Mortality Risk and Urinary Proteome Changes in Acute COVID-19 Survivors in the Multinational CRIT-COV-U Study

Abstract

Background/objectives: Survival prospects following SARS-CoV-2 infection may extend beyond the acute phase, influenced by various factors including age, health conditions, and infection severity; however, this topic has not been studied in detail. Therefore, within this study, the mortality risk post-acute COVID-19 in the CRIT-COV-U cohort was investigated.

Methods: Survival data from 651 patients that survived an acute phase of COVID-19 were retrieved and the association between urinary peptides and future death was assessed. Data spanning until December 2023 were collected from six countries, comparing mortality trends with age- and sex-matched COVID-19-negative controls. A death prediction classifier was developed and validated using pre-existing urinary peptidomic datasets.

Results: Notably, 13.98% of post-COVID-19 patients succumbed during the follow-up, with mortality rates significantly higher than COVID-19-negative controls, particularly evident in younger individuals (<65 years). These data for the first time demonstrate that SARS-CoV-2 infection highly significantly increases the risk of mortality not only during the acute phase of the disease but also beyond for a period of about one year. In our study, we were further able to identify 201 urinary peptides linked to mortality. These peptides are fragments of albumin, alpha-2-HS-glycoprotein, apolipoprotein A-I, beta-2-microglobulin, CD99 antigen, various collagens, fibrinogen alpha, polymeric immunoglobulin receptor, sodium/potassium-transporting ATPase, and uromodulin and were integrated these into a predictive classifier (DP201). Higher DP201 scores, alongside age and BMI, significantly predicted death.

Conclusions: The peptide-based classifier demonstrated significant predictive value for mortality in post-acute COVID-19 patients, highlighting the utility of urinary peptides in prognosticating post-acute COVID-19 mortality, offering insights for targeted interventions. By utilizing these defined biomarkers in the clinic, risk stratification, monitoring, and personalized interventions can be significantly improved. Our data also suggest that mortality should be considered as one possible symptom or a consequence of post-acute sequelae of SARS-CoV-2 infection, a fact that is currently overlooked.

Keywords: PASC; biomarker; long COVID; mortality; peptides; urine.

Figure 1

Age-dependent mortality in the post-acute COVID-19 cohort and healthy controls. Kaplan–Meier survival curves are shown for the cohort who survived acute-phase COVID-19 (A), including zoomed-in view of the acute (without detected deaths) and sub-acute phases, which highlight that only those who survived the acute phase were followed up, and age- and sex-matched COVID-19-negative healthy controls (B). The number of deaths per age group in the post-acute COVID-19 cohort (C) and controls (D) as well as the calculated fold change (FC) between the COVID-19 and controls (E) is given.

Figure 2

ROC curves (displayed as solid lines) for the classification of the deceased and surviving patients in the complete leave-one-out cross-validated discovery cohort (A) and independent validation cohort (B). Dotted lines display 95% confidence bounds.

Figure 3

Performance of the urinary peptide-based death prediction classifier in the independent validation data. The risk of death is significant dependent on the DP201 score (A) although the age dependency can still be observed (B). The hazard ratio for survival probability DP201 classifier (A) and age (C) could be increased using a model including DP201, age, and BMI (D).

Figure 4

Distribution of the classification scores between the deceased and surviving patients for the fibrosis-related classifier FPP BH29 (A) and CKD273 classifier (B).