Expanded Project Abstract:

Intro: Tissue oxygen concentration plays a critical role in determining how tumors will respond to
treatment with radiation. Tumor hypoxia is correlated to malignant characteristics and resistance to
radiation therapy. Information about oxygen concentrations at the tumor site prior to, during, and after
a course of radiation therapy has been shown to be of prognostic value in preclinical and clinical trials.
However, because the conventional technique for oxygen measurement requires use of an invasive
instrument, which penetrates the tissue with a needle causing inflammation, local hemorrhage, and
discomfort, its clinical usefulness is limited in obtaining repeated in vivo measurements of tissue oxygen
concentration over a course of radiation therapy. We are exploring the use of a novel, non-invasive
optical spectroscopy device to measure tissue oxygenation for veterinary patients undergoing
radiotherapy.

Hypothesis: We hypothesized that tumor oxygenation would differ 1) from normal tissue, 2) pre- to
post-radiation exposure, and 3) over the course of the radiotherapy protocol. Three to five
measurements were taken before and after each fraction of radiotherapy at the tumor site and a
corresponding normal tissue site. Tumors undergo a reoxygenation phenomenon where hypoxic cells
which survive a fraction of radiation therapy become oxygenated – and therefore more susceptible to
ionizing radiation – before the next fraction is delivered. If certain tumor types demonstrate patterns of
change in oxygenation, treatment protocols could be designed to treat the tumors when they are most
susceptible to radiation.

Methods: Repeated non-invasive measurements were made possible using a novel optical spectroscopic
device (Zenascope PC-1) able to measure hemoglobin concentration and hemoglobin oxygen saturation
levels in real time using a probe placed directly on the surface of the skin overlying the tumor. The
Zenalux Biological Tissue Spectrometer (Zenascope™) is an ultraviolet spectrometer that achieves
quantitative optical spectroscopy in turbid media and enables a physician to non-destructively evaluate
tissue composition in real time. The system is a specialized, real-time, diagnostic device that shines
white light on opaque target media and then measures and analyzes the reflected signal. The device
uses standard spectroscopic measurement hardware, proprietary software and patented algorithms to
achieve rapid, quantitative and non-destructive analysis of important targeted endpoints. This novel
approach enables a host of new applications for visible spectroscopy in non-ideal, scattering conditions.
Fifteen patients were measured at the tumor site and a corresponding normal tissue site directly before
and after each fraction of radiation they received. Total hemoglobin and hemoglobin oxygen saturation
were directly measured and used to calculate volume averaged tissue oxygen concentration which was
then plotted over time. The zenascope is looking at a volume averaged optical measurement that
includes arterial, venous, and capillary blood, as well as the surrounding tissue. If you multiply the total
hemoglobin concentration by the saturation, you would get the blood oxygen concentration in uM/L,
that is bound to hemoglobin and volume averaged over the sensing volume.

Results: Preliminary results reveal significant differences between normal and tumor tissue oxygenation
in soft tissue sarcomas, squamous cell carcinomas and oral melanomas, and no difference in
osteosarcomas. Soft tissue sarcomas had significantly higher oxygenation at the tumor site compared to
normal tissue, while squamous cell carcinomas had significantly lower oxygenation at the tumor site
compared to normal tissue. Subjectively, most tumors did not undergo consistent change in oxygen
content or hemoglobin saturation over course of treatment. A possible explanation for differences
between tumor and normal tissue is differences in vascularity at the tumor site leading to changes in Hb
concentrations and oxygen concentrations. The lack of significant difference between oxygen levels at
the tumor and normal sites in osteosarcomas invites further investigation, but may mean the zenascope
is not as useful as a diagnostic tool in these tumor types.

Conclusions: The zenascope provides an easy, cheap, repeatable, way to investigate an additional
biomarker of tumor response during radiotherapy. For example, researchers may gain more insight into
why radiotherapy failed a specific patient, and might be able to select relevant patients for studies
hypoxic tumors. Access to tissue oxygen concentration data may allow clinicians to better tailor
radiation doses based on the correlation between hypoxia and tumor responsiveness, providing for
better outcomes for patients with resistant, hypoxic tumors and the ability to minimize the dose
delivered to patients with less hypoxic tumors. Although this was short-term study, if we can follow
patients through future recheck appointments, oxygenation data could be compared to tumor response.
This information could be used to help design future studies using the Zenascope to further explore the
relationship between tissue oxygenation and radiation response/dosing and therapeutic value to
clinicians prescribing radiation for their patients.

Information from the Zenascope manufacturer:

The Science Behind Optical Tissue Spectroscopy:

In diffuse optical spectroscopy, wavelengths of interest span the UV-NIR spectral range – from the
ultraviolet (UV) at ~300 nanometers through to the beginning of near-infrared (NIR) at ~650 nanometers
– a region which is sensitive to the optical absorption and scattering of soft tissues. The shape and
magnitude of the absorption depends on the concentration of the dominant tissue absorbers as well as
their extinction coefficient (an inherent measure of a constituent’s ability to absorb light energy). In
biological tissue, absorbers of interest include oxygenated hemoglobin (HbO2) and deoxygenated
hemoglobin (dHb), beta-carotene, melanin, and proteins in the UV-NIR spectrum. Since diffuse
reflectance spectroscopy can measure both HbO2 and dHb, one can estimate both the total blood
concentration (THb = HbO2 + dHb) and the percent oxygenation saturation (SO2 = 100xHbO2/THb).
Furthermore, the optical scattering coefficient is known to be sensitive to the spatial architecture and
organization of the tissue and therefore can be used as a means to track changes in cellular morphology
and density, in particular proliferation or necrosis. Once measured, the diffuse reflectance must be
processed through rigorous computational models to obtain quantitative information about the
absorption and scattering properties of the tissue. The Zenalux algorithm uses a fast, Monte Carlo
approach that has been developed to extract quantitative absolute optical properties from diffuse
reflectance spectra by employing scaling and similarity relationships that accelerate the modeling. In
short, the Zenalux algorithm quickly compares the measured reflectance spectra to spectra generated
using the Monte Carlo model; when the modeled and experimental reflectance spectra match, the
underlying optical properties of the medium are determined. Once absorption (µa) is determined,
concentration of the absorber can also be determined through the Beer-Lambert law. This forms the
very basis of quantitative optical tissue analysis using the Zenalux Zenascope. An extensive series of
studies have been conducted to test the validity of our model in quantitatively extracting hemoglobin
oxygen saturations, total hemoglobin concentrations, as well as the optical scattering coefficients. In
these studies, diffuse reflectance spectra were obtained using well-controlled, tissue-simulating
phantoms. The performance of the algorithm was tested by measuring the diffuse reflectance data from
liquid tissue phantoms that were prepared by diluting stock solutions of hemoglobin (as the
chromophore absorber) and polystyrene microspheres (to simulate mismatches in refractive index that
cause scattering in tissue) to a final fixed volume. Knowledge of the stock concentrations of hemoglobin
and scatterer solutions allowed calculation of the expected optical properties of absorption and
scattering in the final phantom. Hemoglobin desaturation was achieved by adding yeast (which consume
oxygen) and measuring the partial pressure of dissolved oxygen using an oxygen electrode. The
measured diffuse reflectance was analyzed using our inverse Monte Carlo model in order to extract
optical absorption and scattering from the phantoms.