Serum uric acid, disease severity and outcomes in COVID-19

Abstract

Background: The severity of coronavirus disease 2019 (COVID-19) is highly variable between individuals, ranging from asymptomatic infection to critical disease with acute respiratory distress syndrome requiring mechanical ventilation. Such variability stresses the need for novel biomarkers associated with disease outcome. As SARS-CoV-2 infection causes a kidney proximal tubule dysfunction with urinary loss of uric acid, we hypothesized that low serum levels of uric acid (hypouricemia) may be associated with severity and outcome of COVID-19.

Methods: In a retrospective study using two independent cohorts, we investigated and validated the prevalence, kinetics and clinical correlates of hypouricemia among patients hospitalized with COVID-19 to a large academic hospital in Brussels, Belgium. Survival analyses using Cox regression and a competing risk approach assessed the time to mechanical ventilation and/or death. Confocal microscopy assessed the expression of urate transporter URAT1 in kidney proximal tubule cells from patients who died from COVID-19.

Results: The discovery and validation cohorts included 192 and 325 patients hospitalized with COVID-19, respectively. Out of the 517 patients, 274 (53%) had severe and 92 (18%) critical COVID-19. In both cohorts, the prevalence of hypouricemia increased from 6% upon admission to 20% within the first days of hospitalization for COVID-19, contrasting with a very rare occurrence (< 1%) before hospitalization for COVID-19. During a median (interquartile range) follow-up of 148 days (50-168), 61 (12%) patients required mechanical ventilation and 93 (18%) died. In both cohorts considered separately and in pooled analyses, low serum levels of uric acid were strongly associated with disease severity (linear trend, P < 0.001) and with progression to death and respiratory failure requiring mechanical ventilation in Cox (adjusted hazard ratio 5.3, 95% confidence interval 3.6-7.8, P < 0.001) or competing risks (adjusted hazard ratio 20.8, 95% confidence interval 10.4-41.4, P < 0.001) models. At the structural level, kidneys from patients with COVID-19 showed a major reduction in urate transporter URAT1 expression in the brush border of proximal tubules.

Conclusions: Among patients with COVID-19 requiring hospitalization, low serum levels of uric acid are common and associate with disease severity and with progression to respiratory failure requiring invasive mechanical ventilation.

Keywords: Acute respiratory distress syndrome; Hypouricemia; Mechanical ventilation; Proximal tubule; SARS-CoV-2.

Fig. 1

Flowchart of the study, including discovery and validation cohorts

Fig. 2

Prevalence and kinetics of low serum uric acid levels among patients hospitalized with COVID-19, in discovery and validation cohorts. a In the discovery cohort, the prevalence of hypouricemia was 1/122 (0.8%) under basal conditions, 13/192 (6.8%) at admission, 39/192 (20.3%) at the lowest uric acid level and 5/140 (3.6%) at discharge. b In the validation cohort, the prevalence of hypouricemia increased from 21/325 (6.5%) upon admission to 63/325 (19.4%) during hospitalization. Circles represent individual values and black lines medians. Histograms show the distribution of time from admission to onset of hypouricemia in both cohorts. P values calculated using a mixed-effects model considering repeated measures in individual patients. ***P < 0.001

Fig. 3

Low serum uric acid levels associate with disease severity and outcome in COVID-19 (pooled data from discovery and validation cohorts). a Proportion of patients with hypouricemia (bar graph) and lowest serum uric acid level (scatter plots with median) according to COVID-19 severity. Circles represent individual values and black lines are medians (one-way ANOVA and linear trend, P = 0.002). b Comparison of peak values of highly sensitive C-reactive protein (hsCRP) and lactate dehydrogenase (LDH), and nadir of lymphocytes in patients with or without hypouricemia. Circles represent individual values and lines are medians. ***P < 0.001 using Mann–Whitney test. c Proportion of patients requiring mechanical ventilation, according to the presence (red dots) or not (gray dots) of hypouricemia. d Survival analyses with Kaplan–Meier curves showing time to mechanical ventilation or death (log-rank test, P < 0.001)