Assessment of tumor hypoxia in laryngeal cancer with 18F-fluoroazomycin arabinoside (18F-FAZA)-PET/CT scanning, oxygen-enhanced MRI (OE-MRI) and immunohistochemistry

Tumor hypoxia is considered to be important in determining the prognosis of
patients with head and neck cancer (1). It has been shown that patients with
hypoxic tumors have a higher risk of local recurrences and distant metastases.
Moreover, hypoxic tumors are more resistant to radiation- and chemotherapy (for
review see ref. 2, 3). Therefore, a pretreatment evaluation of tumor
oxygenation status could be useful to select patients which should be treated
differently, e.g. by hypoxic modification. In vivo measurement of tumor
oxygen-saturation can be achieved by either the oxygen electrode method
(Eppendorf electrode) or by imaging techniques. Although the Eppendorf method
is considered gold standard for the detection of tumor hypoxia, its validity
can be questioned, since the spatial heterogeneity always remains a problem
(4). The improvement of nuclear and magnetic imaging technology may solve this
problem.
Recently, animal research and clinical studies proved that positron emission
tomography (PET) may have an important role in visualizing hypoxia in vivo.
Several hypoxia tracers have been tested, including 123I-iodoazomycin
arabinoside (123I-IAZA), 18F-fluoromisonidazole (18F-FMISO), 18F-fluoroazomycin
arabinoside (18F-FAZA), 99Tcm-labelled dioximes (99Tcm-HL91) and 60Cu-labelled
methylthiosemicarbazone (60Cu-ATSM) (for review see ref 5,6 and 7). Among these
tracers, 18F-FMISO has been used most frequently to visualize tumor hypoxia in
head and neck cancer patients. Rajendran et al. (8) found that 18F-FMISO PET
scanning was effective in quantifying regional hypoxia in a series of 73 head
and neck cancer patients. A study by Eschmann et al. (9) concluded also that
18F-FMISO PET has the potential to predict response to radiotherapy. Another
clinical study by the same research group showed the value of correlated
18F-fluorodeoxyglucose (18F-FDG) and 18F-FMISO PET scanning in predicting
treatment response in head and neck cancer patients (10).
The ideal hypoxia tracer [1] is captured specifically by hypoxic cells, [2] is
sufficiently delivered in a microenvironment with altered perfusion, [3]
produces a limited amount of non-specific metabolites and [4] is not present in
the circulation during imaging (11). Although the used hypoxia tracers more or
less meet these requirements, they are far from ideal. The clinical application
of the most frequently applied tracer, the 18F-FMISO is hampered due to the
unfavorable biokinetic properties. The low target-to-background ratio can be
explained by the high lipophilicity that is responsible for the slow
accumulation in hypoxic tissues and slow clearance from normoxic tissues. The
relatively long plasma clearance half-life is responsible for the poor
signal-to-noise ratio in the PET images. 18F-FAZA shows better biokinetics than
18F-FMISO in an animal study, as the tumor-to-background ratio was higher with
faster clearance from blood and non-target tissues (12). Despite of these
favorable properties of 18F-FAZA, there is a very limited number of
publications available about the clinical application of this promising tracer.
A recent publication reports results of 18F-FAZA PET scanning on eleven
untreated head and neck cancer patients (13). The image quality seems to be
feasible for further clinical application of tumor imaging. The biokinetic
parameters (e.g.: tumor-to-muscle ratio) seems to be better than the same
values of other tracers. The authors explained the experienced inhomogeneity of
18F-FAZA uptake by the tumor by a large intratumoral variability of tissue
hypoxia, which seems to be reasonable.
Special magnetic resonance imaging (MRI) techniques are also available to
detect hypoxic regions in vivo. One of these methods is the blood oxygenation
level-dependant magnetic resonance imaging (BOLD-MRI). It benefits the
different paramagnetic properties of oxyhemoglobine (O2Hb) and deoxyhemoglobin
(dHb). In this technique, breathing hyperoxyc gas alters the ratio of O2Hb en
dHb species, modulating tissue effective transverse relaxation rate (R2*) (20).
The largest disadvantage of this method is the unclear relationship between
change in R2* and tumor oxygen physiology, and therefore the imaging result
might be affected not only by the tumor oxygenation status, but also by tumor
blood flow, blood volume and vessel geometry (21). Another interesting method
is the dynamic contract-enhanced MRI (DCE-MRI), which uses sequential MRI at 3
s intervals after i.v. administration of gadolinium using a dynamic scan
sequence (22). The limitations of this technique are the same; it evaluates
more the tumor blood flow, vascular volume and vessel permeability instead of
tumor oxygenation. We would like to study an alternative MRI method, the oxygen
enhanced MRI (OE-MRI), which has been recently described in the literature
(23). Molecular oxygen (O2) is paramagnetic and therefore increases the proton
longitudinal relaxation rate (R1) of water containing dissolved oxygen (24).
The measured change in R1 (= 1/T1, where T1 is the longitudinal relaxation
time) induced by breathing oxygen is proportional to the change in tissue
oxygen concentration, therefore oxygen-induced increase in R1 has the potential
to provide noninvasive measurement of change in tumor oxygen concentration.
The validity of 18F-FAZA PET and OE-MRI scanning in respect of indicating
tissue hypoxia remains an interesting point. Although, in an animal study, a
correlation was found between the 18F-FAZA uptake and the immunohistochemical
detection of hypoxia by pimonidazole binding (14), the value of 18F-FAZA PET in
the detection of tumor hypoxia in humans has not been studied so far. There has
been significant correlation founded between DCE-MRI parameters and
pimonidazole staining in a recent study on patients with head and neck cancer
(25), but the reliability and special accuracy of OE-MRI has not been studied,
yet. There are exogenous and endogenous markers for histological detection of
tissue hypoxia. Exogenous markers are chemicals that accumulate or are
bioreducible under hypoxic conditions. Pimonidazole is one of these exogenous
markers that can be detected by immunohistochemistry. Endogenous markers
(hypoxia-inducible factor 1-* (HIF 1-a), carbonic anhydrase-9 (CA9),
glucose-transported-1 (GLUT-1)) are gene products that are up-regulated under
hypoxic conditions (for review see ref. 3). Therefore, the current study
proposal will focus on the validation of 18F-FAZA PET and OE-MRI with respect
to the accuracy of assessing tumor hypoxia in head and neck cancer patients.

To determine the accuracy of 18F-FAZA-PET/CT and OE-MRI scan in detecting
hypoxic regions within the tumor using histological markers.

A prospective diagnostic study.

The only discomfort for the patients is the extra PET/CT and MRI scanning and
an extra pimonidazole infusion prior to the operation. The risk of the
18F-FAZA-PET/CT scanning means about 7.6 mSv radiation dose, 18F-FAZA has no
known pharmacological risks. Pimonidazole has no registered side effects in
the applied dose (500 mg/m2), whereas in higher doses (above 750 mg/m2)
immediate symptoms of malaise, heat, sweating and disorientation have been
described. OE-MRI includes 16 minutes 100% O2 breathing and the administration
of 0.1 mmol/kg gadodiamid intravenously, which has not got any known
side-effect beside the known very rare (< 0.01%) allergic reaction to the
gadolinium-based contrast.