COVID-19 and Acute Kidney Injury

Abstract

Initial reporting suggested that kidney involvement following COVID-19 infection was uncommon but this is now known not to be the case. Acute kidney injury (AKI) may arise through several mechanisms and complicate up to a quarter of patients hospitalized with COVID-19 infection being associated with an increased risk for both morbidity and death. Mechanisms of injury include direct kidney damage predominantly through tubular injury, although glomerular injury has been reported; the consequences of the treatment of patients with severe hypoxic respiratory failure; secondary infection; and exposure to nephrotoxic drugs. The mainstay of treatment remains the prevention of worsening kidney damage and in some cases they need for renal replacement therapies (RRT). Although the use of other blood purification techniques has been proposed as potential treatments, results to-date have not been definitive.

Key points

• •AKI is common in patients with COVID-19.
• •Increasing age, diabetes, hypertension and CKD are the major risk factors for developing covid-19 associated AKI.
• •No specific therapies are available for treatment of AKI associated with COVID-19 and therefore practitioners should follow accepted local management guidelines.
• •The use of blood purification techniques should be adopted with caution although preliminary data shows promise.
• •The consequences of COVID-19 associated AKI in the longer term are as yet unknown

Introduction

In December 2019, a novel severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) was discovered in Wuhan, China, the rapid spread of which culminated in a global pandemic and critical pressure on health care resources.^{1 , 2} The presentation of COVID-19 varies considerably from asymptomatic individuals and those presenting with mild respiratory symptoms to the more severe spectrum of disease requiring hospitalization. In more severe cases, the development of multi-organ failure may ensue. Overall, mortality from COVID-19 infection is approximately 1% population-wide but may reach 50% or more in those requiring intensive care.^{3 , 4}

Epidemiology of COVID-19-associated acute kidney injury

Initial reports suggested that acute kidney injury (AKI) as defined by the Kidney Disease Improving Global Outcomes (KDIGO) criteria was uncommon following acute COVID-19 infection.^{5 , 6} However, subsequent data from the US and Europe did not support this finding particularly in the critically ill whereby AKI rates in excess of 40% were reported.^{7 , 8} The incidence of C19-AKI continues to demonstrate regional variability among patients hospitalized for COVID-19. For example, a recent international meta-analysis including 49,048 patients found 28.6% of hospitalized individuals with COVID-19 were diagnosed with AKI in Europe and the USA, compared with only 5.5% of inpatients in China.⁹ Similar results have been shown by others,¹⁰ with data from the UK demonstrating C19-AKI rates in intensive care patients of greater than 45% in the period February to July 2020.¹¹ This disparity, may, in part, be explained by the difference in thresholds dictating hospital admission, for example, in China admission of any suspected COVID-19 infection was mandatory, whereas this was not the case in Europe and the USA.⁹

What is clear is that the development of AKI is a poor prognostic factor for individuals with COVID-19 infection with a risk ratio (RR) of 4.6 for mortality when compared with patients with COVID-19 but without AKI.⁹ Cheng and colleagues were able to demonstrate that age over 65, male sex and severe COVID-19 infection were independent risk factors for in-hospital mortality. After adjusting for these, they found a significant increase in mortality with worsening AKI stage, dipstick proteinuria above 1+, and the presence of hematuria.¹²

Risk factors for COVID-19-associated acute kidney injury

Boxes 1 and 2 outline the main risk factors for the development of AKI in patients with COVID-19 infection. Unsurprisingly, there is considerable overlap with factors known to contribute to the development of AKI in patients without COVID-19 infection.¹³ A recent retrospective study from a New York City health system demonstrated a higher incidence of AKI among patients with COVID-19 infection compared with a historical cohort (56.9% vs 25.1%).¹⁴ Factors independently associated with the development of stage 2 or 3 C19-AKI included older age, black race, male sex, diabetes mellitus, nursing home resident, and initial respiratory rate. The median time to development of AKI was 6.5 days in one study in a cohort suffering from severe COVID-19 pneumonitis.¹⁵ Given this delay, predicting those at risk of C19-AKI may influence management and several studies have identified potential candidates for developing C19-AKI including higher levels of ⍺₁-microglobulin excretion.¹⁶

Box 1. Risk factors for the development of C19-AKI.

