A Study of Subcutaneous Bevacizumab in Relapsed / Progressive Glioblastoma

1.1 Primary Aim: To describe MRI response rate as regards edema and enhancement of
glioblastoma and radiation- related brain enhancement when treated with subcutaneous (SQ)
bevacizumab daily. See section 11 for detail on response assessment.

1.2 Secondary Aims: 1.2.1 To characterize toxicities of SQ bevacizumab. 1.2.2 To describe
response rate as per primary aim (above) in study participants who discontinue SQ bevacizumab
for any reason and go on to receive treatment with standard IV bevacizumab or
bevacizumab-containing combination.

2 BACKGROUND 2.1 Bevacizumab is a monoclonal antibody cancer therapeutic which targets
vascular endothelial growth factor (VEGF) and which has been FDA approved for the treatment
of glioblastoma. Clinical trials of bevacizumab in glioblastoma multiforme (GBM) patients
documented edema and enhancement response rates between 75 and 100%, with consequent
improvement in neurologic symptoms and decrease in corticosteroid requirements.(1-3) The FDA
approval was based on response rate; there has not been proof that bevacizumab prolongs
survival in GBM patients. The related VEGF antagonist cediranib can decrease brain edema in a
manner comparable to bevacizumab(4) but failed to prolong survival in GBM patients in a
recent phase III study.(5) Thus bevacizumab remains strictly palliative for GBM patients, a
disease which in any case is not thought to be curable with available treatments.

2.2 Most monoclonal antibody therapeutics remain effective when given subcutaneously instead
of intravenously.(6-9) Preclinical animal data show this to be true for bevacizumab as well,
with 100% bioavailability for SQ as compared to IV administration.(10) In a mouse tumor
model, SQ bevacizumab produced better antitumor effects than intravenous bevacizumab.(11)
Some treatments are also less toxic when given subcutaneously - a recent example is
bortezomib, which produces less neuropathy when given SQ instead of IV but is just as
effective against myeloma.(12) This is presumably because of lower peak doses. There is no
published correlation between dose of bevacizumab and incidence of the common side effects,
which include hypertension and proteinuria, as well as less common but more dangerous
toxicities such as gut perforation. If these toxicities relate to peak doses obtained after
IV administration, it may be that they are less frequent in a dosing scheme in which SQ
administration yields a low sustained and stable dose. It is not clear what SQ dose of
bevacizumab is necessary to obtain benefit in GBM. Although the FDA approved dose is 10mg/kg
every 14 days, earlier European GBM studies used 5mg/kg every 14 days and produced similar
response rates.(1) This uncertainty about the required dose is, in fact, true for all the
diseases in which bevacizumab is used; an example can be seen in two simultaneously published
phase III ovarian cancer studies, one of which used bevacizumab maintenance at 15mg/kg every
3 weeks and the other half that dose at 7.5mg/kg every 3 weeks, with the lower dose study
actually obtaining improved survival for participants as compared to the higher.(13, 14) It
is possible that at least in GBM, lower doses might be better. Some theorize that the hypoxia
produced by antiangiogenic treatments promote a more aggressive, infiltrative tumor
phenotype. There is some thought that a less complete blockade of VEGF would decrease
consequent hypoxia and conversion to this aggressive phenotype.(15) A recent retrospective
series reported that GBM patients who encounter bevacizumab toxicity and undergo dose
reduction have longer survival than contemporaneously treated GBM patients who continue at
10mg/kg every 14 days.(16) Thus it is ethical and perhaps even beneficial to explore lower
doses of bevacizumab in the treatment of GBM.

2.3 The rapid and nearly universal response rate in GBM patients as regards enhancement and
edema together with the direct palliative benefit of bevacizumab monotherapy makes this an
excellent population to study an alternate dosing scheme. In this study the investigators
propose to study subcutaneous bevacizumab for the treatment of patients with GBM and
progressive edema representing tumor progression or symptomatic radiation aftereffects. All
participants will receive the same dose of 25 mg SQ daily, which would translate into 350 mg
over 14 days or 5mg/kg/14 days for a 70 kg patient - the same total dose used in the European
GBM series and the same dose as is used to treat metastatic colon cancer. Bevacizumab
solution is known to be stable in refrigerated prefilled syringes(17), and participants would
take these home, injecting themselves after the manner of low molecular weight heparin.
Participants would receive weekly toxicity assessment, and MRI at 3 weeks looking for
evidence of response in terms of reduction of enhancement and edema; see section 11. Those
participants with response would continue on SQ bevacizumab as long as they continued to
experience clinical benefit. Participants not responding, progressing after initial response
or encountering toxicity specific to the subcutaneous route of injection or discontinuing
study participation for any reason would be offered conversion to standard IV bevacizumab as
standard of care or other treatments as appropriate and available. Improvement in enhancement
/ edema among participants converting to standard IV bevacizumab or bevacizumab containing
combinations will be followed as a secondary endpoint.

