La pathologie pleurale est fréquente dans la pratique médicale. Dans certaines séries, une effusion pleurale est retrouvée chez 10 % des patients hospitalisés en médecine interne. Les épanchements pleuraux peuvent refléter tant une pathologie pleurale princeps que les manifestations dun grand nombre de pathologies pulmonaires et extra-thoraciques. La recherche étiologique de ces pleurésies impose un diagnostic différentiel vaste sous-tendu par des processus physiopathologiques très différents. En particulier, il est de première importance didentifier une pathologie pleurale maligne eu égard à sa prise en charge spécifique et à son pronostic extrêmement réservé. Comme nous lavons rappelé dans une revue de littérature explicitant lexploration pleurale actuellement admise, limagerie conventionnelle ne présente pas de critère spécifique pour le diagnostic différentiel bénin vs malin dune pathologie pleurale. De même, la rentabilité de la thoracocentèse et de la biopsie pleurale percutanée demeure relativement faible et justifie fréquemment le recours à la thoracoscopie médicale. Dans notre travail, nous avons poursuivi trois objectifs. Notre premier objectif a été dévaluer lintérêt de la tomographie à émission de positrons (TEP) utilisant pour traceur le 18F-fluorodéoxyglucose (18FDG) dans le diagnostic étiologique de la pathologie pleurale et, en particulier, dans lidentification des effusions malignes. Dans ce cadre, nous avons aussi évalué la faisabilité dune mesure semi-quatitative de lhyperfixation du 18FDG dans le diagnostic différentiel bénin vs malin dune pleurésie, ainsi que dans la localisation thoracique vs extra-thoracique des pleurésies métastatiques. Nous avons pu établir deux valeurs seuils distinguant ces groupes respectifs. Notre deuxième objectif a été dévaluer la valeur pronostique de lintensité de lhyperfixation pleurale du 18FDG. Nous avons ainsi démontré que la quantification de lhyperfixation du 18FDG au niveau de la pleurésie présentait une valeur pronostique chez des patients présentant une néoplasie pulmonaire non à petites cellules avec pleurésie métastatique. Fort de ces données, nous avons positionné limagerie métabolique dans lapproche diagnostique de la pathologie pleurale et précisé ses indications cliniques. Enfin, notre troisième objectif a été dévaluer la valeur de lanalyse biochimique, en particulier de cytokines ou de facteurs de croissance dans le diagnostic étiologique des épanchements pleuraux. Nous avons étudié lintérêt du dosage pleural de 3 cytokines/facteurs de croissance (Interleukine-6 (IL-6)), Transforming growth factor-β1 (TGF-β1) et Vascular endothial growth factor (VEGF)) dans les pleurésies malignes. Nous avons démontré que, contrairement à lIL-6 et au TGF-β1, seul le VEGF présentait un taux significativement plus élevé dans les pleurésies malignes comparativement aux bénignes. Nous avons montré que le taux de VEGF pleural était corrélé à des marqueurs de lagressivité de la pleurésie (taille de lépanchement, nombre de globules rouges au sein de la pleurésie, taux pleural de glucose et de lactate déshydrogénase (LDH)). Toutefois, malgré ses taux pleuraux accru dans les pleurésies malignes, le dosage du VEGF dans le liquide pleural revêtait une rentabilité diagnostique modeste dans la distinction entre une pleurésie maligne et une bénigne. En outre, nous avons démontré que dans le diagnostic bénin vs malin dune pleurésie, la combinaison des trois critères classiques de Light, définissant un exsudat (rapport des taux pleuraux/sanguins des protéines, des LDH et la valeur absolue du taux de LDH pleuraux supérieur au 2/3 de la norme supérieur), présentait une sensibilité et une rentabilité diagnostiques similaires à celles du VEGF voisine de 60%. En conclusion, limagerie métabolique par TEP est discriminante pour affirmer la malignité dune pathologie pleurale et constitue un facteur pronostique indépendant de survie des patients atteints de pleurésies métastatiques de tumeur pulmonaire non à petite cellule. Dès lors la TEP sintègre dans lalgorithme décisionnel du diagnostic, et de la prise en charge dune pathologie pleurale et permet déviter de recourir systématiquement à une technique dinvestigation pleurale invasive. Quoique significativement plus élevés dans les pleurésies malignes, le taux pleural de VEGF présente un pouvoir discriminant modeste dans le diagnostic