• •
  Patient factors

  • ○
    Obesity¹⁵
  • ○
    Increasing age^{15 , 122 , 123}
  • ○
    Renal transplant recipient¹⁰
  • ○
    Chronic kidney disease¹²⁴
• •
  Disease factors

  • ○
    Invasive mechanical ventilation^{15 , 124}
  • ○
    Severe COVID-19^{122 , 123}
  • ○
    Nephrotoxic drugs exposure¹²³
  • ○
    Vasopressor requirement¹²⁴

Box 2. Causes of renal impairment in COVID-19 infection.

• •Hypotension/hypovolemia
• •
  Vascular

  • ○
    Macrovascular thrombosis
  • ○
    Microthrombi
  • ○
    Endothelialitis
  • ○
    Thrombotic microangiopathies (atypical hemolytic uremic syndrome, thrombotic thrombocytopenic purpura)
• •Acute tubular necrosis
• •Viral infection of renal parenchyma
• •Collapsing glomerulopathy
• •Glomerulonephritis
• •Drug-induced acute interstitial nephritis
• •Drug-induced acute tubular necrosis

Pathophysiology

Given that AKI, rather than being a distinct phenotype, is often multifactorial in nature, C19-AKI may also be due to a variety of concomitant factors with a number of potential pathophysiological processes implicated. These include direct kidney injury as well as indirect mechanisms leading to C19-AKI (Fig. 1 ).^{17 , 18}

Fig. 1.

Pathogenesis of COVID-19 AKI. The pathogenesis of AKI in patients with COVID-19 (COVID-19 AKI) is likely multifactorial, involving both the direct effects of the SARS-CoV-2 virus on the kidney and the indirect mechanisms resulting from systemic consequences of viral infection or effects of the virus on distant organs including the lung, in addition to mechanisms relating to the management of COVID-19.

(From Acute Disease Quality Initiative 25, www.ADQI.org, CC BY 2.0 (https://creativecommons.org/licenses/by/2.0/).)

Tubular and Glomerular Damage

Acute tubular injury is the most frequent finding on autopsy studies reported in C19-AKI although the findings are often mild despite significant serum creatinine elevation with often evidence of preexisting comorbidities such as hypertensive nephrosclerosis associated with kidney disease.^{19 , 20} In keeping with these findings, proteinuria when demonstrated in C19-AKI has a low molecular weight, pointing to a tubular rather than glomerular injury pattern.²¹ In a multicentre study from France including 47 patients who underwent kidney biopsy in those who had severe AKI, the histopathological pattern was almost exclusively tubular injury, whereas none in their comparator group outside the ICU had evidence of acute tubular injury. Interestingly, in those outside the ICU with proteinuria and/or AKI glomerular collapsing glomerulosclerosis characterized by the segmental collapse of the glomerular tuft, parietal cell hypertrophy, or obliteration of the capillary loop or podocyte, was observed. This phenotype has also been described elsewhere predominantly in individuals with high-risk APOL1 genotypes.²² The APOL1 gene encodes apolipoprotein-1 (apol1), part of high-density lipoprotein complex and genetic variants are common in the peoples of western Africa and carriers of APOL1 variants are at higher risk of chronic kidney disease (CKD) including a 17 times higher risk of developing focal segmental glomerulosclerosis.^{23 , 24}

Viral Tropism in the Kidney

Evidence for direct renal tropism by COVID-19 is controversial. Although a few studies have been able to demonstrate evidence of the presence of viral particles in renal tissue many have not.25, 26, 27 Furthermore, the timing of renal biopsies and autopsy studies are often days to weeks after the onset of the associated AKI, putatively beyond the infectious period of SARS-CoV-2. However, the trimeric spike protein of SARS-CoV-2 is a large molecule at approximately 600 kDa which should preclude its filtration in the healthy glomerulus suggesting the infection of the renal tubular cells, the urothelium or filtration occurring through damaged glomeruli.²⁸ Similarly to the related virus SARS-CoV, SARS-CoV-2 enters cells expressing ACE2 and seems to be its principal mechanism of infectivity.^{29 , 30} The cell-free and macrophage-phagocytosed virus can spread to other organs and infect ACE2-expressing cells at local sites, causing multi-organ injury.³¹ Interestingly, in murine models of ischemic tubular injury ACE2 expression may drop.³² This would theoretically reduce the further influx of viral material into the renal epithelial cells. Moreover, SARS-CoV-2 is endocytosed by the kidney injury molecule-1 (KIM-1) glycoprotein expressed on pulmonary and renal epithelial cells. This represents an alternative entry mechanism for the virus into already damaged epithelial cells, further prolonging infectivity.³³

