Imaging the effect of HSP90 inhibitor AUY922 on VEGF by means of 89Zr-bevacizumab PET. A side study to the phase I-II study with AUY922 in adult patients with advanced solid malignancies, or either HER2 or ER positive locally advanced or metastatic breast cancer: protocol CAUY922A2101.

Angiogenesis, the formation of new blood vessels is a critical factor involved
in the development and growth of tumors. New vasculature supplies the tumor
with nutrients and oxygen, regulates disposal of metabolic waste products and
provides route for metastatic spreading. An important factor involved in
angiogenesis is vascular endothelial growth factor (VEGF). The VEGF production
by tumor cells is thought to be regulated by hypoxemia, cytokines and cell
differentiation. Over-expression of VEGF leading to angiogenesis, occurs in
many human tumor types, including breast cancer. Therefore, targeting
angiogenesis is a rational approach in many cancer types. Inhibition of Heat
Shock Protein (HSP) 90, is one way of achieving this. HSP90 is a molecular
chaperone, involved in maintaining the conformation, stability, cellular
localization and activity of several key oncogenic client proteins. It plays a
central role in the basic power of cancer cells to adapt to various forms of
stress. Client proteins of HSP90 include the regulator of VEGF expression
hypoxia-inducible factor 1α (HIF-1α), HER2, hormone receptors, AKT, mutant p53,
and so forth. HSP90 is constitutively expressed to 2- to 10-fold higher levels
in cancer cells compared to their normal counterparts. Furthermore, HSP90 in
tumor cells is present in active multi-chaperone complexes, conferring relative
sensitivity to treatment with HSP90 inhibitors. Targeting multiple survival
pathways by means of HSP90 inhibition may contribute to circumvention of
resistance in cancer cells, to chemotherapeutics but also to trastuzumab and
hormonal therapy. This has already been shown in a recent study, in which HSP90
inhibitor 17-allylamino-17-demethoxy-geldanamycin (17-AAG) was combined with
trastuzumab in trastuzumab refractory patients. Tumor regression was reported
in 4 of 25 of these heavily pre-treated patients.
HSP90 inhibition reduces angiogenesis by HIF-1α inhibition and the consequent
reduction of VEGF secretion. Decreased capillary density and vessel
permeability was seen in xenograft models following HSP90 inhibitor AUY922.
Moreover, a recent study demonstrated that the inhibition of HIF-1α by means of
HSP90 inhibitor 17-AAG, reduced VEGF secretion by 90% in a mouse model. These
results demonstrate the potential to use local VEGF levels in the
micro-environment of the tumor as a surrogate marker for early anti-angiogenic
response on HSP90 treatment. Recently, measurement of these levels of VEGF in
vivo was made possible in our institution. Bevacizumab, radiolabeled with
89Zirconium (Zr) or 111Indium (In), was used for VEGF visualization and
quantification in a xenograft mouse model. Ex vivo biodistribution evaluation
of the tracer showed tumor specific uptake, which could be measured
quantitatively non-invasively. In a xenograft mouse model with a human ovarian
A2780 tumor (which has a naturally high excretion of VEGF), we performed
89Zr-bevacizumab imaging before and after intraperitoneal treatment with the
HSP90 inhibitor AUY922. In vivo VEGF imaging demonstrated an impressive
decrease (~70%) of VEGF levels in the tumor, as evaluated with 89Zr-bevacizumab
uptake. This demonstrates that VEGF is a rational read-out for HSP90 inhibition
effect, and this response can be measured in vivo, by means of 89Zr-bevacizumab
imaging.
In conclusion, HSP90 inhibition is a new, promising treatment modality for
cancer patients, particularly in the setting of resistance. A reliable read out
system (biomarker) for the evaluation of early treatment effect is of great
importance in the development of this treatment modality, and could contribute
to customize this treatment for individual patients. So far, no reliable
biomarker has been described for HSP90 effect. Visualizing the effect of HSP90
on VEGF secretion in vivo in the patient, by whole body 89Zr-bevacizumab
uptake, can be of great importance in this respect, and may contribute to
tailored made cancer treatment.

To evaluate the effect of HSP90 inhibition by AUY922 on VEGF by means of
89Zr-bevacizumab PET.
Primary endpoint: measurement of decreased VEGF compared to baseline. A decline
is defined as a decrease of at least 30% in mean Standardized Uptake Value
(SUV) in a maximum of three lesions.

This feasibility study is designed as a side study to the multicenter,
international phase I-II trial with HSP90 inhibitor AUY922 (protocol
CAUY922A2101), as part of the biomarker assessment. In protocol CAUY922A2101,
section 4, the design of this phase I-II trial is described (p37, 38). Briefly,
a dose-escalation study is performed according to phase I design in adult
patients with advanced solid malignancies. This part is followed by a
dose-expansion study according to a phase II design. In the latter part, breast
cancer patients are enrolled that are either refractory to hormone- or
trastuzumab treatment (both treatment arms, n=40 patients), on the maximal
tolerated dose of AUY922 based on the phase I part of the study. Patients with
ER positive, hormone therapy refractory breast cancer, will receive a
89Zr-bevacizumab PET scan as part of the present side study protocol.
To this end, a 89Zr-bevacizumab PET scan will be performed before (baseline)
and during treatment with HSP90 inhibitor AUY922.
A minimum of six and a maximum of 11 patients will be entered to evaluate
whether the effect of HSP90 inhibition by AUY922 can be detected with a
89Zr-bevacizumab PET scan (see statistical paragraph page 7).

In the present study, radioactive bevacizumab is used for PET scanning. The use
of such a tracer means exposure to ionizing radiation. Twice, an infusion of
radio active bevacizumab is administered: once before start of treatment and
once during treatment with AUY922. The total additional radiation dose for the
patient is 18 mSv at baseline, and 18 mSv at cyclus 1 (ICRP62, category III;
comparable to 1,5 times a CT scan).
The tracer for this study is administered intravenously, which means an
intravenous puncture. This puncture will be combined as much as possible with
the punctures that are requested in the setting of the treatment with HSP90
inhibitor AUY922.