1 Introduction

Acute kidney injury (AKI), part of a diverse range of functional kidney diseases, is an important and complex syndrome caused by multiple mechanisms and associated with considerable morbidity and mortality [1, 2]. In low- and middle-income countries, infection and hypovolemic shock are major causes of AKI [3, 4]. In high-income countries, AKI occurs mostly in hospitalized elderly patients and is associated with sepsis, drugs, or invasive surgery [5, 6]. AKI management in the critical care setting is challenging, including appropriate volume control, management of nephrotoxic medications, and timing and type of renal support. Fluid and electrolyte management is essential. AKI has a poor prognosis in critically ill patients, often requiring renal replacement therapy. Therefore, prevention and early detection of AKI is crucial.

Important progress has been made in recent years in the incidence, pathophysiology, and treatment of renal insufficiency in sepsis. From the initial definition of acute renal failure by RIFLE in 2004 to the refinement of the definition of AKI by KDIGO in 2012, our ability to identify renal injury using traditional assessments of renal function (serum creatinine, urine output) has increased [7, 8]. An increasing number of serum and urine biomarkers of AKI have been identified and have the potential to improve the treatment and clinical outcomes of sepsis-related AKI [9, 10].

In this review, we summarized the published literatures on sepsis-associated acute kidney injury (S-AKI), focusing on the epidemiology, risk factors for developing S-AKI, pathophysiology, management of S-AKI, and prognosis.

2 Definitions

2.1 Sepsis

In 2016, the American Medical Association released the latest definition and diagnostic criteria for sepsis, namely Sepsis 3.0 [11]. The new definition proposes that sepsis is a life-threatening organ dysfunction caused by a dysregulated host response to infection, downplaying the hallmark of sepsis that has been used in the past—systemic inflammatory response syndrome (SIRS), especially emphasizing the difference between the concept of sepsis as a severe infection and general infection. The new criteria propose that patients with infection or suspected infection can be diagnosed with sepsis when the Sequential Organ Failure Assessment (SOFA) score is ≥ 2.

2.2 Acute Kidney Injury

The 2012 KDIGO guidelines defined the diagnostic criteria for AKI [12]. Unlike earlier recommendations, the KDIGO criteria no longer require adequate fluid resuscitation and exclusion of urinary tract obstruction prior to use of the criteria. This definition emphasizes AKI risk assessment and evaluation, while expanding the evaluation criteria to include an assessment of serum creatinine (sCr) of 50% or more above presumed baseline within seven days [12, 13]. However, in clinical practice, serum creatinine and urine output (UO) have certain limitations in the diagnosis of AKI. In animals, sepsis reduces muscle perfusion and thus creatine production, thereby suppressing the increase in serum creatinine concentrations and limiting early detection of AKI [14]. In septic shock, AKI may be missed due to the dilution effect of aggressive fluid resuscitation. Therefore, there is a call for further refinement of AKI based on changes in these traditional functional markers as well as biomarkers of kidney injury [15].

2.3 Sepsis-Related Acute Kidney Injury

Once the diagnostic criteria for both sepsis and AKI are met, a patient is considered to have septic AKI (Sepsis associated acute kidney injury, S-AKI) [16, 17]. In the ICU, sepsis is associated with up to 40–60% of AKI [18, 19]. S-AKI is associated with a six- to eightfold increased risk of in-hospital death [20] and a threefold increased risk of chronic kidney disease among survivors [21]. Furthermore, patients with sepsis-related AKI have significantly increased mortality compared with patients with AKI of other etiologies [22].

3 Epidemiology

Sepsis increases the risk of AKI, but the incidence and trend of S-AKI cannot be accurately estimated [23]. A strict attribution of AKI to sepsis remains difficult, possibly due to the high number of confounding factors affecting the diagnosis of AKI in critically ill patients [24]. Secondly, there is a lack of coordination mechanism in the field of epidemiology among scholars studying sepsis and AKI [23]. Adhikari et al. indicated that there are as many as 19 million cases annually worldwide [25]. The actual incidence may be much higher, as approximately one third of patients with sepsis will develop AKI [26], and the global annual incidence of S-AKI may be approximately 6 million cases, or nearly 1 in 1,000. population. Based on the observation that AKI occurs in approximately two-thirds of septic patients, S-AKI involves six to eleven million cases [22, 23]. In addition, sepsis is also the leading cause of death in AKI patients [27].

