Introduction
Acute kidney injury (AKI) is a clinical syndrome characterized by a sudden decline in kidney function including structural damage and functional impairment of glomeruli and renal tubules. These changes are typically manifested as a decreased glomerular filtration rate (GFR) and reduced urine output [1]. The incidence of AKI among hospitalized patients ranges widely, from 3.6% to 46%, depending on the characteristics of patient population and clinical context [2–4]. AKI is strongly associated with increased in-hospital mortality, prolonged length of stay, and a higher risk of progression to chronic kidney disease (CKD) and end-stage kidney disease [4,5].
Despite the active use of conventional diagnostic tools and various medical management including fluid resuscitation, immunosuppressive agents, and kidney replacement therapy (KRT), AKI still remains as a major clinical challenge. These traditional approaches are limited by delayed detection and a lack of precision in targeting underlying pathophysiological mechanisms.
In recent years, there has been substantial progress in elucidating the complex molecular and cellular mechanisms underlying AKI. This includes the identification of novel biomarkers that enable earlier and more accurate diagnosis, as well as the emergence of innovative therapeutic agents and strategies aimed at mitigating injury and promoting kidney recovery.
In this review, we focus on the evolving understanding of AKI pathophysiology as a foundation for identifying potential therapeutic targets. We also highlight recent advances in pharmacologic and regenerative treatments that offer promise for improving outcomes in patients with AKI.
Pathophysiology of acute kidney injury: potential targets for new treatments
The pathophysiology of AKI is complex, dynamic, and multifactorial. Among the etiologies of AKI, intrinsic renal injury is a predominant category, with acute tubular necrosis (ATN) being the most common subtype. The causes of ATN are typically classified as ischemic, septic, or nephrotoxic, each initiating distinct but overlapping pathways of kidney injury [1] (Fig. 1).
Despite the differentiated etiologies, a central pathogenic mechanism involves ischemia-reperfusion injury (IRI), which sets off a cascade of molecular and cellular events. IRI leads to the generation of reactive oxygen species (ROS), oxidative stress, endothelial dysfunction, and ultimately tubular epithelial cell (TEC) injury and death [6,7]. IRI induces profound metabolic and structural stress in the kidney. Reperfusion following ischemia paradoxically exacerbates injury by triggering excessive ROS production, largely due to mitochondrial dysfunction in stressed kidney cells. This oxidative stress damages lipids, proteins, and DNA, aggravating tubular cell injury and amplifying inflammation [6,8].
In parallel, IRI-induced endothelial injury impairs intrarenal microcirculation, resulting in microvascular congestion, thrombosis, and peritubular capillary rarefaction. This loss of microvascular integrity contributes to sustained hypoxia in the outer medulla, even after perfusion is restored, and has been identified as a major driver of the transition from AKI to CKD [9].
At aspect of cellular level, TEC injury and death play a critical role in determining the extent of kidney dysfunction and recovery. Initially, TECs may undergo apoptosis, a regulated, non-inflammatory form of cell death. However, severe or prolonged injury also activates regulated necrosis pathways, including necroptosis, pyroptosis, and ferroptosis, each contributing to membrane rupture, the release of damage-associated molecular patterns (DAMPs), and amplification of the inflammatory response [6,10].
If the injury is reversible and the inflammatory response is adequately resolved, TECs can regenerate via dedifferentiation, proliferation, and redifferentiation. However, in cases of severe or unresolved injury, maladaptive repair—characterized by persistent TEC dedifferentiation, cell-cycle arrest, and profibrotic signaling—ensues, ultimately leading to interstitial fibrosis and CKD progression. Thus, the pathophysiologic landscape of AKI offers multiple potential therapeutic targets, including the modulation of oxidative stress, preservation of endothelial integrity, and inhibition of regulated necrosis pathways. Therapeutic strategies that promote adaptive tubular repair while suppressing maladaptive responses may offer the most promising path forward in preventing AKI-to-CKD transition (Fig. 2).
Potential or emerging treatments for acute kidney injury
Pharmacological interventions for acute kidney injury
Anti-inflammatory agents
Inflammation plays a pivotal role in both the initiation and propagation of AKI. A key component of the inflammatory cascade is the activation of pathogen-recognition receptors, such as toll-like receptors and NOD-like receptors, which are expressed on immune cells including monocytes, macrophages, and dendritic cells. Upon recognizing pathogen-associated molecular patterns or DAMPs, these receptors activate downstream signaling pathways such as nuclear factor kappa B, culminating in the transcriptional upregulation of proinflammatory cytokines like interleukin-6 (IL-6) and tumor necrosis factor-alpha (TNF-α) [11,12].
In the context of sterile inflammation induced by ischemia, trauma, or nephrotoxins, DAMPs released from necrotic TECs further amplify the innate immune response, recruiting additional inflammatory cells to the injured renal parenchyma [12,13]. This uncontrolled immune activation not only contributes to the initial tubular damage but also drives maladaptive repair and interstitial fibrosis if unresolved.
Among anti-inflammatory strategies, IL-6 signaling blockade has emerged as a promising therapeutic avenue. IL-6 acts through both classical signaling (via membrane-bound IL-6 receptor) and trans-signaling (via soluble IL-6 receptor), both of which contribute to kidney inflammation and injury. Preclinical models have demonstrated that inhibition of IL-6 can attenuate kidney injury, reduce leukocyte infiltration, and improve histological and functional outcomes in AKI [14]. These findings support IL-6 as a key inflammatory mediator and a viable therapeutic target.