différentiel bénin vs malin dune pleurésie exsudative. Toutefois, ces observations aident à la compréhension physiopathologique des pleurésies tumorales et ouvrent la porte à de nouvelles thérapeutiques. Summary: Pleural involvements are common in various respiratory diseases including inflammatory, infectious, occupational or neoplastic pathological entities and pleural thickening and pleurisy are usual radiological presentation. Pleural pathology is found in 10 % of the patients hospitalized in a department of internal medicine. Pleural effusions can be induced both by primary pleural disease and by lung and extra-thoracic pathologies. Finding their etiology imposes a vast and sometimes difficult differential diagnosis as different physiopathological processes may be involved in thickening and effusions. In particular, its of first importance to identify malignant pleural disease in consideration of its specific treatment and its dire prognosis. Here we review the currently accepted pleural diseases investigation methods. There are no specific diagnosis criteria for malignancy with conventional pleural imaging by radiology, ultrasound, scanning and nuclear magnetic resonance. The analysis of chemistry, bacteriology and cytology of pleural fluid obtained by thoracocentesis makes a significant contribution to the diagnostic approach even though its diagnostic sensitivity does not exceed 62 %. Similarly, although allowing tissue examination, the sensitivity of closed-needle pleural biopsies does not exceed 51 %. Therefore the invasive thoracoscopy is often justified to find out the accurate diagnosis of pleural diseases, achieving a diagnostic sensitivity greater than 95%. Our goals in this thesis were threefold. Our first objective was to study the accuracy of positron emission tomography (TEP) using the 18F-fluorodeoxyglucose (18FDG) to distinguish between benign and malignant disease in exudative lymphocytic pleural effusions and pleural thickening. We evaluated the diagnostic performance of 18FDG TEP imaging with semi-quantitative analysis for differentiating benign from malignant pleural effusions and for guiding the search for the primary tumor (extra-thoracic vs thoracic location) of pleural metastases. We established two threshold values distinguishing benign from malignant pleural lesions on the one hand and extra-thoracic from thoracic malignant pleurisies on the other hand. Our second objective was to estimate the value of the pleural 18FDG uptake intensity as an independent survival prognosis factor in non small cell lung cancer with pleural extension (T4 NSCLC). We showed an inverse relationship between survival and pleural metabolic activity while there was no relationship between survival and the primary tumor metabolic activity. Finally, our third objective was to assess the value of biochemical analysis and in particular the value of cytokine (IL-6) and growth factors ((TGF-ß 1, VEGF) in differentiating benign from malignant pleural effusions. No significant difference was found between malignant and benign effusions with respect to IL-6 and TGF- pleural levels. In contrast to IL-6 and TGF-pleural VEGF concentrations were significantly higher in malignant than in benign effusions. Furthermore VEGF levels correlated with the effusion volume, pleural LDH and glucose levels and red cell counts. Nevertheless, although being greater in malignant pleural effusions, pleural VEGF levels are of a rather limited clinical interest in distinguishing between malignant and benign pleural effusions. VEGF levels did not perform better than the use the 3 combined Lights criteria in this attempt with a quite similar sensitivity around 60%. In conclusion, the metabolic imaging by TEP is an accurate method for differentiating benign from malignant pleural diseases and the intensity of the pleural 18FDG uptake seems to be an independent survival prognostic factor in non small cell lung cancer with pleural malignant effusion. Our results suggest TEP could be integrated in the investigation algorithm of pleural diseases and facilitate decision-making as to when begin invasive procedures. Raised pleural VEGF levels in malignant effusions are related to the aggressiveness of the tumors and, although of limited diagnostic value, should prompt the evaluation in the near future of new treatment strategy using anti-VEGF.