In a multicenter prospective cohort study of critically ill patients in 24 European countries, the incidence of AKI in sepsis patients was 54% and ICU mortality was 41% [28]. A retrospective study covering Chinese patients found that AKI was present in 47.1 percent of sepsis cases [29].

4 Pathophysiology of S-AKI

The pathophysiology of S-AKI remains unclear, mainly for the following reasons. The first point is that although progress has been made in the diagnosis of S-AKI, relying on serum creatinine and urine output has significant limitations in the timely diagnosis of AKI, and cannot provide specific causes of injury [30]. The second point is that kidney biopsies are difficult to obtain because sepsis patients, especially those with S-AKI, are usually in critical condition, which makes it difficult to obtain data on the pathological evolution of S-AKI. A third point is that more than 50% of patients with septic shock develop AKI before receiving medical care [20], making it difficult to determine which comes first. However, scientists have made considerable efforts in this area, and recent evidence indicates that the underlying pathophysiological mechanisms of S-AKI are mainly caused by inflammation, peritubular and glomerular microvascular dysfunction, cell cycle arrest and apoptosis, and the metabolic response of renal tubular epithelial cells to injury (Fig. 1). These current theories are derived from animal models or autopsy results [23].

Fig. 1
figure 1

Pathophysiology of sepsis-associated acute kidney injury. Abbreviations: DAMPs, damage-associated molecular patterns; PAMPs, pathogen-associated molecular patterns; OXPHOS, oxidative phosphorylation; TIMP-2, metalloproteinase 2; IGFBP7, insulin-like growth factor binding protein 7

4.1 Inflammatory Cascade

During sepsis, the Pattern recognition receptor (i.e. toll like receptors or TLRs) expressed on the surface of immune cells recognize the released pathogen related molecular patterns (PAMPs) and injury related molecular patterns (DAMPs), and start the intracellular molecular cascade, which is manifested as an inflammatory response to infection.[31]. In addition, renal tubular epithelial cells also express TLRs, especially TLR-2 and TLR-4. Thus, once PAMPs or DAMPs are filtered through the glomerulus, similar pathways are activated, leading to increased oxidative stress, production of reactive oxygen species, and mitochondrial damage [27, 32,33,34,35].

4.2 Macrovascular and Microvascular Dysfunction

Studies have shown that renal efferent arteriole vasodilation during sepsis results in decreased glomerular flow filtration [35]. At the same time, renal arteriole vasodilation and intrarenal shunt lead to large vessel dysfunction. Large vessel dysfunction diverts renal blood flow from the medulla to the cortex, resulting in reduced perfusion and oxygenation of the medulla, further worsening renal function [36]. Tissue perfusion is critical for the adequate functioning of any organ. Studies have shown that microcirculatory alterations occur during sepsis even in the absence of hemodynamic instability [37]. These observations suggest that altered microcirculation plays a critical role in the development of organ damage. Sepsis causes changes in local microcirculatory blood flow characterized by increased heterogeneity in regional blood distribution and increased capillary insufficiency (i.e., intermittent or stopped flow) [38,39,40]. Multiple mechanisms underlie this characteristic microcirculatory disturbance, including endothelial dysfunction, impaired erythrocyte deformability, damage and shedding of the glycocalyx layer, increased leukocyte activation, adhesion and recruitment, platelet adhesion, and activation of the coagulation cascade and fibrin deposition [41, 42]. Microcirculation dysfunction can lead to changes in local blood flow distribution, which in turn leads to nephron ischemia and loss of autoregulation, and exacerbates renal tubular epithelial cell damage and dysfunction [37, 43, 44]. Tubular epithelial dysfunction activates glomerular feedback by increasing non-reabsorbed chloride concentrations in the densified plaques, resulting in decreased glomerular filtration rate (GFR), increased sCr, and decreased UO.