Similarly, TNF-α is a central cytokine involved in early tubular injury and the progression to CKD. TNF-α is secreted by activated neutrophils, monocytes, macrophages, and dendritic cells in both soluble and membrane-bound forms. It exerts its effects via tumor necrosis factor receptor 1 (TNFR1) and TNFR2, promoting apoptosis, necroptosis, and profibrotic signaling. Inhibiting TNF-α has been shown in experimental models to reduce inflammation, preserve renal structure, and suppress profibrotic immune cell infiltration, thereby mitigating the AKI-to-CKD transition [15]. Targeting TNF-α may be particularly useful in subtypes of AKI characterized by excessive immune activation, such as sepsis-associated AKI.
While these anticytokine strategies have shown encouraging results in preclinical models, clinical translation remains limited. Challenges include the heterogeneity of AKI etiologies, the timing of therapeutic intervention, and concerns regarding immunosuppression in critically ill patients. Nevertheless, anti-inflammatory therapies, especially those targeting upstream cytokine signaling pathways, represent an important and evolving frontier in AKI management.
Antioxidants and reactive oxygen species inhibitors
While reperfusion is essential to restore blood flow to ischemic tissues, the abrupt reoxygenation paradoxically triggers reperfusion injury, characterized by a surge in ROS generation [16]. In the setting of AKI, especially following IRI, excessive ROS production plays a critical role in promoting oxidative stress, mitochondrial damage, and TEC death [12].
The elevation in ROS levels induces direct oxidative damage to mitochondrial lipids, proteins, and DNA, compromising cellular integrity and bioenergetics [17]. In addition, ROS can trigger the release of cytochrome C from mitochondria, leading to caspase activation and apoptotic cell death. Concurrently, ROS and mitochondrial injury can promote the release of DAMPs, thereby activating immune responses and further amplifying kidney inflammation [7].
Among various antioxidant strategies, resveratrol, a naturally occurring polyphenol, has shown renoprotective effects in preclinical models of AKI. Resveratrol exerts its action by scavenging free radicals, preserving mitochondrial function, and upregulating endogenous antioxidant defense systems such as superoxide dismutase and catalase [8]. These mechanisms contribute to reduced tubular injury and improved histological outcomes in AKI models.
Another potent antioxidant, salvianolic acid B (SAB), derived from Salvia miltiorrhiza, has demonstrated protective effects against oxidative stress-induced kidney injury. SAB reduces the expression of epithelial-mesenchymal transition markers and mitigates renal fibrosis, suggesting its utility in preventing long-term progression of AKI [18] to CKD. Furthermore, SAB formulated as nanoparticles was shown to efficiently restore lysosomal function in proximal tubule cells under oxidative stress conditions, highlighting a novel drug-delivery platform to enhance therapeutic efficacy [19].
In addition, curcumin, a bioactive compound from turmeric, has been extensively studied in animal models of AKI. Curcumin reduces the severity of TEC injury in both septic and ischemic AKI by inhibiting proinflammatory cytokine production, reducing lipid peroxidation, and modulating key intracellular signaling pathways associated with oxidative damage [20,21].
Although these antioxidant compounds showed significant renoprotective effects in preclinical models, clinical evidence remains limited. Most findings are derived from small-scale animal studies, and their translation to human AKI treatment is yet to be validated through robust clinical trials. Nonetheless, the consistency of results across various models underscores the therapeutic potential of dietary and pharmacologic antioxidants, particularly as adjunctive therapies to mitigate oxidative injury in AKI.
Growth factors and nephroprotective agents
The dynamic repair process following AKI is highly dependent on the orchestration of growth factors and signaling molecules that modulate angiogenesis, cell survival, epithelial regeneration, and fibrosis. Several of these pathways, traditionally associated with kidney development or vascular homeostasis, have been actively explored for their potential as therapeutic targets in AKI.
Among these, vascular endothelial growth factor A (VEGFA) plays a central role in angiogenesis and endothelial cell survival. In preclinical models, VEGFA administration during the early phase of AKI has been shown to preserve peritubular capillary networks, mitigate tubular hypoxia, and attenuate renal injury by reducing inflammatory cell infiltration and capillary rarefaction [22,23]. These effects contribute to improved microvascular integrity and functional recovery. Interestingly, administration timing appears critical: while early proangiogenic stimulation is protective, inhibiting VEGFA during the fibrosis phase has been found to reduce maladaptive fibrotic signaling, suggesting that VEGF modulation must be temporally tailored to the stage of AKI [24].
The hypoxia-inducible factor (HIF) pathway is another promising target, especially given the profound tissue hypoxia seen in AKI. HIF-1α and HIF-2α are upregulated in response to hypoxic stress and serve as master regulators of genes involved in angiogenesis, metabolism, erythropoiesis, and cell survival. HIF-1α is predominantly expressed in tubular and glomerular epithelial cells, while HIF-2α localizes to glomerular endothelial cells, peritubular capillaries, and interstitial fibroblast-like cells [25]. Animal studies have shown that earlier and more robust expression of HIF-1α and HIF-2α during the repair phase of AKI correlated with better structural and functional outcomes [26,27].
Based on these findings, HIF prolyl-hydroxylase inhibitors, originally developed for anemia, have been repurposed for renal protection. For example, roxadustat, a HIF stabilizer, significantly reduced tubular injury and inflammatory cell infiltration in mice subjected to IRI [28]. These agents promote endogenous erythropoietin production, improve oxygen delivery, and upregulate cytoprotective genes, making them attractive candidates for AKI therapy.
