Presentation and outcomes of chronic kidney disease patients with COVID-19

Abstract Introduction: COVID-19 is currently a global health issue and an important cause of mortality. Chronic kidney disease (CKD) is one of the risk factors for infection, morbidity and mortality by SARS-CoV-2. In our study, we aimed to evaluate the clinical presentation and outcomes of CKD patients with COVID-19, as well as identify predictors of mortality. Methods: This was a retrospective study of CKD patients admitted in a tertiary-care Portuguese hospital between March and August of 2020. Variables were submitted to univariate and multivariate analysis to determine factors predictive of in-hospital mortality. Results: 130 CKD patients were analyzed (median age 73.9 years, male 60.0%). Hypertension (81.5%), cardiovascular disease (36.2%), and diabetes (54.6%) were frequent conditions. Cough, dyspnea, fever and respiratory failure were also common. Almost 60% had anemia, 50% hypoalbuminemia, 13.8% hyperlactacidemia and 17% acidemia. Mean serum ferritin was 1531 µg/L, mean CRP 8.3 mg/dL and mean LDH 336.9 U/L. Most patients were treated with lopinavir/ritonavir, hydroxychloroquine or corticosteroids and only 2 with remdesivir. Eighty percent had acute kidney injury and 16.2% required intensive care unit admission. The 34 patients who died were older and more likely to have heart failure. They had higher neutrophils/lymphocytes ratio, ferritin, lactate, and LDH levels. Multivariate analysis identified an association between older age [OR 1.1 (CI 1.01-1.24), p=0.027], higher ferritin [OR 1.0 (CI 1.00-1.00), p=0.009] and higher LDH levels [OR 1.0 (CI 1.00-1.01), p=0.014] and mortality. Conclusion: In our cohort of CKD patients with COVID-19, older age, higher ferritin, and higher LDH levels were independent risk factors for mortality.


IntRoductIon
Coronavirus disease (COVID- 19) was first discovered in Wuhan, Hubei Province, China, in December 2019 and declared a global pandemic by March 11th 2020 1 ,with 88,828,328 confirmed cases globally by January 11th 2021 2 . Illness severity varies greatly, ranging from asymptomatic and mild clinical course (80%) to severe disease requiring ventilatory support (3-5%) 3 . According to the literature, comorbidities such as cardiovascular disease, hypertension, obesity, diabetes mellitus, cancer and chronic lung disease may contribute to severe disease 4 .
The role of chronic kidney disease (CKD) in COVID-19 was not clear initially. On the one hand, immunosuppression associated to this condition might attenuate the hyperinflammatory state described in COVID-19 5 . On the other hand, immunity dysfunction and high prevalence of comorbidities (including cardiovascular disease and diabetes mellitus) may contribute to a worse clinical course 5,6 . Recent studies have associated CKD with severe COVID-19, higher risk of hospitalization, and higher mortality 7-12 . In summary, given the frequent contact of CKD patients with medical care, which exposes these patients to a higher risk of SARS-CoV-2 infection 13 and worse disease course, we sought to evaluate the clinical presentation and outcomes of CKD patients with COVID-19, as well as identify predictors of mortality.

mAteRIAls And methods
This study is a retrospective analysis of patients admitted between March and August of 2020 in a unit dedicated for COVID-19 patients in the Centro Hospitalar Universitário Lisboa Norte (CHULN) in Lisbon, Portugal named Unidade de Internamento de Contingência de Infeção Viral Emergente (UICIVE). The Ethical Committee approved the study, in agreement with institutional guidelines. Informed consent was waived, given the retrospective and noninterventional nature of the study.

ParticiPants
All adult patients (≥18 years of age) with chronic kidney disease (CKD) and a positive SARS-CoV-2 real time polymerase chain reaction test from nasopharyngeal exudate sample admitted in UICIVE from March 1st to August 31st of 2020 were eligible. For patients who had multiple qualifying hospital admissions, only the first hospitalization was considered. Patients who had less than 2 determinations of serum creatinine (SCr) (a) and patients who were discharged or died less than two days after admission (b) were excluded from the study.

