No combate às neoplasias, diferentes técnicas radioterápicas têm surgido, apoiadas em avanços tecnológicos, com o objetivo de otimizar a eliminação das células tumorais produzindo o menor dano a tecidos sadios dos pacientes. Os planejamentos de tratamentos radioterápicos visam o estabelecimento de parâmetros técnicos de irradiação de forma que as doses prescritas nos volumes de tratamento sejam atingidas. Enquanto que a prescrição das doses se baseiam em considerações biológicas de radiosensibilidade dos tecidos, os cálculos físicos do planejamento levam em conta parâmetros dosimétricos associados aos feixes de radiação e às características físicas dos tecidos irradiados. A incorporação de informações de sensibilidade de tecidos aos cálculos radioterápicos pode auxiliar na particularização de tratamentos e no estabelecimento de critérios de comparação e escolha de técnicas radioterápicas, contribuindo para o controle tumoral e sucesso do tratamento. Para tanto, modelos biológicos de resposta celular à radiação ionizante devem ser bem estabelecidos. Este trabalho visou estudar a aplicabilidade do uso de modelos biológicos em cálculos de planejamento radioterápico com objetivo de auxiliar na avaliação de técnicas radioterápicas. A probabilidade de controle tumoral (TCP) foi estudada para duas formulações do modelo linear-quadrático, com e sem consideração de repopulação celular, em função de parâmetros de planejamento, como dose por fração, e de parâmetros radiobiológicos, como a razão ?/?. Além disso, o uso de critérios biológicos para comparação de técnicas radioterápicas foi testado através da simulação de um planejamento de próstata utilizando simulação Monte Carlo com o código PENELOPE. Posteriormente, planejamentos radioterápicos de tumores de próstata de cinco pacientes do Hospital das Clínicas da Faculdade de Medicina de Ribeirão Preto, USP, utilizando-se três técnicas de irradiação diferentes, foram comparados através do critério de probabilidade de controle tumoral. Para tanto, matrizes de dose obtidas do sistema de planejamento radioterápico XiO foram utilizadas para obter as distribuições de TCP e histogramas TCP-volume. Os estudos realizados permitiram concluir que variações em parâmetros radiobiológicos podem influenciar significativamente cálculos de controle tumoral e que a análise de histogramas TCP-volume pode fornecer informações importantes para avaliação de tratamentos radioterápicos. Entretanto, o estabelecimento de fatores quantitativos de comparação através de critérios radiobiológicos passa pelo estabelecimento de protocolos de prescrição clínica baseados nesses critérios. Além disto, valores radiobiológicos sofreram grandes alterações na literatura recentemente e, portanto, a inclusão destes parâmetros nos cálculos de planejamentos requer grandes cuidados. / In the fight against cancer, different irradiation techniques have been developed based on technological advances and aiming to optimize the elimination of tumor cells with the lowest damage to healthy tissues. The radiotherapy planning goal is to establish irradiation technical parameters in order to achieve the prescribed dose distribution over the treatment volumes. While dose prescription is based on radiosensitivity of the irradiated tissues, the physical calculations on treatment planning take into account dosimetric parameters related to the radiation beam and the physical characteristics of the irradiated tissues. To incorporate tissue\'s radiosensitivity into radiotherapy planning calculations can help particularize treatments and establish criteria to compare and elect radiation techniques, contributing to the tumor control and the success of the treatment. Accordingly, biological models of cellular response to radiation have to be well established. This work aimed to study the applicability of using biological models in radiotherapy planning calculations to aid evaluating radiotherapy techniques. Tumor control probability (TCP) was studied for two formulations of the linear-quadratic model, with and without repopulation, as a function of planning parameters, as dose per fraction, and of radiobiological parameters, as the ?/? ratio. Besides, the usage of biological criteria to compare radiotherapy techniques was tested using a prostate planning simulated with Monte Carlo code PENELOPE. Afterwards, prostate plannings for five patients from the Hospital das Clínicas da Faculdadede Medicina de Ribeirão Preto, USP, using three different techniques were compared using the tumor control probability. In that order, dose matrices from the XiO treatment planning system were converted to TCP distributions and TCP-volume histograms. The studies performed allow the conclusions that radiobiological parameters can significantly influence tumor control calculations and that the TCP-volume histograms can provide important information for treatment techniques evaluation. However, the establishment of quantitative comparison parameters using radiobiological criteria demands the establishment of prescription protocols based on these same parameters. Also, the literature recently showed large variations in radiobiological parameters, meaning that the inclusion of those in treatment planning calculations should require a careful endeavor.