Another key pathway under investigation is the epidermal growth factor receptor (EGFR) axis. EGFR signaling plays an essential role in kidney development, tubular proliferation, and epithelial regeneration following injury [29]. In AKI models, EGFR expression and activation are upregulated early after IRI. Genetic deletion of EGFR or pharmacological inhibition with erlotinib during this early phase resulted in delayed kidney recovery and worsened histologic injury, supporting the importance of EGFR-mediated signaling in adaptive repair [30,31]. However, prolonged EGFR activation may have adverse effects, such as promoting myofibroblast proliferation, extracellular matrix deposition, and the development of interstitial fibrosis—a hallmark of AKI-to-CKD progression [31]. These findings underscore the dual nature of EGFR signaling, being beneficial in the acute phase of AKI but potentially deleterious if sustained. This temporal biphasic effect mirrors that of VEGFA and highlights the need for precise modulation in future therapies.
Collectively, VEGF, HIF, and EGFR pathways represent key molecular targets that govern vascular preservation, epithelial regeneration, and fibrosis modulation in AKI. While these strategies remain experimental, ongoing research continues to elucidate their roles in adaptive versus maladaptive repair, laying the groundwork for stage-specific or time-sensitive interventions that could revolutionize the pharmacologic treatment of AKI.
Hemodynamic modulators
Maintaining adequate renal perfusion is fundamental to the prevention and management of AKI, particularly in critically ill patients. Ischemia-induced hemodynamic instability is a key contributor to both the onset and progression of AKI, and pharmacologic agents that modulate renal blood flow have garnered increasing attention as potential therapeutic options.
Nitric oxide (NO) plays a critical role in vascular homeostasis by activating soluble guanylate cyclase, leading to increased cyclic guanosine monophosphate and subsequent vasodilation via reduction of intracellular calcium levels [32]. Based on this mechanism, NO has been utilized as an inhaled vasodilator, and agents such as phosphodiesterase type 5 inhibitors (e.g., sildenafil) have been applied clinically in pulmonary hypertension [33,34].
In the context of kidney IRI, sildenafil has demonstrated renoprotective effects in animal models. It increases expression of inducible NO synthase, enhances NO bioavailability, and attenuates tubular damage [35,36]. Conversely, pharmacologic inhibition of NO synthase has been shown to increase kidney susceptibility to ischemic injury, further supporting the protective role of NO signaling in AKI [37].
Clinical evidence also supports this strategy. In cardiac surgery-associated AKI, the use of cardiopulmonary bypass (CPB) contributes to systemic inflammation and renal hypoperfusion, promoting postoperative AKI [38,39]. A randomized trial demonstrated that administration of inhaled NO during CPB significantly reduced the incidence of AKI, possibly through improved renal perfusion and modulation of inflammatory responses [40]. These findings suggest that augmenting NO bioavailability during ischemic stress could serve as a viable approach to mitigate kidney IRI.
Another promising hemodynamic modulator is the angiotensin II type 2 receptor (AT2R), which serves as a functional antagonist to the vasoconstrictive angiotensin II type 1 receptor [41]. AT2R activation induces vasodilation, natriuresis, diuresis, and blood pressure reduction, contributing to renal hemodynamic stability [42–44]. In murine models of renal IRI, the AT2R agonist (compound C21) was reported to reduce the expression of proinflammatory cytokines such as monocyte chemoattractant protein-1, TNF-α, and IL-6, while simultaneously increasing NO production [45,46]. Furthermore, AT2R stimulation increased the population of IL-10–producing CD4+ T cells and regulatory T cells in the early repair phase of AKI, suggesting an additional immunomodulatory benefit [46]. These results highlight AT2R as a novel target that integrates both vascular and immune mechanisms to improve outcomes in AKI.
Vasopressor therapy remains essential in the management of distributive shock-related AKI, where impaired systemic perfusion induces renal ischemia. In the FINNAKI prospective cohort study, a mean arterial pressure (MAP) below 73 mmHg was independently associated with AKI progression in patients with severe sepsis [47], underscoring the importance of maintaining adequate perfusion pressure.
Among vasopressors, vasopressin analogs such as terlipressin showed beneficial effects in patients with hepatorenal syndrome, improving kidney perfusion and delaying the need for dialysis [48]. Additionally, synthetic angiotensin II (Giapreza, La Jolla Pharmaceutical Company) was reported to achieve superior MAP responses compared to placebo in patients with vasodilatory shock in a randomized controlled trial. Notably, patients in the angiotensin II group were more likely to discontinue KRT within 7 days, suggesting a potential renoprotective effect through restoration of systemic vascular resistance and renal perfusion [49].
In summary, hemodynamic modulation remains a cornerstone in AKI therapy, particularly in critical care settings. Strategies aimed at optimizing renal perfusion pressure, enhancing NO signaling, or activating protective renin-angiotensin pathways (such as AT2R) represent promising therapeutic directions. Nevertheless, these interventions must be carefully tailored to the hemodynamic profile and inflammatory status of each patient, and further clinical trials are warranted to define optimal timing, dosing, and target populations.
