The COV50 classifier predicts frailty and future death, independent from SARS-CoV-2 infection
=============================================================================================

* Felix Keller
* Joachim Beige
* Justyna Siwy
* Alexandre Mebazaa
* Dewei An
* Harald Mischak
* Joost P. Schanstra
* Marika Mokou
* Paul Perco
* Jan A. Staessen
* Antonia Vlahou
* Agnieszka Latosinska

## Summary

COV50, a urinary proteomic classifier, predicts disease progression and death from SARS-CoV-2 at early stage, suggesting it might predict pre-established vulnerability. This study investigated the value of COV50 in predicting non-COVID-19 associated death. Urinary proteomic data were extracted from the Human Urinary Proteome Database. In the ICU group (n=1719), an increase in the COV50 score of one unit resulted in a 20% higher relative risk of death (adj. HR 1·2 [95% CI 1·17-1·24]). The same increase in COV50 in non-ICU patients (n=7474) resulted in a higher relative risk of 61% (adj. HR 1·61 [95% CI 1·47-1·76]), in line with adjusted meta-analytic HR estimate of 1·55. A higher COV50 scoring was observed in frail patients (p<0·0001). The COV50 classifier is predictive of death, and is associated with frailty suggesting that it detects pre-existing vulnerability. These data may serve as basis for proteomics guided intervention, reducing the risk of death and frailty.

## INTRODUCTION

In the recently reported CRIT-COV-U study, a urinary peptide-based classifier that predicts a critical/ lethal course of COVID-19 was developed and validated 1,2. This classifier, termed COV50, is based on 50 specific urinary peptides, all significantly associated with critical/lethal outcome. The most prominent changes observed were reduction of peptides derived from collagen alpha 1(I), polymeric immunoglobulin receptor and CD99 antigen, and an increase in peptides derived from alpha-1-antitrypsin 3. A support-vector machine-based algorithm was applied to combine 50 peptides into a classifier that provides a dimensionless variable predicting outcome. In the validation study, COV50 could predict critical disease course, death, or both, with an area under the receiving operating curve (AUC) of 0·81.

Prediction of critical disease course and/or death was possible at the earliest possible date, i.e., the first positive indication of a SARS-CoV-2 infection. This result suggested that outcome prediction may not be solely based on molecular events associated with SARS-CoV-2 infection, but may be partly due to already present pre-established vulnerability. This would indicate that prediction of severe disease course may be possible even before the virus infection due to susceptibility to a “second hit” – here, SARS-CoV2 infection. A study to investigate this hypothesis does not appear feasible, as it would require the recruitment of an excess of 105 subjects to include a sufficient number of patients that will experience SARS-CoV-2 infection and a critical or lethal disease course. However, it appeared reasonable to assume that, if the COV50 classifier detects a vulnerable population, a significantly higher number of this vulnerable population (defined by a COV50 score above the threshold) should experience death, compared to the population with a lower score. This hypothesis was investigated in the present study, testing the association of mid- and long-term mortality with COV50.

## RESULTS

First, we evaluated the hypothesis that the COV50 classifier defines the vulnerable population at a molecular level, irrespective of SARS-CoV-2 infection. For that purpose, we investigated datasets from subjects from the FROG-ICU study 4, as this study was more comparable to the CRIT-COV study (patients in ICU) and had available a large number of endpoints. We identified 1719 datasets to be included in this study, for which follow-up and information on relevant covariables were available5.

To further support our analysis, we also investigated if COV50 enables death prediction in subjects, not in ICU. Studies with more than 50 individuals and available follow-up information were selected from the Human Urinary Proteome Database 6,7, applying the same criteria as above regarding the availability of covariables.

The demographic information on the subjects included in the study split by ICU and death is given in **Table 1**. Among the risk factors for death, we found significant differences at the aggregate level in both the ICU and the non-ICU subgroups, according to expectation. The median COV50 score is significantly (p <0·001) higher in patients that were to experience death during the observation time, as also displayed in it’s distributions in Panel A of **Figure 1**.

View this table:
[Table 1:](http://medrxiv.org/content/early/2023/05/01/2023.04.28.23289257/T1)

Table 1: 
Descriptive statistics for the ICU and non-ICU samples analysed within this study.

