L’optimisation hémodynamique péri-opératoire est une stratégie qui vise à maximaliser le transport artériel en oxygène et/ou le volume d’éjection systolique lors de chirurgie à risque. Ce concept a beaucoup évolué lors de ces trente dernières années, vers une approche plus simple, plus réalisable en pratique clinique et moins invasive. Les principales thérapeutiques utilisées dans les différents protocoles d’optimisation hémodynamique sont le remplissage vasculaire, l’administration d’agents inotropes et de vasopresseurs. Cependant, les conséquences physiopathologiques de l’agression chirurgicale peuvent impacter grandement les modalités d’administration et l’efficacité des thérapeutiques précitées. Dans la première étude, nous avons décrit l’impact de l’acidose respiratoire et métabolique (fréquemment rencontrées lors de chirurgie majeure et/ou de coeliochirurgie) sur l’efficacité des agents α et β-adrénergiques sur le myocarde sain de rat. Dans un deuxième travail nous avons mis en évidence que le remplissage vasculaire ne pouvait pas être guidé par des indices dynamiques de précharge dépendance lors du clampage chirurgicale de l’aorte abdominale sous-rénale, dans un modèle porcin. Enfin, dans la troisième étude, nous avons montré dans un modèle clinique, que le positionnement en décubitus ventral lors d’une chirurgie du rachis entrainait des modifications majeures des interactions cardiorespiratoires et que les indices dynamiques devaient être interprétés avec prudence pour guider le remplissage vasculaire dans ce contexte. Ces études translationnelles soulignent trois situations fréquentes impactant l’efficacité et/ou les modalités d’administration des thérapeutiques nécessaires à une optimisation hémodynamique peropératoire / The aim of perioperative haemodynamic optimization is to maximize oxygen delivery and/or stroke volume during high risk surgery. This concept has evolved during the last thirty years, to a simpler, more feasible and less invasive approach. Main treatments used in different hemodynamic optimization protocols are fluid loading, inotropes and vasopressors administration. However, pathophysiological consequences of surgical stress can greatly impact the mode of administration and the efficacy of the above therapeutics. In the first study, we described the impact of respiratory and metabolic acidosis (frequently encountered during major surgery and/or laparoscopic surgery) on the effectiveness of α and β-adrenergic agents in healthy rat myocardium. In a second work, we demonstrated that intravenous fluids cannot be guided by dynamic indices of preload dependency during surgical clamping of the infrarenal abdominal aorta in a porcine model. Finally, in the third study, we demonstrated in a clinical model, that positioning in prone position during spine surgery induced major changes in cardiorespiratory interactions and dynamic indices should be interpreted with caution to guide fluid therapy in this context. These translational studies highlight three common situations impacting the effectiveness and/or administration of therapeutic necessary for intraoperative hemodynamic optimization.

This thesis explored whether haemodynamic responses to psychological stress test predict future blood pressure (BP) levels: the Reactivity Hypothesis. The research included a systematic review and two prospective cohort studies in the UK and Thai samples. In addition, the Blunted Reactivity Hypothesis, which posits that cardiovascular reactivity is inversely related to symptoms of anxiety and depression, was examined in cross-sectional analyses. A systematic review with meta-analysis and meta-regression with 41 prospective cohort studies (from 1950 to 2012) examined whether cardiovascular responses to psychological stress tests predict future BP levels, hypertension status, preclinical coronary heart disease (CHD) and cardiac events. Three possible moderators were included in analyses: type of task (active versus passive coping), age group (children versus adults), and duration of follow-up (short versus long-term follow-up). The review found that systolic BP reactions to psychological stress tests predict future systolic BP levels and that there was better prediction in child samples with shorter follow-up periods. Similarly, diastolic BP reactions to psychological stress predicted future diastolic BP levels. Cardiovascular reactions to psychological stress tests did not predict hypertension, preclinical CHD, or cardiac events. Cross-sectional analysis of two studies conducted in the UK and Thailand provided some evidence that anxiety and depressive symptoms were negatively associated with cardiovascular reactivity: these findings supported the Blunted Cardiovascular Hypothesis. However, these relationships were observed in the UK sample, but not in the Thai sample. Further, Thai participants responded to psychological stress task with large cardiovascular reactions, of a similar magnitude to the UK participants and observed in previous studies of Europeans and North Americans. Finally, prospective analyses revealed that systolic BP responses to mental arithmetic predict future systolic BP levels after one year of follow-up in both UK and Thai individuals, after controlling for baseline cardiovascular activity and traditional risk factors. In contrast, haemodynamic responses did not predict future BP. These results provide support for the “Reactivity Hypothesis” although the effect sizes were relatively small. However, responses to only one of the three stressors, mental arithmetic, predicted future BP implicating beta-adrenergically mediated cardiovascular responses. However, there was no physiologic evidence (i.e., cardiac output responses) that suggested beta-adrenergic mechanisms. Accordingly, future studies should examine alternate mechanisms (e.g., platelet aggregation and endothelial function) and cardiovascular responses in larger samples with a longer follow-up to further clarify the predictive value of reactivity in the development of hypertension, along with potential mechanisms.