Variables and OutcOmes
Data was obtained from individual electronic clinical records. The following variables were collected: demographic characteristics (age, gender); clinical presentation (cough, fever, dyspnea and respiratory failure); comorbidities [CKD (and previous renal replacement treatment (RRT) need), diabetes mellitus, hypertension, cardiovascular disease (CVD), heart failure, chronic obstructive pulmonary disease (COPD), cirrhosis and/or active malignancy]; current treatment with angiotensin-converting enzyme inhibitors or angiotensin receptor blockers (RAAS inhibitors); disease severity according to Brescia-COVID Respiratory Severity Scale (BCRSS) at admission 14 ; laboratory values at admission [serum hemoglobin, hematocrit, neutrophil and lymphocyte count and their ratio (N/L ratio), serum albumin, serum ferritin, SCr (baseline and admission), C-reactive protein (CRP), lactic acid dehydrogenase (LDH), serum sodium, serum chloride, prothrombin time (TP), activated partial thromboplastin time (aPTT), arterial blood gas and pH analysis, and serum lactate]; exposure to nephrotoxins during the first week of admission [non-steroidal anti-inflammatory drugs (NSAIDS), radiocontrast, vancomycin, aminoglycosides]; need for intensive care unit (ICU) admission, mechanical ventilation and vasopressors; acute kidney injury (AKI) development during hospitalization; need for RRT; treatment options used for COVID-19 (hydroxychloroquine, lopinavir/ ritonavir, corticosteroids, tocilizumab); length of stay (LOS) and in-hospital mortality. definitiOns COVID-19 diagnosis was established according to the World Health Organization provisional guidelines 15 .
Baseline SCr was defined as a pre-admission value within the previous three months. The estimated glomerular filtration rate (eGFR) of non-dialysis patients was calculated using the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) creatinine equation 16 . Presence of CKD was defined as an eGFR lower than 60 mL/min/1.73m 2 . AKI was defined and stratified using SCr criteria of Kidney Disease Improving Global Outcomes (KDIGO) classification 17 .
Diabetes mellitus was defined in accordance with the American Diabetes Association Guidelines 18 . Arterial hypertension was diagnosed according to the European Society of Cardiology and European Society of Hypertension Guidelines 19 . COPD included emphysema and chronic bronchitis. If a history of cerebrovascular disease, chronic heart failure of any cause, ischemic heart disease and/or peripheral arterial disease was documented, CVD was considered. Acidemia was defined as blood gas pH below 7.35.

statistical methOds
Categorical variables were described as total number and percentage of each category, while continuous variables were described as mean ± standard deviation. Continuous variables were compared using Student's t-test and categorical variables were compared using Chi-square test.
All variables were submitted to univariate analysis to find statistically significant factors that could be predictive of in-hospital mortality. Subsequently, variables with a significant association underwent multivariate analysis using the Cox-logistic regression method. Data are reported as odds ratios (OR) with 95% confidence intervals (CI). Statistical significance was established as a P-value lower than 0.05. The statistical software SPSS for Windows (version 21.0) was used for data analysis.

Results
A total of 130 CKD patients were admitted to UICIVE with a diagnosis of COVID-19.

baseline characteristics
Baseline characteristics of this cohort are described in Table 1. Mean age was 73.9 ±12.2 years and the majority of patients were male (60.0%). There was a large prevalence of hypertensive (81.5%), CVD (54.6%), heart failure (51.5%) and diabetic (36.2%) patients. Twenty-four patients (18.5%) were on hemodialysis and 3 patients (2.3%) had a kidney transplant. Forty five percent of the patients were taking RAAS inhibitors. Mean baseline serum creatinine (SCr) was 1.7±0.9 mg/dL and mean baseline eGFR was 42.5±15.6 mL/min/1.73m 2 .
clinical PresentatiOn and labOratOrial findings at admissiOn The main clinical presentation was cough and dyspnea in 46.2% of patients each, respiratory failure in 43.8%, and fever in 37.7% of patients. Thirteen percent of patients had a Brescia score greater than 2.
One hundred and five patients (80.8%) developed AKI during hospital stay and 28 required dialysis, although only twelve percent of patients were exposed to nephrotoxins during hospitalization.
in-hOsPital mOrtality LOS was 35.0±45.9 days and 26.2% of patients died in hospital (n=34). Patient characteristics according to in-hospital mortality are described in Table 1