Complement Activation

The immune/inflammatory response to COVID-19 infection has been implicated in the development of C19-AKI. For example, complement activation has been demonstrated within the kidney with evidence of complement deposition and membrane attack complex in nephron vessels and the tubular basement membrane.³⁴ The activation of the complement cascade has previously been shown to lead to chronic renal inflammation and subsequent tubulointerstitial fibrosis.³⁵ This has led to studies administering the complement C5a inhibitor eculizumab in patients with COVID-19.³⁶ Although preliminary results show promise, these are proof of concept studies with insufficient numbers to demonstrate any significant effects on C19-AKI or the need for RRT.

Cytokine Activation

The inflammatory response to COVID-19 infection has been described as a “cytokine storm” contributing to organ dysfunction. Although poorly defined, cytokine storm syndrome (macrophage activation syndrome) is a life-threatening inflammatory response involving high levels of circulating cytokines and immune cell hyperactivation triggered by multiple mechanisms including sepsis. The proinflammatory cytokines interleukin (IL)-1, IL-6, and tumor necrosis factor-alpha (TNFα) are 3 of the most important cytokines of the innate immune system and are implicated in the development of a cytokine storm. Although in COVID-19 elevated levels of monocytes and macrophages have been demonstrated, particularly in the lungs, and are thought to account for the high levels of IL-1, IL-6, and TNFα observed in some individuals other data suggest that levels of circulating cytokines are often lower in patients with COVID-19 than in patients with acute respiratory distress syndrome (ARDS) due to causes other than COVID-19.37, 38, 39 Nevertheless, monoclonal antibodies against IL-6 have been trialed in patients with the RECOVERY trial demonstrating that the anti-IL-6 monoclonal antibody tocilizumab had a positive effect in moderate COVID-19 pneumonitis, contradicting the results of a smaller multicentre Italian study which found no benefit.^{40 , 41} However, in the critically ill, the data are more contentious. The REMAP-CAP trial demonstrated a reduction in the duration of cardiorespiratory support in an intensive care population when administering tocilizumab or sarilumab, another anti-IL-6 monoclonal antibody.⁴² The excess cytokine production resulting in ARDS maybe associated with disease severity in COVID-19; however, its role in the contribution toward kidney damage is vague. IL-6 has been implicated in the development of AKI given that elevated IL-6 levels may induce renal endothelium cells to secrete other proinflammatory cytokines and chemokines contributing to microvascular dysfunction.⁴³ Moreover, in patients with a greater than 100-fold increase in IL-6 levels increased rates of AKI have been observed although this is not a consistent finding.^{44 , 45}

COVID-19-Associated Coagulopathy

The extrapulmonary clinical manifestations of COVID-19-infection are likely to be related to associate widespread vascular pathology given prominent pulmonary as well as systemic endotheliitis represents a distinguishable and distinct feature of COVID-19 infection.⁴⁶ The prothrombotic nature of COVID-19-associated sepsis has been well described.⁴⁷ Platelet-rich thrombi have been observed in the microvasculature of the heart, brain, kidney, and liver and renal infarction secondary to arterial thrombi have also been described.^{48 , 49} Although prophylactic anticoagulation with low-molecular-weight subcutaneous heparin or enoxaparin (a low-molecular-weight heparin) was shown to provide a mortality benefit in inpatients with COVID-19 from the US Veterans database this finding was confirmed only in moderate COVID-19 pneumonitis, failing to show benefit in the critical care population.^{50 , 51} Furthermore, no effect on AKI rates was observed. Thrombotic microangiopathy characterized by thrombocytopenia and microthrombi which may lead to ischemic tissue injury has been observed both in the pulmonary vasculature and kidneys of patients with severe COVID-19.52, 53, 54 In addition, significant alterations of the von Willebrand factor (VWF)-ADAMTS13 axis in patients with COVID-19 have been observed with an elevated VWF:Ag to ADAMTS13 activity ratio being strongly associated with disease severity. Such an imbalance enhances the hypercoagulable state of patients with COVID-19 and their risk of microthrombosis.⁵⁵

