The role of the tumor microenvironment of pancreatic cancer to predict treatment outcome

Pancreatic cancer is a highly lethal disease. Patients with resectable or
borderline resectable disease may benefit from preoperative radiochemotherapy.
However, only a subset of patients will respond to this potentially toxic and
expensive treatment. Therefore, novel predictive markers are needed to
determine treatment efficacy at an early stage. Preferably, these markers could
be determined non-invasively and provide insight into the biology of pancreatic
cancer.

Pancreatic cancers are heterogeneous tumors. The tumor microenvironment is
often characterized by large amounts of stroma, hypovascularization, and
hypoxia. As these three factors can all contribute to treatment resistance, a
quantitative assessment of these markers may aid in the prediction of response
to preoperative radiochemotherapy. Moreover, these assessments may have
prognostic value. Finally, further insight into the interrelation of these
aspects of the tumor microenvironment can contribute to the evaluation of new
targeted treatment options.

Tumor cellularity and extracellular matrix composition can be assessed
non-invasively in vivo by diffusion weighted magnetic resonance imaging (DWI)
and tumor vascularity can be assessed by dynamic contrast enhanced magnetic
resonance imaging (DCE-MRI). Finally, tumor hypoxia can be evaluated by T2* MRI
and PET-CT, using the 18F-labeled hypoxic marker HX4.

The primary aim of the study is to assess whether DWI, DCE-MRI, T2*, and
18F-HX4-PET/CT predict overall survival in patients with pancreatic cancer
treated with surgery and adjuvant chemotherapy or with neoadjuvant
radiochemotherapy, surgery and adjuvant chemotherapy. Secondary aims of the
study include the assessment of the predictive value of DWI, DCE-MRI, T2*, and
18F-HX4-PET/CT for pathological response to neoadjuvant chemoradiation, the
correlation of DWI, DCE-MRI, T2*, and 18F-HX4-PET/CT with histopathological
assessment of tumor stroma, vascularization, and hypoxia, and the assessment of
the predictive value of these histopathological markers for overall survival.

The target population will be recruited from the the Acdemic Medical Centre
(AMC) and Erasmus MC. First, to assess reproducibility, patients with
pancreatic cancer will undergo MRI twice, once in the AMC and once in the EMC.
Next, 40 consecutive patients that will undergo surgery+adjuvant treatment will
have MRI and 18F-HX4-PET/CT measurements once to assess the value of the
techniques to predict outcome of standard treatment. 40 patients who will
undergo preoperative radiochemotherapy will have MRI and 18F-HX4-PET/CT at
baseline, and 1 week before surgery. We will assess the relative contribution
of each imaging method as well as the integrated use of these methods as
predictive markers for survival and pathological response to treatment. Tumor
tissue from resected patients will be analyzed for markers of tumor
vascularization (CD31, VEGF), hypoxia (HIF1alfa, GLUT1, CA9), and stromal
activation (smooth muscle actin, markers for Hedgehog pathway activity).
Results will be correlated with imaging parameters, as well as patient outcome.

* The patient will not have a direct benefit from the study.
* The administration of Gadolinium involves a very small risk of an acute
allergic reaction.
* The administration of Buscopan involves a risk of an adverse reaction. The
most common adverse reactions are accommodation disorders, tachycardia, vertigo
and a dry mouth. Patients who have a contraindication for IV administration of
Buscopan (mega-colon, ileus, myasthenia gravis, glaucoma, prostate hypertrophy
with urine retention, intestinal stenosis and tachycardia), will undergo the
scans without administration of Buscopan.
* The proposed 18F-HX4 dose is chosen based on the phase 1 study with
18F-HX4.75 In this study no toxicities were observed except for a mild, grade I
headache one day after 18F-HX4 injection, which was considered unlikely to be
related to the injection. In view of previous experiences with 18F-HX4,
conventional PET-CT and other nitroimidazole drugs, we expect no side effects.
Nevertheless, it cannot be excluded that patients will experience an acute
allergic reaction to 18F-HX4. Therefore, all patients will be monitored
carefully during and directly after administration of the labelled 18F-HX4 by
trained caregivers.
* Participation in the study will involve exposure to radiation that is
estimated to be 16.4 mSV for the maximum of two 18F-HX4-PET/CT scans together.
The theoretical chance for radiation-induced cancer induction is 5% per
Sievert. For a radiation exposure of 10 mSv this implies a chance of 1 in 2000.
This number applies to patients of 30 years of age. The risk reduces with ~50%
for a typical oncological population of patients aged 55-70 years. Moreover, a
substantial portion of patients will be treated with radiotherapy. Radiation
exposure due to the 18F-HX4 scans is negligible compared to radiation exposure
because of the clinically delivered radiotherapy.
* Whenever possible, MRI and 18F-HX4-PET/CT scans will be performed on one day.
Nevertheless, this will most probably involve an extra visit to the hospital
for most patients. Also, patients participating in the reproducibility part of
the study and patients treated with chemoradiation will undergo MRI and
18F-HX4-PET/CT scanning twice and, therefore, will have to come to the hospital
twice.