• Citer

Dysregulation of Cerebral Perfusion In Sepsis-Associated Brain Dysfunction

par Crippa, Ilaria Alice
Président du jury Schiffmann, Serge N.
Promoteur Creteur, Jacques
Co-Promoteur Taccone, Fabio
Publication Non publié, 2023-08-29

Thèse de doctorat

• Accès en ligne
• Détails
• Contenu
• Statistiques

Résumé :

Sepsis-associated brain dysfunction (SABD) is the acute brain dysfunction that occurs during sepsis as a wide range of neurologic deficits - from intermittent episodes of disorientation to coma. SABD has been associated with mortality, but it is not clear whether brain dysfunction increases the risk of death per se in critically ill patients or is just a marker of sepsis severity. Furthermore, the neurologic status of septic patients is often fluctuating, but whether the trend in neurological condition has a prognostic value is not known. The impact of brain dysfunction on in-hospital mortality in critically ill patients has been investigated in the first part of this research project. Data from 7192 (2096 septic and 5096 non-septic) patients from multiple centres were analysed; brain failure was defined as neurological sequential organ failure assessment (nSOFA) score 3–4 (i.e., GCS ≤ 9). At multivariable analysis, sepsis and brain failure entailed an increased risk of death compared with non-septic patients with normal brain function, with an adjusted risk ratio (RR) and 95% confidence interval (95%CI) of 1.66 (1.31–2.09) and 4.85 (3.33–7.07), respectively. In case of simultaneous presence of sepsis and brain failure the RR was 5.61 (3.93–8.00). Brain deterioration – defined as an increased in nSOFA over the first 3 days of intensive care unit (ICU) stay - was associated with increased in-hospital mortality in both septic and non-septic patients. However, the risk of death septic patients who experienced brain improvement – defined as a decrease in nSOFA over the first 3 days of ICU stay - was the same as that of non-septic patients without brain dysfunction. Although pathogenesis of SABD is still unclear, cerebral perfusion alterations likely play a role. Cerebral perfusion pressure (CPP) is the difference between mean systemic arterial pressure (MAP) and intracranial pressure (ICP), thus MAP reductions or ICP increases might lead to reduced CPP. MAP is frequently decreased during severe sepsis and septic shock, while data on ICP in septic patients are scarce. Moreover, sepsis is associated to endothelial and cellular alterations that can lead to blood-brain barrier disruption and vasogenic and cytotoxic oedema, thus potentially increasing ICP. In physiological conditions, several mechanisms cooperate to maintain CBF within a functional range. Cerebral pressure autoregulation (CAR) buffers MAP variations to maintain cerebral blood flow (CBF) nearly constant over a finite, although wide, range of MAP. Extreme hemodynamic instability and sepsis-related inflammation effect on cerebral vasculature may override CAR buffer, exposing the brain to altered CBF and risk of hypoperfusion even in case of minor alterations in MAP. Furthermore, a number of factors regulate vasomotor basal tone or vasoreactivity, thus having a potential impact on CAR. Arterial carbon dioxide partial pressure (PaCO2) is a major determinant of cerebral vascular tone and is often altered during sepsis, because of either associated compensating mechanisms, organ failure or medical treatment. Intracranial hemodynamics of septic patients was investigated in the second part of this research project. The aims were determining whether CPP and ICP are altered, and their alterations are associated to SABD; determining whether alterations in CAR is associated to SABD and determining whether CAR is influenced by modifications in PaCO2. CPP and ICP were non-invasively estimated (eCPP and eICP) and the possible association between either reduced eCPP or elevated eICP and SABD was investigated in a monocentric cohort of 132 septic patients, among which 49% developed SABD during ICU stay. Median [interquartile range, IQR] eCPP was 63 [58-71] mmHg and 33% of patients had eCPP<60 mmHg; median [IQR] eICP was 8 [4-13] mmHg and 4% of patients had eICP>20 mmHg. SABD occurrence did not differ between groups. Notably, only 65% of patients had normal values of eCPP (i.e., eCPP>60 mmHg) and eICP (i.e., eICP<20 mmHg), while 35% had abnormal values of either eCPP, eICP or both. CAR was non-invasively assessed in a multicentre cohort of 100 patients. CAR was altered in 50% and SABD was diagnosed in 57% of patients. At multivariable analysis, lower MAP and a history of chronic kidney disease were associated to altered CAR; while altered CAR, mechanical ventilation and a history of vascular disease were associated to SABD development. The effects of PaCO2 on CAR were studied in 15 mechanically ventilated patients in which respiratory settings were modified to obtain two different levels of PaCO2, (36 [34-38] mmHg vs 45 [41-49] mmHg, p<0.01). Moving from lower to higher PaCO2 values resulted in a significant increase in CBF estimated from the increase in blood flow velocity in intracranial arteries; however, PaCO2 changes did not significantly affect CAR. Autonomic nervous system (ANS) modulates basal vascular tone and stress response of cerebrovasculature during hemodynamic challenges. ANS alteration is associated with worse CAR in several conditions, including acute brain injury and severe liver failure. ANS is often deregulated during sepsis. ANS function can be assessed through pupillary light reflex (PLR), which can be evaluated bedside using an automated pupillometer. In this research project, the association between CAR and PLR was evaluated in 40 septic patients. Despite a normal global PLR, pupillary dilation velocity (DV) was lower in patients with altered compared to those with intact CAR (1.3 [1.2–1.9] vs 2.6 [1.8–3.2] mm/s; p < 0.01). The area under the receiver operating characteristic (AUROC) for DV to predict altered CAR was 0.78 (95% confidence interval, 95%CI 0.63–0.94)%; values < 2.2 mm/s had sensitivity [inter-quartile range] 85 [65–95]% and specificity 69 [46–84]%. eCPP and eICP were non-invasively estimated: global PLR function was lower in patients with low compared to normal eCPP (3.7 [3.5–4.1] vs 4.6 [4.5–4.6], p<0.01, respectively) and in patients with high eICP compared to normal eICP (3.5 [3.5–3.6] vs 4.5 [4.3–4.6]; p<0.01); global PLR function correlated with eCPP (r = 0.77, p<0.01) and eICP (r = -0.87, p<0.01). CAR response to modifications in CPP is not instantaneous; the latency of CAR activation has been studied in animals and humans, but whether is identifiable and relevant in septic patients remains unknown. The last part of this research project focused on the possibility to identify and quantify the latency in CAR response to modification in CPP in septic patients. CAR was assessed in the time-domain with mean flow index (Mxa), based on transcranial Doppler assessment of intracranial arteries blood flow velocity (FV) and invasive arterial blood pressure (ABP). The relationship between FV and ABP was explored to identify a delay in FV modifications with regard to ABP fluctuations, possibly indicating a delay in CAR response. Of 154 patients, 54% showed normal CAR, 26% showed a potentially delayed CAR response, and 28% showed altered CAR. The risk to develop SABD progressively increased in patients with normal, delayed or altered CAR. ICU mortality was significantly higher in patients with altered CAR compared to patients with normal or delayed CAR.