2.4 The study would be done as a Simon 2 stage, with an expected response rate as regards
enhancement / edema of 90%. If fewer than 5 of an initial 6 participants show response the
study will end and the SQ route will be deemed ineffective. If >4 of 6 respond, then accrual
will expand to a total of 13 patients. See section 10.1, below.

2.5 The convenience of a SQ regimen would benefit some patients with GBM. The current
requirement for IV therapy requires the patient to travel to an oncology infusion center and
have IV access placed every other week, which is arduous and cumbersome. In addition, if
bevacizumab were effective SQ for patients with GBM, this would offer the prospect of SQ
bevacizumab being studied in other diseases with the goal of devising all-outpatient regimens
and liberating patients from the infusion center. Examples include ovarian cancer (for which
maintenance bevacizumab monotherapy improves survival), hereditary hemorrhagic
telangiectasia, and in combination with other subcutaneous or oral medicines (e.g.
capecitabine, interferon) for the treatment of metastatic solid tumors.

Inclusion Criteria:

- Diagnosis: Participants with glioblastoma are eligible for this study. These will
include

- Those with a histologically proven diagnosis of glioblastoma who have developed
new changes on MRI following primary treatment.

- Those who received primary treatment for a histologically proven lower grade (2
or 3) glioma and who now progress with radiographic characteristics of
transformed glioma.

- Disease status: Patients must have abnormal enhancement on contrast enhanced MRI of
the brain. They must be patients for whom bevacizumab is indicated and appropriate, as
drug will be charged to insurance.

- Participants with newly detected enhancement are eligible, with bevacizumab
treatment hoped to prevent symptoms.

- Participants with stable enhancement / edema are eligible if they require
corticosteroids to control symptoms, and it is thought bevacizumab treatment
might allow lowering of the corticosteroid dose or improvement of symptoms.

- Performance status: Eastern Cooperative Oncology Group (ECOG) 0-3.

- Age: Greater than 18 years.

- Life expectancy: > 3 months.

- Prior therapy:

- Anti-VEGF treatments: 3 months must have elapsed between any prior anti-VEGF
treatment (for example, given as a component of primary treatment) and study
participation. These therapies include bevacizumab, cediranib, axitinib,
sunitinib as well as other therapeutics targeting VEGF.

- Other anti-cancer treatments: Treatments in this category include chemotherapy
and targeted therapies not targeting VEGF. 14 days must have elapsed since
discontinuation of prior chemotherapeutic treatments for glioma and study
treatment.

- Radiation: Radiation is integral to the primary treatment of glioma. All participants
on this study must have had prior radiation to the brain. Radiation must have been
completed 14 days prior to first study treatment.

- Surgery: 14 days must have elapsed since prior major surgery.

- Organ function requirements:

- Creatinine clearance of 50 or greater.

- Absolute neutrophil count of 1500 or greater.

- Blood pressure requirements: Participants must have a baseline blood pressure of
160/90 or less. Participants requiring medicines to control blood pressure are
eligible.

- Informed consent: All patients must sign a written informed consent.

- Urine pregnancy test: Women of childbearing age must have a negative urine pregnancy
test.

Exclusion Criteria:

- People who progress with only nonenhancing tumor on MRI are ineligible. Patients must
have some component of abnormal enhancement on contrast enhanced MRI of the brain.
Combinations of nonenhancing and enhancing tumor are eligible.

- Pregnancy or breast-feeding: Pregnant or breast-feeding women will not be entered on
this study.

- Renal insufficiency: Patients with a creatinine clearance of less than 50 are
ineligible.

- Proteinuria: Patients with 2+ proteinuria/Moderate or more at baseline are ineligible.

- Comorbidities: Patients may not have any baseline comorbidities or laboratory
abnormalities which would be of grade 3 or worse if graded as toxicities by CTCAE
(excepting alopecia). An exception is also made for neurologic comorbidities (eg
ataxia, aphasia) arising as a consequence of the brain tumor; symptoms severe enough
to warrant medical treatment as is offered on this study are by definition grade 3.

- Patients who in the opinion of the investigator may not be able to comply with the
safety monitoring requirements of the study.

- Illness or any other circumstances (as defined by the investigator), which would
preclude safe performance of study procedures or compromise the ability of the patient
to consent to study.