosis. The FET features of radiation necrosis were a less pronounced signal intensity and different kinetics compared to vital tumor tissue (47). However, an MRI performed at the same time point did not strongly confirm the presence of tissue necrosis. However, regardless of the ambivalent findings, we decided to rate this case as radiation necrosis. Table 3. Safety profile of bevacizumab (N = 20) according to Common Toxicity Criteria Toxicity No. of patients (%) and grade Fatigue 1 (5%) Grade 2 CNS hemorrhage 0 (0%) Hypertension 1 (5%) Grade 2 Wound healing complication 1 (5%) Grade 4 Deep vein thrombosis 1 (5%) Grade 3 Abbreviation: CNS = central nervous system.

6 72 I. J. Radiation Oncology d Biology d Physics Volume 82, Number 1, 2012 Table 4. Treatment groups and characteristics Characteristic RT + bevacizumab (n = 20) RT only (n = 10) p value Sex (M/F) 12/8 5/5 NS (p = 0.705) Median age (y) NS* KPS (<70, $70) 6/14 5/5 NS (p = 0.425) Surgery (yes/no) 16/4 8/2 NS (p = 1.0) MGMT (meth./not meth.) 8/7 3/3 NS (p = 1.0) TMZ concomitant (yes/no) 12/8 4/6 NS (p = 0.442) WHO grade (II/III/IV) initial 3/3/14 0/3/7 NS (p = 0.413) WHO grade (III/IV) relapse 5/15 3/7 NS (p = 0.548) Adj. chemotherapy (yes/no) 10/10 3/7 NS (p = 0.440) Abbreviations: KPS = Karnovsky performance status; MGMT = O-6-methylguanine-DNA methyltransferase; TMZ = temozolomide; WHO = World Health Organization; NS = not significant; meth. = methylated; adj. = adjuvant; * = statistical trend only (p = 0.097) Comparisons were carried out with Fisher s exact test or asymptotic chi-square tests; as a nonparametric test the Mann-Whitney U test was carried out. In the third case, progressive leukoencephalopathy was diagnosed on MRI; however, PET imaging revealed clear tumor progress. In several other cases, a mixed tumor response pattern was observed (residual tumor and necrotic areas), none of which fulfilled the criteria for radiation necrosis. Thus, imaging revealed at maximum 2 patients with changes compatible with radiation necroses. In none of the cases was leukoencephalopathy outside the radiation portals seen. Comparison between reirradiation with and without bevacizumab Fisher s exact test and the Mann-Whitney U test were used to test whether there was an imbalance between these cohorts (Table 4, characteristics of both groups). No significant differences regarding sex, KPS, previous surgery, MGMT, concomitant TMZ application at first radiotherapeutic treatment, initial WHO grade, WHO grade at relapse, or adjuvant chemotherapy were detectable. Patients treated with bevacizumab were younger, on average, but this turned out to be nonsignificant; thus, this retrospective analysis is relatively well balanced although no randomization took place. When comparing the outcomes it becomes visible that the addition of bevacizumab was associated with improved PFS/ survival rates. The corresponding Kaplan-Meier plots are shown in Fig. 3; log-rank testing revealed a significant result (p = 0.03). The mos after reirradiation alone was 175 days (mpfs, 143 days). The mean survival was days after reirradiation alone compared to days after reirradiation plus bevacizumab; the median PFS was 244 days compared to 143 days with reirradiation alone (Fig. 3). The PFS-6 was 72% for reirradiation and bevacizumab compared to reirradiation alone, with 24%. From seven adjuvant bevacizumab-based chemotherapies, 6 patients had received Re-RT plus bevacizumab previously. Among these 6 patients, survival and PFS were not significantly different from those in patients who did not receive additional bevacizumab-based adjuvant regimens Fig. 3. Kaplan-Meier progression-free survival/survival curves for patients retreated with radiotherapy (RT) and bevacizumab (N = 20) vs. patients who received only reirradiation (= 10); p values derived from log-rank test.

