Association of Urinary Indices and Urine Microscopy with the Progression of Acute Kidney Injury and Mortality in Hospitalized Patients with Severe COVID-19

Introduction

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) belongs to a ssRNA virus family and causes Coronavirus disease 2019 (COVID-19), which is a respiratory infectious disease.^1,2 During the pandemic, AKI became one of the most common complications of COVID-19 and an important cause of morbimortality.^3-5 Currently, AKI diagnosis relies on assessing serum creatinine levels and urine output according to the KDIGO criteria.⁶ While urinary indices and urine microscopy can be useful in differentiating AKI into hemodynamic AKI and acute tubular necrosis (ATN), they could also be employed as complementary parameters for both diagnosis and prognosis.^7-9

In the context of sepsis and COVID-19, urinary biochemistry could be altered depending on the magnitude of the kidney lesion.^10-14 With onset of SARS-CoV-2 viremia and sepsis, arterial vasodilation is associated with the overactivation of the renin-angiotensin system, arginine vasopressin release, cytokine storm, and a prothrombotic state, leading to microclots in the glomerular vessels.^1,4,15 Together, these factors could modify the fractional excretion of sodium (FeNa) and urean nitrogen (FeUN), and blood urea nitrogen (BUN), urine osmolality, as well as the release of renal tubular epithelial (RTE) cells and casts.^1,14-17 Thus, we hypothesize that these urinary biomarkers could predict the worsening of renal function and be associated with mortality.

Given the substantial costs associated with AKI, there is great interest in predicting a more severe AKI progression during hospitalization and assessing mortality.^2,9 Ongoing research is actively exploring methods to anticipate these critical clinical outcomes. Therefore, the present study’s objective is to determine whether urinary indices and urine microscopy are associated with AKI progression and mortality during hospitalization in patients with severe COVID-19.

Materials and Methods

A retrospective observational study was performed using medical files from patients hospitalized at the Internal Medicine department of the Hospital General Dr. Manuel Gea Gonzalez (HGDMGG) from April 2020 to December 2021. We report our findings abiding by the STROBE guidelines for cross-sectional studies (Available at: https://www.strobe-statement.org). The patient inclusion criteria were as follows: age >18 years, serum creatinine measurement taken at least 3 months before admission, confirmed SARS-CoV-2 infection through a positive PCR test, hospitalization at the emergency department for <24 hours, and patients with a 30-day follow-up. Severe COVID-19 infection was defined as clinical signs of dyspnea, respiratory frequency >30/min, oxygen saturation <93%, arterial oxygen partial pressure/fractional inspired oxygen (PaO₂/FiO₂) ratio <300, and/or lung infiltrates >50% of the lung field within 24-48 hours.² The AKI diagnosis was based solely on serum creatinine levels, as urinary output data were not obtained. Therefore, AKI was defined as follows: (i) Stage 1 (increase in serum creatinine level of 0.3 mg/dL within 48 hours or a 1.5-1.9 times increase in serum creatinine level from baseline within 7 days), (ii) Stage 2 (2-2.9 times increase in serum creatinine level within 7 days), and (iii) Stage 3 (≥3 times increase in serum creatinine level within 7 days or initiation of dialysis).⁶ The exclusion criteria included incomplete 30-day follow-up and loss of variables of interest (i.e., creatinine serum levels and AKI stage).

The present study was approved by the HGDMGG Research Committee and Research Ethics Committee (REF 14-17-2022), and patient anonymity was guaranteed according to the 1975 Declaration of Helsinki. Upon medical admission, the patient or a family member signed an informed consent permitting using his/her medical file information for didactic, research, and publication purposes.

AKI progression was defined as the initial AKI stage at admission being KDIGO stage 1, and progression to stage 2 or 3 after hospitalization. Clinical and biochemical data were obtained at admission. The KIDGO guidelines defined chronic kidney disease (CKD) as a glomerular filtration rate (GFR) <60 mL/min/1.73m² for >3 months, structural renal changes, or when the patient self-reported a previous diagnosis.⁶ The GFR was estimated using the 2021 CKD Epidemiology Collaboration equation.

