A Dose Escalation Phase I Study Of hrBMP4 Administrated Via Convection-
Enhanced Delivery In Patients With Progressive And/Or Multiple Recurrent
Glioblastoma Multiforme.

Gliomas are the most common supratentorial primary brain tumour in adults and
children. Malignant gliomas constitute at least 35% of all primary brain
tumours and are the 3rd leading cause of death from cancer in persons 15 to 34
years. Despite advances in conventional treatments, the prognosis for most
patients with gliomas remains poor. For patients with the most common form,
glioblastoma multiforme, the median survival is approximately 1 year, and at
most, only 5% of people will survive 5 years after diagnosis. In patients with
recurrent GBM following conventional therapy, the median overall survival time
is 25-27 weeks (Wong et al 1999 and Huncharek & Muscat 1998). This has led to
an intensive search for alternatives.

Conventional treatment of malignant gliomas is based on surgery (whenever
possible), and on therapies, such as radiotherapy and chemotherapy, which are
toxic to cells, primarily targeting cells that proliferate rapidly. However,
despite advances in conventional treatments, the prognosis for most patients
with gliomas remains poor.
The diminished efficacy seen in many such treatments is not related to the
inability of the cytotoxic compounds to kill tumour cells but is more related
to the sub-optimal dose given to patients in order to avoid systemic toxicity
and to the induced resistance of tumour cells to these therapies.

Based on the ability of hrBMP4 to cause terminal differentiation of BTSCs
without apparent toxicity, our underlying hypothesis is that hrBMP4 can be
safely and effectively delivered into and around malignant glioma tissue by
means of Convection Enhanced Delivery (CED) to deliver BMP4 into brain tissue,
including tumours, allowing convective distribution of high drug concentrations
over large volumes of the target tissue, while avoiding systemic toxicity
(Bobo, et al., 1994) (Lieberman, Laske, Morrison, Bankiewicz, & Oldfield,
1995), making CED perfectly suited for delivery of BMP4 to recurrent GBM. As
detailed above, infusion of hrBMP4 was able to extend hGBM-bearing mice
survival, starting at the dosage of a total of 40 µg of protein delivered over
two weeks. hrBMP4 can therefore severely lessen the growth of GBM tumours in
vivo using technical settings similar to those proposed for human patients.

Primary:
To evaluate the feasibility and safety of intra-tumour and peri-tumour therapy
with GMP hBMP4 in increasing doses in patients with progressive and/or multiple
recurrent GBM, identify DLT, and to determine whether there is a maximum
tolerated dose (MTD)

Secondary:
* To assess the radiological response to GMP hrBMP4 following intra-tumour and
interstitial administration by convection enhanced delivery (CED).
* To assess the systemic exposure to GMP hrBMP4 following intra-tumour l and
interstitial administration by CED.

This multicentre, open-label, dose escalating, Phase I study will enrol
approximately 18 patients with progressive and/or multiple recurrent GBM, who
after failure of standard therapies will receive GMP hrBMP4 via intra-tumour
and interstitial delivery by CED.
5 dose levels will be tested in a 3+3 design; one dose level will be assigned
to a cohorts of 3 patients, if one patients develops an SAE another 3 patients
in this dose level will be treated. If no SAEs occur the next dose level will
be tested. If another SAE will occur this level will be considered as a DLT and
the MTD will be determined as one dose level beneath this level.

The intervention will consist of intratumoral slow delivery by CED of the study
drug. For this the patient will be admitted during 6 days . In a surgical
procedure under general anesthesia 3 catheters will be placed in and around the
tumor and a needle biopsy will be performed, all using neuronavigational
techniques.
After this the study drug will be infused in and around the tumor during 4 days
with micro infusion pumps.

The clinical advantages in the use of hrBMP4 is an alternative approach to
cytotoxic therapy based on inducing stem and progenitor cell differentiation,
causing them to lose their stem and proliferative qualities as well as their
tumour/initiating capacity.

It is believed the risks for patients treated with hrBMP4 in this Phase I study
will be minimal for the following reasons: (1) hrBMP4 will be delivered
directly to the tumour reducing systemic exposure to hrBMP4 and possible
peripheral effects, (2) BMP2, which is a similar ligand as hrBMP4 that binds to
the same receptor and has nearly identical actions on blocking proliferation
of tumour cells (Lee et al, 2008) is FDA approved for peripheral human
clinical use, (3) over expression of hrBMP4 in a rodent produced no changes in
the number of neurons, myelination patterns or gross structural abnormalities
(Gomes, Mehler, & Kessler, 2003), and (4) our published (Piccirillo et al,
2006) and unpublished pre-clinical data have found no toxicity related issues
with infusion of up to 100 *g of hrBMP4 either subcutaneously or intracranially.

Patients will be admitted for 6 days in which a surgical procedure takes place
and the study drug will be infused in and around the tumor by micro infusion
pumps during 4 days. During the surgery 3 catheters will be placed in and
around the tumor and a needle biopsy of the tumor will be taken. The risks of
needle biopsy is well established in literature, it consists of approx 1 % of
causing a bleeding, A number of bleedings, but not all maybe symptomatic
causing neurological deficit, some of which may be permanent. Very rarely a
bleeding may be fatal .
Infusion of studydrug into the brain may cause increased pressure leading to
headache, worsening of existing neurological deficit or seizures. These are all
treatable with corticosteroid medication. Mostly these symptoms are temporary,
ceasing when infusion is ended. Our group was involved in another CED study in
which 60 catheters were placed , without causing bleedings or tissue damage