Diffuse Optical Spectroscopy for the Quantitative Assessment of Acute Ionizing Radiation Induced Skin Toxicity Using a Mouse Model

Acute skin toxicities from ionizing radiation (IR) are a common side effect from therapeutic courses of external beam radiation therapy (RT) and ...

Ontario Cancer Institute / Campbell Family Institute for Cancer Research

Extent of resection is a major goal and prognostic factor in the treatment of gliomas. In this study we evaluate whether quantitative ex vivo tissue measurements of δ-aminolevulinic acid-induced protoporphyrin IX (PpIX) identify regions of increasing malignancy in low- and high-grade gliomas beyond the capabilities of current fluorescence imaging in patients undergoing fluorescence-guided resection (FGR). Surgical specimens were collected from 133 biopsies in 23 patients and processed for ex vivo neuropathological analysis: PpIX fluorimetry to measure PpIX concentrations (C(PpIX)) and Ki-67 immunohistochemistry to assess tissue proliferation. Samples displaying visible levels of fluorescence showed significantly higher levels of C(PpIX) and tissue proliferation. C(PpIX) was strongly correlated with histopathological score (nonparametric) and tissue proliferation (parametric), such that increasing levels of C(PpIX) were identified with regions of increasing malignancy. Furthermore, a large percentage of tumor-positive biopsy sites (∼40%) that were not visibly fluorescent under the operating microscope had levels of C(PpIX) greater than 0.1 µg/mL, which indicates that significant PpIX accumulation exists below the detection threshold of current fluorescence imaging. Although PpIX fluorescence is recognized as a visual biomarker for neurosurgical resection guidance, these data show that it is quantitatively related at the microscopic level to increasing malignancy in both low- and high-grade gliomas. This work suggests a need for improved PpIX fluorescence detection technologies to achieve better sensitivity and quantification of PpIX in tissue during surgery.

δ-aminolevulinic acid-induced protoporphyrin IX concentration correlates with histopathologic markers of malignancy in human gliomas: the need for quantitative fluorescence-guided resection to identify regions of increasing malignancy.

Evaluation of imaging for cancer detection and localization can be achieved by correlation of gold-standard histopathology with imaging data. Usage of a 3D biomechanical-based deformable registration for correlation of the histopathology of whole-tissue specimens with ex vivo imaging necessitates measurement of the distribution of biomechanical properties in the ex vivo tissue specimen and changes that occur during pathology fixation. To measure high-resolution 3D distributions of Young's modulus (E) prefixation and postfixation, a quasi-static magnetic resonance elastography method was developed at 7 T. Use of echo-planar imaging allowed for shorter imaging times, in line with limited time frames allowable for pathology specimens. The finite element modeling algorithm produced voxel-wise E measures, and mechanical indentation was used for comparison. An initial preclinical evaluation with canine prostate specimens (n = 5) demonstrated a consistent increase in E with fixation (P < 0.002) by a factor of 4 (± 1). Increases were a function of distance from the tissue edge and correlated with fixation time (ρ = 1, P < 0.02). The technique will be used to generate population-averaged data of E from clinical ex vivo specimens prefixation and postfixation to inform registration of whole-mount histopathology with in vivo imaging.

Quasi-static magnetic resonance elastography at 7 T to measure the effect of pathology before and after fixation on tissue biomechanical properties.

To propose a novel technique to experimentally validate deformable dose algorithms by measuring 3D dose distributions under the condition of deformation using deformable gel dosimeters produced by a novel gel fabrication method.

A novel technique to enable experimental validation of deformable dose accumulation.

Quantum dot (QD) contrast-enhanced molecular imaging has potential for early cancer detection and image guided treatment, but there is a lack of quantitative image contrast data to determine optimum QD administered doses, affecting the feasibility, risk and cost of such procedures, especially in vivo. Vascular fluorescence contrast-enhanced imaging is performed on nude mice bearing dorsal skinfold window chambers, injected with 4 different QD solutions emitting in the visible and near infrared. Linear relationships are observed among the vascular contrast, injected contrast agent volume, and QD concentration in blood. Due primarily to differential light absorption by blood, the vasculature is optimally visualized when exciting in the 435-480 nm region in 81% of the cases (89 out of 110 regions of interest in 22 window chambers). The threshold dose, defined here as the quantity of injected nanoparticles required to yield a vascular target-to-autofluorescence ratio of 2, varies from 10.6 to 0.15 pmol g(-1) depending on the QD emission wavelength. The wavelength optimization maximum and broadband gain, defined as the ratio of threshold doses estimated for optimal and suboptimal (worst wavelength or broadband) spectral illumination, has average values of 4.5 and 1.9, respectively. This study demonstrates, for the first time, optimized QD imaging in vivo. It also proposes and validates a theoretical framework for QD dose estimation and quantifies the effects of blood absorption, QD emission wavelength, and vessel diameter relative to the threshold dose.

Estimation of minimum doses for optimized quantum dot contrast-enhanced vascular imaging in vivo.