4.3 Cell Cycle Arrest and Apoptosis

There is increasing evidence that mitochondria are closely related to the regulation of the cell cycle [45], and cell cycle arrest is an important cellular defense strategy in the context of sepsis [46]. Sepsis can induce significant mitochondrial damage, activation of mitochondrial mass regulation processes [23]. Since one of the most energy-intensive processes a cell undergoes is replication, when mitochondrial function is compromised and energy supply is insufficient, the cell will enter cell cycle arrest to avoid energy exhaustion and apoptosis. Thus, during sepsis, cells appropriately downregulate aerobic glycolysis and increase oxidative phosphorylation to defend against S-AKI [47, 48]. Two markers of cell cycle arrest, tissue inhibitor of metalloproteinase 2 (TIMP-2) and insulin-like growth factor binding protein 7 (IGFBP7), have been found to be the best predictors of S-AKI [49].

4.4 Metabolic Reprogramming

Metabolic reprogramming is a conserved defense mechanism that cells utilize to optimize and adjust energy expenditure [50, 51]. Prioritization of energy expenditure during sepsis is critical for cell survival during the acute phase of sepsis and may determine the repair phenotype during the recovery phase [52]. Studies have shown that in the early stages of sepsis, immune cells, stimulated by inflammatory signals, switch their metabolism from oxidative phosphorylation (OXPHOS) to aerobic glycolysis (ie, OXPHOS or Warburg metabolism in the presence of sufficient oxygen) [47, 53]. This metabolic shift drives the differentiation of T cells and monocytes towards pro-inflammatory phenotypes such as Th-17 and M1 macrophages, respectively [54]. Waltz et al. showed that metabolomic results in mice with sepsis indicated a shift in renal tissue metabolism from OXPHOS to aerobic glycolysis [50]. This metabolic reprogramming change protects against AKI and increases survival in animal models of sepsis [55]. Activation of OXPHOS regulators, such as Sirt1 or peroxisome proliferator-activated receptor (PPAR)-γ coactivator-1α (PGC-1α), also reduces sepsis mortality, suggesting that OXPHOS has a positive effect on sepsis. protective effect [48, 56].

5 Diagnosis, Screening and Prevention

Once the diagnostic criteria for both sepsis and AKI are met, a patient is considered to have septic AKI [57]. However, the diagnosis of sepsis is based on the SOFA score, which does not take into account underlying chronic kidney disease (CKD) when evaluating the renal system and thus cannot differentiate between new-onset AKI, underlying CKD, or acute or chronic AKI [30]. Mehta et al. found that 40% of critically ill patients developed sepsis after AKI [58], making it difficult to pinpoint the exact timing of the onset of both syndromes.

Based on the above reasons, finding biomarkers of renal injury is of great significance for the diagnosis of S-AKI. Studies have shown that both proenkephalin and cystatin C are highly correlated with AKI and GFR, and serum creatinine increases before severe sepsis patients [59,60,61], which is helpful for the early prediction and diagnosis of S-AKI. Studies have shown that inhibitors of metalloproteinase 2 (TIMP2) and insulin-like growth factor binding protein 7 (IGFBP7) in urine have the highest sensitivity and specificity for the prediction of S-AKI [62, 63]. Neutrophil gelatin Enzyme-associated lipocalin (NGAL) is expressed in many cell types, including prostate, uterus, salivary gland, lung, trachea, stomach, colon, and kidney [64]. Research has shown that even without an increase in serum creatinine, an increase in plasma NGAL can effectively predict the risk of severe AKI in critically ill patients [65]. When the critical value of plasma NGAL of 454 ng/mL is taken in sepsis, the sensitivity and specificity for diagnosing AKI can reach 72% and 74%, respectively[66]. Kidney injury molecule-1 (KIM-1) is expressed in the proximal tubular membrane upon ischemic or inflammatory injury. Meta-analysis suggests that urinary KIM-1 has a good predictive ability for AKI (area under the curve 0.86, sensitivity 74%, specificity 86%) [67].

6 Management

Most S-AKI therapies remain nonspecific, and early appropriate antibiotic therapy and avoidance of secondary renal injury remain the mainstay of sepsis treatment, which also prevents further renal injury.