Cellular and regenerative therapies for acute kidney injury
Stem cell therapy in acute kidney injury
Mesenchymal stromal cells (MSCs) have emerged as a leading candidate for regenerative therapy in AKI due to their potent immunomodulatory, anti-inflammatory, and pro-reparative properties. In preclinical models, MSCs exert their therapeutic effects primarily through paracrine signaling, rather than engraftment or differentiation into renal cells. They modulate the host immune response by suppressing T and B cell proliferation, upregulating regulatory T cells, and downregulating proinflammatory cytokines such as IL-6 and TNF-α [50].
In murine models of IRI, administration of MSCs led to functional improvement in renal parameters, preservation of tubular architecture, and reduced histological injury compared to control or extracellular vesicle (EV)-treated groups [51,52]. These findings support the hypothesis that MSCs promote renal repair by dampening the inflammatory milieu and enhancing epithelial cell recovery in the early post-injury phase.
Despite these compelling preclinical data, translation to human studies has yielded more equivocal results. In the ACT-AKI trial, a multicenter, double-blind, randomized phase II study, patients who developed AKI after cardiac surgery were treated with a single dose of allogeneic MSCs [53]. While the therapy was found to be safe and well-tolerated, it did not result in significant improvement in clinical outcomes, such as time to renal function recovery, need for dialysis, or short-term mortality when compared with placebo. Despite promising preclinical findings, these disappointing clinical results highlight several challenges in stem cell translation for AKI. Potential contributing factors include timing of administration, dosing strategy, cell source and viability, and patient heterogeneity. Furthermore, the inflammatory and hemodynamic context of AKI in humans, particularly post-cardiac surgery, may differ substantially from the controlled settings of animal models.
Nevertheless, the ACT-AKI trial provided valuable insights into the feasibility and safety of MSC-based therapies in critically ill populations and laid the groundwork for improved trial designs. Moving forward, strategies such as enhancing MSC potency through preconditioning, using cell-free MSC-derived products (e.g., exosomes), and selecting appropriate AKI phenotypes may improve the therapeutic efficacy of stem cell-based approaches in clinical settings.
Extracellular vesicle-based therapy in acute kidney injury
EVs represent a promising cell-free alternative to stem cell-based therapies for AKI. EVs are a heterogeneous population of membrane-bound vesicles that are released either from the endosomal compartment (exosomes) or directly from the plasma membrane (ectosomes) [54]. These vesicles carry a rich cargo of proteins, lipids, messenger RNAs, and microRNAs, which mediate intercellular communication and influence recipient cell behavior.
Similar to MSC therapy, EVs derived from stem cells have demonstrated protective effects in preclinical models of AKI. In murine IRI models, administration of MSC-derived EVs led to reduced interstitial fibrosis and lymphocyte infiltration, enhanced proliferation, and suppressed apoptosis of TECs [55–57]. These findings suggest that EVs can recapitulate many of the immunomodulatory and proregenerative functions of parental MSCs, primarily via paracrine signaling mechanisms, without the risks associated with live cell infusion.
Beyond preclinical evidence, early human data also support the feasibility of EV therapy. In a randomized phase II/III clinical trial, 40 patients with CKD received MSC-derived EVs intravenously. The treated group showed a significant improvement in estimated GFR (eGFR) and a reduction in plasma TNF-α levels over a 12-week period compared to the placebo group [58]. While this study was conducted in patients with CKD rather than AKI, it provides encouraging preliminary evidence for the systemic safety and biological activity of EV-based therapies in human kidney disease.
Several challenges remain before EV therapy can be widely adopted in AKI. These include the lack of standardized protocols for EV isolation, purification, quantification, and characterization. Nonetheless, EVs offer several theoretical advantages over whole-cell therapies: they are less immunogenic, easier to store and transport, and pose no risk of uncontrolled proliferation or differentiation. Given their favorable safety profile and demonstrated efficacy in preclinical AKI models, EV-based therapy represents a next-generation regenerative strategy, and continued investigation in AKI-specific clinical trials is warranted.
Biomarker-guided precision medicine
Early diagnostic biomarkers in acute kidney injury
The current consensus criteria for diagnosing AKI are primarily based on changes in serum creatinine and urine output [59]. While serum creatinine remains the most widely used surrogate marker of kidney function, it has significant limitations. Creatinine is a functional biomarker, and its concentration rises only after a substantial reduction in GFR, making it a delayed and insensitive indicator of early or evolving kidney injury [60].
To address these limitations, a variety of structural or stress biomarkers were developed to detect tubular injury earlier than the elevation of serum creatinine. Among these, urinary neutrophil gelatinase-associated lipocalin (NGAL) has emerged as a leading candidate. NGAL is produced by injured TECs and is detectable in urine within hours of injury onset. Its diagnostic and prognostic accuracy has been validated in multiple clinical settings, including cardiac surgery, sepsis, and contrast-induced nephropathy [61,62].
In addition to NGAL, the combination of two G1 cell-cycle arrest markers, tissue inhibitor of metalloproteinases-2 (TIMP-2) and insulin-like growth factor–binding protein 7 (IGFBP7), has shown excellent performance for early risk stratification of patients at high risk for developing severe AKI, particularly in intensive care and perioperative settings. These biomarkers reflect tubular cell stress, rather than overt damage, and are detectable before serum creatinine rises, facilitating earlier identification and potential intervention [63].