Indirect Kidney Injury

Indirect mechanisms include damage from the therapeutic interventions to manage critical illness as well as the systemic effects of COVID-19 infection (see Fig. 1). Insensible losses leading to hypovolemia and kidney injury, through an increased work of breathing, pyrexia, and gastrointestinal manifestations of infection may be significant in individuals with COVID-19. Moreover, hemodynamic instability and fluid restrictive strategies in patients with ARDS may further exacerbate kidney injury. The relationship between the underlying pulmonary pathology and the kidney may also exacerbate C19-AKI. The consequence of respiratory failure and subsequent hypoxemia on the kidney are well documented with increased renovascular resistance, exacerbated by hypercapnia, leading to a reduction in glomerular filtration rate (GFR).⁵⁶ Furthermore, increases in pulmonary artery and intrathoracic pressure may lead to right ventricular dysfunction and renal venous congestion, effects exacerbated by the use of mechanical ventilation and the application of positive end-expiratory pressure (PEEP).^{57 , 58} This effect may be exaggerated in severe COVID-19, whereby high PEEP levels and high peak and plateau pressures are often required to achieve adequate oxygenation in the context of COVID-19 ARDS.⁵⁹ Administration of nephrotoxic medications, such as antibiotics, may also contribute to the development of AKI in patients already at high risk.

Assessment and investigation of acute kidney injury

Initial assessment of any patient with COVID-19 infection should include a full medical history focusing on those comorbidities and patient factors identified as being associated with higher risk for AKI development.^{13 , 46 , 60} Clinical examination should include the evaluation of volume status and whereby appropriate, hemodynamic assessment. AKI in COVID-19 is defined and classified by the KDIGO criteria-based serum creatinine and urine output changes.⁵ Urinalysis is mandatory as it may help to differentiate various causes of AKI and give an indication of potential glomerular involvement, even whereby alterations in kidney function as defined by KDIGO criteria are absent. This has been observed in one study whereby urinalysis was positive for proteinuria in 65.8% and hematuria in 41.7%, while only 4.7% of the patients met KDIGO criteria for AKI.⁶¹

Biomarkers of Acute Kidney Injury

Novel biomarkers of AKI in the evaluation of C19-AKI have been evaluated in several studies. Neutrophil gelatinase-associated lipocalin (NGAL) is produced in the distal nephron and its synthesis is upregulated in response to kidney injury and may predict the need for RRT requirement and in-hospital mortality.^{62 , 63} A small observational trial of 17 patients with COVID-19-positive admitted to a Japanese ICU showed that elevated urinary NGAL on admission to the ICU was associated with the development of AKI during their stay.⁶⁴ Of note, patients with elevated urinary NGAL had a longer duration of mechanical ventilation and ICU length of stay which may reflect the effect of AKI; however, increased NGAL levels have also been observed in ventilator-associated lung injury.⁶⁵ The type-1 transmembrane glycoprotein KIM-1 is expressed in proximal tubular epithelial cells and has been shown to be associated with AKI development.⁶⁶ A recent study has shown that KIM-1 was significantly elevated in patients with COVID-19 with, compared with those without AKI (P = .005) and was significantly elevated in the patients with COVID-19 that had to be transferred to the ICU.⁶⁷ The use of other biomarkers such as tissue inhibitor of metaloproteinases-2 (TIMP-2) and insulin-like growth factor binding protein-7 (IGFBP-7) has also been proposed in assessing patients with COVID-19 and a recent study demonstrated that the use of this biomarker combination may identify patients with AKI and infection early.^{68 , 69} Also, increased requirement for RRT in individuals with C19-AKI and high levels of [TIMP-2]x[IGFBP-7], has been observed. While a further study in a cohort of 352 patients found admission soluble urokinase plasminogen activator receptor (suPAR) levels to be predictive of in-hospital AKI and the requirement for RRT.^{16 , 70}