7 Irradiation and bevacizumab in recurrent malignant glioma d M. NIYAZI et al. 73 Table 5. Univariate analysis, influence on survival (model parameters) Variable p value Age (< 60 y, $ 60 y) NS (p = 0.245) KPS (< 70, $ 70) NS (p = 0.054) surgery (yes/no) NS (p = 0.090) Bevacizumab (yes/no) p = MGMT (meth./not meth.) NS (p = 0.624) TMZ concomitant (yes/no) NS (p = 0.096) Initial WHO grade (<IV/IV) NS (p = 0.115) WHO grade at relapse (III/IV) p = Adjuvant chemotherapy (yes/no) NS (p = 0.087) Abbreviations: KPS = Karnovsky performance score; MGMT = O-6-methylguanine-DNA methyltransferase; TMZ = temozolomide; WHO = World Health Organization; NS = not significant, meth. = methylated. (OS-6 80% vs. 85.7%, p = 0.536; PFS not significant, p = 0.932). Because of the limited number of patients, a final conclusion on adjuvant chemotherapy may not be drawn. Univariate and multivariate analyses To define prognostic and/or predictive factors for overall survival, univariate and multivariate testing was performed. The results are shown in Tables 5 and 6. Univariate testing revealed that bevacizumab and WHO grade at relapse were the only variables with significant influence on survival (p = and p = 0.042, respectively). The corresponding Kaplan-Meier plot for bevacizumab is shown in Fig. 3. In parallel, both factors also turned out to be a significant variable within the full multivariate Cox model (Table 6), ( p = and p = 0.007, respectively; the hazard ratios were 0.139/ Interestingly, a lower WHO grade at relapse was associated with decreased survival (Grade IV median 311 days vs. Grade III 229 days), which might be explained by the heavy pretreatments during the longer overall treatment course. Age and MGMT status were found to be nonsignificant variables in the univariate analysis (p = and p = 0.624) (Table 5). The influence of previous resection, adjuvant chemotherapy, or concomitant TMZ at first RT treatment had only a marginal influence on the duration of survival after retreatment (univariate analysis: p = 0.090, p = 0.087, and p = 0.096) (Table 5). Log-rank test performed on the survival curves resulted in marginally significant p values of (Fig. 4), 0.071, and.078. Fig. 4. Kaplan-Meier survival curves for patients who either underwent surgery or were primarily irradiated; p value derived from logrank test (N = 30). The KPS was marginally significant after univariate analysis (p = 0.054, Table 5), but the two subgroups showed significantly different survival patterns according to the Kaplan-Meier plots (Fig. 5) (p = 0.04). After multivariate testing, the influence of KPS turned out to be significant (p = 0.011) (Table 6). DISCUSSION Although chemotherapy is one of the most common salvage strategies in recurrent MG, prolongation of survival with low toxicity rates can be achieved with modern precision RT techniques in certain subgroups of patients. Our aim was to improve local efficacy by the addition of the highly active drug bevacizumab. Although this study is limited by its retrospective nature, our data clearly support the interpretation that reirradiation in combination with bevacizumab is safe and effective. In this regard, the outcome of our trial compares nicely with data presented by Gutin et al. (45). After matching with an external cohort, Gutin et al. also reported an improved efficacy after the addition of bevacizumab. Table 6. Multivariate Cox proportional hazards analysis, influence on survival (model parameters) (N = 30) Variable Hazard ratio (full model) p value KPS (< 70, $ 70) p = Bevacizumab (no/yes) p = WHO grade at relapse (III/IV) p = Abbreviations: KPS = Karnovsky performance status; WHO = World Health Organization; NS = not significant. Fig. 5. Kaplan-Meier survival curves for patients within different Karnovsky performance status (KPS) categories; p value derived from log-rank test (N = 30).