The integration of serum creatinine with tubular injury biomarkers (e.g., NGAL, TIMP-2•IGFBP7) has enabled a more improved diagnostic approach, capable of identifying cases of “subclinical AKI.” This term refers to patients who demonstrate molecular or structural evidence of kidney injury in the absence of elevated serum creatinine or oliguria [64]. Subclinical AKI often reflects a disconnection between glomerular filtration and tubular injury, which can occur when a subset of nephrons is damaged but the overall GFR is preserved by compensatory function from remaining healthy nephrons, a phenomenon known as the “kidney functional reserve” [65].
Recognizing subclinical AKI is important, as it has been associated with adverse outcomes including prolonged hospitalization, persistent renal dysfunction, and progression to CKD. Therefore, the use of early damage biomarkers not only improves diagnostic sensitivity but also enhances clinical decision-making by enabling timely risk stratification, early intervention, and potential enrollment into biomarker-driven trials.
Prognostic biomarkers for acute kidney injury outcomes
In addition to early diagnosis, prognostic biomarkers are increasingly recognized as critical tools in predicting the persistence, severity, and long-term outcomes of AKI. These biomarkers help differentiate between transient renal dysfunction and progressive injury that may require KRT or deteriorate to CKD.
The RUBY study represents a landmark investigation in this field, aiming to identify biomarkers associated with persistent stage 3 AKI [66,67]. Among the candidates evaluated, urinary C-C motif chemokine ligand 14 (CCL14) emerged as the strongest predictor of adverse renal outcomes. Patients with CCL14 levels >13 ng/mL had more than a 10-fold increased risk of progressing to persistent stage 3 AKI, initiating KRT, or experiencing mortality, compared to those with lower levels [67]. These findings establish CCL14 as a robust prognostic biomarker capable of guiding clinical decisions, such as identifying high-risk patients who may benefit from early nephrology consultation or enrollment in intervention trials.
Another notable example is derived from a biomarker substudy of the VA NEPHRON-D trial, which evaluated urinary biomarkers to elucidate the renal effects of combination renin-angiotensin-aldosterone system (RAAS) blockade in diabetic kidney disease [68]. Although patients receiving dual RAAS blockade exhibited a higher incidence of AKI as defined by serum creatinine elevation, closer examination of tubular injury and fibrosis markers showed distinct data. Specifically, levels of YKL-40 (also known as CHI3L1) and albuminuria, both associated with chronic tubular damage, declined in the combination therapy group compared to monotherapy [68,69]. These findings suggest that biomarkers can distinguish true structural kidney injury from transient functional or hemodynamic changes that merely manifest as serum creatinine elevation.
By integrating injury-specific biomarkers into routine assessment, clinicians may be better equipped to determine which AKI events are biologically significant (i.e., associated with ongoing tubular damage and higher risk of CKD progression) and which are clinically benign, reflecting reversible hemodynamic shifts without structural harm. As biomarker-guided phenotyping becomes more sophisticated, it may also support risk-adapted management strategies, helping clinicians decide when to intensify monitoring, escalate therapy, or de-escalate unnecessary interventions. Ultimately, prognostic biomarkers hold great promise in transforming AKI management from reactive to proactive, personalized care.
Biomarker-guided therapeutic interventions in acute kidney injury
Beyond their diagnostic and prognostic roles, biomarkers are now being used to guide therapeutic interventions in AKI, ushering in a new era of precision medicine. Randomized controlled trials and large observational studies have demonstrated that biomarker-guided care strategies can improve clinical outcomes by enabling early risk identification and timely intervention.
Notable studies investigated the TIMP-2•IGFBP7 as urinary biomarkers, which reflect early tubular cell stress. In a subanalysis of the ProCESS (Protocol-based Care for Early Septic Shock) trial, patients who showed TIMP-2•IGFBP7-negative at baseline but became positive (>0.3 [ng/mL]2/1,000) after initial resuscitation had a three-fold increased risk of adverse outcomes including stage 3 AKI, need for KRT, or death compared to those who remained negative of these biomarkers [70,71]. Similarly, after cardiac surgery, a TIMP-2•IGFBP7 value >2.0 [ng/mL]2/1,000 at postoperative 4 hours identified patients at the highest risk for subsequent AKI [72]. The predictive utility of TIMP-2•IGFBP7 has also been demonstrated in noncardiac surgical patients, where higher levels correlated with a greater incidence of all-stage AKI [73]. These findings support the clinical utility of biomarker-guided AKI care bundles: early identification of at-risk patients through biomarker testing followed by targeted interventions such as appropriate fluid management, the avoidance of nephrotoxic agents, and intensified monitoring timed to biomarker elevation.
In the broader context of precision medicine, biomarkers are also being employed to tailor therapies according to the subtypes of AKI. For instance, urinary chemokine C-X-C motif chemokine ligand 9, an interferon-γ–inducible chemokine, has emerged as a potential diagnostic biomarker for acute interstitial nephritis, a treatable cause of AKI that is often underdiagnosed [74]. Identification of such disease-specific biomarker profiles may allow targeted immunosuppressive therapy in appropriate patients, avoiding unnecessary exposure in others.
Moreover, in a post-hoc analysis of vasopressor trials in septic shock, a distinct AKI endotype characterized by elevated angiopoietin-2/1 ratio and soluble TNFR-1, markers of endothelial injury, was identified [75]. Remarkably, patients with this biomarker-defined phenotype had significantly improved 90-day survival when treated with norepinephrine plus early vasopressin, as compared to norepinephrine alone. This suggests that specific biological subtypes of AKI respond differently to therapies, and that biomarker-guided phenotyping could help treatment selection in critically ill patients.