6.1 Antibiotics to Control Infection

Early and appropriate initiation of antibiotic therapy is critical to prevent AKI and reduce mortality. Delayed use of antibiotics in septic shock is associated with early AKI development [19]. Care should be taken in antibiotic treatment, and drugs that have less impact on renal function should be used as much as possible. Use cautiously with drugs such as vancomycin, especially with antibacterial drugs such as piperacillin–tazobactam, aminoglycosides, or amphotericin B, and avoid concomitant use with other nephrotoxic drugs such as contrast media [12, 68].

6.2 Fluid Resuscitation and Vasoactive Drugs

Fluid resuscitation and vasopressors are the cornerstones of septic shock treatment. For septic patients at risk of developing AKI, balanced crystalloids are recommended for resuscitation, and colloids such as albumin, hydroxyethyl starch, and gelatin are not recommended [69, 70]. In septic patients, the use of hydroxyethyl starch and gelatin solutions increases the risk of AKI and mortality [71]. Clinical randomized controlled studies have shown that the use of chloride-rich solutions in sepsis increases the risk of AKI [72].

The 2016 Save Sepsis Campaign (SSC) recommended norepinephrine as the preferred vasopressor for septic shock, which increases mean arterial pressure (MAP) and improves renal perfusion [73]. Dopamine is not recommended for renoprotection and is associated with more adverse events than norepinephrine [74, 75]. Angiotensin II (ATII) is a potent vasoconstrictor [76], and sepsis causes ATII to be relatively scarce [77]. ATII infusion restored blood pressure and increased urine output and creatinine clearance [78]. A recent randomized controlled study of the effect of ATII in the treatment of high-output shock (80% of patients diagnosed with septic shock) showed increased blood pressure and reduced need for other vasopressors [79]. A post hoc analysis of patients with AKI requiring renal replacement therapy found that 28-day mortality was reduced on ATII, and patients receiving ATII were less likely to receive renal replacement therapy [79]. Therefore, ATII may be beneficial for S-AKI patients with shock.

6.3 Diuretics and Alkaline Phosphatase

There is currently no evidence that diuretics improve or alleviate AKI [80]. In particular, furosemide not only fails to improve AKI, but may be detrimental to critically ill patients [81]. Therefore, the routine use of diuretics to prevent or treat S-AKI is not recommended. However, the utility of diuretics in regulating and maintaining fluid balance is used for short periods of time in the treatment of critical illness with close monitoring of renal function.

Clinical studies have demonstrated a protective effect of systemic alkaline phosphatase against S-AKI [82, 83]. Alkaline phosphatase is an endogenous detoxifying enzyme that dephosphorylates pathogen and injury-associated molecular patterns, thereby reducing inflammation and organ dysfunction and improving survival [84]. A randomized controlled study in patients with S-AKI showed that although recombinant alkaline phosphatase did not significantly improve short-term renal function, it significantly improved long-term renal function and reduced 28-day mortality in patients with S-AKI [85].

6.4 Renal Replacement Therapy (RRT)

Current studies have shown that early initiation of RRT can improve the prognosis of patients with AKI. However, the timing of RRT in severe AKI patients is still controversial, and there is currently no unified treatment standard. A 2008 meta-analysis of 23 studies (5 randomized controlled trials, 1 prospective study, and 16 retrospective studies) showed that early RRT improves survival in patients with AKI [86]. Multiple randomized multicenter studies have shown no significant difference in mortality between early and delayed renal replacement therapy strategies [87, 88]. Most current guidelines recommend that in the setting of S-AKI, the dose of continuous RRT should be in the range of at least 30–35 mL/kg/h to ensure adequate dosing [89]. Higher doses (70 mL/kg/h) of continuous RRT did not improve patient survival [90].

7 Conclusions

S-AKI is an early, common, life-threatening complication of sepsis, and despite advances in our understanding of the pathophysiology of S-AKI, it remains a common complication of critical illness. Therefore, patients at risk for S-AKI should be actively screened for early identification and implementation of interventions. Better understanding of S-AKI pathophysiological processes (inflammatory cascade, macrovascular and microvascular dysfunction, cell cycle arrest and apoptosis, etc.). It may provide new ideas for therapies to prevent or reverse S-AKI.