Blood samples from the patients were collected after admission to the emergency department. The measurements were carried out with commercially available standardized methods. Creatinine (serum and urine), BUN (serum and urine), urine sodium, urine potassium, C-reactive protein, and lactic dehydrogenase (LDH) were measured using the DxC 700 AU Chemistry Analyzer (Beckman Coulter, Fullerton CA). Plasma ferritin concentrations were estimated using ELISA (Beckman Coulter DxC 600i, Fullerton, CA). D Dimer levels were estimated using an ACL Top 550 CTS (Werfen Company, Spain).

As part of our institution’s protocol, when a patient meets the KDIGO AKI criteria, fresh urine samples are systematically obtained for AKI assessment. Fresh urine samples were examined after spontaneous voiding when possible. In patients with indwelling bladder catheters, urine was collected from a tube to avoid using old urine that had been sitting in the bag. Internal medicine residents transported the urine samples to the central laboratory within 1 to 2 hours of collection to prevent cell and cast degradation. Afterward, trained laboratory staff at HGDMGG objectively analyzed urine microscopy. To maximize yield, 10 mL of urine was centrifuged for at least 5 minutes at a minimum speed of 1,500 rpm. After removing 9.5 mL of supernatant urine by suction or careful decanting, the test tubes were gently manually agitated, or the sediment was gently suctioned and expelled. A single drop of the urine sediment was placed on a standardized glass slide, and a coverslip was placed over it. The sediment field was examined under low (original magnification ×10) and high power (original magnification ×40) using brightfield or phase-contrast microscopy, with a minimum of 10 fields observed under each magnification.

Following this analysis, the results of urinalysis and urine sediment were reported in the system. Moreover, we utilized the Perazella Score based on the number of granular casts and RTE cells, originally designed to evaluate the accuracy of urine microscopy in differentiating ATN from prerenal AKI.^7,9 Perazella Score 1 was defined as RTE cells 0 and granular casts 0; Perazella Score 2 as RTE cells 0 and granular casts 1 to 5, or RTE cells 1 to 5 and granular casts 0; and Perazella Score 3 as RTE cells 1 to 5 and granular casts 1 to 5, or RTE cells 0 and granular casts 6 to 10, or RTE cells 6 to 20 and granular casts 0.⁷

Results

Overall, records from 1057 hospitalized patients with COVID-19 diagnosis were analyzed. Of them, 644 patients were excluded due to incomplete medical records. A flow diagram of the patient selection process has been presented in Figure 1. Thus, the study included 217 patients with severe COVID-19 infection and AKI stage 1. Table 1 presents the patients’ baseline clinical and biochemical characteristics at admission. The mean age of the subjects was 56.4 ± 14 years, and 32.7% (n= 71) were females. Out of the 217 patients with AKI stage 1, 67.3% (n= 146) remained at stage 1 and 13.8% (n=30), and 18.9% (n=41) progressed to stages 2 and 3, respectively. A total of 144 patients (66.4%) had community-acquired AKI, while 73 patients (33.6%) developed AKI during hospitalization. The 28-day mortality rate was 27.2% for all patients. Clinical and biochemical data of the patients, stratified by AKI progression, have been presented in Table 1. Patients progressing to AKI stages 2 and 3 were significantly older than those who did not progress. They had longer hospital stays and AKI durations, as well as a higher prevalence of mechanical ventilation and vasopressor use. Additionally, these patients exhibited higher levels of BUN, D-dimer, and ferritin. Both groups showed similar median basal serum creatinine. Only patients who progressed to AKI stage 3 required kidney replacement therapy (KRT) during hospitalization. They exhibited lower admission hemoglobin. Supplementary Table 1 presents the characteristics of the patients stratified by 28-day mortality. In general, those who survived were younger and had a lower frequency of mechanical ventilation, vasopressor use, and RRT. Surviving patients exhibited lower levels of serum BUN, D-Dimer, and C-reactive protein and a shorter duration of AKI during hospitalization.

Close

Figure 1:

Patient selection flow diagram. AKI: Acute kidney injury.