Il s’agit d’étudier les effets du TEL du rayonnement, d’un soluté (oxygène, antioxydant, scavengers...) et d’une géométrie de confinement sur les processus physiques et chimiques. Nos études sur la production de radicaux libres en fonction de la pression d’oxygène et du TEL du rayonnement nous ont permis de noter une similitude entre la production des radicaux O2- / HO2 et l’effet oxygène décrit en radiothérapie. Ces résultats confortent l’idée que ce couple de radicaux pourraient jouer un rôle dans l’effet oxygène (radiothérapie) et expliqueraient la disparition de cet effet à haut TEL (hadronthérapie). Nous avons aussi montré que cette similitude persistait en présence de Glutathion (antioxydant majoritaire dans le noyau cellulaire) / -

It has long been known that oxygen has the potential to be toxic to biologic systems and that this toxicity is not due to oxygen itself, but due to the production of oxygen radicals. One of these potentially toxic radicals, superoxide (O₂⁻) can be generated as a result of ionizing radiation, and if not adequately removed can proceed to cause cell damage. Superoxide dismutase (SOD) is one of the key enzymes involved in the defence against oxygen toxicity. SOD activity can be inhibited by diethyldithiocarbamate (DOC), a powerful copper chelator. If inhibition of SOD by DOC increases the lifetime and effectiveness of radiation induced O₂⁻, it follows that the potential exists for DOC to enhance the effect of radiation. DOC is however also a thiol compound, and thus may act as a radioprotector by modifying tissue oxygenation status or by free radical scavenging. This study has concerned itself primarily with the inhibition of superoxide dismutase by diethyldithiocarbamate in order to sensitize tumours to ionizing radiation. The use of DOC as an inhibitor of SOD has however meant that any sensitization resulting from SOD inhibition could be masked by a radioprotective effect by DOC. The inhibition of SOD by DDC was confirmed in a murine rhabdomyosarcoma, and it was shown that this inhibition can be maintained for up to twenty-four hours after DDC administration. It was hypothesised that there was a potential for the radioprotective effect of DDC to be overcome, if the levels of DDC were low enough at the time of irradiation. Indeed, if DDC was removed from the growth medium of B16 mouse melanoma cells in culture prior to irradiation, a significant sensitization was demonstrated. It was shown that DDC could act as both a radiosensitizer and as a radioprotector in the same experiment. The dominant action of DDC was found to be dependent on the time allowed between DDC administration and irradiation. If this time was approximately 4 hours, it was possible to show a radiosensitizing effect by means of a tumour growth delay assay. This time modulation effect of DOC was shown in larger tumours, rather than smaller tumours, which could indicate that tumour oxygenation is an important criterion in determining the response to radiation of DOC treated cells. It was shown that B16 mouse melanoma cells exposed to 43°C after DDC pre-treatment were sensitized to thermal damage. This work suggests that some caution should be exercised when DDC is put forward as either a radiosensitizer or a radioprotector in the clinic, but that DDC may have potential as a thermosensitizer.

