A dose prediction model for treatment planning was generated using U-Net architecture. The model was generated for advanced stage cancer patients. The U- Net architecture was created with depth=6 and kernel=6. The model architecture was successful to reduce the input image size (192X192) to feature map (6X6) which helped to extract the low level features. The dose prediction of the model was trained with depth=6, kernel=6, MSE loss, Adam optimizer, 1000 epochs and a batch size of 4. The predicted dose was rescaled for gamma analysis to quantify accuracy of the model. The renormalized predicted dose was quantified using gamma analysis with a 3mm, 3% dose tolerance. The gamma map was generated to visualize the regions where dose distributions failed. The gamma percentage values obtained on the training set were acceptable. The mean and standard deviation values of gamma pass percentage obtained on training dataset were 97.5% and 1.24% respectively, which concluded that training process was successful and was an almost perfect match of true dose and predicted dose. However, gamma pass percentage values obtained on validation set was not a good representation of the true dose. Nevertheless, the validation dataset was able to predict the approximate highest dose region. A gamma analysis with a 5mm, 5% dose tolerance was performed to test the the level of discrepancy between the predicted and true dose in the validation set. This increased the gamma pass percentage compared to the 3mm, 3% analysis to a mean gamma pass percentage of 26.2 ± 7.47%. Although the predicted dose was not of sufficient accuracy for clinical use, there technique studied in this work show promise for further development. / Thesis / Master of Science (MSc)

京都大学 / 新制・課程博士 / 博士(医学) / 甲第24788号 / 医博第4980号 / 新制||医||1066(附属図書館) / 京都大学大学院医学研究科医学専攻 / (主査)教授 伊達 洋至, 教授 松田 道行, 教授 後藤 慎平 / 学位規則第4条第1項該当 / Doctor of Medical Science / Kyoto University / DFAM

京都大学 / 新制・課程博士 / 博士(医学) / 甲第24502号 / 医博第4944号 / 新制||医||1064(附属図書館) / 京都大学大学院医学研究科医学専攻 / (主査)教授 長船 健二, 教授 朝長 啓造, 教授 伊達 洋至 / 学位規則第4条第1項該当 / Doctor of Medical Science / Kyoto University / DFAM

<p>  </p> <p>Non-small cell lung cancer (NSCLC), the most common type of lung cancer, is essentially the leading cause of cancer related deaths in United States. Only 24% of NSCLC patients survive 5-years post diagnosis, largely attributed to the lack of efficient treatment strategies at the metastatic stage. Thus, understanding the biological mechanisms that promote NSCLC metastasis is critical for the development of effective cancer-specific therapeutic agents. The development of cancer metastasis is greatly driven and influenced by intercellular communication. Key mediators of cell-to-cell communication are extracellular vesicles (EVs). For the past years, the study of EVs released by tumor cells have gained attention, given their impact in modulating the tumor immunity, supporting tumorigenesis, and contributing to the development of metastasis. However, the mechanisms though which tumor EVs contribute to tumor development are still understudied. In this study, we isolated and characterized small EVs, also referred as exosomes, from NSCLC cell lines (H358, Calu6, H460, SKMES-1) and investigated their release, uptake, and impact in non-tumorigenic lung epithelial cells recipient cells (BEAS-2B and HBEC). Our results demonstrated that EVs from NSCLC can induce migration and invasion of non-tumorigenic epithelial cells, and impair epithelial barrier permeability, suggesting their role in supporting tumorigenesis and metastasis. Furthermore, we assessed the immunomodulatory effects of NSCLC EVs on anti-tumor immune cells, particularly T cells. Our findings revealed a suppressive effect of EVs derived from mutant KRASG12C NSCLC (H358) on T-cell proliferation, suggesting their contribution to immune evasion mechanisms in mutant KRAS tumors. To dissect the underlying mechanisms, we employed a dual approach utilizing genetic manipulation (shRNA knockdown) and a small molecule inhibitor (ARS-1620) targeting the oncogenic KRASG12C. Our data demonstrated that targeting KRASG12C impaired the EV-driven cancer phenotypes, highlighting the pivotal role of KRAS oncogenic signaling in tumorigenesis and immune suppression mediated by EVs. Overall, our study sheds light on the crucial involvement of tumor derived EVs in NSCLC progression, both in terms of promoting cellular migration and invasion, as well as dampening anti-tumor immune responses. By elucidating the mechanisms underlying EV-driven tumorigenesis and highlighting the therapeutic potential of targeting KRAS signaling, our findings pave the way for the development of novel and effective therapeutic agents for NSCLC.</p>

Critical care patients are often connected to ventilators, to support or replace their breathing. The ventilators deliver a mixture of gas to the patient by applying a specific volume or pressure, and then the patient exhales passively. This thesis is based of the hypothesis that a slower reduction of the expiration pressure could benefit intensive care patients connected to a ventilator. To enable research within the area, a device which can control the expiration is needed. In this thesis project, an expiration valve was controlled to create different pressure patterns during expiration. To facilitate the research and the usage of the expiration control, an application software was created with the purpose to simulate relevant pressure, flow and volume curves. The prototype is an expiration cassette created for the ventilator Servo-i by Maquet Getinge Group. To enable flexibility, the prototype is external and no information is transmitted from or to the ventilator. The prototype has its own flow and pressure sensors. The different pressure patterns which the prototype uses are designed as a linear decrease and as if a constant resistance was added to the system. The user can also create their own pressure pattern, by deciding 20 pressure points in the duration of two seconds. The simulation application was designed with the ability to simulate the same pressure patterns available with the prototype. By using a lung model, it is possible to simulate the ideal pressure, flow and volume in the lungs which can be expected from the chosen expiration control. During the implementation, two different types of lung models were evaluated in order to determine the specificity required. The prototype was tested with settings which were chosen to challenge the performance of the control. Some problematic areas were detected, such as high pressures or large volumes. However, the prototype was judged to perform well enough to be used in animal trials. The lung model used for the simulation application was a simple model of the lung, consisting of a resistor and a capacitor in series. The simulations were compared with the real system with the purpose to get an indication on the difference between theory and reality. The application presents the expected behavior when using the expiration control. However, it should be kept in mind by the user that the application represents a theoretical model.