Together, these findings underscore the transformative potential of biomarker-guided therapeutic strategies, not only in early intervention and prevention of AKI progression, but also in developing individualized treatment approaches based on molecular signatures. As evidence continues to accumulate, this approach will likely play a central role in future management of AKI.
Multiomics approaches for precision nephrology in acute kidney injury
The goal of precision medicine is to develop preventive and therapeutic strategies tailored to individual biological variability: spanning genes, cellular states, and molecular pathways [76]. This is especially relevant in AKI, a condition marked by substantial heterogeneity in etiology, pathophysiology, and outcomes.
The Kidney Precision Medicine Project (KPMP) represents a flagship initiative in this domain. KPMP is conducting deep molecular profiling of human kidney biopsies from patients with AKI or CKD using a multiomics framework [77]. This includes genomics, transcriptomics, proteomics, epigenomics, metabolomics, and all integrated with digital pathology and clinical metadata. The overarching goal is to construct a “reference kidney atlas” and uncover distinct mechanistic disease subtypes, cellular phenotypes, and molecular circuits that drive kidney injury and repair.
A compelling example from KPMP involves the identification of nuclear factor kappa B subunit 1 positive failed-repair proximal tubule cells, which exhibit a senescence-associated secretory phenotype and are characterized by proinflammatory and profibrotic signaling [78]. Using single-cell transcriptomics, these cells were shown to persist following AKI and are believed to play a key role in the AKI-to-CKD transition. Such findings underscore the value of cell-level resolution in capturing pathological cell states that are invisible in bulk-tissue analysis.
Multiomics approaches also facilitate the discovery of novel biomarkers and therapeutic targets, including pathways related to mitochondrial dysfunction, immune dysregulation, and cell death. For example, the integration of metabolomic and proteomic datasets has highlighted disruptions in mitochondrial energy metabolism, particularly involving the nicotinamide adenine dinucleotide (NAD⁺) biosynthetic pathway. One key finding is the downregulation of peroxisome proliferator-activated receptor gamma coactivator 1-alpha (PGC1α), a master regulator of mitochondrial biogenesis, in AKI. Loss of PGC1α activity results in local renal NAD⁺ depletion, impairing cellular energy homeostasis and repair mechanisms. Conversely, overexpression of PGC1α in experimental models led to increased NAD⁺ levels and was associated with reduced renal injury [79]. Building on these findings, oral niacinamide (a NAD⁺ precursor) supplementation successfully restored renal NAD⁺ pools in AKI mouse models and attenuated injury severity. Notably, a small phase 1 clinical trial in patients undergoing cardiac surgery provided early evidence that niacinamide therapy may confer kidney protection, potentially via restoration of NAD⁺ metabolism [80,81].
The integration of multiomics technologies with clinical phenotyping and histopathologic data represents a transformative advance in precision nephrology. By revealing the molecular architecture of AKI, these approaches are shifting the diagnostic paradigm from syndromic classification toward mechanism-based stratification. Ultimately, this enables the development of tailored diagnostics, biomarker-driven therapies, and personalized interventions that promise to improve patient outcomes in AKI.
New modalities of kidney replacement therapy
Prolonged intermittent kidney replacement therapy
In critically ill patients with AKI, KRT becomes necessary when metabolic demand surpasses kidney clearance capacity or when complications, such as hyperkalemia, volume overload, or metabolic acidosis, require urgent correction [82]. While continuous KRT (CKRT) and intermittent hemodialysis (iHD) remain the mainstays of treatment, prolonged intermittent KRT (PIKRT) has emerged as a hybrid modality offering clinical and logistical advantages.
PIKRT refers to renal replacement delivered over 6 to 12 hours per session, combining features of both CKRT (24-hour slow clearance) and iHD (shorter, high-efficiency treatments). It is conceptually aligned with sustained low-efficiency dialysis, which has traditionally been performed using conventional dialysis machines but can also be implemented via CKRT platforms that incorporate convective clearance through hemofiltration [83]. The strength of PIKRT lies in its hemodynamic stability, which is comparable to CKRT and superior to conventional iHD—making it suitable for patients who are hemodynamically tenuous but do not require 24-hour therapy. Additionally, PIKRT offers adequate solute and fluid clearance, particularly when prescribed and delivered using protocolized approaches. PIKRT is more cost-efficient than CKRT, requiring fewer disposables and shorter machine time, which is especially advantageous in resource-constrained or high-volume intensive care unit (ICU) environments [84,85].
Operationally, PIKRT supports interdepartmental collaboration between ICU teams and dialysis units. It can alleviate staffing burdens during times of nursing shortage, as dialysis unit nurses can assist with therapy planning while ICU nurses oversee on-site treatment delivery. This model proved especially effective during surges in demand, such as the coronavirus disease 2019 pandemic [83,86].
Clinically, PIKRT is well-suited as a transitional therapy, initiated in patients recovering from shock who are no longer dependent on vasopressors but remain unstable for standard iHD. Furthermore, nocturnal PIKRT schedules can free up daytime hours, enabling patient mobilization, imaging, and surgical procedures, thus supporting broader goals of early rehabilitation and ICU liberation [83].
However, caution is warranted in specific high-risk populations. In patients with traumatic brain injury or intracranial hypertension, the potential for osmolar shifts and impaired intracranial pressure control remains a concern. In such cases, CKRT providing more consistent fluid and solute balance is still considered as the preferred modality until further safety data on PIKRT become available.
