The impact of transient and persistent acute kidney injury in hospital mortality in COVID-19 patients

Abstract Introduction: Acute kidney injury (AKI) has been described in Coronavirus Disease 2019 (COVID-19) patients and is considered a marker of disease severity and a negative prognostic factor for survival. In this study, the authors aimed to study the impact of transient and persistent acute kidney injury (pAKI) on in-hospital mortality in COVID-19 patients. Methods: This was a retrospective observational study of patients hospitalized with COVID-19 in the Department of Medicine of the Centro Hospitalar Universitario Lisboa Norte, Lisbon, Portugal, between March 2020 and August 2020. A multivariate analysis was performed to predict AKI development and in-hospital mortality. Results: Of 544 patients with COVID-19, 330 developed AKI: 166 persistent AKI (pAKI), 164 with transient AKI. AKI patients were older, had more previous comorbidities, had higher need to be medicated with RAAS inhibitors, had higher baseline serum creatine (SCr) (1.60 mg/dL vs 0.87 mg/dL), higher NL ratio, and more severe acidemia on hospital admission, and more frequently required admission in intensive care unit, mechanical ventilation, and vasopressor use. Patients with persistent AKI had higher SCr level (1.71 mg/dL vs 1.25 mg/dL) on hospital admission. In-hospital mortality was 14.0% and it was higher in AKI patients (18.5% vs 7.0%). CKD and serum ferritin were independent predictors of AKI. AKI did not predict mortality, but pAKI was an independent predictor of mortality, as was age and lactate level. Conclusion: pAKI was independently associated with in-hospital mortality in COVID-19 patients but its impact on long-term follow-up remains to be determined.


IntRoductIon
In late December 2019, a surge of atypical severe pneumonia was detected in Wuhan, China. The initial cases were all associated with the local wholesale food market and caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) 1 . This disease became known as Coronavirus Disease 2019 . The infection spread rapidly around the world and it was declared a pandemic by the World Health Organization on March 11, 2020 2 . Bu the end of January 2021, almost 100 million cases of COVID-19 had been reported worldwide, resulting in more than two million deaths 3 .
Most patients present with mild symptoms such as fever, dyspnea, cough, headache, and diarrhea or are asymptomatic. However, more severe cases of pneumonia can lead to acute respiratory distress syndrome (ARDS), septic shock, multiple organ failure, and death 4,5 .
Acute kidney injury (AKI) has been described in COVID-19 patients and is considered a marker of disease severity and a negative prognostic factor for survival 6,7 . However, which factors predict mortality in patients with COVID-19 with AKI is still unknown.
The authors studied the impact of transient and persistent acute kidney injury on in-hospital mortality in COVID-19 patients.

PAtIents And methods
This study was a retrospective analysis of hospitalized patients admitted to the Dedicated Unit for COVID-19 patients (UICIVE) at the Department of Medicine of the Centro Hospitalar Universitario Lisboa Norte (CHULN), in Lisbon, Portugal, between March and August, 2020. The Ethical Committee approved this study in agreement with institutional guidelines and informed consent was waived, given its retrospective and non-interventional nature of the study.
Eligible patients were adults (≥18 years of age) who tested positive for COVID-19 by polymerase chain reaction testing of a nasopharyngeal sample and were admitted at the UICIVE from March 1st to May 31st of 2020. For patients who had multiple qualifying hospital admissions, we included only the first hospitalization.
Exclusion criteria were (a) chronic kidney disease (CKD) patients on renal replacement therapy, (b) patients who underwent renal replacement therapy one week prior to admission, (c) patients who had less than 2 determinations of SCr and (d) patients who were discharged or died less than two days after admission.
AKI that developed during hospital stay was defined according to the Kidney Disease Improving Global Outcomes (KDIGO) classification, using the serum creatinine (SCr) criteria, as follows: Stage 1: increase in SCr by 0.3 mg/dL within 48 hours or a 1.5-1.9-fold increase in SCr from baseline within 7 days; Stage 2: 2.9-fold increase in SCr within 7 days; Stage 3: 3-fold or greater increase in SCr within 7 days or initiation of renal replacement therapy (RRT) 10 . Patients were stratified according to the highest AKI stage reached during their hospital stay. Persistent AKI (pAKI) was defined as continued AKI beyond 48 h according to the KDIGO criteria as per the consensus report of the ADQI 16 Workgroup 11 . Transient AKI (tAKI) was defined as AKI of less than 48 h duration 11 .
Pre-admission SCr (SCr within the previous three months) was considered as baseline value. The estimated glomerular filtration rate (eGFR) for patients with previous baseline SCr was calculated using the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) creatinine equation 12 . When unavailable, baseline SCr was estimated from the MDRD equation, accepting the lower limit of a normal baseline GFR of 75 mL/min/1.73 m 2 , as previously proposed 10 . Presence of CKD was estimated according to the baseline SCR as an eGFR of lower than 60 mL/min/1.73 m 2 13,25 .
Diabetes mellitus was diagnosed according to the American Diabetes Association criteria 14 . Hypertension was diagnosed according to the 2018 European Society of Cardiology (ESC) and European Society of Hypertension Guidelines 15 . COPD comprised emphysema and chronic bronchitis. CVD was considered whenever a history of cerebrovascular disease, chronic heart failure of any cause, cardiac ischemic disease, and/ or peripheral arterial disease was documented. Acidemia was defined as blood gas pH <7.35. N/L ratio at admission was calculated as: neutrophil count / lymphocyte count.
Analyzed outcomes were the development of AKI during the first week of admission and in-hospital mortality.