8 74 I. J. Radiation Oncology d Biology d Physics Volume 82, Number 1, 2012 Vordermark et al. (7) reirradiated patients with recurrent MG without additional chemotherapy. The mos in their study was 9.3 months from the time of hypofractionated stereotactic RT for the whole cohort and 7.9 months for GBM patients. The results of Gutin et al. combining bevacizumab with radiation were superior (12.5 months mos for GBM patients), which is in line with our results (45). But in contrast to the study by Gutin et al., our study was monocentric and, to our knowledge, is to date the only one directly comparing survival rates of reirradiated patients with or without bevacizumab; thus, it may be regarded as being less biased concerning institutional differences. The survival rate of the combined treatment is promising and is higher than most reirradiation survival rates found in the literature (5 8, 10, 11). The combined treatment approach was exceptionally well tolerated. Overall toxicity in our study was not higher than that with the use of bevacizumab alone or in combination with other agents in patients with MG (17, 29, 32, 34 36, 48 51). The only adverse event with a putative potential association with the application of bevacizumab was the occurrence of a wound dehiscence in 1 patient. This side effect of bevacizumab has been described previously (46). Because of the limited number of patients, it is not possible to determine how far the occurrence of a wound dehiscence might be aggravated by the additional use of radiation. To what extent other approaches for retreatment may be superior or should be used earlier in the sequence of retreatments is unclear at present. Reresection, hypofractionated stereotactic radiosurgery, brachytherapy, or singlechemotherapy/polychemotherapy schedules, including new intensified or alternative TMZ treatment protocols, other nitrosourea-based agents, and novel agents like bevacizumab may all be used in this situation (52). Inasmuch as all of these approaches have limited activity and are associated with relevant and sometimes severe toxicity, we consider reirradiation with bevacizumab to be a very effective and safe approach for those patients in whom a second course of RT is feasible. Nevertheless, because of the retrospective nature of this trial, several shortcomings have to be discussed. The sample size was relatively small. Thus, the statistical power is limited. However, the finding that regardless of the small sample size the use of bevacizumab was significantly associated with overall survival, even in a multivariate analysis, underlines the importance of this observation. A second shortcoming is the fact that in nine cases the MGMT status was not known. Although MGMT status is one of the most important prognostic outcome markers in the primary setting, its influence in retreatment settings has not been validated (53, 54). How far our approach is superior to other approaches is difficult to asses at present because randomized trials comparing different retreatment options are lacking. At present, bevacizumab-based chemotherapy protocols are considered to be one potential standard of care in some countries, whereas this approach has not been approved in others. The most successful bevacizumab-containing chemotherapy combination in glioblastoma patients to date is bevacizumab/irinotecan with a PFS-6 of 50.3%; the same study revealed that patients treated with bevacizumab monotherapy had a PFS-6 of 42.6% (55). All other bevacizumabbased approaches resulted in inferior PFS-6 rates (range, %) (33, 35, 56, 57); a comprehensive review has been published by Chamberlain (58). The patients in our series had a PFS-6 of 72%, which is nominally more than the best bevacizumab/chemotherapy result. However, based on the high selection bias, the direct comparison allows only very limited conclusions. All in all, further randomized trials are warranted to finally answer the question whether bevacizumab adds substantial benefit to a radiotherapeutic retreatment setting. REFERENCES 1. Bashir R, Hochberg F, Oot R. Regrowth patterns of glioblastoma multiforme related to planning of interstitial brachytherapy radiation fields. Neurosurgery 1988;23: Jansen EP, Dewit LG, van Herk M, et al. Target volumes in radiotherapy for high-grade malignant glioma of the brain. Radiother Oncol 2000;56: Wallner KE, Galicich JH, Krol G, et al. Patterns of failure following treatment for glioblastoma multiforme and anaplastic astrocytoma. Int J Radiat Oncol Biol Phys 1989;16: Stupp R, Hegi ME, Mason WP, et al. Effects of radiotherapy with concomitant and adjuvant temozolomide versus radiotherapy alone on survival in glioblastoma in a randomised phase III study: 5-year analysis of the EORTC-NCIC trial. Lancet Oncol 2009;10: Ernst-Stecken A, Ganslandt O, Lambrecht U, et al. Survival and quality of life after hypofractionated stereotactic radiotherapy for recurrent malignant glioma. J Neurooncol 2007;81: Henke G, Paulsen F, Steinbach JP, et al. Hypofractionated reirradiation for recurrent malignant glioma. Strahlenther Onkol 2009;185: Vordermark D, Kolbl O, Ruprecht K, et al. Hypofractionated stereotactic re-irradiation: Treatment option in recurrent malignant glioma. BMC Cancer 2005;5: Bauman GS, Sneed PK, Wara WM, et al. Reirradiation of primary CNS tumors. Int J Radiat Oncol Biol Phy 1996;36: Rimmer YL, Burnet NG. Is it worth reirradiating primary brain tumours? Cambridge experience. Br J Cancer 2002;86:S55 S Combs SE, Gutwein S, Thilmann C, et al. Stereotactically guided fractionated re-irradiation in recurrent glioblastoma multiforme. J Neuro-Oncol 2005;74: Combs SE, Widmer V, Thilmann C, et al. Stereotactic radiosurgery (SRS): Treatment option for recurrent glioblastoma multiforme (GBM). Cancer 2005;104: Biswas T, Okunieff P, Schell MC, et al. Stereotactic radiosurgery for glioblastoma: Retrospective analysis. Radiat Oncol 2009;4: Fuller CD, Choi M, Forthuber B, et al. Standard fractionation intensity modulated radiation therapy (IMRT) of primary and recurrent glioblastoma multiforme. Radiat Oncol 2007;2:26.