In summary, PIKRT represents a flexible, resource-conscious alternative to traditional modalities of KRT. It offers comparable efficacy to CKRT in appropriate patients and facilitates operational efficiency in ICU settings. However, patient selection remains crucial, and PIKRT should be integrated into care plans through a multidisciplinary, individualized approach.
Adsorptive therapies for cytokine/endotoxin removal
Extracorporeal blood purification (EBP) therapies have been developed to mitigate dysregulated systemic inflammation in critically ill patients by removing circulating endogenous cytokines and exogenous endotoxins. Commercially available adsorption devices include Toraymyxin (Toray Industries, Inc.), CytoSorb (CytoSorbents Corp.), and Oxiris (Baxter International Inc.), each designed to target specific mediators or pathogen-associated molecules [87].
In vitro and preclinical studies demonstrated that EBP devices effectively remove TNF-α, IL-6, IL-18, and endotoxins from circulation [88]. However, translation into clinical benefit has been inconsistent. Two randomized controlled trials evaluated Oxiris and CytoSorb filters integrated into CPB circuits during high-risk cardiac surgeries [89,90]. In both studies, there was a reduction in inflammatory biomarkers, including TNF-α and IL-18, across treatment arms, but no significant differences in secondary outcomes such as severe AKI, vasopressor requirements, mechanical ventilation duration, ICU length of stay, or mortality [89,90]. Nonetheless, one of these trials showed a significantly lower incidence of postoperative AKI within 7 days in the Oxiris group compared to the control group under standard care [90]. This represents the first randomized evidence that cytokine adsorption during surgery may reduce the occurrence of AKI, although long-term outcomes remained unchanged.
Given the safety profile and biochemical efficacy of EBP techniques, further research is warranted to determine which patient subsets such as those with cytokine storm, high inflammatory burden, or specific immunophenotypes might truly benefit from these therapies. Future trials should focus on biomarker-guided selection of patients and clinically relevant endpoints to fully assess the therapeutic potential of extracorporeal cytokine and endotoxin removal in AKI.
Novel dialysis membranes (high-cutoff, bioengineered filters)
The accumulation of uremic toxins, compounds normally excreted by the kidneys, is a hallmark of kidney failure and contributes to the clinical manifestations of AKI and CKD. These toxins are broadly classified into three categories [91,92]: 1) small water-soluble solutes (e.g., urea, creatinine; <500 Da), 2) protein-bound solutes (e.g., p-cresyl sulfate), and 3) middle molecules (≥ 500 Da), including β2-microglobulin (approximately 12 kDa), light chains (25 kDa), myoglobin (17 kDa), and proinflammatory cytokines like IL-1β (32 kDa), IL-6 (25 kDa), and leptin (16 kDa) [91,93]. Conventional hemodialysis effectively clears small solutes but provides limited removal of protein-bound and middle-molecular-weight toxins, with reduction rates often under 30% to 35% [94,95]. To address this limitation, high-cutoff (HCO) membranes have been developed, allowing for the clearance of larger solutes (>60 kDa) such as cytokines and myoglobin, which are implicated in septic AKI, rhabdomyolysis, and multiple myeloma-associated renal injury [96,97]. Clinical trials of HCO filters in CKRT have shown biochemical improvements including a reduction in IL-6 levels. However, these benefits have not consistently translated into improved clinical outcomes such as reduced vasopressor duration, ICU length of stay, or mortality [98–100]. This discrepancy underscores the need for better patient phenotyping, as the benefits of enhanced middle molecule clearance may be limited to specific AKI subgroups with high inflammatory or toxic burden.
In summary, advanced dialysis membranes offer a theoretically superior method for toxin removal in AKI, particularly in inflammation-driven or toxin-mediated injury; however, stronger evidence linking biochemical improvements to tangible patient outcomes is required for routine application of these membranes.
Artificial intelligence and digital health for acute kidney injury
Artificial intelligence-based early prediction models
Traditional static prediction models for AKI rely on baseline demographic and clinical characteristics, often employing conventional regression techniques to generate risk scores. One such example includes a logistic regression model incorporating four variables (age, baseline eGFR, diabetes mellitus, and heart failure) to predict AKI risk [101]. In contrast, artificial intelligence (AI)-driven dynamic models, particularly those utilizing machine learning, integrate both baseline and real-time clinical data to enhance predictive performance. These models can continuously learn from data inputs such as vital signs, laboratory results, medication exposures, and electronic health record activity, enabling near continuous risk stratification of AKI [102,103]. For instance, a study from the University of Chicago developed a gradient boosted machine model using 59 variables and achieved strong performance in predicting stage 2 AKI within a 48-hour window, validated across internal and external cohorts [104]. Similarly, a recent cohort study developed a machine learning model using 24 clinical variables to predict survival in patients with postoperative AKI, demonstrating robust performance across diverse surgical populations [105].
Despite their promise, a systematic review of 46 machine learning-based AKI prediction models published between 2012 and 2020 revealed significant heterogeneity in model performance, with area under the receiver operating characteristic curve values ranging from 0.49 to 0.99 [103]. Common limitations included poor calibration, limited external validation, and lack of interpretability, which hindered clinical translation. To fully realize their potential, future machine learning models must prioritize methodological rigor, including robust validation, transparent algorithms, and incorporation of clinician-friendly interfaces that can be safely and effectively used at patients’ bedsides.
