Randomized, Double-Blind, Placebo-Controlled Trial of Bevacizumab Therapy for Radiation Necrosis of the Central Nervous System

Abstract & Commentary

By Samir P. Kanani, MD, Associate Clinical Professor of Neurosurgery and Radiation Oncology, George Washington University; Radiation Oncology, Inova Fairfax Hospital, Falls Church, VA. Dr. Kanani reports no financial relationships relevant to this field of study.

Synopsis: In a small, single-institutional, prospective, randomized, placebo-controlled trial, 14 eligible patients (median age 47 years) with clinical and radiographic evidence of radiation necrosis secondary to prior head-and-neck or CNS irradiation were randomized to receive intravenous saline of bevacizumab at 3-week intervals. Patients were followed with serial MRI scans, neurologic examinations, and formal neuropsychological examinations. No patients receiving placebo responded by MRI scans while five of five (100%) of patients who received bevacizumab responded on MRI with a decrease in necrosis as estimated on T2-weighted fluid–attenuated inversion recovery scans and T1-weighted gadolinium-enhanced scans. Crossover was allowed in patients receiving placebo who did not respond after two cycles; seven of seven patients crossed over and all of them were found to have a radiographic response. All bevacizumab-treated patients — and none of the placebo-treated patients — showed clinical improvement. At a median of 10 months after the last bevacizumab dose, two patients had a recurrence of radiation necrosis and both of those patients were retreated with 1-2 doses of bevacizumab with success. This provided class I evidence of the efficacy of bevacizumab in the treatment of symptomatic central nervous system radiation necrosis.

Source: Levin VA, et al. Randomized double-blind placebo-controlled trial of bevacizumab therapy for radiation necrosis of the central nervous system. Int J Radiation Biol Phys 2011;79:1487-1495.

Chemoradiotherapy is the standard of care in the management of in high-grade gliomas and advanced-stage head and neck cancers, while fractionated radiotherapy and high-dose stereotactic radiotherapy often are used as an alternative to neurosurgical resection in benign CNS diseases. The rates of symptomatic radiation necrosis are generally less than 5-10% and correlate to the dose of radiotherapy administered as well as the volume of tissue irradiated. The mechanism of action responsible for radiation necrosis is currently felt to be triggered by local cytokine release resulting in an increase in capillary permeability and extracellular edema ultimately leading to tissue hypoxia, demyelination of neurons, and necrosis. Treatments to date have included corticosteroids, antiplatelet agents, anticoagulants, hyperbaric oxygen, high-dose vitamins, and surgery.1 Preclinical studies have demonstrated that vascular endothelial growth factor (VEGF) plays a role in the process of disrupting the blood-brain barrier and the cascade of events that follow resulting in radiation necrosis. The investigators hypothesized that blocking VEGF with bevacizumab would reduce the movement of plasma water through "leaky" brain capillary endothelium.

This Phase 3 trial was conducted at the University of Texas M.D. Anderson Cancer Center. Fourteen patients were randomized to either receiving bevacizumab or placebo infusions. Patients were considered eligible for the trial if there was MRI evidence of radiation necrosis and/or biopsy confirmation. In addition, all patients were required to have neurologic symptoms consistent with the location of necrosis and no prior bevacizumab therapy. Concurrent dexamethasone was allowed in the study as long as the dose was stable for at least a week prior to enrolling. Concurrent warfarin was allowed on study as long as INR was therapeutic. Patients were considered ineligible if they had clinically significant cardiovascular disease such as hypertension, CVA, MI, arrhythmias, unstable angina, or CHF < 6 months of study entry. Patients in Group A received IV bevacizumab at a dose of 7.5 mg/kg at 3-week intervals for two treatments and if patients were responding clinically and radiographically they received another two doses separated by 3 weeks. Patients in Group B received placebo (normal saline). Since none of the placebo-treated patients had a response, all seven were allowed to crossover and received bevacizumab. Failure was defined as an increase in radiation necrosis on MRI, no reduction on MRI, or insufficient reduction and progression of neurologic symptoms. MRI studies included pre- and post-gadolinium administered sequences on a 3.0T scanner. Post-contrast images included 3D dynamic contrast-enhanced MRI images (DCE-MRI). All images were reviewed by a blinded team of neuroradiologists. Serial neurologic examinations, formal neurocognitive testing, dexamethasone dosing records, and self-reports of symptoms using the M. D. Anderson Symptom Inventory (MDASI) were also performed on all patients.

