Introduction

Acute kidney injury (AKI) is one of the most common extrapulmonary complications in patients hospitalized with COVID-19 infection and is associated with a significant increase in patient morbidity and mortality.¹ Early in the pandemic, reports from China demonstrated AKI rates ranging from 0.3 to 25%, contrasting with rates in the United States, Korea, and European nations, which ranged from 4.7% to as high as 72%.^2–7 The rate of AKI increases with increases in the severity of illness and with the need for escalation to critical care.^3–5,8 Differences in the incidence of AKI between China and other geographical locations are most likely due to differences in the age, race, severity of illness, genetic factors, and comorbidities of cohorts.^1,4,9 Cheng et al prospectively evaluated 701 hospitalized patients with COVID-19 and reported an incidence of AKI of 5.1%.¹⁰ Proteinuria and hematuria were present in 44 and 23% of their patients, respectively.¹⁰ In their study, the high prevalence of proteinuria and hematuria on admission may indicate glomerular involvement or chronic kidney disease (CKD) in those patients. Many patients were hospitalized after having symptoms for approximately 10 days, supporting the theory that direct virally induced tubular injury may have occurred prior to admission.¹⁰ On biopsy, the most common pathologic finding is acute tubular necrosis (ATN), predominantly affecting the proximal tubule. In most series, a marked increase in mortality was observed in patients who developed ATN, with mortality increasing as the severity of AKI increased.^5–7,10 Patients requiring renal replacement therapy (RRT) have mortality rates as high as 70%.^1,11 Of those patients on RRT who survived hospitalization, many were discharged from the hospital still requiring dialysis, with no recovery of renal function.^5–7

Overall, AKI in COVID-19 involves several mediators of injury. Risk factors for the development of AKI include elevated serum creatinine on admission, proteinuria, hematuria, elevated inflammatory markers, male sex, and black race.^1,3,6 Additional risk factors for AKI development are comorbidities such as hypertension and diabetes mellitus, hypoxemia, the need for vasopressor support, and mechanical ventilation (MV).^1,3,6 Initially, there had been considerable controversy regarding the continued use of angiotensin-converting enzyme 2 (ACE2) inhibitors and angiotensin 2 receptor antagonists, with some observations demonstrating an increased risk for AKI. Currently, however, there is overwhelming evidence that the use of these agents does not increase the risk of AKI.¹² Some evidence implicates the use of nonsteroidal anti-inflammatory drugs (NSAIDs) and vancomycin as independent risk factors for AKI development.^12,13 It seems likely that COVID-19-associated AKI is the result of multiple hits. These include the usual hemodynamic and nephrotoxic factors predisposing to ATN, pulmonary–renal crosstalk as observed in acute respiratory distress syndrome (ARDS), tubular and endothelial injury induced by the host response to severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), and direct virally induced damage. The role played by each of these potential hits may vary with the phase of the disease and is phase specific. By way of example, during the viremic phase, viral tropism for renal tubular epithelial cells and podocytes as well as prerenal factors may predominate, whereas during the early pulmonary phase and the later hyperinflammatory phase of the infection, pulmonary–renal crosstalk, cytokine storm–induced shock, endothelial injury with associated vascular dysregulation, and microvascular thrombosis may predominate (Fig. 8.1).

Viral Tropism and Potential Viral Cytopathic Effects

The relatively large amount of cellular ACE2 in renal tissue makes the kidney a potential nonpulmonary target of SARS-CoV-2.¹⁴ Many patients hospitalized with COVID-19 infection present with proteinuria and hematuria on admission.^10,15–17 Whether SARS-CoV-2 kidney infection occurs and results in viral cytopathic effect has been a matter of debate, with some studies demonstrating viral tropism and injury and others not.^18–22 Recent evidence supports viral renal tropism. The presence of SARS-CoV-2 in glomeruli and proximal tubules has been demonstrated in several ways, including immunostaining and hybridization of viral RNA.^21,22 Moreover, autopsy studies have retrieved replication-competent SARS-CoV-2.^23,24Finally, electron microscopic imaging has revealed possible viral spike protein in renal tubular epithelial cells and glomerular visceral epithelial cells (podocytes). Recently, investigators demonstrated evidence of proximal tubular dysfunction, including low-molecular-weight proteinuria, abnormalities in urate and phosphate excretion, and aminoaciduria, in a subset of patients with COVID-19 infection.²⁵ Among patients infected with COVID-19, renal tissue obtained at autopsy or by biopsy has shown injury to glomerular podocytes.^21,26 Collapsing focal segmental glomerulosclerosis (CFSGS), not uncommonly observed with HIV and other viral infections, has been observed in some patients.^21,26 While the incidence of CFSGS is markedly increased in patients of African origin, especially in those with homozygosity in the high-risk ApoL1 allele, CFSGS is not restricted to this population (Fig. 8.2a, b).^26,27