Artificial intelligence-based hemodynamic monitoring
Hemodynamic instability is a major contributor to AKI, especially in perioperative and intensive care settings. AI tools are increasingly being developed to predict and preempt hypotensive episodes, thereby reducing renal hypoperfusion and subsequent injury.
A commercially available example is the hypotension prediction index (HPI), which leverages arterial waveform analysis to forecast a MAP <65 mmHg lasting ≥1 minute, up to 15 minutes in advance [106]. In a meta-analysis of 12 intraoperative studies, use of the HPI was associated with a reduction in the incidence, duration, and severity of hypotension during surgery [107]. Beyond blood pressure prediction, AI-enhanced echocardiography is gaining traction as a tool for real-time hemodynamic assessment. Machine learning algorithms have been developed to automate measurements of left ventricular ejection fraction, subaortic velocity-time integral, inferior vena cava variability, and mitral annular plane systolic excursion, all of which offer critical insights into fluid responsiveness and cardiac output [106].
While the use of AI-based hemodynamic models for direct prediction or prevention of AKI remains in its early stages, these tools could significantly enhance clinicians’ capacity to detect early hemodynamic deterioration, optimize volume management, and implement preventive strategies, all of which are essential components for mitigating AKI.
Clinical decision support systems for acute kidney injury management
Clinical decision support systems (CDSS) are computerized tools designed to enhance patient care by delivering patient-specific insights, clinical guidelines, and real-time alerts to healthcare professionals at the point of care [108].
In the context of AKI, CDSS has been developed to provide early detection alerts, management recommendations, and consultation prompts, with the goal of improving outcomes through timely recognition and intervention. Several studies showed that CDSS implementation leads to earlier AKI diagnosis, more prompt nephrology referrals, and improved adherence to evidence-based care bundles, which may result in better renal recovery [109–111]. However, the clinical effectiveness of CDSS remains mixed. While CDSS can influence clinician behavior and enhance process metrics, recent randomized controlled trials have reported no significant improvement in hard clinical outcomes such as serum creatinine trajectory, dialysis requirement, or mortality [111–113]. This discrepancy highlights a familiar challenge in digital health: tools that modify care processes do not always translate into measurable outcome improvements.
Despite these limitations, the scope of CDSS in AKI is expanding with exploratory applications in special populations such as kidney transplant recipients, pregnant patients, and individuals receiving iodinated contrast media [110,114–116]. Since timely identification of AKI is especially critical in these high-risk groups, tailored CDSS protocols may offer unique advantages.
The impact of CDSS on AKI care could be substantially enhanced through refinement and integration with other tools such as biomarker data, risk prediction algorithms, and structured AKI care bundles. The promise of CDSS lies not merely in detection but in its potential to change clinician behavior, optimize decision-making, and standardize high-quality care that are critical for improving outcomes of patients at risk for or experiencing AKI.
Future directions and conclusion
AKI remains a major clinical challenge associated with significant morbidity, mortality, and long-term renal complications. Despite advances in early detection and supportive therapies, no pharmacologic intervention has yet been definitively proven to reverse or halt the course of AKI. However, emerging research over the past few years has illuminated several promising directions for the future of AKI management.
First, biomarker-driven precision medicine is poised to redefine how AKI is diagnosed, risk-stratified, and treated. Novel biomarkers such as TIMP-2•IGFBP7, CCL14, and panels derived from multiomics analyses offer the ability to detect injury earlier and differentiate subtypes of AKI with distinct pathophysiological pathways. Future clinical trials must incorporate biomarker stratification to identify which therapies are most effective in specific AKI subphenotypes, moving beyond the traditional “one-size-fits-all” approach.
Second, innovative pharmacological interventions targeting inflammation, oxidative stress, and endothelial dysfunction are undergoing active investigation. Therapies based on HIF stabilizers, stem cell-derived paracrine factors, and immunomodulators hold considerable promise. Importantly, therapeutic strategies should aim not only to attenuate acute injury but also modulate maladaptive repair processes that contribute to CKD progression following AKI.
Third, the field of cellular and regenerative medicine offers transformative possibilities. MSC therapies, EV-based interventions, and kidney organoid models are advancing rapidly, providing novel platforms for both treatment and mechanistic understanding. Although clinical translation remains a challenge due to issues of scalability, manufacturing, and regulatory approval, early human trials suggest that regenerative approaches could be pivotal in promoting renal recovery.
Fourth, innovations in KRT—particularly new CKRT modalities, adsorption technologies, and bioengineered dialysis membranes—are enhancing the ability to support patients through severe AKI episodes. While extracorporeal removal of cytokines or endotoxins and high-cutoff/medium-cutoff membranes have shown promise in improving solute clearance, future research must rigorously determine their impact on patient-centered outcomes such as mortality, dialysis dependence, and quality of life.
Finally, AI and digital health technologies are becoming integral to the AKI landscape. Machine learning models capable of real-time risk prediction, CDSS embedded in electronic health records, and telemedicine platforms for post-AKI care have shown encouraging results. However, widespread adoption of these digital tools will require careful validation, user-centered design, and attention to issues of data privacy and equity.
In conclusion, the next era of AKI management will likely be characterized by early and precise diagnosis, biologically targeted interventions, personalized supportive care, and seamless integration of digital technologies. Wide clinical application of these advances will depend on continued multidisciplinary collaboration, rigorous clinical trials incorporating precision frameworks, and an ongoing commitment to translating scientific discovery into tangible clinical benefit for patients suffering from AKI.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Download Citation
Print
