Low-dose Intra-arterial Bevacizumab for Edema and Radiation Necrosis Therapeutic Intervention (LIBERTI)

BACKGROUND Radiation Necrosis: Stereotactic radiosurgery has become integral in treatment of brain tumors and arteriovenous malformations (AVM). In up to 10% of cases, this can lead to radiation necrosis (RN) with significant surrounding vasogenic edema and mass effect. Medical treatment for RN includes steroids, vitamin E, pentoxiphylline, and hyperbaric oxygen. Up to 20% of cases however, are medically refractory and experience progressive neurological decline and disabling headaches.

Surgical resection and laser interstitial thermal therapy (LITT) are sometimes used for treatment of medically refractory radiation necrosis. These invasive options carry risks of serious complications such as infection, seizures, neurological deficits, and hemorrhage, and may have a failure rate as high as 39% in patients with radiation necrosis of the brain.

Bevacizumab: Bevacizumab (Avastin, Genentech BioOncology, South San Francisco, CA) is a recombinant humanized version of a murine anti-human vascular endothelial growth factor (VEGF) monoclonal antibody. Recently, bevacizumab was shown in a small randomized controlled trial (n=14) to be effective in treatment of refractory radiation necrosis after radiation therapy in brain tumors1. Patients received 7.5 mg/kg IV-Bevacizumab every 3 weeks for 4 cycles. All patients receiving Bevacizumab and none of the patients receiving placebo had significant clinical and radiographic improvement.

PRE-CLINICAL DATA Role of vascular endothelial growth factor (VEGF) in radiation necrosis VEGF has been implicated in the pathophysiology of radiation necrosis. Reactive astrocytes immediately surrounding the necrotic core in RN are strongly VEGF-positive by immunohistochemistry. It is postulated that radiation causes microvascular injury leading to hypoxia. Hypoxia-induced VEGF up-regulation then drives an increase in vascular permeability, leading to the extensive vasogenic edema seen in RN. Bevacizumab binds circulating VEGF receptors with high specificity, blocking the down-stream signaling cascade.

CLINICAL DATA:

Bevacizumab was originally developed and tested as an anti-angiogenic treatment for various solid tumors. More recently, IV-Bevacizumab was shown in a small, randomized controlled trial (n=14) to be very effective in treatment of refractory radiation necrosis after radiation therapy in brain tumors1. Patients received 7.5 mg/kg IV-Bevacizumab every 3 weeks for 4 cycles. All patients receiving Bevacizumab and none of the patients receiving placebo had significant clinical and radiographic improvement. This improvement was durable at 10 months in 8 of 11 patients (4 patients crossed over from the control group). There was however, a very high rate of adverse events (60%), and major adverse events (30%). Major adverse events included venous sinus thrombosis, pulmonary embolus, and aspiration pneumonia.

The investigators of the present study recently published a case series of two pediatric patients with highly symptomatic steroid refractory radiation necrosis in the brain after stereotactic radiosurgery for treatment of cerebral arteriovenous malformations3. Both patients were refractory to all accepted medical therapies. Both were steroid dependent for a prolonged period and severely cushingoid. Both had suffered a significant decline in quality of life with severe headache and needed to withdraw from school. In both instances, the patients made a remarkable progressive clinical and radiographic improvement after receiving a single 2.5 mg/kg dose of intra-arterial bevacizumab, which was durable one-year later. To increase bevacizumab penetration into the brain, the investigators used intra-arterial Mannitol to disrupt the blood-brain barrier immediately prior to targeted intra-arterial drug administration.

RATIONALE:

RATIONALE: CURRENT IV BEVACIZUMAB REGIMEN FOR RADIATION NECROSIS AND ITS ASSOCIATED MORBIDITY:

Current IV-bevacizumab regimens use a dose of 7.5 mg/kg every 3 weeks for 4 cycles. There are significant known side effects of bevacizumab including gastrointestinal perforation, deep venous thrombosis, venous sinus thrombosis, pulmonary embolus, intracranial hemorrhage, wound dehiscence, and severe hypertension. These complications are common to the anti-angiogenic class of drugs and reflect systemic exposure to bevacizumab. In our initial clinical experience, the investigators utilized a combination of intra-arterial (IA) route of delivery and BBB disruption to reduce bevacizumab dose while maintaining efficacy. This is supported by the durable clinical and radiographic response in our patients after a single 2.5 mg/kg dose of bevacizumab. The investigators believed that this approach would reduce the incidence of serious systemic toxicities compared to the IV-bevacizumab regimens (7.5-15 mg/kg every 2-3 weeks for several weeks to months).

There are multiple recent reports of patients with radiation necrosis who improved with IV-bevacizumab, only to relapse months later. In fact 3/11 patients in the randomized controlled trial discussed above required repeat treatment with IV-bevacizumab because of RN symptom progression1. In contrast, the two patients in our series who received IA-bevacizumab continue to show progressive clinical and radiographic improvement more than one year later. The investigators believe that the increased penetration of bevacizumab into the brain because of the intra-arterial administration after blood-brain barrier disruption results in binding of virtually all VEGF molecules. The fact that the results are durable and progressively improving suggests that massive blocking of VEGF activity could have stopped a positive feedback loop of inflammation. Therefore, IA-bevacizumab may result in more effective and durable control of radiation necrosis compared to traditional IV-bevacizumab treatment.