![Figure 1:](http://medrxiv.org/https://www.medrxiv.org/content/medrxiv/early/2023/05/01/2023.04.28.23289257/F1.medium.gif)

[Figure 1:](http://medrxiv.org/content/early/2023/05/01/2023.04.28.23289257/F1)

Figure 1: 
A) Density of the COV50 distribution in ICU and non-ICU subjects. B) Mortality per person-years for FROG and non-ICU cohorts given age and COV50. C) Mortality as share [0-1] from a logistic regression for FROG and non-ICU cohorts given age and COV50.

As age is a crucial risk factor for death, we described the relationship between COV50 and mortality split by age groups in panels B and C of **Figure 1**. Panel B describes mortality in person-time in COV50 groups, whereas panel C relates mortality as a percentage with continuous COV50. In both subgroups, an increase in COV50 accompanies higher mortality, and the effect is more pronounced in higher age groups.

Crude HRs in **Figure 2 A** for all studies on average show an association of higher relative risk of death with increasing COV50 scores, with all but 5 studies showing a significantly elevated relative risk. For all subgroups in the meta-analysis, besides the diabetes-related studies, the combined estimate for the HR is significantly different from 1, as indicated by the 95% CIs. In general, adjustment for risk factors lowered the COV50 HR estimates, which aligns with expectations,, as adjustment generally improves comparability by accounting for observed between study heterogeneity on the patient level. However, in studies (particularly PersTIgAN, STOPIgAN) with low numbers of events, variance increased drastically with the adjustment. The estimates from the meta-analysis resulted in an unadjusted HR of 1·77 [95% CI 1·58-1·97] and an adjusted HR of 1·55 [95% CI 1·39-1·73].

![Figure 2:](http://medrxiv.org/https://www.medrxiv.org/content/medrxiv/early/2023/05/01/2023.04.28.23289257/F2.medium.gif)

[Figure 2:](http://medrxiv.org/content/early/2023/05/01/2023.04.28.23289257/F2)

Figure 2: 
Random effects meta analyes based on the log-HR and the standard errors from the separate cox regressions. A) Unadjusted, B), adjusted for sex, age, kidney function and BMI. The size of dot symbols is proportional to weight and weight is inverse proportional to HR standard error.

Though appearing on a trend level within and between the subgroups, neither heterogeneity nor subgroup differences were significant (**Figure 2**).

With a 95% prediction interval for the unadjusted HR ranging from 1·47 to 2·13 we can expect future studies to find a predictive effect of COV50 on death in a broader population.

Subgroup HR estimates from the adjusted meta-analysis in **Figure 2 B** are robust, since they are close to corresponding estimates from the pooled adjusted Cox regression in **Table 2**. In the ICU group, an increase in the COV50 score of one unit results (on average, ceteris paribus) in a 20% higher relative risk of death (adj. HR 1·2 [95% CI 1·17-1·24]). As the absolute risk of death is considerably lower in non-ICU patients, the same increase in COV50 in non-ICU patients results in a higher relative risk of 61% (adj. HR 1·61 [95% CI 1·47-1·76]). This is in line with the adjusted HR estimate of 1·55 from the meta-analysis.

View this table:
[Table 2:](http://medrxiv.org/content/early/2023/05/01/2023.04.28.23289257/T2)

Table 2: 
Estimates from the pooled adjusted Cox regression.

Frailty may be considered an extensive vulnerability. Based on these considerations, we examined the distribution of COV50 scoring in a cohort of frail subjects and controls where CE-MS data from previous analysis were available 8. The mean COV50 scoring in 196 controls was -1·041, while in 73 frail subjects it was -0·677. The difference in the means of 0,365 was highly significant (p<0·0001), supporting our hypothesis that COV50 is associated with frailty.

Since COV50 is a composite score based on 50 distinct urinary peptides, we examined which of these 50 peptides served as individual predictors of death in the cohorts investigated (ICU, non-ICU). Specifically, we compared the distribution of the 50 peptides in the datasets from survivors with those from subjects that died. The outcome was assessed at 1 and 5 years for ICU and non-ICU cohorts, respectively.

The results of this analysis are shown in **Table 3**. A high degree of concordance was found when comparing the peptides regulation trend in the context of COVID-19, death in or after ICU, or death without ICU stay. The most notable and significant changes were associated with future fatal events are the reductions of specific collagen fragments, most of collagen alpha I(I).