Avec la chimiothérapie et la chirurgie, la radiothérapie fait partie intégrante de l'arsenal thérapeutique pour lutter contre le cancer. Afin de potentialiser l’efficacité des rayonnements ionisants, la radiochimiothérapie s’est développée mais en raison des résultats insuffisants de cette stratégie, de nouvelles voies permettant une modulation de la radiosensibilité tumorale sont évaluées. C’est dans ce contexte d’amélioration de l’efficacité de la radiothérapie que s’inscrit ce travail de thèse. Nous avons évalué l’intérêt thérapeutique de l’association d’inhibiteurs de la réparation de l’ADN à la radiothérapie sur un modèle orthotopique de cancer bronchique et sur un modèle orthotopique de cancer de la tête et du cou. En raison de son rôle prépondérant dans la réparation des cassures simple brin, PARP1 a été ciblé dans un premier temps pour éprouver cette stratégie, à l’aide d’un inhibiteur chimique l’Olaparib. Le rationnel consistait à inhiber la réparation de dommages induits par l’irradiation, pouvant ainsi conduire à la mort des cellules tumorales. Les résultats obtenus in vitro ont montré que l’inhibition de PARP1 permettait en effet de potentialiser les effets de la radiothérapie. Cette association thérapeutique a, par la suite, été évaluée in vivo et a montré une très faible radiosensibilisation, limitée par une toxicité induite par cette association. Afin d’augmenter l’efficacité de cette stratégie thérapeutique, un inhibiteur d’ATR (AZD6738), une des protéines majeures de la réponse aux dommages de l’ADN et au stress réplicatif, a été ajouté à la combinaison initiale. Il a en effet été montré que Chk1, la cible principale d’ATR, était activée dans les cellules traitées avec l’Olaparib et/ou irradiées. Nous avons démontré in vitro et in vivo, que l’AZD6738 améliorait l’efficacité de la combinaison irradiation et Olaparib dans nos deux modèles tumoraux, suggérant le potentiel de cette triple combinaison en clinique. Enfin, en raison du rôle de l’irradiation et de PARP1 dans différents processus immunitaires, nous avons étudié de façon préliminaire l’influence de nos différentes combinaisons thérapeutiques sur l’infiltrat immunitaire tumoral. Sachant que l’efficacité de l’association Olaparib/irradiation avait été démontrée dans des modèles tumoraux implantés en sous-cutané, ce travail de thèse montre l’importance et la pertinence de modèles précliniques plus proches de la pathologie humaine, comme les modèles orthotopiques. En effet, il est très probable que les toxicités observées au cours de ce travail soit la conséquence d’une détérioration avancée des muqueuses présentes dans le champ d’irradiation et que celles-ci ne puissent être observées lors d’irradiation localisée de tumeurs implantées en sous-cutané. / With chemotherapy and surgery, radiotherapy is part of anti cancer therapeutic strategy. To increase ionizing radiations effects, radiochemotherapy has emerged, but because of inefficient results, new pharmacological strategies for modulation of radiosensitization has been assessed. My thesis project is part of this context of improvement of radiotherapy efficiency. We have evaluated therapeutic interest of association of DNA repair inhibitors and radiotherapy on lung cancer model and head and neck cancer model. Because of its implication in single strand break repair, PARP1 has been first, targeted to assess this strategy, with the chemical inhibitor Olaparib. The rational was to inhibit radio-induced damages, leading to cellular death. In vitro, we have demonstrated that Olaparib was promising for enhancing radiation efficacy, but has an in vivo limited radiosensitization, plus we observed with this association a toxicity. Non toxic association has been found by decreasing Olaparib dose, but association efficiency has been limited, meaning that Olaparib, in our model, has a restrained therapeutic index.To increase the efficiency of this combination, we have added an ATR inhibitor (AZD6738), one of the key proteins implicated in response to DNA damages and replicative stress. In fact it has been demonstrated, that ATR main target, Chk1, was activated in Olaparib-treated and/or irradiated cells. We have demonstrated in vitro and in vivo, that AZD6738 improved efficiency of Olaparib and radiotherapy combination in both models, suggesting the potential of this triple combination in clinic.Finally, because of effects of PARP1 and radiation on different immune processes, we have preliminary studied, the influence of this different combinations on immune infiltrate.Knowing that efficiency of the association Olaparib and radiotherapy has already been demonstrated in subcutaneous models, this work has shown the importance and relevance of preclinical models, closer to human pathologies, as orthotopic models. In fact, it is likely that toxicities observed during this work, are the consequence of mucous membrane damaging in the field of irradiation, which cannot be observed with localized irradiation of subcutaneous tumors.

