Hello everyone, welcome to the Society of Critical Care Medicine Congress 2023. My name is Joaquin Cantos, I'm an ICU physician from the Hospital Italiano de Buenos Aires, Argentina. And I'm going to present our study named Allactive Base Excesses and Independent Projector of Death in sepsis, a propensity score analysis. Our study, rationally, is based on two main concepts. One of them is Allactic Base Excess, a novel concept introduced by Gattinoni and collaborators in 2019. In sepsis, Allactic Base Excess quantifies the role of renal function in the acid-base balance. Allactic Base Excess is calculated as a sum of standard base excess and lactate. This variable focuses on the role of fixed acids other than lactate in septic patients, and helps in the rapid discrimination between metabolic acidosis secondary to lactate accumulation from that caused by an increase in fixed acids due to kidney failure. The other concept on which we base our research is sepsis-associated acute kidney injury. Its diagnosis is based, according to Cádigo Criteria, on increased serum creatinine and or decreased urine output in association with sepsis diagnosis. The sepsis-associated acute kidney injury pathophysiology involves multiple mechanisms. However, recent studies found that tubular dysfunction behaves as an early adaptive mechanism to stress and that it is closely related to glomerular function loss. Dissociation between structural and functional changes has motivated research on novel biomarkers to detect kidney stress or damage. These novel biomarkers can predict adverse outcomes and could provide an early acute kidney injury diagnosis. A recent consensus statement proposed by the AdKey Group proposed a widened acute kidney injury definition that included these novel biomarkers to identify subclinical acute kidney injury, which is an independent risk factor of mortality. Given this background, negative allactic base excess in patients without acute kidney injury criteria could reflect a subclinical acute kidney injury, since it is secondary to fixed acid retention. However, if negative allactic base excess reflects a subclinical acute kidney injury, it should be related to poor outcomes independently of acute kidney injury criteria. Therefore, we made ourselves the following question. Do patients without acute kidney injury criteria and negative allactic base excess have worse clinical outcomes than patients without acute kidney injury and non-negative allactic base excess? This study's objectives were to evaluate if acid-base changes reflected by allactic base excess predict unfavorable outcomes in septic patients independently of kidney function markers. Also, to evaluate whether allactic base excess changes in septic patients behaved as predicted of in-hospital mortality, even in patients with glomerular filtration rate greater than 60 milliliters per minute. Lastly, our objective was to evaluate whether lactic acidosis is associated with acidemia in septic patients with and without acute kidney injury. To answer these questions, we carried out a retrospective cohort study in two medical surgical ICUs from two university teaching hospitals. Our inclusion period was between January 2015 to January 2020. We followed up patients until discharge or hospital death. Our inclusion criteria were adult patients, 18 years and older, with a diagnosis of sepsis or septic shock at ICU admission, according to sepsis 3 definition. Excluded patients were those with a history of kidney replacement therapy and or kidney transplant history at ICU admission. Also, patients without data to calculate allactic base excess or SOFA score. After inclusion, patients were divided in three groups according to the groups proposed by Gattinoni. They were negative allactic base excess, neutral allactic base excess, and positive allactic base excess. With respect to data analysis and outcomes, we established our primary outcome as in-hospital mortality related to allactic base excess groups, adjusting by acute kidney injury and other confounders. As secondary outcomes, we evaluated in-hospital mortality associated with allactic base excess groups in patients with glomerular filtration rate greater than 60 milliliters per minute and association between lactic acidosis and acidemia in septic patients with and without acute kidney injury. With respect to data analysis, we made Cox regression to evaluate the hazard ratio of mortality in the different allactic base excess groups, adjusting by inverse probability weighting with propensity score. We made sensitivity analysis with adjusted Cox regression. This is our inclusion and exclusion criteria flowchart in which we show that we included 1,178 patients with a diagnosis of sepsis or septic shock that were included for follow-up and the posterior formation of the three different groups. This is our demographic data table in which we highlighted the negative allactic base excess population in which we found worst Apache, SOFA, and Charleston scores, greater proportions of septic shock and mechanical ventilation, as well as longer ICU length of stay. In the case of clinical acid base and electrolyte data table, we found that also this population had lower pH values, higher lactate values, greater proportions of metabolic acidosis, and elevated chloride and unmeasured anions in plasma. This last table also shows that this group of patients has had statistically significant greater plasma creatinine values, as well as lower glomerular filtration rate. They also had greater proportions of chronic kidney disease, as well as acute kidney injury. They had less urine output, as well as greater proportions of kidney replacement theory. With respect to our primary outcome, in the Cox regression analysis weighted by inverse probability weighting, we observed that patients with negative allactic base excess had higher in-hospital mortality than patients with neutral allactic base excess. And patients with neutral allactic base excess had the same mortality as patients