View this table:
[Table 3:](http://medrxiv.org/content/early/2023/05/01/2023.04.28.23289257/T3)

Table 3: 
List of 50 urinary peptides included in the COV50 classifier and their respective regulation trend in investigated cohorts (ICU and non-ICU).

In the ICU subjects, 28 of the 50 peptides were found to be significantly associated with future death. Of these, 26 showed a regulation trend in a similar direction as for critical/lethal COVID-19. At the same time, two peptides, one from alpha-1-acid glycoprotein 1 and one from sodium/potassium-transporting ATPase subunit gamma, had an opposing regulation. When investigating the most prominent peptides derived from collagen, all significant changes are concordant between death in COVID-19 or ICU. In the non-ICU subjects, 36 of the 50 peptides were significantly associated with future death. Of these, 27 showed a regulation concordant with the one in critical/lethal COVID-19, while 9 peptides changed in an opposing direction. The latter are the peptides derived from alpha-1-acid glycoprotein 1, polymeric immunoglobulin receptor, neurosecretory protein VGF, CD99 antigen, hornerin, collagen alpha-1(I), and collagen alpha-2(I).

A major difference in comparison to the distribution in COVID-19 patients was observed for CD99 antigen and polymeric immunoglobulin receptor. While in critical COVID-19 patients a consistent and significant reduction of multiple CD99 antigen and polymeric immunoglobulin receptor peptides was associated with severe/lethal disease, this was not observed in ICU and non-ICU populations not infected with SARS-CoV2.

## DISCUSSION

The data demonstrate that the COV50 classifier not only predicts an unfavourable outcome of a COVID-19 episode but apparently identifies “vulnerable subjects”, who are likely at substantially higher risk of severe or lethal COVID-19. At the same time, this vulnerability also appears relevant in other clinical situations (e.g., non-SARS-CoV2 infections) leading to ICU admission. Moreover, it increases the risk of death in most pathological conditions. This concept of the “second hit phenomenon” exaggerated on the grounds of pre-established chronic diseases, and inappropriate innate immune response to injury and trauma has been related to perfusion changes of critical compartments and pathological sequelae 9.

The most prominent and consistent findings are the reduction of several specific urinary collagen fragments, most from collagen alpha-I(I). The reduction of these collagen fragments most likely indicates reduced collagen degradation in the extracellular matrix, which is expected to result in increased organ fibrosis. Fibrosis is generally associated with vulnerability to “second-hit events” either in e.g. infectious or general (cardiovascular) scenario. It is reasonable to assume that the “second hit” in the context of a SARS-CoV-2 infection is depicted via peptides deregulated in severe COVID-19 only, like CD99 antigen and Polymeric immunoglobulin receptor.

The concordance of significant changes observed due to critical/lethal COVID-19 appears to be higher in the context of ICU than in non-ICU subjects. While an objective measure to assess significant differences does not seem to exist, a concordance (based on up- or down-regulation) of 93% (in the case of ICU) compared to 75% (in the case of non-ICU) is at least indicative.

As expected, similarities between changes in biomarkers in patients developing the critical condition, irrespective of the underlying pathology and disease etiology, could be observed. At the same time it becomes evident that specific changes, a decrease of peptides from CD99 antigen and polymeric immunoglobulin receptor, are associated explicitly with critical COVID-19, and cannot be associated with all-cause death, neither in nor outside ICU.

