The high-grade malignant brain tumors, glioblastoma multiforme (GBM) and anaplastic astrocytoma (AA), comprise the majority of all primary brain tumors in adults. Initial therapy consists of either surgical resection, external beam radiation or both. All patients experience a recurrence after first-line therapy, so improvements in both first-line and salvage therapy are critical to enhancing quality-of-life and prolonging survival. It is unknown if currently used intravenous (IV) therapies even cross the blood brain barrier (BBB). Superselective Intraarterial Cerebral Infusion (SIACI) is a technique that can effectively increase the concentration of drug delivered to the brain while sparing the body of systemic side effects. One currently used drug called, Cetuximab (Erbitux) has been shown to be active in human brain tumors but its actual CNS penetration is unknown. This phase I clinical research trial will test the hypothesis that Cetuximab can be safely used by direct intracranial superselective intraarterial infusion up to a dose of 500mg/m2 to ultimately enhance survival of patients with relapsed/refractory GBM/AA. By achieving the aims of this study the investigators will determine the the toxicity profile and maximum tolerated dose (MTD) of SIACI Cetuximab. The investigators expect that this study will provide important information regarding the utility of SIACI Cetuximab therapy for malignant glioma, and may alter the way these drugs are delivered to the investigators patients in the near future.

There is no current standard of care for recurring GBM after patients receive Bevacizumab (Avastin) intravenously (IV) at 10mg/kg with CPT-11 (Irinotecan). At that point, these patients are deemed treatment failures and are given another experimental treatment. Because of the blood brain barrier (BBB) where IV drugs do not penetrate the blood vessel walls well to get into the brain, no one knows for sure if these IV drugs actually get into the brain after infusion. Previous studies have shown that if you want to increase your penetration of drug to the brain, that intra-carotid artery (intraarterial) delivery is superior to standard intravenous delivery. Previous techniques using intraarterial (intracarotid) infusion still were non-selective as drug delivery still went to all blood vessels in the brain, so patients still had significant adverse events, such as blindness. Newer techniques in interventional neuroradiology have allowed for a more selective delivery of catheters higher up into the arterial tree where agents such as chemotherapies, can be delivered without the risk of adverse affects such as blindness. In fact, studies here at Cornell have developed very new and exciting super selective intraarterial delivery treatments for Retinoblastoma, eye tumors with little toxicity and a clinical trial of intraarterial delivery of Avastin is currently underway for GBM. Therefore, this trial will ask one simple question: Is it safe to deliver a dose of Cetuximab intraarterially using these super selective delivery techniques instead of the standard intravenous route of administration? This should not only increase the amount of drug that gets to the tumor but also spare the patient any adverse effects from a less selective delivery. Prior to that single dose of intraarterial Cetuximab, the patient will also receive a dose of mannitol that opens up the blood brain barrier to improve delivery of the agent to the brain. After that single dose of Mannitol and Cetuximab intraarterially, the patient will be evaluated for 4 weeks to assess for toxicity. After this point, the patient is done with the "experimental" aspects of the protocol. If no toxicity at this point, then the patient will go on and get their chosen chemotherapy as determined by their treating oncologist. In summary, this is a Phase I trial that is designed to test the safety of a single dose of intraarterial delivery of Mannitol and Cetuximab , prior to starting the patients next round of chosen chemotherapy

To summarize:

Current Standard of Care Therapy : None

Experimental portion of this proposal:

Day 0: Intraarterial Mannitol to open the blood brain barrier followed by Intraarterial Cetuximab single dose (starting at 100mg/m2 and up to 500mg/m2)

Therefore the experimental aspects of this treatment plan will include:

1. Subjects will first be treated with Mannitol prior to chemotherapy infusion (Mannitol 25%; 3-10 mL/s for 30seconds) in order to disrupt the blood brain barrier. This technique has been used in several thousand patients in previous studies for the IA delivery of chemotherapy for malignant glioma.
2. To add a single intraarterial delivery (SIACI) of the Cetuximab for patients with recurring or relapsing high grade glioma. After a one cycle observation period to assess for toxicity from the IA infusion, the subject will receive a regimen of chemotherapy to be decided by their treating oncologist The dose escalation algorithm is as follows: We will use a single intracranial superselective intraarterial infusion of Cetuximab, starting at a dose of 100mg/m2 in the first three patients. Assuming no dose limiting toxicity during the next 28 days after the infusion, the patient will then begin their standard chemotherapy regimen which is to be determined by their treating oncologist. The doses will be escalated from 100, to 200, 300, 400 and finally 500mg/m2 in this Phase I trial.