RATIONALE: INTRA-ARTERIAL (IA) ROUTE OF BEVACIZUMAB ADMINISTRATION SIGNIFICANTLY INCREASES DRUG DELIVERY TO THE BRAIN:

IA-therapy decreases volume dilution of the drug in the circulation and reduces first-pass degradation via proteolytic catabolism, resulting in higher drug delivery to target brain tissue. Super-selective IA-injection of 99mTc-HMPAO (Ceretec®) into human cerebral arteries achieves a concentration of radiotracer in brain tissue 50 times higher than with IV injection. In clinical studies of cerebral chemotherapy, the concentration delivered to the tumor by using intra-arterial injection versus intravenous administration of chemotherapeutic agents was five times higher with hydrosoluble drugs and up to 50 times higher with liposoluble drugs. The investigators will infuse bevacizumab in the artery that supplies the territory affected by RN, such as cervical internal carotid artery and/or cervical vertebral artery.

RATIONALE: BLOOD-BRAIN-BARRIER BREAKDOWN PRIOR TO INTRA-ARTERIAL THERAPY FURTHER ENHANCES DRUG DELIVERY TO THE BRAIN:

The blood-brain-barrier is a selective permeability barrier that block entry of many drugs into the brain. Bevacizumab is a monoclonal antibody with a high molecular weight (149 kDa). There is convincing evidence in the literature that the concentration in the brain of high molecular weight molecules can be significantly increased after osmotic BBB disruption. Several tumor clinical trials have shown that localization of monoclonal antibodies to the brain is poor without BBB disruption (0.0006%-0.0043% of the injected dose/g of tumor). There is also evidence of a 20-fold increase in permeability to immunoreactive IgM Mab with BBB disruption in rats. The investigators believe that using blood-brain-barrier disruption significantly increases delivery of Bevacizumab to the affected brain. The investigators will use the protocol described by Neuwelt and colleagues, using infusion of 25% Mannitol over 30 seconds. This protocol has been shown to temporarily disrupt the blood brain barrier, peaking at 15 minutes and dissipating in 4 hours. IA-chemotherapy following BBBD has been shown to be feasible and safe across multiple centers with low incidence of complications27. The efficacy and safety profile was reproducible across multiple centers. In fact, safety of this protocol has been established in more than 6000 patients treated worldwide with BBBD for intra-arterial chemotherapy infusion. The main possible complication is seizure, which occurs in <6% of cases. It is important to note that these seizures generally occurred in patients with widespread malignant pathology such as Glioblastoma and CNS lymphoma who were treated with very toxic chemotherapy agents immediately after BBBD. Recent refinements to the osmotic BBBD protocol have incorporated the use of general anesthesia, as well as prophylaxis with an anti-epileptic agent and Valium to reduce seizure threshold and the chance of seizures.

SAFETY OF CEREBRAL INTRA-ARTERIAL BEVACIZUMAB TREATMENT:

Safety of IA-Bevacizumab treatment after hyperosmotic BBBD was recently established in a series of malignant glioma patients. This was done through super-selective injection of intracranial tumor arterial pedicles for purpose of anti-tumor effects. Dose-escalation was performed from 2 mg/kg to 15 mg/kg without reaching maximal tolerated dose. There was a significant decrease in the contrast enhancing and FLAIR signal characteristics of the tumor and surrounding brain at one month after treatment. Overall toxicity for this cohort was comparable to previous reports for IV Bevacizumab therapy. Specifically, hyperosmotic BBB-breakdown followed by IA-Bevacizumab administration did not cause any direct neurotoxicity; there were no cases of intracranial hemorrhage. Multiple other reports of BBBD followed by intra-arterial bevacizumab treatment for other pathologies such as vestibular schwannoma, ependymoma, and malignant brainstem glioma have also demonstrated good safety profile with no obvious neurotoxicity.

TREATMENT PLAN:

VASCULAR ACCESS, CEREBRAL ANGIOGRAM, AND OSMOTIC BLOOD-BRAIN-BARRIER DISRUPTION:

The investigators will use the protocol described by Neuwelt and colleagues, using infusion of 25% Mannitol over 30 seconds. The safety of this protocol has been established in more than 6000 patients treated worldwide with BBBD for intra-arterial chemotherapy infusion.

The patients are to be premedicated with 6 mg Dexamethasone and 1000 mg Keppra. General endotracheal anesthesia will be induced. The femoral artery will be accessed using the Seldinger technique. A 5-French diagnostic catheter will be used to catheterize the cervical internal carotid artery ipsilateral to the area of radiation necrosis. Baseline internal carotid angiogram will be performed.

The anesthesiologist will be instructed to maintain SBP >120 or at pre-operative baseline, whichever value is higher. This is important for efficient bulk flow of drug through the blood brain barrier opening. The catheter is positioned at C1-2 level in the cervical internal carotid artery and C6-7 for a vertebral artery infusion. Optimal rate of Mannitol infusion will be determined by performing injection of contrast at 4 ml/sec for 3 seconds into vessel. If there is no reflux of contrast into the external carotid artery, the injection rate injection will be increased by 2 ml/sec to maximum of 12 ml/sec. The lowest rate at which there is reflux into the external carotid artery will be chosen (the rate to just exceed cerebral blood flow.

Next, 5 mg IV Valium and 0.2 mg IV Atropine are to be administered. Warm (37 degrees Co) 25% Mannitol is filtered through a 5-micron filter, and then infused into the ipsilateral cervical carotid artery at the rate determined above for a total of 30 seconds.

INTRA-ARTERIAL BEVACIZUMAB ADMINISTRATION:

Test injection of contrast will be done in the artery. If there is any evidence of catheter-induced vasospasm, the catheter may be withdrawn more proximally within the artery. Repeat test injection of contrast will be done to document resolution of vasospasm. Within 5 minutes of Mannitol infusion, 2.5 mg/kg bevacizumab in a volume of 100 ml will be administered into the artery over 10 minutes. Repeat angiogram will be performed to document BBBD, as well as to rule out thromboembolic phenomenon.