This study’s findings agree with previous research that reported an association of urinary peptides (or classifiers based on these) with unfavourable outcome. A search in Pubmed for the keywords (urine OR urinary) AND (peptidom* OR proteom*) AND (death OR mortality) in the title or abstract resulted in 96 publications. Upon manual inspection by three authors, 11 manuscripts were found to be relevant, describing studies investigating the association of urinary peptides with mortality in humans, including the manuscripts describing the development of COV50 1,2,10. Currie et al. described a significant value of CKD273, a classifier based on 273 urinary peptides, in predicting mortality in 155 microalbuminuric type 2 diabetic patients 11. Similar results were also presented by Verbeke et al., where CKD273 was significantly associated with mortality in 451 chronic kidney disease patients 12. Nkuipou-Kenfack et al. reported the association of urinary peptides with death, and the development of a classifier based on these to predict mortality after ICU stay in 1243 patients 5. In 2021, Martens et al. described the association of multiple urinary peptides, many of these collagen fragments, with biological age, and mortality 13. Batra et al. presented a proteomics-based mortality signature in COVID-19 and acute respiratory distress syndrome patients 14. In the context of hepatocellular carcinoma, Bannaga et al. described several urine peptides being significantly associated with death 15. Very recently, Wei et al. reported on the detection of urinary peptides related to pulse-wave velocity also being associated with mortality 16. In a very well-powered study on 1170 patients that underwent cardiac surgery, Piedrafita and colleagues reported the identification of 204 urinary peptides associated with acute kidney injury 17. A classifier based on these 204 peptides was validated in an independent cohort of 1569 ICU patients, demonstrating good performance and significant association with mortality. In almost all of these studies, collagen peptides were among the most prominent biomarkers, with reduced abundance being associated with increased risk of death, as also demonstrated recently by He and colleagues in the context of heart failure 18. The data on quite large cohorts in ICU and subjects not in critical condition at the time of sampling indicate that urinary peptides and classifiers derived thereof hold significant predictive value for a patient-relevant endpoint: death. In complete agreement with previous studies, prediction of death appears to be based mainly on collagen fragments, which may indicate attenuation of collagen degradation, consequently progressing fibrotic processes. Evidently, the COV50 classifier was not developed to predict death in the general population. Also, based on the observation in this study that several peptides contained in this classifier show opposite regulation on predicting critical COVID-19 or death from any cause, it is to be expected that a classifier developed explicitly for prediction of death, based on only those peptides significantly associated with death, could be of substantial value in guiding death-preventing interventions. It is to be expected that such classifier would be based mainly on collagen fragments.

In this study we demonstrate that the urinary COV50 classifier is significantly associated with future death in ICU patients as well as in non-ICU patients. Further research is needed to assess if specific, personalized intervention guided by urinary collagen fragments can significantly improve outcomes, reducing future death.

## LIMITATIONS OF THE STUDY

The study has limitations. It has been performed based on previously generated datasets. However, the large number of datasets, the high number of endpoints assessed, and the very high significance level of the findings strongly support that the results can be generalized. Along these lines, a strength of the study is the inclusion of datasets from multiple studies, indicating a robust basis for the assessment.

## Data Availability

Anonymised data and code used in conducting the analyses will be made available upon request directed to the corresponding author. Proposals will be reviewed and approved by the authors with scientific merit and feasibility as the criteria. After approval of a proposal, data can be shared via a secure online platform after signing a data access and confidentiality agreement. Data will be made available for a maximum of 5 years after a data sharing agreement has been signed

## Authors’ contributions

AL, HM, AV, JS and FK conceptualised the study. JS, AL, MM, JPS and HM collected and curated the urine proteomic data. AL, FK, PP and HM performed the statistical analysis. JB, AM, JAS and DA provided access to the patient data. JB and JAS had access to and verified the data reported in this study. AL, HM and FK wrote the first draft of the manuscript. All authors interpreted the results, commented on and revised successive drafts of the manuscript, and approved the final version. All authors had full access to all the data in the study and had final responsibility for the decision to submit for publication.

## Declaration of interest

HM is the cofounder and co-owner of Mosaiques Diagnostics (Hannover, Germany) and AL, MM, and JS are employees of Mosaiques Diagnostics. PP is also employed by Delta4 GmbH. AM reports grants or contracts from 4TEEN4, Abbott, Roche and Sphyngotec, and consulting fees from Roche, Adrenomed, Corteria, Fire1 and payment or honoraria from Merc and Novartis. All other authors declare no competing interests.

## METHODS

### Patients

Intensive care unit (ICU): patients from the medical, surgical, or mixed ICUs at 14 university hospitals from the FROG-ICU study were used 19. Inclusion criteria were mechanical ventilation or administration of vasoactive agents for at least 24□h. The exclusion criteria were age under 18, severe head injury with a Glasgow Coma Scale below 8, brain death or persistent vegetative state, pregnancy or breastfeeding, transplantation in the past 12□months, moribund status, and lack of social security coverage. All capillary electrophoresis coupled to mass spectrometry (CE-MS) datasets where 1-year follow-up and information on relevant covariables (age, body mass index (BMI), sex, blood pressure, estimated glomerular filtration rate (eGFR), presence of diabetes, kidney, cardiovascular disease, hypertension) was available were included in the present study without pre-selection.