with positive allactic base excess. For sensitivity analysis, we performed Cox regression adjusted by the same confounders, including the propensity score. In this analysis, we observed that patients with negative allactic base excess had higher in-hospital mortality than patients with neutral allactic base excess. And patients with neutral allactic base excess had the same mortality as patients with positive allactic base excess. Regarding our secondary outcome, patients with glomerular filtration rate greater than 60 and negative allactic base excess had higher in-hospital mortality than patients with neutral allactic base excess in the multivariate analysis. On the other hand, patients with glomerular filtration rate greater than 60 and positive allactic base excess had the same in-hospital mortality as patients with neutral allactic base excess, both in the univariate and multivariate analysis. As a secondary outcome, we evaluated the relation between lactic acidosis and acidemia among patients with kidney injury. We observed that hyperlactatemia between 2.5 and 5.6 was not associated with acidemia in patients without acute kidney injury. A common finding among critically ill patients under hypermetabolic stress. On the other hand, patients with lactate levels higher than 5.6 had acidemia independently of the presence of acute kidney injury. Furthermore, as seen in the figure, pH values in non-acute kidney injury patients start to decrease with lactate levels higher than 4 millimoles per liter. It could also be added that the lactate values are those of the first laboratory analysis at the time of diagnosis of sepsis or septic shock, for which it is presumed that the patients have not been resuscitated up to that time, which in turn could explain the hyperperfusion related hyperlactatemia. Arguably, this is an expected finding, given that the latter group could represent those patients with more severe illness and tissue-equal perfusion. The main finding of our study was that negative lactic base excess was an independent risk factor for mortality in critically ill septic patients, even in patients with preserved glomerular function. Additionally, we observed that septic patients with negative lactic base excess values presented a higher glomerular filtration rate than the one who hypothetically should induce fixed acid retention, that should be more or less less than 15. The fixed acid retention without acute kidney injury criteria is secondary to the kidney's incapability to acidify urine as a compensatory mechanism during metabolic acidosis. This fixed acid retention can be attributed to an acute tubular dysfunction, which reduces acid tubular secretion. In effect, as is consistent with Gattinoni and collaborators' findings, we observed that septic patients with negative lactic base excess exhibited greater rates of high anion gap metabolic acidosis, as well as augmented plasmatic unmeasured anion concentrations. Thus, these acid-base disturbances, probably attributed to an acute tubular dysfunction, could be monitored by lactic base excess. It is also worth noting that 46.2% of patients on the negative lactic base excess group had chronic kidney disease, which not only constitutes an acute kidney injury risk factor, but can also worsen tubular transporter dysfunction. Additionally, due to the relationship between chronic kidney disease and both tubular transporter dysfunction and mortality in critically ill patients, chronic kidney disease was considered as a potential confounder. Nevertheless, the relationship between death and negative lactic base excess was independent of chronic kidney disease presence. In conclusion, negative lactic base excess reflects acid-base disturbances in septic patients with preserved glomerular function. As this is a retrospective study useful for hypothesis generation, we express our hypothesis regarding the potential usefulness of lactic base excess in septic patients. In this sense, we propose that negative lactic base excess, even in the absence of acute kidney injury criteria, could be a reflection of renal tubular dysfunction, not yet expressed by glomerular function markers. Accordingly, patients with negative lactic base excess without acute kidney injury criteria should present urinary alterations such as insufficient urinary acidification and positive damage biomarkers in plasma. Moreover, we would expect that the incapability of normalizing negative lactic base excess values will mean progression to acute kidney injury in septic patients. This being said, we conceive dynamic lactic base excess monitoring as a complementary tool to detect subclinical acute kidney injury, given that it is easy to calculate and it is a widespread resource. The main strengths of our study were that we included a large septic population of more than 1,000 patients, we determined acute kidney injury definition with category criteria which were not addressed by Gattinoni, and we included data regarding chronic kidney disease in our population. On the other hand, our study had the following limitations. We used cut-off values to build a lactic base excess groups instead of using sex styles of distribution reported by Gattinoni. We did not have the proportion of patients with diuretic therapy history before hospitalization which could have had an impact on metabolic alkalosis rates. Data on nephrotoxic agents was lacking, for example, antibiotics. Finally, we had missing data related to some confounders, especially chronic kidney disease. Therefore, in the inverse probability weighting analysis, complete case analysis, we evaluated 70% of the population. In spite of this, we achieved the minimum number of patients with diuretic therapy history In spite of this, we achieved the minimum sample size calculated. This was the bibliography used to build this presentation. Thank you very much for your attention. I hope you enjoyed our presentation.

Allactic Base Excess

septic patients

kidney function markers

mortality

acid-base balance