La radiochirurgie par Gamma Knife constitue une modalité thérapeutique reconnue de certaines affections cérébrales. Le traitement se base sur l’administration d’un rayonnement focalisé au niveau d’une cible intracrânienne. L’efficacité de ce traitement repose sur la délivrance d’une dose d’irradiation efficace au sein d’un volume-cible associé à la délivrance d’une dose d’irradiation négligeable à l’extérieur de ce même volume-cible. En pratique, la dose d’irradiation administrée à l’intérieur du volume-cible n’est pas distribuée de manière homogène, et la dose d’irradiation reçue par les tissus situés en-dehors du volume-cible n’est pas nécessairement faible. Notre travail est basé sur l’hypothèse que l’imperfection de la distribution de la dose d’irradiation au sein du volume-cible et en-dehors de celui-ci peut être responsable des échecs et des complications rencontrées en radiochirurgie. Dans deux modèles cliniques de traitement radiochirurgical, le schwannome vestibulaire et la névralgie du trijumeau, nous avons montré qu’il existait une relation entre les paramètres de distribution de dose d’irradiation et certains résultats du traitement radiochirurgical par Gamma Knife de ces pathologies. Nous avons développé deux modèles expérimentaux d’irradiation radiochirurgicale de rats, l’un ciblé sur le striatum et l’autre sur le nerf trijumeau, permettant d’analyser les conséquences histologiques des variations de la distribution de dose à l’intérieur du volume-cible ainsi qu’à distance de celui-ci. Nous avons démontré que la réponse radiobiologique des tissus irradiés était fortement dépendante de ce paramètre dosimétrique, et que ce dernier constituait une donnée de la planification chirurgicale aussi importante que la dose de prescription. Nous avons corrélé ces résultats avec certaines observations réalisées dans d’autres indications de traitement radiochirurgical ainsi que dans l’analyse histologique de tumeurs traitées par Gamma Knife. Ces études mettent en évidence le rôle important joué par l’optimalisation de la distribution de la dose d’irradiation dans l’amélioration des résultats cliniques du traitement radiochirurgical. Les valeurs optimales de la distribution de dose dans les différentes indications de traitement radiochirurgical doivent être recherchées, et les différentes méthodes mises à notre disposition lors de la planification dosimétrique pour améliorer la distribution de dose doivent être utilisées avec discernement pour obtenir la dosimétrie radiochirurgicale la plus parfaite possible. / Doctorat en sciences médicales / info:eu-repo/semantics/nonPublished

Thesis (PhD)--Stellenbosch University, 2001. / ENGLISH ABSTRACT: The factors which control radiosensitivity are of vital importance for the understanding of cell inactivation and for cancer therapy. Cell cycle blocks, total induced DNA damage, DNA repair, apoptosis and chromatin structure are likely to playa role in the responses leading to cell death. I have examined aspects of irradiation-induced G2/M blocks in DNA damage and repair. In HT29, L132 and ATs4 cells the total amount of induced DNA damage by isodoses of 4.5 Gy, 5 Gy and 2 Gy was found to be 14 %, 14 % and 12 % respectively. Most of the DNA repair was completed before the G2/M maximum and only 3 % of DNA damage remains to be restored in the G2/M block. The radiosensitivity in eleven cell lines was found to range from SF2 of 0.02 to 0.61. By FADU assay the undamaged DNA at 5 Gy was found to range from 56% to 93%. The initial DNA damage and radiosensitivity were highly correlated (r2=0. 