The majority of the patients enrolled in the trial had CNS malignancies and presenting symptoms of radiation necrosis varied from headaches to hemiparesis to decreased vision. Of the seven patients randomized to placebo, five had progression of neurologic symptoms and two demonstrated only progression on MRI. All patients then crossed over and received bevacizumab. For patients randomized to bevacizumab, all demonstrated MRI responses and none progressed clinically. The median increase in volume of FLAIR abnormality was +14% in patients receiving placebo and the median decrease in volume of FLAIR abnormality was 59% in those assigned to bevacizumab (P = 0.0149). The median increase in volume of enhancing tissue was +17% in patients receiving placebo, while patients in the bevacizumab arm had a median decrease in volume of enhancing issue of -63% (P = 0.0058). At a median follow-up of 10 months after the end of treatment, three of 12 patients (25%) required retreatment with one to four doses of bevacizumab and all patients demonstrated radiographic improvement. Of the entire cohort, five patients were on dexamethasone at baseline and four of five were able to reduce their dexamethasone. The one patient who was not able to reduce dexamethasone was found to have progression of his astrocytoma to glioblastoma. Six of 11 patients receiving bevacizumab experienced an adverse event and three were classified as serious: one aspiration pneumonia, one PE secondary to DVT, and one superior sagittal sinus thrombosis. Neurocognitive testing demonstrated a trend toward improvement in the aspects of learning and memory after 6 weeks of therapy despite increasing deficits in memory retrieval. Also, there was some evidence for a reduction in severity of symptoms and improved everyday functioning.


This study provides excellent data and justification for the use of bevacizumab in the treatment of symptomatic radiation necrosis. One salient point is the data are drawn from a population with symptomatic not asymptomatic radiation necrosis. A number of patients will develop radiographic changes on an MRI after CNS directed chemo-radiotherapy. It is important to distinguish mere MRI changes and symptomatic radiation necrosis. Too often in my experience, patients are receiving scans for vague or mild symptoms 2-4 weeks after completing radiotherapy. Invariably the reading neuroradiologist will describe progression of disease and/or radiation necrosis in the radiographic findings. Differentiating radiation necrosis can be difficult. Other imaging, such as FDG PET, methionine PET, or thallium chloride-201 PET, can help differentiate tumor growth from necrosis.2 In addition, magnetic resonance spectroscopy can differentiate necrosis from tumor growth.3 The concept of "pseudoprogression" has been well documented in a number of cases for patients treated with temozolomide and radiotherapy for GBM and should not be confused with radiation necrosis and/or progression of disease. It should be pointed out that all patients in this trial were at least 6 months out from radiotherapy. Administering bevacizumab 4-8 weeks after chemoradiotherapy should be done with great caution. Not only does bevacizumab result in adverse events (6/11 on this trial alone), but it precludes patients from enrolling in a number of clinical trials investigating new agents for the treatment of refractory gliomas.

Complications of therapy included superior saggital sinus thrombosis, DVT, and ischemic changes. The authors hypothesize that low-dose anticoagulation might be appropriate with bevacizumab. This could potentially help with some of the discussed complications but may increase risk of hemorrhage. The optimal regimen has yet to be established, but I think potentially de-escalating the dose of bevacizumab to 5 mg/kg and incorporating low-dose anticoagulants would be an appropriate regimen to investigate. Decisions regarding the management of symptomatic radiation necrosis should be made in a multidisciplinary fashion with consultation from neurosurgery, medical oncology, radiology, hyperbaric medicine, and radiation oncology.


1. Schellart NA, et al. Hyperbaric Oxygen treatment improved neurophysiologic performance in brain tumor patients after neurosurgery and radiotherapy: A preliminary report. Cancer 2011;117:3434.

2. Thiel A, et al. Enhanced accuracy in differential diagnosis of radiation necrosis by PET-MRI coregistration: Technical case report. Neurosurgery 2000;46:232.

3. Rock JP, et al. Associations among magnetic resonance spectroscopy, apparent diffsion coefficients, and image-guided histopathology with special attention to radation necrosis. Neurosurgery 2004;54:1111-1117.