Management of C19-acute kidney injury

As the syndrome of C19-AKI has multiple etiologies, no generalized single management plan can be proposed for use in all cases and there is no evidence that the treatment of C-19 AKI should be managed differently to other causes of AKI in hospitalized patients.¹³ Patients admitted with COVID-19 are often intravascularly deplete and fluid resuscitation until euvolemic with vasopressor support whereby required, should be administered according to usual best practice and individualized whereby possible. This is in keeping with recent evidence showing that targeted resuscitation through dynamic hemodynamic assessment reduces the risk of both AKI and respiratory failure.⁷¹ Fluid choice for initial resuscitation should be crystalloid, preferably balanced in those who are critically ill. It has been shown that a composite outcome of death, new RRT, or persistent kidney dysfunction among critically ill patients was reduced with the administration of balanced crystalloids over 0.9% saline. Similar findings in noncritically ill patients were also generated.^{72 , 73} Although subsequent meta-analysis failed to demonstrate a definite benefit for balanced crystalloids over 0.9% saline, other indications, such as hypochloremia or hypernatremia may guide the clinician toward using balanced solutions.⁷⁴ Although recent data from a randomized trial on over 11,000 patients in Brazil did not demonstrate a difference in mortality between saline and balanced solutions these data are not directly transferable to severely ill patients with COVID-19 with AKI.⁷⁵ These data were from patients with lower acuity (median APACHE II score 12 and SOFA 4) and 40% of the patients were not hypotensive. Median volumes of trial fluid administered were low (mean < 1 L/d) and 68% of all patients received fluid before randomization with significant crossover. Furthermore, following randomization, approximately 30% of the total fluid received by day 3 was nonstudy crystalloid. General management should follow the KDIGO guidelines and include glucose monitoring and control, relevant given the potential association between diabetes, insulin resistance, and COVID-19 infection.⁷⁶ Preferably pharmacy lead medication review should be undertaken and pharmacokinetics and drug clearance should be considered as a dose adjustment may be required in AKI for both COVID-19 specific acute therapies as well as other medications. General guidance for nutritional assessment and support in critically ill patients with AKI should be followed especially as COVID-19 infection which is associated with an inflammatory hypercatabolic state, reduced oral intake, and immobilization predisposing to malnutrition and muscle wasting.⁷⁷ Where mechanical ventilation is needed lung-protective low tidal volume ventilation strategies as per general ARDS management should be followed.78, 79, 80 Prone ventilation has been reported as beneficial in patients with COVID-19 pneumonitis, and at present no evidence suggests that any effect on intraabdominal pressure and renal blood flow impact on the risk of AKI.81, 82, 83

COVID-19-specific therapies

Several therapeutic agents have emerged as potentially beneficial in COVID-19 infection. Remdesivir, an inhibitor of the viral RNA-dependent RNA polymerase was studied in the Adaptive COVID-19 Treatment Trial (ACTT-1) and demonstrated that compared with placebo, remdesivir shortened the time to recovery although no significant mortality benefit was seen.⁸⁴ Of note, however, patients with AKI or CKD were excluded and as such, the clinical effect of remdesivir in C19-AKI remains largely unknown. The RECOVERY trial examined the use of the glucocorticoid dexamethasone at a dose of 6 mg/d for up to 10 days. The use of dexamethasone reduced the mortality of hospitalized patients receiving invasive mechanical ventilation or oxygen therapy at the time of randomization with further evidence supporting this approach from meta-analysis of systemic corticosteroids in COVID-19.^{85 , 86} The RECOVERY trial also demonstrated a reduced need for RRT in patients not requiring RRT at the time of randomization, with a risk ratio of 0.61 (95% CI 0.48–0.76). More recently a small pilot study from France has also shown a reduction in AKI in COVID-19 infection.⁸⁷

As outlined, there may be significant systemic inflammation complicating COVID-19 infection. This has led to the use of immunomodulatory therapies such as Tocilizumab, an anti-IL-6 receptor monoclonal antibody. The use of Tocilizumab improved survival in critically ill patients receiving organ support in intensive care with an observed reduction in the need for RRT, while the RECOVERY trial demonstrated a reduced 28-day mortality, probability of discharge at 28 days and reduced progression to the composite outcome of mechanical ventilation and death in those not already ventilated in hospitalized patients with hypoxia and systemic inflammation (defined as CRP ≥ 75 mg/L).^{40 , 42} Although other therapies have been considered for COVID-19 at the time of writing there is insufficient evidence of clinical efficacy in the management ofC19-AKI.