dIscussIon
After more than a year since the first COVID-19 reported case, our knowledge of the disease characteristics in the general population have grown substantially. However, the clinical presentation of SARS-CoV-2 infection in CKD patients is not as described in the literature, although there seems to be an association of this comorbidity with mortality and severe presentation 11,12 .
In our study, cough, dyspnea, and fever were frequent symptoms, which is in line with what has been previously reported in CKD patients (cough in 35-69%, dyspnea in 6.25-57%, and fever in 43-71%) 3,13,[20][21][22][23][24] , with the exception of the ERACODA series that documented considerably lower incidences (cough in 9.55%, dyspnea in 6.65%, and fever in 11%). This could be explained by the inclusion of CKD patients followed and treated as outpatients and, therefore, more likely to be asymptomatic.
Our findings are also in line with what is seen in the general population 25-28 . Regarding laboratorial findings at hospital admission, we found that more than half of our patients had anemia, although with a higher mean hemoglobin than reported in previous cohorts 3,21,24,29 . As these studies included mainly hemodialysis patients, the insufficient endogenous production of erythropoetin and chronic inflammatory status, as well as the exposure to heparin during treatment may explain the discrepancy 21 . Hypoalbuminemia was common in our cohort, but with lower mean values than previously described (3.4-3.7 g/dL) 3,9,22,24 . This might indicate a worse nutritional status of our cohort. As expected, inflammation markers (ferritin and CRP) and LDH were elevated 3,4,9,20,[22][23][24]26,29 .
During the first half of our investigation period, national guidelines recommended treating COVID-19 patients with pneumonia and/or respiratory failure with hydroxychloroquine and/or lopinavir/ ritonavir (or remdesivir in the ICU setting) 30 . This, in conjunction with the potential risks associated with the administration of remdesivir in patients with a eGFR lower than 30 mL/min, explains the frequencies of treatment described in our cohort 5 .
Several studies have recorded a broad range of AKI incidence in COVID-19 in the general population (0.5-46%) 31 . CKD is a risk factor for AKI, which explains why it was frequent in our patients 31 . The high percentage of lopinavir/ritonavir use and severe presentations could have also contributed to this fact 32 . As this cohort included CKD patients already under RRT, the effect of the need for de novo RRT in mortality might have been underestimated.
There is a wide range of ICU admission (12-39.4%) 4,9,23 and mechanical ventilation (4-31.5%) 4,9,13,[22][23][24]29 rates in the literature depending on the series and the relative percentage of CKD patients under RRT. In our cohort, 16% of CKD patients where admitted in ICU, 8.5% were mechanically ventilated, and 3.8% needed vasopressors. These relatively low percentages could be explained by the older age of our patients and the resulting higher burden of comorbidities and clinical frailty as well as by the inclusion of CKD patients who were already under RRT at admission, possibly underestimating the effect of mechanical ventilation on mortality.
We found that the deceased were significantly older, had higher NL ratios and ferritin, lactate and LDH levels. There was also an association between death and heart failure. However, after a multivariate analysis, only older age, higher ferritin, and higher LDH levels were independent risk factors for mortality. Older age is a well-documented risk factor for infection, morbidity, and mortality by SARS-CoV-2, as it negatively affects lung function and immunity response 8,10 . SARS-CoV-2 infection induces a pro-inflammatory state, which can lead to cytokine storm response and subsequently to secondary tissue damage and a poorer prognosis 37,38 . As ferritin acts as an acute phase protein and LDH as a marker for tissue damage, we hypothesize there may be a connection between their elevation and hyperinflammatory response in COVID-19.
Chawki S. et al. also found reduced mortality in CKD patients treated with RAAS inhibitors 33 . This is a controversial subject in the literature, with description of potential upregulation of angiotensin-converting enzyme 2 receptor, as well as its blockage 31,33 . Most studies in the general population have failed to show an association between RAAS and mortality, and professional societies continue recommending its use 31,33 . In our cohort no association was found.
We must take into account several limitations of our study. First, this was a single-center, retrospective study, which limits generalization of results. The small size of our cohort and the lack of some laboratorial results may have compromised, at least in part, our conclusions. Additionally, causes of CKD were not assessed. Regardless of these potential biases, there are some strengths worth noting. The most important is the study population, which included all stages of CKD, not only end-stage kidney disease as in most published series. Furthermore, to the best of our knowledge, this is the largest study of COVID-19 in CKD patients in Portugal 39,40 .
To conclude, in this cohort of CKD patients with COVID-19, older age and higher ferritin and LDH levels at admission were independent risk factors for mortality, suggesting their potential use as predictors of poorer prognosis.