9 Irradiation and bevacizumab in recurrent malignant glioma d M. NIYAZI et al Combs SE, Debus J, Schulz-Ertner D. Radiotherapeutic alternatives for previously irradiated recurrent gliomas. BMC Cancer 2007;7: Fokas E, Wacker U, Gross MW, et al. Hypofractionated stereotactic reirradiation of recurrent glioblastomas: A beneficial treatment option after high-dose radiotherapy? Strahlenther Onkol 2009;185: Nieder C, Grosu AL, Molls M. A comparison of treatment results for recurrent malignant gliomas. Cancer Treat Rev 2000; 26: Buie LW, Valgus J. Bevacizumab: A treatment option for recurrent glioblastoma multiforme. Ann Pharmacother 2008;42: Dietrich J, Norden AD, Wen PY. Emerging antiangiogenic treatments for gliomas: Efficacy and safety issues. Curr Opin Neurol 2008;21: Kreisl TN. Chemotherapy for malignant gliomas. Semin Radiat Oncol 2009;19: Mrugala MM. Bevacizumab for recurrent malignant gliomas: Efficacy, toxicity, and patterns of recurrence. Neurology 2009;72:773. author reply Wong ET, Hess KR, Gleason MJ, et al. Outcomes and prognostic factors in recurrent glioma patients enrolled onto phase II clinical trials. J Clin Oncol 1999;17: Ameri A, Poisson M, Chauveinc L, et al. Treatment of recurrent malignant supratentorial gliomas with the association of carboplatin and etoposide: A phase II study. J Neurooncol 1997;32: Kappelle AC, Postma TJ, Taphoorn MJ, et al. PCV chemotherapy for recurrent glioblastoma multiforme. Neurology 2001; 56: Omuro AM. Exploring multi-targeting strategies for the treatment of gliomas. Curr Opin Investig Drugs 2008;9: Jain RK, Di Tomaso E, Duda DG, et al. Angiogenesis in brain tumours. Nat Rev Neurosci 2007;8: Plate KH, Breier G, Weich HA, et al. Vascular endothelial growth-factor is a potential tumor angiogenesis factor in human gliomas in vivo. Nature 1992;359: Argyriou AA, Giannopoulou E, Kalofonos HP. Angiogenesis and anti-angiogenic molecularly targeted therapies in malignant gliomas. Oncology 2009;77: Sathornsumetee S, Cao Y, Marcello JE, et al. Tumor angiogenic and hypoxic profiles predict radiographic response and survival in malignant astrocytoma patients treated with bevacizumab and irinotecan. J Clin Oncol 2008;26: Bokstein F, Shpigel S, Blumenthal DT. Treatment with bevacizumab and irinotecan for recurrent high-grade glial tumors. Cancer 2008;112: Chamberlain MC. Bevacizumab plus irinotecan in recurrent glioblastoma. J Clin Oncol 2008;26: author reply Desjardins A, Reardon DA, Herndon JE 2nd, et al. Bevacizumab plus irinotecan in recurrent WHO grade 3 malignant gliomas. Clin Cancer Res 2008;14: de Groot JF, Yung WK. Bevacizumab and irinotecan in the treatment of recurrent malignant gliomas. Cancer J 2008;14: Kreisl TN, Kim L, Moore K, et al. Phase II trial of single-agent bevacizumab followed by bevacizumab plus irinotecan at tumor progression in recurrent glioblastoma. J Clin Oncol 2009;27: Poulsen HS, Grunnet K, Sorensen M, et al. Bevacizumab plus irinotecan in the treatment patients with progressive recurrent malignant brain tumours. Acta Oncol 2009;48: Vredenburgh JJ, Desjardins A, Herndon JE 2nd, et al. Phase II trial of bevacizumab and irinotecan in recurrent malignant glioma. Clin Cancer Res 2007;13: Desjardins A, Vredenburgh JJ, Gururangan S, et al. A phase II study of bevacizumab plus etoposide among recurrent malignant glioma patients. Neuro-Oncology 2009; Joint Meeting of the Society-for Neuro-Oncology/American-Association-of-Neurological -Surgeons/Congress of Neurological Surgeons; Oct New Orleans, LA. 37. Reardon D, Desjardins A, Vredenburgh JJ, et al. Bevacizumab plus etoposide among recurrent malignant glioma patients: Phase II study final results. ASCO Annual Meeting; May 29- June Orlando, FL. 38. Soffietti R, Ruda R, Trevisan E, et al. Phase II study of bevacizumab