Renal replacement therapy

The indications for RRT in C19-AKI do not differ from AKI complicating other conditions and should consider both patient and disease factors. However, resource limitations in the setting of the pandemic may require further consideration of potential for benefit at the individual patient level.¹³ Vascular access should be through the internal jugular and femoral sites with ultrasound directed placement as this increases success rate and reduces complications.^{5 , 13 , 88 , 89} Internal jugular access may be associated with lower infection rates compared with femoral in patients with elevated body mass index, but left internal jugular access is associated with higher rates of vascular access dysfunction.^{90 , 91} Internal jugular access may also be preferable in patients whereby prone ventilation is anticipated.¹³ In the absence of an emergent indication, multiple trials have failed to demonstrate any impact on mortality using either early/accelerated versus delayed initiation of RRT, and indeed, premature start may be associated with adverse outcomes.^{92 , 93} However, it must be remembered that the ELAIN trial and more recently, the AKIKI2 trial found that an overly delayed strategy may be associated with harm (references) This implies that the exact timing of initiation of RRT in COVID-19 should be on a patient by patient basis considering the full clinical context, not just the degree of kidney dysfunction as measured by conventional means.^{94 , 95} Use of maximal medical management, whereby safe, including loop diuretics, potassium binders, and sodium bicarbonate should be considered before committing to RRT, especially whereby resources may be limited in surge situations. Continuous RRT, prolonged intermittent renal replacement therapy (PIRRT), and intermittent hemodialysis may all be considered depending on local familiarity and resources given there is no evidence for superior outcomes with any one modality of RRT over another. However, continuous RRT may allow more fluid removal and tends to cause less hemodynamic instability, which may be a consideration in critically ill patients with COVID-19.⁹⁶ During peak admissions associated with the COVID-19 pandemic, the demand for ICU care and RRT was stretched, and shortages of RRT devices, disposables, and dialysis fluid were described.⁹⁷ Approaches to mitigate this included moderating RRT intensity to conserve fluids, running accelerated high clearance RRT or PIRRT to allow machine sharing, in-house preparation of dialysis fluid, and early transition to IHD.98, 99, 100 Peritoneal dialysis (PD) is rarely used in critical care due to concerns regarding unpredictable fluid balance, variable dialysis adequacy, potential peritoneal infection, and compromised ventilation due to diaphragmatic restriction. Also from a practical standpoint, there may be substantial challenges involved in delivering PD to patients who are ventilated in the prone position and intra-abdominal pressure may be increased. Despite these reservations during surge conditions PD was successfully implemented under certain conditions.101, 102, 103

Anticoagulation is recommended for RRT unless contra-indicated especially given the proinflammatory and prothrombotic nature of COVID-19 infection. This is especially relevant given that reduced RRT circuit life has been widely reported which has implications on the dose delivered as well as increasing the consumption of consumables and potential exposure of staff to infection risk.^{13 , 104 , 105} Regional citrate anticoagulation has been shown to be superior to systemic heparin for anticoagulation with RRT and although some centers have reported reduced effectiveness of citrate in patients with COVID-19 others have suggested superiority over heparin.^{106 , 107} Choice of anticoagulation regime is likely to be center dependent, but it is important that if issues with filter lifespan are identified a stepwise approach to optimizing anticoagulation is taken, with the consideration of a shift in modality to IHD, PIRRT, or acute peritoneal dialysis if possible if issues persist.