It remains controversial whether tubular and podocyte injury is related to direct viral cytopathic effects or to cytokine-driven toxicity.^23,24 Su et al, employing electron microscopic analysis of postmortem renal tissue, demonstrated cytoplasmic vesicular structures resembling coronavirus viral particles in both proximal tubular epithelial cells and podocytes.²¹ The investigators concluded that these vesicular structures supply direct evidence of viral tropism.²¹ However, direct viral tropism and viral cytopathic effects have been challenged by other investigators, who have shown that clathrin-coated pits in noninfected patients can resemble viral particles.^{18,21,23,28,29}

Employing in situ hybridization and immunohistochemistry, Diao et al demonstrated the presence of viral RNA, nucleocapsid protein, and spike protein deposits in proximal tubular epithelial cells.³⁰ Furthermore, there was infiltration of the interstitium by CD68+ macrophages, along with deposition of the terminal complement complex C5–C9 localized to proximal tubules.³⁰ Immunohistochemistry revealed colocalization of viral nucleocapsid and spike proteins, as well as the hypoxia-inducible proteins prostaglandin synthase and DP2 (prostaglandin D2 receptor-2).³⁰ Thus, SARS-CoV-2 potentially induces hypoxia-mediated tubular injury.³⁰ These findings support the concept of direct viral infection of proximal tubular epithelial cells with complement- and hypoxia-mediated initiation of inflammatory cascades.

In healthy individuals, macrovascular and microvascular homeostasis is maintained by a balance of opposing effects, including vasoconstrictive/vasodilatory, anti-inflammatory/proinflammatory, and antithrombotic/prothrombotic elements.^31,32 The major driver of vascular homeostasis at both a systemic and a local level is the renin–angiotensin system (RAS). Within this system, angiotensin II (AII) is the major effector.^31–34 An intact RAS system is essential for renal microcirculatory autoregulation, glomerular filtration, and overall homeostasis.^34–36 Among other properties, AII is a powerful vasoconstrictive and proinflammatory agent.^34,37,38 These effects^21,39 are mediated by binding of AII to its type 1 cell-surface receptor found on endothelial cells. Cell-surface ACE2 catalyzes the conversion of AII into angiotensin 1-7 (Ang 1-7). Ang 1-7 acting via its G-coupled receptor, MasR, mediates vasodilatory and anti-inflammatory effects, thereby counterbalancing the effects of AII^34,40 (Fig. 8.3a, b). The binding of SARS-CoV-2 to cell-surface ACE2 leads to the downregulation of ACE2. The resulting decreased conversion of AII to Ang 1-7 may produce an imbalance in the amounts of AII and Ang 1-17, with an increased ratio of AII to Ang 1-7 favoring the vasoconstrictive, proinflammatory, and prothrombotic effects of AII.^41–43 Such an imbalance is most likely to occur during the initial viremic phase of COVID-19, and the resulting homeostatic disruption of the renal microcirculation could predispose the kidney to the development of AKI.

In addition to ATN, AKI may also be the result of intrarenal thrombotic microangiopathy. Severe SARS-CoV-2 infection is associated with a dysregulated thromboinflammatory state characterized by thrombosis in microvascular and macrovascular systems.^44–49 Postmortem studies have revealed endotheliitis and extensive thrombosis in the microcirculation of multiple organs.^39,50,51 Several lines of evidence suggest that the microvascular bed may be involved in COVID-19-associated AKI. Varga et al demonstrated endotheliitis and possible viral inclusion bodies in glomerular endothelial cells.³⁹ The postmortem study by Su et al described endothelial cell swelling and foamy degeneration in 5 out of 26 patients with AKI and glomerular fibrin thrombi in 3 out of 26 patients.²¹ In addition, these same investigators described significant erythrocyte stagnation in glomerular capillary loops.²¹ Together, these findings are consistent with virally mediated injury to the vascular endothelium. Similar findings of inflammatory injury, organ failure, and thrombosis have been observed in patients with disease states characterized by elevated levels of proinflammatory cytokines, such as macrophage activation syndrome and hemophagocytic lymphohistiocytic syndromes.^52–54