The focus of this thesis is on practice changing clinical studies which impact upon the day to day treatment of patients with pleural infection, answering specific questions on several aspects of patient management. Specific areas of assessment in this thesis include: Assessment of the current evidence for optimal drain size choice in patients with pleural infection; Analysis and statistical modelling of a previous cohort of patients with pleural infection, in order to assess optimal drain size choice in pleural infection; The design, conduct and analysis of a 2 x 2 factorial multi-centre randomised, placebo controlled trial to assess the efficacy of two novel intrapleural agents (tPA and DNase) in aiding drainage in patients with pleural infection (The 2nd Multi-centre Intrapleural Sepsis Trial, referred to from here on as MIST2); Validation work informing the primary outcome measure of MIST2, assessing the relationship between chest radiograph imaging of infected pleural effusion and CT measured volume of pleural fluid using novel digital measurement strategies.

[Truncated abstract] The function of the diaphragm as a volume pump has not been adequately evaluated because there are no accurate methods to measure the volume displaced by diaphragm motion (ΔVdi). As a consequence, the work done, power output and efficiency of the diaphragm have not been measured. Efficiency of the diaphragm could be measured by relating the power output of the diaphragm to its neural activation. The aims of this thesis were to (a) develop a new biplanar radiographic method to measure ΔVdi and use this to evaluate the effect of costophrenic fibrosis and emphysema on ΔVdi, (b) develop a new fluoroscopic method to enable breath-by-breath measurements of ΔVdi, (c) evaluate a method for quantifying neural activation of the diaphragm, and (d) combine measurements of transdiaphragmatic pressure, ΔVdi, inspiratory duration and neural activation of the diaphragm to quantify the neuromechanical efficiency of the diaphragm

Thesis (PhD (Medicine. Internal medicine))--University of Stellenbosch, 2007. / 1.1. Background Tumors arising from the lung, pleura, or chest wall are a frequent problem in clinical pulmonary medicine. Most lesions are either infectious, neoplastic or granulomatous in nature, but a variety of other differential diagnoses must be considered. An accurate diagnosis is important because the available treatments differ substantially, and because any delay will impair the prognosis in potentially curable patients with lung carcinoma. The investigations involve the disciplines of radiology, pulmonology, surgery, microbiology, and anatomical pathology and consume a respectable amount of resources. The aim of the work covered in this thesis was to optimize the available diagnostic methods for the routine use in a health care setting with limited resources. 1.2. Methods The general idea of this work was to identify conventional sampling methods that could be developed further to become more useful for the diagnosis of chest tumors in a low resource health care setting. The key method was research performed: a) to revise and expand the indication for a sampling method, b) to technically improve the sampling process, and c) to optimize sample transport, preparation and analysis in collaboration with the analytical laboratory. 1.3. Results A list of invasive diagnostic procedures, imaging methods and analytical processes were developed, evaluated and integrated into clinical practice. A) transbronchial needle aspiration, B) transthoracic cutting needle biopsy, C) transthoracic fine needle aspiration, D) transthoracic ultrasound, and E) rapid on-site evaluation of needle aspirates by a cytopathologist. Five studies pertaining to this thesis were published in international peerreviewed journals: â ¢ Safety and yield of ultrasound-assisted transthoracic biopsy performed by pulmonologists (Respiration 2004;71:519-22) This paper established that ultrasound-assisted transthoracic biopsy performed by pulmonologists is feasible, safe, practical, low-cost and has a high yield. â ¢ Utility of rapid on-site evaluation of transbronchial needle aspirates (Respiration 2005;72:182-8) This paper demonstrated the economical advantages of on-site evaluation of transbronchial specimens in a low-resource setting. â ¢ Transbronchial needle aspirates: comparison of two preparation methods (Chest 2005;127:2015-8) This paper demonstrated that preparing smears on-site has a far better yield than pooling samples into a vial. This means that the yield is improved over the current standard at no additional cost. â ¢ Transbronchial needle aspirates: how many passes per target site? (European Respiratory Journal 2007;29:112-6) This paper investigated the most economical and effective approach to serial sampling with transbronchial needle aspiration during flexible bronchoscopy. â ¢ Ultrasound assisted transthoracic biopsy: fine needle aspiration or cutting needle biopsy? (European Respiratory Journal 2007;29:357-62) This paper compared two common methods of sampling and demonstrates that the less expensive method is sufficient in the majority of cases. 1.4. Conclusion This work has impacted on current practice in multiple ways. Conventional methods have been optimized by improving technical factors and with the integration of interdisciplinary collaboration. The initiated research is ongoing with the aim to achieve continued technical and economical improvements in the diagnosis of chest tumors.