and nitrosourea in patients with recurrent malignant glioma: A multicenter Italian study. ASCO Annual Meeting; May 29-June Orlando, FL. 39. Sathornsumetee S, Desjardins A, Vredenburgh JJ, et al. Phase II study of bevacizumab plus erlotinib for recurrent malignant gliomas. ASCO Annual Meeting; May 29-June Orlando, FL. 40. Thompson EM, Dosa E, Kraemer DF, et al. Bevacizumab plus carboplatin increases survival in patients with recurrent malignant glioma. Neuro-Oncology 2009;11: Joint Meeting of the Society-for Neuro-Oncology/American-Association-of- Neurological -Surgeons/Congress of Neurological Surgeons; Oct New Orleans, LA. 41. Jenab-Wolcott J, Giantonio BJ. Bevacizumab: Current indications and future development for management of solid tumors. Expert Opin Biol Ther 2009;9: Norden AD, Drappatz J, Wen PY. Novel anti-angiogenic therapies for malignant gliomas. Lancet Neurol 2008;7: Lai A, Filka E, McGibbon B, et al. Phase II pilot study of bevacizumab in combination with temozolomide and regional radiation therapy for up-front treatment of patients with newly diagnosed glioblastoma multiforme: Interim analysis of safety and tolerability. Int J Radiat Oncol Biol Phys 2008;71: Combs SE, Bischof M, Welzel T, et al. Radiochemotherapy with temozolomide as re-irradiation using high precision fractionated stereotactic radiotherapy (FSRT) in patients with recurrent gliomas. J Neurooncol 2008;89: Gutin PH, Iwamoto FM, Beal K, et al. Safety and efficacy of bevacizumab with hypofractionated stereotactic irradiation for recurrent malignant gliomas. Int J Radiat Oncol Biol Phys 2009;75: Cohen MH, Shen YL, Keegan P, et al. FDA drug approval summary: Bevacizumab (Avastin) as treatment of recurrent glioblastoma multiforme. Oncologist 2009;14: Mehrkens JH, Popperl G, Rachinger W, et al. The positive predictive value of O-(2- F-18 fluoroethyl)-l-tyrosine (FET) PET in the diagnosis of a glioma recurrence after multimodal treatment. J Neurooncol 2008;88: Ali SA, McHayleh WM, Ahmad A, et al. Bevacizumab and irinotecan therapy in glioblastoma multiforme: A series of 13 cases. J Neurosurg 2008;109: Badruddoja MA. Bevacizumab plus temozolomide for recurrent malignant glioma: Update. Neuro-Oncology 2009;11: Joint Meeting of the Society-for Neuro-Oncology/ American-Association-of-Neurological -Surgeons/Congress of Neurological Surgeons; Oct New Orleans, LA. 50. Chamberlain MC. Bevacizumab for recurrent malignant gliomas: Efficacy, toxicity, and patterns of recurrence. Neurology 2009;72: author reply Gilbert MR, Wang M, Aldape K, et al. RTOG 0625: A phase II study of bevacizumab with irinotecan in recurrent glioblastoma (GBM). ASCO Annual Meeting; Nieder C, Adam M, Molls M, et al. Therapeutic options for recurrent high-grade glioma in adult patients: Recent advances. Crit Re Oncol Hematol 2006;60: Nagane M, Kobayashi K, Ohnishi A, et al. Prognostic significance of O6-methylguanine-DNA methyltransferase

10 76 I. J. Radiation Oncology d Biology d Physics Volume 82, Number 1, 2012 protein expression in patients with recurrent glioblastoma treated with temozolomide. Jpn J Clin Oncol 2007;37: Weller M, Stupp R, Reifenberger G, et al. MGMT promoter methylation in malignant gliomas: Ready for personalized medicine? Nat Rev Neurol 2010;6: Friedman HS, Prados MD, Wen PY, et al. Bevacizumab alone and in combination with irinotecan in recurrent glioblastoma. J Clin Oncol 2009;27: Vredenburgh JJ, Desjardins A, Herndon JE 2nd, et al. Bevacizumab plus irinotecan in recurrent glioblastoma multiforme. J Clin Oncol 2007;25: Chamberlain MC, Johnston SK. Salvage therapy with single agent bevacizumab for recurrent glioblastoma. J Neurooncol;96: Chamberlain MC. Emerging clinical principles on the use of bevacizumab for the treatment of malignant gliomas. Cancer;116: