Increasing angiotensin-converting enzyme concentrations and absent angiotensin-converting enzyme activity are associated with adverse kidney outcomes in pediatric septic shock

Septic shock is common in the pediatric intensive care unit and associated with increased morbidity and mortality [1, 2]. Children with septic shock often develop acute kidney injury (AKI), which confers increased risk of poor outcomes [3]. Unfortunately, the treatment of septic shock and sepsis-associated AKI remains limited to supportive care, as no specific disease-modifying therapies exist [4]. A better understanding of the mechanisms of organ dysfunction in septic shock is necessary to improve outcomes.

There is increasing evidence that the renin–angiotensin–aldosterone system (RAAS) is dysregulated in critical illness [5‐9], and may represent a modifiable target for treatment of patients with septic shock and AKI [6, 9, 10]. Specifically, it is postulated that endothelial injury results in ACE dysfunction with subsequent decreased conversion of angiotensin I (AngI) to angiotensin II (AngII) (a potent vasoconstrictor), exacerbating vasodilatory shock (Fig. 1) [6, 9]. AngII also has direct effects on intrarenal circulation, causing preferential vasoconstriction of efferent arterioles to increase glomerular perfusion pressure, and thus decreased levels may also exacerbate AKI [10]. This hypothesis is supported by evidence demonstrating increased AngI/AngII ratios in adults with catecholamine resistant vasodilatory shock (CRVS) [6, 9], however, few studies have directly measured ACE concentrations [11] and none have measured ACE activity. The importance of testing this hypothesis directly is underscored by the availability of synthetic AngII for treatment of CRVS.

We performed a pilot study measuring serum ACE concentrations and activity in a cohort of children with septic shock. We hypothesized children with septic shock would have low ACE concentrations and activity, and that these values would be associated with increased renin concentrations and adverse kidney outcomes.

Further details are provided in the Additional file 2.

The finding of diminished ACE activity in children with septic shock provides direct evidence in support of the hypothesis that ACE dysfunction occurs in these patients, presumably secondary to endothelial injury (Fig. 1) [6, 9]. While it is proposed that this dysfunction leads to reduced AngII production and exacerbation of vasodilatory shock and AKI, we were unable to measure AngI/AngII concentrations, which is a limitation of this study. Nevertheless, the association of absent ACE activity with poor outcomes is consistent with adult literature, which has shown similar associations in patients with CRVS, elevated renin concentrations (a proposed upstream surrogate for AngII deficiency that is more easily measured), and increased AngI/AngII ratios [6, 10]. Importantly, in these studies, investigators demonstrated that use of synthetic angiotensin II improved outcomes in patients with presumed AngII deficiency by elevated renin concentrations, including improved renal recovery [6, 10]. These data, combined with our recent data demonstrating extreme elevation of renin + prorenin concentrations in pediatric septic shock [13], suggest that further study is warranted to elucidate if AngII may be particularly beneficial in this population.

While the median ACE concentration for the cohort was lower than expected in healthy children, the finding that higher concentrations were associated with poor outcomes is unexpected. In this cohort, having an ACE concentration above the median was associated with severe persistent AKI and KRT use. This is contrary to a study in adults with sepsis, which demonstrated that lower ACE concentrations on Day 1 were associated with worse outcomes [11]. However, these values continued to decrease over the first 72 h [11], and thus we postulate that extensive endothelial injury may result in release of ACE reservoirs early in sepsis, with subsequent exhaustion over time (Fig. 1). Accordingly, data limited to Day 1 would prevent us from seeing this expected decrease in ACE concentrations. Additionally, it is possible that ACE activity and ACE levels may be uncoupled in this population, secondary to an intrinsic defect in ACE function [19] or circulating ACE inhibitors [20]. Further study is needed to examine these hypotheses.

Our study has several limitations. First, our sample size is small, and these findings need to be corroborated in a larger cohort. This was a secondary analysis of an existing study not designed to examine our hypotheses, introducing the possibility of bias. The type of shock at presentation (i.e., cold versus warm) was unknown, though 54/63 (86%) who required vasoactives received agents suggesting vasodilatory shock (high dose epinephrine, norepinephrine and/or vasopressin). We were unable to trend these concentrations over time and with changes in hemodynamics, an important limitation given the dynamic nature of critical illness. Finally, the lack of direct renin and AngI/AngII data makes direct comparison to the existing adult literature challenging. These issues should be addressed in future prospective studies.

We demonstrate ACE activity is low in children with septic shock, apparently uncoupled from ACE concentrations, and associated with poor outcomes. If validated in a larger cohort, these data provide clarity on the potentially modifiable mechanisms of RAAS derangement in these patients.