81). After 5 Gy irradiation and 12 hours repair two groups of cell lines emerged. The group 1 cell lines restored undamaged DNA to a level ranging from 94 % to 98 %. The group 2 cell lines restored the undamaged DNA to a level ranging from 77 % to 82 %. No correlation was seen between residual DNA damage remaining after 12 hours repair and radiosensitivity. In CHO-K1 cells chromatin condensation induced by Nocodazole was found to marginally increase the radiosensitivity as shown by the change of the mean inactivation dose (D) from 4.446 to 4.376 Gy. Nocodazole also increased the initial DNA damage, induced by 5 Gy, from 7 % to 13 %. In xrs1 cells these conditions increased the radiosensitivity from D of 1.209 to 0.7836 Gy and the initial DNA damage from 43 % to 57 %. Disruption of chromatin structure with a hypertonic medium was found to increase radiosensitivity in CHO-K1 cells from D of 4.446 to 3.092 Gy and the initial DNA damage from 7 % to 15 %. In xrs1 cells these conditions caused radiosensitivity to decrease from D of 1.209 to 1.609 Gy and the initial DNA damage from 43 % to 36 %. Repair inhibition by Wortmannin increased the radiosensitivity in CHO-K1 from a D of 5.914 Gy in DMSO controls to a D 3.043 Gy. In xrs1 cells repair inhibition had no effect on radiosensitivity. Significant inhibition of repair was seen in CHO-K1 at 2 hours (p<0.0001) and at 20 hours (p=0.0095). No inhibition of repair was seen in xrs1 cells at 2 hours (p=0.6082) or 20 hours (p=0.6069). While DNA repair must be allocated to the post-irradiation period, the G2/M block seen in p53 mutants reaches a maximum only 12 hours post-irradiation when most of the repair is completed. As the G2/M block resolves and cells reenter cycle 28 hours after the G2 maximum it appears that repair processes cannot be the only reason for the G2IM cell cycle arrest. At low doses of irradiation initial DNA damage correlates with radiosensitivity. This suggests that the initial DNA damage is a determinant for radiosensitivity. Repair of DNA double-strand breaks by the non-homologous end joining (NHEJ) mechanism, identified by inhibition with Wortmannin, was shown to influence residual DNA damage and cell survival. Both the initial DNA damage and DNA repair were found to be influenced by chromatin structure. Chromatin structure was modulated by high salt and by Nocodazole, and has heen identified as a parameter which influences radiosensitivity. / AFRIKAANSE OPSOMMING: Die faktore wat betrokke is in die meganisme van stralings-sensitisering is van hoogs belang vir die begrip van sel inaktiveering en kanker terapie. Sel siklus blokke, totale geïnduseerde DNS skade, DNS herstel, apoptose en chromatien struktuur is moontlike rol vertolkers in die sellulêre response wat ly tot seldood. Ek het die aspekte van stralings-geïnduseerde G2/M blokke in DNS skade en DNS herstelondersoek. Die hoeveelheid geïnduseerde DNS skade, deur ooreenstemmende stralings-dosisse, in HT29, L132 en ATs4 selle is 14 %, 14 % en 12 %. Meeste van die DNS herstel is klaar voordat die G2/M maksimum beryk word en net 3 % DNS skade blyoor om herstel te word in die G2/M blok. Die stralings-sensitiwiteit in elf sel lyne varieer tussen 'n SF2 van 0.02 en 0.61. Deur die gebruik van die FADU metode is gevind dat die onbeskadigde DNS na 5 Gy bestraling varieer tussen 56 % en 93 %. Die totale geïnduseerde DNS skade en stralings-sensitiwiteit was hoogs gekorreleer (r2=0.81). Na 5 Gy bestraling en 12 ure herstel kan die sel lyne in twee groepe gegroepeer word. Die groep 1 sellyne herstel die onbeskadigde DNS terug na 'n vlak wat varieer tussen 94 % en 98 %. Die groep 2 sel lyne herstel die onbeskadigde DNS terug tot op 'n vlak wat varieer tussen 77 % en 82 %. Geen korrelasie is gesien tussen oorblywende DNS skade en stralings-sensitiwiteit na 12 ure herstel nie. In die CHO-K1 sel lyn, chromatien kompaksie geïnduseer deur Nocodazole, vererger die stralings- sensitiwiteit soos gesien deur die gemiddelde inaktiveerings dosis (D) wat verlaag het van 4.446 tot 4.376. Nocodazole