To clarify the role of charge in protein movement across the pulmonary endothelial barrier, we simultaneously measured the permeability-surface area product (PS) for native [isoelectric point (pI) 4.4-5.1] and cationic (pI 7.2-8.0) albumin in isolated rabbit lungs perfused with and without protamine sulfate. We focused our measurement on the initial (endothelial) barrier by using a technique that is based on the very rapid (3 min) uptake of tracer. This allowed us to distinguish the charge properties of the endothelium separate from other barriers. In control studies, PS was greater for cationic than for native albumin (8.67 ± 0.93 vs. 2.55 ± 0.20 x 10-2 ml · min-1 · g dry lung-1). In the presence of 1 mg/ml protamine sulfate, cationic albumin permeability was not different from control (7.34 ± 0.49 x 10-2 ml · min-1 · g dry lung-1), whereas PS for anionic albumin increased to 8.82 ± 1.32 x 10-2 ml · min-1 · g dry lung-1. Thus the protamine sulfate eliminated the difference between native and cationic albumin PS. This selective increase in anionic albumin permeability is presumably due to the cation, protamine sulfate, binding to the anionic charges on the endothelium and reducing the anionic charge-charge repulsion. If protamine sulfate had produced a general endothelial injury, the PS for both albumins would have increased. Our results suggest that the normal pulmonary endothelium is an anionic charge barrier restricting the transcapillary movement of negatively charged molecules.

No description available.

No description available.

Receptors for advanced glycation end-products (RAGE) are multi-ligand cell surface receptors highly expressed in the lung that modulate pulmonary inflammation during disease. However, the contributions of RAGE signaling are unknown during pulmonary organogenesis. In order to test the hypothesis that RAGE misexpression adversely affects lung morphogenesis, conditional transgenic mice were generated that over-express RAGE in alveolar type II cells of the lung. When RAGE is over-expressed throughout embryogenesis, severe lung hypoplasia ensues, culminating in perinatal lethality. Flow cytometry and immunohistochemistry employing cell-specific markers for various distal cell types demonstrated anomalies in key epithelial cell populations resulting from RAGE up-regulation through embryonic (E) 18.5. Electron microscopy also identified significant morphological disturbances to distal cell types including separation from the basement membrane. Possible mechanisms leading to the disappearance of pulmonary tissue by increased RAGE expression were then evaluated. A time course of lung organogenesis commencing at E12.5 demonstrated that increased RAGE expression primarily alters lung morphogenesis beginning at E16.5. TUNEL immunohistochemistry and immunoblotting for active caspase-3 confirm a shift toward apoptosis in lungs from RAGE over-expressing mice when compared to wild type controls. Assaying for NF-κB also revealed elevated nuclear translocation in lungs from transgenic mice compared to controls. An RT-PCR assessment of genes regulated by NF-κB demonstrated elevated expression of Fas ligand, suggesting increased activity of the Fas-mediated signal transduction pathway in which ligand-receptor interaction triggers cell death. These data provide evidence that RAGE expression must be tightly regulated during organogenesis. Furthermore, additional elucidation of RAGE signaling potentially involved in branching morphogenesis and cell cycle abnormalities may provide insight into the progression of RAGE-mediated lung diseases.

Receptors for advanced glycation end-products (RAGE) are multi-ligand cell surface receptors of the immunoglobin family expressed by epithelium and macrophages. RAGE expression increases following ligand binding and when diverse cells are exposed to a variety of insults including cigarette smoke extract (CSE). The current research sought to characterize the pro-inflammatory contributions of RAGE expressed by alveolar macrophages (AMs) following CSE exposure. Acute exposure of mice to CSE via nasal instillation revealed diminished bronchoalveolar lavage (BAL) cellularity and fewer AMs in RAGE null mice compared to controls. Primary AMs were obtained from BAL, exposed to CSE in vitro, and RNA, DNA, and protein were analyzed. CSE significantly increased RAGE expression by wild type AMs. Employing ELISAs, wild type AMs exposed to CSE had increased levels of active Ras, a small GTPase that perpetuates pro-inflammatory signaling. Conversely, RAGE null AMs had less Ras activation compared to wild type AMs after exposure to CSE. In RAGE null AMs, assessment of p38 MAPK and NF-κB, important intracellular signaling intermediates induced during an inflammatory response, revealed CSE-induced inflammation occurs at least in part via RAGE signaling. For example, activated p38 was diminished in RAGE null AMs compared to controls and assessment of phosphorylated NF-κB in CSE exposed RAGE null AMs suggest lessened nuclear translocation of NF-κB compared to wild type AMs exposed to CSE. Importantly, quantitative RT-PCR revealed that mRNA expression of pro-inflammatory cytokines including TNF-α and IL-1β were detectably decreased and analysis of secreted proteins by ELISA displayed diminished IL-1β in RAGE null AMs exposed to CSE compared to CSE-exposed wild type AMs. These results reveal that primary AMs orchestrate CSE-induced inflammation, at least in part, via RAGE-mediated mechanisms.