Point quantique (QD) contraste accentu&eacute; imagerie mol&eacute;culaire poss&egrave;de un potentiel de d&eacute;tection pr&eacute;coce du cancer et le traitement de l&#39;image guid&eacute;, mais il y a un manque de donn&eacute;es de contraste d&#39;image quantitative pour d&eacute;terminer QD optimale re&ccedil;u des doses, affectant la faisabilit&eacute;, le risque et le co&ucirc;t de telles proc&eacute;dures, surtout in vivo. Fluorescence vasculaire, imagerie de contraste accentu&eacute; est effectu&eacute;e sur des souris Nudes portant dorsale pli cutan&eacute; fen&ecirc;tre chambers, inject&eacute; avec 4 diff&eacute;rentes solutions QD &eacute;mettant dans le visible et infrarouge &agrave; proximit&eacute;. On observe des relations lin&eacute;aires entre le contraste vasculaire, volume agent de contraste inject&eacute;e et QD concentration dans le sang. Due principalement &agrave; diff&eacute;rentielle absorption de lumi&egrave;re par le sang, le syst&egrave;me vasculaire est optimale visualis&eacute; quand passionnant dans la r&eacute;gion de 435-480 nm dans 81 % des cas (89 sur 110 des r&eacute;gions d&#39;int&eacute;r&ecirc;t dans les 22 chambres de fen&ecirc;tre). La dose seuil, d&eacute;finie ici comme la quantit&eacute; de nanoparticules inject&eacute;s requis pour produire un rapport cible-&agrave;-autofluorescence vasculaire de 2, varie de 10,6 &agrave; 0,15 pmol g(-1) selon la longueur d&#39;onde d&#39;&eacute;mission de QD. La longueur d&#39;onde optimisation maximale et &agrave; large bande gain, d&eacute;fini comme le rapport des doses de seuil estim&eacute; pour optimal et sous-optimal (pires longueur d&#39;onde ou large bande) illumination spectrale, poss&egrave;de des valeurs moyennes de 4,5 et 1,9, respectivement. Cette &eacute;tude montre, pour la premi&egrave;re fois, QD optimis&eacute;e d&#39;imagerie in vivo. Il propose aussi et valide un cadre th&eacute;orique pour l&#39;estimation de dose QD et quantifie les effets de l&#39;absorption de sang, longueur d&#39;onde d&#39;&eacute;mission du QD et le diam&egrave;tre du vaisseau par rapport &agrave; la dose seuil.

Estimation des doses minimales pour optimis&eacute; quantum dot contraste accentu&eacute; imagerie vasculaire in vivo.

Quantum dot (QD) contrasto-enhanced imaging molecolare ha il potenziale per la diagnosi precoce del cancro e trattamento immagine guidata, ma c&#39;&egrave; una mancanza di dati di contrasto immagine quantitativa per determinare l&#39;ottimale QD somministrato dosi, che interessano la fattibilit&agrave;, il rischio e il costo di tali procedure, specialmente in vivo. Fluorescenza vascolare contrasto-enhanced imaging viene eseguita su topi nudi cuscinetto dorsale skinfold finestra chambers, iniettato con 4 diverse soluzioni QD emissione nel visibile e vicino infrarosso. Relazioni lineari sono osservate tra il contrasto vascolare, volume agente di contrasto iniettato e QD concentrazione nel sangue. Dovuta principalmente al differenziale assorbimento della luce da sangue, sistema vascolare viene visualizzato in modo ottimale quando emozionante della regione 435-480 nm nel 81% dei casi (89 di 110 regioni d&#39;interesse di 22 camere di finestra). La dose soglia, qui definita come la quantit&agrave; di nanoparticelle iniettate necessarie per produrre un rapporto target-autofluorescenza vascolare 2, varia da 10.6 a 0,15 pmol g(-1) a seconda della lunghezza d&#39;onda dell&#39;emissione di QD. Il lunghezza d&#39;onda ottimizzazione a banda larga e massimo guadagno, definito come il rapporto di dosi soglia stimato per ottimale e non ottimali (peggiore lunghezza d&#39;onda o a banda larga) illuminazione spettrale, ha valori medi di 4.5 e 1.9, rispettivamente. Questo studio dimostra, per la prima volta, QD ottimizzata di imaging in vivo. Inoltre propone e convalida un quadro teorico per la stima di dose QD e quantifica gli effetti di assorbimento del sangue, QD emissione lunghezza d&#39;onda e diametro del vaso rispetto alla dose soglia.

Stima delle dosi minime per ottimizzato quantum dot enhanced contrasto vascolare imaging in vivo.