het ook die totale DNS skade verhoog van 7 % tot 13 %. Onder dieselfde kondisies, in die xrs1 sel lyn, is 'n verergering van stralings-sensitiwiteit (D) gesien van 1.209 tot 0.7836 en verhoog ONS skade van 43 % tot 57 %. Die ontwrigting van die chromatien struktuur deur die gebruik van hipertoniese medium het die stralings-sensitiwiteit (D) vererger in CHO-K1 selle van 4.446 tot 3.092. Die totale ONS skade is verhoog van 7 % tot 15 %. Onder dieselfde kondisies, in die xrs1 sellyn, verbeter die stralings-sensitiwiteit (D) van 1.209 tot 1.609 en die totale ONS skade verminder van 43 % tot 36 %. ONS herstel inaktiveering in die teenwoordigheid van Wortmannin het die stralings-sensitiwiteit (D) in CHO-K1 selle vererger van 5.914 in DMSO verwysings kondisies tot 3.043. Die ONS herstel inaktiveering in xrs1 selle het geen uitwerking gehaat op stralingssensitiwiteit nie. Noemenswaardige inaktiveering van ONS herstel is gesien in CHO-K1 selle na 2 ure (p<0.0001) en na 20 ure (p=0.0095). Geen inaktiveering is gesien in xrs1 selle na 2 ure (p=0.6082) of na 20 ure (p=0.6069) nie. TerwylONS herstel moet plaasvind na die bestralings periode, beryk die G2/M blok in p53 gemuteerde selle sy maksimum 12 ure na bestraling terwyl meeste van die ONS herstel alreeds voltooi is. Aangesien die G2/M blok eers 28 ure later begin sirkuleer moet die G2/M blok nog 'n funksie vervul anders as ONS herstel. By lae dosisse van bestraling korreleer die totale geïnduseerde ONS skade met stralings-sensitiwiteit. Dit dui daarop dat die totale ONS skade 'n bepalende faktor moet wees in stralings-sensitiwiteit. Die herstel van ONS skade deur die nie-homoloë eindpunt samevoeging (NHES) meganisme, geïdentifiseer deur inaktiveering deur Wortmann in, het 'n invloed op oorblywende ONS skade en sellulêre oorlewing. Beide die totale ONS skade en ONS herstel was beïnvloed deur die chromatien struktuur. Chromatien struktuur was gemoduleer deur hoë sout konsentrasies en deur Nocodazole, en is geïdentifiseer as a belangrike parameter wat stralings-sensitiwiteit beïnvloed.

No description available.

Ionizing radiation (IR) triggers many signaling pathways stemming from DNA damage, and, independently, from extra-nuclear events. To generate radio-mimetic DNA double-strand breaks (DSBs) without and minimizing the effects on extra-nuclear radiation targets, human (p53+) glioma and carcinoma cells containing bromodeoxyuridine (BrdU)- substituted DNA were treated with Hoechst 33258 followed by long wave-length UV (UV-A) (BrdU photolysis). BrdU photolysis resulted in well-controlled, dose-dependent generation of DSBs equivalent to 0.2 - 20 Gy of IR, as detected by pulse-field gel electrophoresis, accompanied by dose-dependent H2AX phosphorylation at ser-139 and ATM phosphorylation at ser-1981, indicating ATM activation. Furthermore, BrdU photolysis increased phosphorylation of Chk2 (at thr-68) and p53 (at ser-15). p53 phosphorylation was reduced by the ATM inhibitor caffeine, and H2AX phosphorylation was greatly reduced in AT cells, confirming that phosphorylation was primarily ATM-dependent. We also examined the effects of BrdU photolysis on the major cellular signaling ERK pathways. Interestingly, low-dose (≤ 2 Gy-equivalents) BrdU photolysis stimulated ERK1/2 phosphorylation whereas higher doses (≥ 5 Gy eq.) resulted in Em1/2 dephosphorylation. ERK1/2 phosphorylation was ATM-dependent, whereas dephosphorylation was ATM-independent and DSBs dose-dependent. Thus ERK1/2 appear to be both positively and negatively regulated by ATM depending on the severity of the insult to DNA. In summary, few DSBs trigger ATM-dependent ERK1/2 pro-survival signals whereas more DSBs result in ERK1/2 dephosphorylation consistent with a switch from pro-survival to anti-survival signaling that might affect DSBs repair.