Prerenal Insults, Other Causes of AKI, and Pulmonary Renal Crosstalk

During the symptomatic phase of COVID-19 infection, the patient may develop fever, loss of appetite, loss of taste and smell, and prominent gastrointestinal symptoms, resulting in a volume-depleted state. In this setting, the concurrent use of NSAIDs, diuretics, and agents that impact the renin–angiotensin–aldosterone system (RAAS), such as ACE2 inhibitors and angiotensin 2 receptor blockers, may lead to a prerenal state.^14,55–59 The risk and severity of prerenal states may be particularly pronounced in patients with preexisting endothelial dysfunction, as occurs with advanced age or in patient with comorbidities such as diabetes mellitus, CKD, or congestive heart failure.^1,3,60,61 Patients with COVID-19 are at increased risk for the development of a wide variety of nephrotoxic AKI.^1,14,58,62 In particular, AKI may occur because of rhabdomyolysis.^21,58

The development of severe AKI usually coincides with the development of respiratory failure during the pulmonary phase of the illness.^55,63 Many patients with COVID-19-associated ARDS (CARDS) demonstrate the L phenotype of lung injury, characterized by low lung water (dry lungs) and low elastance.^64,65 These patients are either euvolemic or volume-depleted, and administration of diuretics in this setting may be detrimental.⁵⁹ In contrast, strategies employing prolonged aggressive volume expansion may lead to volume overload and the development of pulmonary edema, especially in association with cytokine-mediated inflammation and capillary leak. In this setting, volume overload can result in venous congestion, increases in intra-abdominal pressure, renal vein congestion, renal parenchymal interstitial edema, and further endothelial injury.^60,63,66,67 Persistent renal interstitial edema may result in loss of renal autoregulation, both of renal blood flow and of glomerular filtration rate (GFR).^68,69

To prevent the deleterious effects of excessive volume expansion, we recommend clinical assessment of volume status, such as passive leg raising maneuvers, central venous pressure monitoring, assessment of capillary refill, and other available measures. Judicious fluid challenges of 500 to 1,000 mL of crystalloid solution should be administered to patients who are fluid responsive, particularly during the early phases of CARDS when many display the L phenotype. In fluid-responsive patients, volume expansion may be used to manage hypotension. During the later stages of illness, when patients demonstrate decreasing pulmonary compliance and recruiting ability, more restrictive fluid management should be employed, and active diuresis or ultrafiltration may need to be initiated.

Although patients with severe COVID-19 requiring hospitalization may present with renal dysfunction, severe AKI usually develops with the onset of respiratory failure progressing to CARDS and the need for MV. MV causes many physiologic changes, which impact renal hemodynamics and function.^{63,66,68–71} MV increases intrathoracic pressure, leading to an increase in right ventricular afterload and a decrease in right ventricular preload. As a result, patients may develop fluid-responsive hypotension. Increasing intrathoracic pressures, particularly when positive end-expiratory pressure (PEEP) is employed, also results in increase release of antidiuretic hormone, aldosterone, and renin, leading to decreased urine output and GFR.^{63,66,68,70–72} Theoretically, the release of renin induced by increasing intrathoracic pressure may worsen the imbalance between AII and Ang (1-7) in an already dysregulated renal microcirculation.^71,73 As patients develop worsening CARDS with stiff lungs, MV and PEEP can increase intra-abdominal pressure, resulting in renal venous congestion and decreasing GFR.^63,66,68 Lung-protective ventilatory strategies result in permissive hypercapnia, often employed as a lung-protective strategy. It has the potential to impair autoregulation further.^66,74 Finally, worsening hypoxia can induce the release of hypoxia-inducible factors, resulting in proximal tubular injury.⁷²

Crosstalk between the kidneys and lungs can result in worsening failure of both organs. The development of AKI increases proapoptotic pulmonary endothelial cell gene expression and apoptosis. The release of proinflammatory cytokines (interleukin 6 [IL-6], IL-1β, IL-12, IL-10, and granulocyte colony-stimulating factor) and vascular endothelial growth factor (VEGF) enhances pulmonary endothelial permeability, with worsening pulmonary edema.^75–80 Similarly, as in H1N1 influenza–associated ARDS, CARDS can increase the circulating levels of such proinflammatory chemokines and cytokines as VEGF, monocyte chemo-attractant protein-1 (MCP-1), IL-6, and IL-8, which have deleterious downstream renal effects.^77,79

In summary, AKI is a multifactorial/multihit process that in some degree parallels the phases of SARS-CoV-2 infection. During the viremic and early symptomatic phase, viral tropism and cytopathic effects predominate and initiate injury. During the pulmonary phase, cytokine storm, ventilator-associated increases in intrathoracic pressure and hypotension, crosstalk between the lungs and kidneys, widespread inflammation, and thrombotic events exacerbate injury. Other “hits” include viral myositis and possibly rhabdomyolysis.^21,58,81,82 Superimposed nephrotoxic tubular injury has also been reported.^55,57,58,62 As organ failure progresses, patients with COVID-19-associated AKI develop oliguria and severe electrolyte abnormalities.