Optimization is an important but relatively unexplored aspect of contrast-enhanced fluorescence imaging, since minimizing contrast agent usage reduces the associated cost and potential toxicity. In a previous study, the authors developed a quantitative experimental approach to optimize quantum dot (QD)-based imaging using homogenized liver as a model tissue. In this follow-up study, the authors further extend and validate the approach using eight different tissues and five QDs emission wavelengths, and introduce quantitative imaging performance metrics, namely the threshold QD concentration and wavelength optimization gain. These metrics allow quantification of the improvements through spectral optimization in terms of reduced QD dose and identify the conditions that make the optimization process worthwhile. The authors show that, for most tissues, the most important parameter to optimize is the emission wavelength, yielding improvements of up to four orders of magnitude, followed by the excitation wavelength (up to 20-fold improvement) and the excitation filter bandwidth (up to 50% improvement). The authors also observe, by means of the optimization gain metric, that tissues exhibiting both high autofluorescence and strong pigmentation are generally better candidates for excitation wavelength optimization. This work contributes to the development of robust and quantitative dosimetry for QD-based fluorescence imaging near to the tissue surface.

Effect of tissue optics on wavelength optimization for quantum dot-based surface and subsurface fluorescence imaging.

Effective therapies for malignant gliomas are still elusive and limited survival improvements are provided only by Temozolomide or fluorescence guided resection. The efficacy of photodynamic therapy (PDT) in this indication is limited by the higher sensitivity of normal brain structures compared to glioma necessitating a modulation of its sensitivity. We evaluate the influence of hypothermia and the tyrosine kinase inhibitor Erlotinib on cell's ability to synthesize PPIX following the administration of ALA which was not previously investigated. We demonstrate that both hypothermia and Erlotinib are favorable in PPIX selectivity as only glioma cell lines demonstrate an increased PPIX synthesis, whereas the neuronal and astrocytic synthesis is remaining unaffected. The results are encouraging to consider hypothermia and Erlotinib as adjuvant therapies to increase the PDT therapeutic index between GBM and normal intracranial tissues, as well as to improve contrast in fluorescence guided resection.

Modulation of PPIX synthesis and accumulation in various normal and glioma cell lines by modification of the cellular signaling and temperature.

Monitoring the onset of erythema following external beam radiation therapy has the potential to offer a means of managing skin toxicities via biological targeted agents - prior to full progression. However, current skin toxicity scoring systems are subjective and provide at best a qualitative evaluation. Here, we investigate the potential of diffuse optical spectroscopy (DOS) to provide quantitative metrics for scoring skin toxicity. A DOS fiberoptic reflectance probe was used to collect white light spectra at two probing depths using two short fixed source-collector pairs with optical probing depths sensitive to the skin surface. The acquired spectra were fit to a diffusion theory model of light transport in tissue to extract optical biomarkers (hemoglobin concentration, oxygen saturation, scattering power and slope) from superficial skin layers of nude mice, which were subjected to erythema inducing doses of ionizing radiation. A statistically significant increase in oxygenated hemoglobin (p < 0.0016) was found in the skin post-irradiation - confirming previous reports. More interesting, we observed for the first time that the spectral scattering parameters, A (p = 0.026) and k (p = 0.011), were an indicator of erythema at day 6 and could potentially serve as an early detection optical biomarker of skin toxicity. Our data suggests that reflectance DOS may be employed to provide quantitative assessment of skin toxicities following curative doses of external beam radiation.

Quantitative monitoring of radiation induced skin toxicities in nude mice using optical biomarkers measured from diffuse optical reflectance spectroscopy.

Polarization second harmonic microscopy was used for collagen imaging in human non-small cell lung carcinoma and normal lung tissues ex vivo and revealed significant differences in the nonlinear susceptibility component ratio, demonstrating potential use in cancer diagnosis.

Characterization of collagen in non-small cell lung carcinoma with second harmonic polarization microscopy.

Protoporphyrin IX (PPIX) produced following the administration of exogenous 5d-aminolevulinic acid is clinically approved for photodynamic therapy and fluorescence-guided resection in various jurisdictions around the world. For both applications, quantification of PPIX forms the basis for accurate therapeutic dose calculation and identification of malignant tissues for resection. While it is well established that the PPIX synthesis and accumulation rates are subject to the cell’s biochemical microenvironment, the effect of the physical microenvironment, such as matrix stiffness, has received little attention to date. Here we studied the proliferation rate and PPIX synthesis and accumulation in two glioma cell lines U373 and U118 cultured under five different substrate conditions, including the conventional tissue culture plastic and polyacrylamide gels that simulated tissue stiffness of normal brain (1 kPa) and glioblastoma tumors (12 kPa). We found that the proliferation rate increased with substrate stiffness for both cell lines, but not in a linear fashion. PPIX concentration was significantly higher in cells cultured on tissue-simulating gels than on the much stiffer tissue culture plastic for both cell lines. These findings, albeit preliminary, suggest that the physical microenvironment might be an important determinant of tumor aggressiveness and PPIX synthesis in glioma cells.

Polyacrylamide gel substrates that simulate the mechanical stiffness of normal and malignant neuronal tissues increase protoporphyin IX synthesis in glioma cells.

Carolyn Niu

Ontario Cancer Institute / Campbell Family Institute for Cancer Research

Diffuse Optical Spectroscopy for the Quantitative Assessment of Acute Ionizing Radiation Induced Skin Toxicity Using a Mouse Model