Usefulness of a particular clinical assay is directly correlated with its ability to extract highest possible signal from available data. This is particularly relevant for personalized radiation therapy since early plan modifications confer greater benefits to treatment outcome. Recent studies have demonstrated capability of single-cell Raman microscopy to detect cellular radiation response at clinical (below 10Gy) doses, but only in certain strongly responding cell lines and after at least two day incubation. One possible cause is rather unoptimized signal processing used. This work investigates application of several advanced multivariate methods - weighted principal component analysis (WPCA), robust PCA, probabilistic PCA, and nonlinear PCA to increase radiation response signal. Representative datasets from strongly (H460 - human lung) and weakly (LNCaP - human prostate) responding cell lines were analysed in 0-50Gy and 0-10Gy dose ranges and results quantified to determine relative and absolute algorithm performance. It was found that with careful tuning, significant improvements in sensitivity and better signal separation could be achieved as compared to conventional PCA. / Graduate

Les travaux effectués au cours de cette thèse ont été réalisés dans le cadre du développement d'une plateforme d'irradiation protons dédiée aux études en radiobiologie (Radiograaff) à partir de l'accélérateur Van de Graaff 4 MV de l'Institut de Physique Nucléaire de Lyon. Ceci, grâce à une collaboration étroite et interdisciplinaire entre les physiciens et ingénieurs de l'Institut de Physique Nucléaire de Lyon (IPNL) et des biologistes du Laboratoire de Radiobiologie Cellulaire et Moléculaire (LRCM) du CHU de Lyon Sud. Afin de présenter le développement de la plateforme Radiograaff, ce manuscrit se découpera en trois chapitres. Le premier chapitre présente les fondements de l'hadronthérapie et le contexte scientifique du projet. Nous traiterons, dans une première partie, de l'interaction ions-matière vivante afin de mieux cerner les concepts théoriques associés à l'hadronthérapie. Et nous présenterons, dans un second temps, les intérêts scientifiques et les objectifs du projet Radiograaff. Le deuxième chapitre concernera les développements instrumentaux de la ligne d'irradiation. Après avoir discuté du mode de délivrance et d'extraction du faisceau ainsi que des méthodes de simulation que nous avons employées pour le dimensionnement théorique la ligne d'irradiation, nous présenterons le système de contrôle de la dosimétrie. Le chapitre 2 sera conclu par les résultats de l'évaluation et de la qualification de la plateforme. Enfin, dans le troisième chapitre, après une brève présentation de quelques aspects de biologie et de radiobiologie, nous présenterons les premiers résultats obtenus concernant l'efficacité biologique relative des protons délivrés par la plateforme Radiograaff. Nous discuterons des protocoles, des conditions expérimentales ainsi que des méthodes d'analyse des résultats / The work done in this thesis has been made in the development of a proton irradiation platform dedicated for radiobiological studies from the 4 MV Van de Graaff accelerator of the Institut de Physique Nucléaire de Lyon (IPNL). These developments have been made through a close interdisciplinary collaboration between physicists and engineers from IPNL and biologists from the Laboratoire de Radiobiologie Cellulaire et Moléculaire (LRCM, CHU Lyon Sud). To present the development of Radiograaff platform, this manuscript will split into three chapters. The first chapter introduces the basics of hadrontherapy and the scientific context of the project. In the first part of this chapter, we present the interactions between ion and matter in order to better understand the theoretical concepts associated with the hadrons. Then, we present the scientific interests and objectives of the Radiograaff project. The second chapter concerns the instrumental developments of the beam line. After discussing the delivery mode and beam extraction in air and simulation methods that we used to design the theoretical line radiation, we present the dosimetric monitoring system. The chapter 2 is concluded by the results of the evaluation and qualification of the platform. Finally, in the third chapter, after a brief presentation of some aspects of biology and radiobiology, we present the first results on the relative biological effectiveness of protons delivered by the Radiograaff platform. We discuss the protocols, experimental conditions and methods for analysing results