Export to PPT

The results of urinary indices and urine microscopy classified by AKI progression have been shown in Table 2. Compared with patients who did not progress, those who did sh more RTE cells and granular casts on urine microscopy, with a higher Perazella Score stage 3. All the urinary indices were not statistically relevant for AKI progression. In addition, we analyzed urinary indices and urine microscopy and their association with 28-day mortality [Supplementary Table 1]. We found no bivariate association between any urinary index and 28-day mortality. However, those who survived had lower granular casts and RTE cells at urine microscopy and had a lower prevalence of a Perazella Score > 1. Supplementary Table 2 shows possible factors associated with prerenal AKI. Overall, the frequency of patients with diarrhea due to COVID-19 and the use of parenteral and/or pleural catheters was low. However, among those who progressed to AKI stage 2-3, there was a significantly higher frequency of both. Finally, in a secondary subanalysis, patients who initially presented with AKI stage 2–3 showed no differences in urine indices or microscopy between survivors and non-survivors [Supplementary Table 3].

The AKI progression distribution [Figure 2] differed among Perazella Score stages (χ²=8.125, p= 0.017). Factors associated with a higher risk of AKI progression were analyzed using Cox regression. Only a Perazella score stage 3 (χ²= 6.910, HR: 2.630, 95% CI [1.279 - 5.407]; p = 0.009) was associated with progression. Interestingly, mechanical ventilation was associated with a reduced risk of progression (χ²= 3.957, HR: 0.470, 95% CI [0.224 - 0.989]; p = 0.047).

Close

Figure 2:

Time-to-event analysis for Perazella score.

Export to PPT

Subsequently, a second Cox regression analysis was performed to evaluate the association between 28-day mortality and the variables of interest. In this model, the Perazella score stage 2 (χ²= 5.246, HR: 3.465, 95% CI [1.196 - 10.034]) and stage 3 (χ²= 7.114, HR: 2.857, 95% CI [1.321 - 6.178]) were associated with increased mortality.

Discussion

AKI is one of the most common complications of COVID-19, widely associated with adverse outcomes.² Our study highlights three critical findings: first, urinary indices such as FeNa, BUN/Cr, and urine osmolarity were not associated with AKI progression or 28-day mortality in patients with severe COVID-19. Second, the Perazella score derived from urine microscopy has proven to be a valuable tool for predicting both AKI progression and mortality, demonstrating its utility as a complementary diagnostic marker. Third, KDIGO criteria, while foundational for AKI diagnosis, may benefit from the incorporation of urine microscopy to provide a more comprehensive assessment of renal injury, particularly in severe COVID-19 cases. This arrangement of diagnostic tools supports integrating urine sediment analysis into clinical practice for improved stratification and prognosis of AKI severity. While these results may not be directly applicable to the entire AKI population due to potential differences in non-COVID AKI cases, urine microscopy could serve as a valuable tool for assessing AKI severity. AKI diagnosis currently relies on assessing serum creatinine levels and urine output according to the KDIGO criteria.⁶ While urinary indices and urine microscopy can serve as useful tools to differentiate AKI into traditional categories, such as prerenal AKI and ATN, they could also be employed as complementary parameters for both diagnosis and prognosis.^7,12-14

FeNa and BUN/Cr ratios are traditional biomarkers distinguishing prerenal AKI from intrinsic causes like ATN.^13,18 FeNa values <1% typically suggest prerenal AKI due to volume depletion or reduced renal perfusion, while higher values indicate impaired tubular function, such as in ATN. Similarly, BUN/Cr ratios above 20:1 indicate prerenal AKI, as urea reabsorption increases with hypoperfusion. However, these indices may have limited diagnostic utility in critically ill patients, including those with COVID-19, due to the multifactorial nature of AKI in this population.^13,18

Urine microscopy has historically been viewed as the ”liquid renal biopsy,” providing insights into the kidneys beyond what serum markers reveal.^12,13 Although prerenal AKI is frequently observed among hospitalized patients, tubular injury induced by ischemic, toxic, direct viral damage, or combined insults became more prevalent during the pandemic.^12,13,17 Consequently, urine sediment analysis has proven valuable in identifying AKI etiologies and progression, particularly in hospitalized patients.^7,9,13 A distinctly prerenal cause of AKI typically manifests as a urinary sediment that is either bland or marked by hyaline casts.¹³ The prerenal causes of AKI in the context of severe COVID-19 infection are believed to stem from dehydration triggered by factors such as fever, nausea, vomiting, or diarrhea, which were common symptoms accompanying COVID-19 during the first waves.^1,15 Additionally, hypotension due to sepsis, dehydration, or myocardial dysfunction contributes to prerenal AKI in COVID-19 patients.^1,12

