Are you an EPFL student looking for a semester project?
Work with us on data science and visualisation projects, and deploy your project as an app on top of GraphSearch.
Despite the experience gained over several decades in various types of light-based medical treatments, the optimization of the corresponding therapeutic protocols and accurate forecasting of their outcome have not yet been achieved in many cases. The difficulty often arises from the heterogeneity of living tissues, their variable optical properties, and from the heterogeneous distribution of the photoactive or photosensitive substances – whether naturally present in the tissue or artificially adde. Our work focuses on the individualization and control of irradiation parameters, in order for the physician to be able to elicit a predictable clinical response in the irradiated tissues. In this thesis, we present three separate studies in which we tried to evaluate the possibility of individualizing and optimizing the corresponding clinical outcomes by measuring or monitoring certain, particular parameters. In a first clinical study, performed at the medical practice of Dr Vezzola, MD, in Saló, Italy, the human eye's retinal reflectance was measured and mapped, in the framework of subthreshold thermal laser therapy, using an excitation wavelength identical to that of the treatment laser, i.e. at 810 nm. The specific goal of this study was to correlate the occurrence of retinal burns with the measured infrared retinal reflectance. This study was performed using a modified fundus camera to record infrared reflectance images of the retina, and by recording the slit-lamp based laser therapy parameters (irradiation parameters and spot location) in such a way so as to overlay the map of the laser treatment spots on the corresponding reflectance fundus image. The clinical study demonstrated the expected existence of spatial variations in light reflectance at 810 nm (probably due to changes of the tissue absorption), which we then tried to relate to the occurrences of retinal burns observed during the laser treatment. The analysis of the results obtained with the applied conditions, did not however show a clear correlation between the local retina reflectance, the laser beam parameters, and the occurrence of retinal burns. Therefore we postulate that either the absorbing structures of the retina cannot be seen with the imaging device we used, possibly due to its limited resolution, or other important elements play a role in the laser-tissue interactions during this type of lasers-light interaction with the local retinal tissue. The second study was pre-clinical, performed at EPFL, and aimed at monitoring in real-time the tissular oxygen concentration during photodynamic therapy (PDT). It was performed in vivo on the chicken embryo's chorio-allantoic membrane (CAM) model, which was submitted to aminolevulinic acid (ALA)-based PDT. The molecular oxygen, which is thought to be an essential actor in the cascade of reactions leading to the tissular PDT effect, is actually the main molecule responsible for the photosensitizer's (PS) triplet state quenching. Therefore, the delayed fluorescence lifetime of the photosensitizer, protoporphyrin IX (PpIX), was measured with a specially designed and assembled, optical fiber-based, time-resolved spectrofluorometer, and used as a proxy for tissular pO2. Simultaneously, vascular damages caused by PDT were characterized and quantified, to check for correlation between the two parameters. Using the PS's delayed fluorescence lifetime to evaluate tissular pO2 proved to be a quite reasonable strategy, due to the fact that it is possible to measure the pO2 at the location of the PS molecule. The study's results demonstrate a robust, linear correlation between tissular pO2 reduction and vascular damage extent. They also suggest that the amount of oxygen consumed during PDT could be a useful, measurable parameter for assessing and/or controlling the PDT's therapeutic effect. The third, clinical, study aimed at measuring the fluorescence photobleaching of the PpIX photosensitizer due to PDT. This study was performed at the "Hôpitaux Universitaires de Genève – HUG", in Geneva, Switzerland, in collaboration with Dr Denis Salomon, MD, and Dr Behrooz Kasraee, MD, in the framework of a series of standard clinical PDT treatments of aktinic keratoses (AK), which is a pre-cancerous skin lesion. The lesions' fluorescence intensity was quantified with a specially adapted quantitative imaging device, using a homogeneous and constant intensity fluorescence excitation light. The specific goal was to check for a relation between the extent of photobleaching of the PS (in this case PpIX, which was induced by the administration of Metvix®) and the clinical outcome (disappearance of AK) evaluated several months after the treatment. The study's results show that the amount of photobleached PS is strongly and linearly correlated to the fluorescence measured before the treatment. Likewise, the preliminary assessment of the clinical outcomes confirms the existence of a correlation of these outcomes with the PS's fluorescence bleaching, thus making it possible, in principle, to select and optimize the PDT's irradiation parameters before starting the treatment. The results of this study demonstrate the relevance of measuring the PS's photo-bleaching for optimizing PDT and forecasting of its outcome.
Georges Wagnières, Hubert van den Bergh, Filippo Piffaretti