Extracorporeal blood purification

There has been considerable interest in the use of extracorporeal blood purification (EBP) therapies to modify or remove circulating inflammatory mediators with the aim of mitigating organ damage, including AKI.¹⁰⁸ Given the inflammatory profile associated with COVID-19 this provides the rationale for such a treatment, but, as discussed earlier, the degree of cytokine production is generally not as pronounced in COVID-19 infection as in other causes of ARDS or bacterial sepsis which may confound this approach. Despite these reservations, several extracorporeal blood purification filters received emergency use authorization from the US FDA for the treatment of severe COVID-19 pneumonia in patients with respiratory failure not specifically for AKI. To-date, several single-center case series have been produced with variable results. In a time-series analysis of 44 consecutive COVID-19 cases treated with the AN69ST (oXiris) cytokine adsorbing hemodiafilter a decrease in acute phase proteins was demonstrated with a reduction in IL-6 levels and an observed mortality of 36.3% across the cohort.¹⁰⁹ In a further study on 5 patients using the AN69ST filter a reduction in cytokines levels and improvement of hemodynamic status was also observed and similarly in 37 patients in a further single-center study a reduction in expected mortality was also seen (8.3% compared with the expected rate as calculated by APACHE IV).^{110 , 111} Several studies have reported benefits on the use of the hemadsorption filter Cytosorb whereby improvements in catecholamine use as well as decreases in inflammatory markers were seen.¹¹² A multi-centre study enrolled 61 patients with COVID-19 treated with the Seraph 100 microbind affinity sorbent hemoperfusion filter which contains polyethylene beads coated with immobilized heparin and allows for broad-spectrum pathogen removal.¹¹³ An overall mortality of 37.3% was observed compared with 67.4% in historical controls (P = .003). In addition, multivariable logistic regression analysis yielded an odds ratio of 0.27 (95% confidence interval 0.09–0.79, P = .016) in favor of the treatment. However, there are important caveats principally that this was a retrospective analysis with differing local criteria for initiating extracorporeal blood purification therapy and hence the potential for significant selection bias.¹¹⁴ However, not all EBP interventions have demonstrated such positive findings as a recent randomized controlled pilot study examined cytokine adsorption during the first 72 h after the initiation of venovenous ECMO in severe COVID-19 demonstrates. Of the 34 patients assessed for eligibility, 17 (50%) were treated with cytokine adsorption but cytokine adsorption did not result in reduced interleukin-6 concentrations after 72 h, compared with the control group. One patient in each group died before 72 h. Survival after 30 days was 3 (18%) of 17 with cytokine adsorption and 13 (76%) of 17 without cytokine adsorption (P = 0·0016) These findings were in contrast with the hypothesis of a treatment benefit for patients in the cytokine adsorption group although the study was not powered to detect a mortality benefit, the results are of interest.^{115 , 116} These variable results show that although control of inflammation in the critically ill through immunomodulation may hold promise, more data from large, multicentre trials with robust yet pragmatic endpoints are required.

Long-term outcomes

Early observational data suggest that approximately 50% of patients who have had AKI associated with COVID-19 infection had not recovered to baseline by the time of hospital discharge.^{61 , 117} Similarly, emerging data suggest that C19-AKI may be associated with an increased decline in GFR postdischarge than patients who had AKI from other causes.¹¹⁸ Data from New York showed that in survivors from AKI who required RRT, 30.6% remained dialysis dependent on discharge with a history of CKD being the only independent risk factor for this association (adjusted OR, 9.3 [95% CI, 2.3–37.8]).¹¹⁹ In another US-based cohort study from 67 hospitals, 1 in 5 patients developed AKI-RRT, 63% of whom died during hospitalization. Among those who survived to discharge, 1 in 3 remained dialysis dependent at discharge, and 1 in 6 remained dialysis dependent 60 days after ICU admission.¹²⁰ Similar results were observed in a German study whereby 67% of patients who had required RRT were dialysis free at hospital discharge and encouragingly at a mean follow-up of 151 days over 90% were dialysis independent.¹²¹

Summary

Despite early reports, AKI complicating COVID-19 infection is common in hospitalized patients. The development of AKI increases the risk of mortality significantly and therefore, efforts should be made to minimize the occurrence of AKI and limit the progression to more severe stages. Treatment should follow accepted practice guidelines for the general management of AKI given the heterogeneous nature of the potential causes of AKI in this group. To-date, no specific therapies have demonstrated a benefit for patients with C19-AKI. Extracorporeal blood therapies show promise but should be adopted with caution and preferably within a clinical trial. Long-term outcomes from C19-AKI may not be as poor as initially suggested although data are still accumulating.

Clinics care points

• •COVID-19 associated AKI resembles AKI due to other causes and therefore these should be excluded where possible.
• •Although no specific therapies are available for the treatment of AKI associated with COVID-19 practitioners should follow accepted management guidelines for the treatment of AKI.
• •Where possible follow up of patients who have sustained severe AKI should occur to minimise longer term sequelae such as CKD.