Both hematologic and non-hematologic toxicity from the IA infusion of Cetuximab will be determined and scored according to the NCI Common Toxicity Criteria (version 3.0). Monitoring will also include an MRI of the brain at 4 weeks post infusion.

Most patients with GBM are also monitored every two months with serial history, neurological and physical examinations together with serial blood counts, prothrombin time (PT), partial thromboplastin time (PTT) and chemistries. In addition, most patients with GBM have an MRI performed every two cycles or approximately every two months to assess for tumor progression. .

Since this is a Phase I trial, response is not a primary endpoint of the trial. However, we will evaluate response to the one time IA Cetuximab therapy with a a MRI with the injection of contrast about 4 weeks after infusion. Follow-up of all patients in the trial regardless of the chemotherapy regimen they try after the IA Cetuximab therapy will continue until disease progression or death. Survival will be measured from the time of the dose of IA Cetuximab®. We expect patients in the trial to monitored for 12 months.

This treatment may be harmful to a fetus . female subjects of childbearing age, will be asked to practice birth control methods while participating in this research study and for 3 months following her treatment. These methods include oral contraceptives, contraceptive shots, and barrier methods, such as condom use, sponges, and diaphragms. Fertile males are required to use these barrier methods.

The patient may be responsible for any additional costs associated with enrollment in the trial. All costs of the IA delivery and the cost of the drug will be submitted to the patient's insurance provider.

Inclusion Criteria:

• Male or female patients of greater than or equal to 18 years of age.
• Patients with a documented histologic diagnosis of relapsed or refractory (malignant tumors that recur or resist treatment) glioblastoma multiforme (GBM), anaplastic astrocytoma (AA) or anaplastic mixed oligoastrocytoma (AOA).
• Patients must have at least one confirmed and evaluable tumor site. A confirmed tumor site is one in which is biopsy-proven. NOTE: Radiographic procedures (e.g., Gd-enhanced MRI or CT scans) documenting existing lesions must have been performed within three weeks of treatment on this research study.
• Patients must have a Karnofsky performance status greater than or equal to 60% (or the equivalent ECOG level of 0-2) (see Appendix A; Performance Status Evaluation) and an expected survival of greater than or equal to three months.
• No chemotherapy for two weeks prior to treatment under this research protocol and no external beam radiation for two weeks prior to treatment under this research protocol.
• Patients must have adequate hematologic reserve with WBC greater than or equal to 3000/mm3, absolute neutrophils greater than or equal to 1500/mm3 and platelets greater than or equal to 100,000/ mm3. Patients who are on Coumadin must have a platelet count of greater than or equal to 150,000/ mm3
• Pre-enrollment chemistry parameters must show: bilirubin less than 1.5X the institutional upper limit of normal(IUNL); AST or ALT less than 2.5X IUNL and creatinine less than 1.5X IUNL.
• Pre-enrollment coagulation parameters (PT and PTT) must be equal to or less than 1.5X the IUNL.
• Concomitant Medications
• Steroids Systemic corticosteroid therapy is permissible in patients with CNS tumors for treatment of increased intracranial pressure or symptomatic tumor edema. Patients with CNS tumors who are receiving dexamethasone must be on a stable or decreasing dose for at least 1 week prior to study entry. We do not believe that study procedures place subjects with increased intracranial pressure at any additional risk.
• Study Specific Patients on enzyme-inducing anticonvulsants or non-enzyme inducing anticonvulsants will be allowed on the study. Patients receiving proton pump inhibitor or H2 blockers will be allowed on study. Patients taking antacids will be allowed on study.
• Patients must agree to use a medically effective method of contraception during and for a period of three months after the treatment period. A pregnancy test will be performed on each premenopausal female of childbearing potential immediately prior to entry into the research study.
• Patients able to understand and give written informed consent and those patients that are cognitively impaired (which is common in GBM) are eligible for the trial. Informed consent must be obtained at the time of patient screening (prior to Day 0 of the procedure) either by the patient or a legalized authorized representative (LAR) of the patient ( health-care proxy).

Exclusion Criteria:

• Women who are pregnant or lactating.
• Women of childbearing potential and fertile men will be informed as to the potential risk of procreation while participating in this research trial and will be advised that they must use effective contraception during and for a period of three months after the treatment period.
• Patients with significant inter-current medical or psychiatric conditions that would place them at increased risk or affect their ability to receive or comply with treatment or post-treatment clinical monitoring.