StatiStical analySiS
Categorical variables were described as the total number and percentage for each category, whereas continuous variables were described as the mean ± standard deviation. Continuous variables were compared with the Student's t-test and categorical variables were compared with the Chi-square test. All variables underwent univariate analysis to determine statistically significant factors that may have contributed to AKI development and in-hospital mortality. Subsequently, significant variables were included in the multivariate analysis using the logistic regression method. Data were reported as odds ratios (ORs) with 95% confidence intervals (CIs). Statistical significance was defined as a P-value <0.05. Statistical analysis was performed with the SPSS for windows statistical software package (version 21.0).

ParticiPantS
From March 1st to August 30st, 544 patients were admitted to UICIVE with a diagnosis of COVID-19. Patients' demographic and clinical data are described in Table 1.
More than 20% of hospitalized patients (n=120) required ICU admission mostly due to respiratory failure, 56 patients fulfilled ARDS criteria, 69 patients required mechanical ventilation, and 18 patients required vasopressor support.

dIscussIon
In this retrospective analysis we report a high incidence of AKI associated with COVID-19, and that persistent AKI was independently associated with mortality.
The studies that reported a lower incidence of AKI on COVID-19 patients, as in the study by Wang et al., who reported an incidence of 5.1% 17 and Cui et al., who reported an incidence of AKI of 18.1% 19 , had younger patients and with fewer comorbidities than our patients.