Another emerging marker for evaluating kidney function and predicting AKI outcomes is the albumin-to-creatinine ratio (ACR), especially in COVID-19 patients.^19,20 Schnabel et al. identified higher ACR levels as an independent predictor of AKI, suggesting that pre-existing renal damage may increase susceptibility to virus-induced injury.²⁰ Similarly, Na et al. reported that severe COVID-19 cases exhibited significantly elevated ACR values, correlating with a higher incidence and severity of AKI.¹⁸ Combining ACR with urine microscopy findings, such as the Perazella score, could enhance the stratification of AKI risk and improve diagnostic accuracy in severe COVID-19 cases.

Histopathological studies have identified ATN as the most common etiology of AKI in patients with COVID-19.^10,11 In this context, proximal tubular cells are particularly susceptible to injury due to their high Angiotensin-converting enzyme 2 expression, which facilitates viral entry.^1,17 Our findings support this by demonstrating that patients with AKI progression exhibited higher numbers of granular casts and RTE cells, markers for tubular injury. Werion et al. described tubular luminal expansion with cellular debris and alterations in brush border membranes as characteristic features of COVID-19-associated kidney injury, alongside moderate proteinuria.¹⁷ Furthermore, evidence of proximal tubule dysfunction, such as elevated urinary β2-microglobulin or albumin, low-molecular-weight proteinuria (70-80%), neutral aminoaciduria (46%), and impaired handling of uric acid (46%) or phosphate (19%), aligns with our observations of tubular damage in progressive AKI.¹⁷ The cytokine storm induced by SARS-CoV-2, with elevations in interleukin 6 and tumor necrosis factor α, likely contributes to hemodynamic instability and direct nephrotoxic effects, further exacerbating ATN in severe cases.^1,10

Tubular cells from different nephron sections exhibit diverse morphologies, and their presence in urine sediment, alongside granular casts, indicates ischemic and/or toxic damage to the renal tubules.^7,11,13 These findings are often associated with intrinsic renal lesions, such as ATN. However, granular casts may also be observed in conditions like acute interstitial nephritis or thrombotic microangiopathy.^7,9,13 Our results demonstrated that patients with AKI progression exhibited significantly higher numbers of granular casts, RTE cells, and Perazella scores >1, emphasizing their role as tubular injury markers. Hernandez-Arroyo et al. similarly reported granular casts in 85% of patients with COVID-19-associated AKI, with a mean Perazella score of 2. At the same time, Morita et al. found that granular cast counts increased with COVID-19 severity.^11,12 These findings collectively support the utility of urine microscopy, particularly the Perazella score, in differentiating prerenal and ATN etiologies and predicting AKI progression in hospitalized COVID-19 patients.

This study has several limitations. First, as a single-center retrospective study conducted in Mexico, the findings may not be generalizable to broader or more diverse populations, necessitating validation in larger, multicenter studies. Second, the lack of renal biopsy data limits our ability to confirm the definitive diagnosis or etiology of AKI, such as distinguishing between ATN and other forms of intrinsic renal injury. Third, variability in urine sediment analysis must be considered; while trained nephrologists and internal medicine residents conducted some analyses, others were performed by laboratory staff. Although the central lab staff was blinded to clinical data to reduce bias, this variability may have influenced the consistency of findings. Despite these limitations, the study provides valuable insights into the severe COVID-19-associated AKI progression and highlights the utility of urine microscopy in this context.

In conclusion, this retrospective cohort study of hospitalized patients with AKI and severe COVID-19 found that urinary indices are not useful for predicting AKI progression or 28-day mortality. However, urine microscopy revealed that patients who experienced AKI progression had a higher number of RTE cells and granular casts, along with elevated Perazella Scores. The study provides valuable insights into the progression and outcomes of severe COVID-19-related AKI. While the findings suggest that combining urine sediment analysis with biochemical biomarkers might enhance precision in predicting AKI outcomes, further validation in larger, multicenter studies is necessary to confirm this approach.