Characteristic Mortality
Unadjusted OR (95% CI) P-value Adjusted OR (95% CI) P-value The largest cohort of hospitalized patients with COVID-19, which included 5449 patients, reported an AKI incidence of 36.6% which developed mainly early in the course of COVID-19 infection and 46.5% of AKI patients had AKI KDIGO stage 1 6 . This is in accordance with our study, in which most patients developed AKI within the first 48 hours. Most studies have reported that most patients present with lower (KDIGO 1) or higher severity (KDIGO 3) [20][21][22] . In our study, most patients were KDIGO stage 3. There was no difference in AKI severity between transient and persistent AKI. In fact, the prevalence of stage 3 AKI was slightly higher in tAKI, which could be due to dehydration secondary to vomiting or diarrhea, which could resolve rapidly after hospital admission. Some patients probably had AKI before hospital admission, as reflected by a higher SCr at hospital admission compared with baseline SCr.
The etiology of AKI in COVID-19 patients appears to be multifactorial. It can be related to fluid balance disturbances secondary to gastrointestinal symptoms (nausea, vomiting and diarrhea), renal venous congestion secondary to myocardiopathy or acute viral myocarditis 7 , toxic tubular damage following cytokine release syndrome or rhabdomyolysis 8 , direct cytopathic effect of SARS-CoV-2 26 , endothelitis, thrombotic events and intravascular coagulation 27,28 , nephrotoxicity from drugs such as lopinavir/ritonavir, nucleoside analogues, remdesivir, tenofovir, chloroquine phosphate and hydroxychloroquine sulfate 29 , and interaction between SARS-COV-2 and angiotensin II receptors (apparently the patients with low D allele polymorphism have high mortality) 7 .
In our study, CKD and serum ferritin were independent predictors of AKI development. In previous studies, age, CKD, hypoalbuminemia, lymphopenia and neutrophil/lymphocytes ratio, lactate dehydrogenase, d-dimers, C-reactive protein, and need for mechanical ventilation or vasopressor support were reported as independent predictors of AKI devolopment 6,16,20,25,30 . Despite the considerable focus on the use of RAAS inhibitors and severity of COVID-19 and a recent study of Soleimani et al. which reported the association of RAAS inhibitors and AKI development, this was not found in our cohort 31 . Interestingly, in that study, the discontinuation of RAAS was associated with a greater risk of mortality, of invasive ventilation, and of AKI. Another two studies found that RAAS inhibitors were not associated with increased mortality in COVID-19 patients [32][33] and, in contrast, one of those studies found that the discontinuation of RAAS inhibitors was associated to a high mortality of COVID-19 patients 26 .
Other studies have tried to find biomarkers predictive of AKI in COVID- 19 19 . A recent study by Chan et al., in a larger cohort of 3993 hospitalized patients with COVID-19, found that AKI was associated to a higher mortality, as 50% of AKI patients died versus 8% of non-AKI (p<0.001) 22 . Hirsch et al. described an important relationship between AKI and respiratory failure. First, they found that most of the cases of severe AKI occurred in close temporal proximity to intubation and mechanical ventilation and secondly, patients on ventilators had a higher AKI rate and more severe AKI stages 6 . Interestingly, in our cohort of patients the development of acidemia was a predictive factor for AKI but lactate level was not, which possibly reflected disease severity associated to COVID-19, mainly respiratory acidemia. In fact, the mean lactate level of our cohort was below 20 mg/dL. Some studies that only analyzed the mortality associated with AKI in critically ill COVID-19 patients reported higher mortality in those patients. Fominskiy et al., in a study with patients with COVID-19 admitted in the UCI requiring invasive mechanical ventilation, found that patients with AKI had 40% mortality and patients that required continuous RRT had a 50% mortality 20 . Xu et al., in a retrospective multicenter observational study with 671 patients with COVID-19 admitted in the UCI, reported a higher mortality at 28 days in patients with AKI (72 vs 42%, p<0.001) 21 .
In our study, AKI was not predictive of mortality in COVID-19 patients but the persistence of AKI for more than 48 hours was. None of the previously mentioned studies evaluated persistent AKI in mortality. According to previous studies in patients without COVID-19, persistent AKI affected mortality [36][37][38] . To date, we are not aware of any study that evaluated the real impact of AKI duration on vital prognosis of COVID-19 patients. Thus, the question of whether mortality associated with AKI in COVID-19 patients is mainly influenced by the duration of AKI or by AKI development itself remains to be clarified. In fact, this question is extremely important, as Rubin et al. analyzed 77 critically ill patients with COVID-19 and demonstrated that persistent AKI was present in 93% of these patients 39 . Another important point is the follow-up of COVID-19 patients with persistent AKI. These patients should have a reassessment of renal function and cardiovascular risk within 30 days of follow-up, as proposed by Kellum in patients with pAKI not associated with COVID-19 40 .
Previous studies in COVID-19 patients with AKI demonstrated that age, AKI severity, and high SOFA score were independent predictors of mortality 16,18,19,22,25 . In our study, Brescia score ≥2 was not predictive of mortality, maybe because this score does not include age and lactate level, which were factors predictive of mortality in our cohort.
The current study has some noteworthy limitations. First, the single-center retrospective nature limits the generalizability of our results. We did not analyze laboratorial parameters that were available for majority of the patients, such as urinalysis, which might have added important diagnostic and prognostic information. Finally, we did not analyze the exact mechanisms contributing to AKI and mortality.
Nevertheless, our study has some important merits. This is one of the first studies to evaluate the impact of AKI duration on mortality in COVID-19 patients. AKI was defined and stratified according to the KDIGO classification using SCr criteria. Both AKI severity and AKI duration were assessed to evaluate their impact on prognosis. Despite the retrospective design, the variables studied were routinely recorded in daily practice.

conclusIon
To conclude, we demonstrated that AKI was frequent in hospitalized patients with COVID-19 and that persistent AKI was independently associated with in-hospital mortality. Older age and higher lactate levels were also predictors of mortality in this cohort. This study highlights the need to improve early detection of AKI in order to initiate timely therapeutic strategies, as rapid recovery of renal function within 48 hours is associated with a better prognosis. The impact of AKI duration on the longterm follow up of COVID-19 patients remains to be determined. interpretation of data; writing or critical review of the manuscript; and final approval of the version to be published.