Non-ICU: The assessment of COV50 in the non-ICU population was based on 7474 datasets from the Human Urinary Proteome Database 7 with available information on age, sex, eGFR, blood pressure, BMI, presence of diabetes, kidney disease, cardiovascular disease, hypertension, and a follow-up data. The additional dataset from frail patients and controls included all samples investigated in the FRAILOMIC project 8.

All datasets were from previously published studies and fully anonymized. Ethical review and approval were waived for this study by the ethics committee of the Hannover Medical School, Germany (no. 3116-2016), due to all data being fully anonymized. The number of subjects per study and patient characteristics are listed in **Table 1**.

### Urinary proteome/peptidome data

The urinary proteome is well characterized and reference standards are available 20. Urine proteome analysis was performed on urine samples collected at study inclusion, bio-banked until assayed. Detailed information on urine sample preparation, proteome analysis by CE-MS, data processing, and sequencing of the urinary peptides allowing identification of parental proteins is available in previous publications 13,21,22.

### Outcome

In the FROG-ICU study, information on vital status was collected 3, 6, and 12□months after ICU discharge, as previously described 4. For the non-ICU patients, vital status and outcome was assessed as described in the specific original studies 12,13,18,23-36.

### Statistics

As descriptive statistics for the ICU and non-ICU samples, shown in **Table 1**, median and 1st and 3rd quartile (IQR) were used for continuous variables and absolute (N) and relative frequencies (%) for categorical variables. Hypotheses of no differences in scale or distribution of patient characteristics between the death and non-death groups were tested with Wilcoxon–Mann–Whitney tests for continuous and with χ2-homogeneity tests for categorical variables.

Kernel density estimates of the distribution of COV50 scores split by ICU and mortality groups are displayed in **Figure 1 A**. Mortality per person-time stratified by age and COV50 groups, seen in **Figure 1 B**, is estimated as the ratio of the number of the deceased to the sum of all patients’ observation times within each group scaled to 100 person-years. The corresponding mortality probabilities and their 95% confidence intervals (CI) for each group, as seen in **Figure 1 C**, represent estimates from a logistic regression including all 9193 patients.

For each study, separate unadjusted Cox regressions for the effect of the COV50 score on experiencing death were performed, as listed in **Figure 2 A**. In **Figure 2 B** these models were additionally adjusted for age, female, log(BMI), mean arterial pressure (MAP) and log(eGFR). All regressors besides female and COV50 were normalized (mean 0, sd 1). The natural logarithms of the estimated hazard ratios (logHR) and their standard errors were combined in meta-analyses to determine the effect of the COV50 score on mortality. A random effects model was estimated after the assumption that included studies are heterogeneous, i.e., coming from different populations. Study weights are based on the logHR estimates’ uncertainty, namely their standard errors. Studies were categorized into more homogenous subgroups, and estimates for each subgroup are displayed in **Figure 2**. Overall and group-wise between-study heterogeneity is presented with τ2 and assessed by Higgins & Thompson’s I2 statistic. χ2-Tests for heterogeneity and subgroup differences are based on Cochran’s Q. Random effects meta-analysis estimates are presented with 95% CIs and a 95% prediction interval for the overall effect is given. One Cox regression stratified by study pooling all 9193 patients was used as a benchmark to the meta-analytic approach. As displayed in **Table 2**, the model’s adjustment specification is identical to the adjusted separate study regressions (**Figure 2 B**). To be comparable to the adjusted meta-analytic estimate, HRs for COV50 interacted with ICU and non-ICU, as well as for the above-mentioned non-ICU subgroups, were estimated. Standard errors were clustered on the study level for more robust inference and due to unobserved heterogeneity between studies. The models log-likelihood, it’s Wald test and concordance are reported in **Table 2**. We allow for a type 1 error of 5%, all hypotheses are two-sided. All analyses were carried out using R 4.2.2.

## Role of the funding source

The funders had no role in the design of the study; collection, analyses, or interpretation of data; in the writing of the manuscript; or in the decision to publish the results.

## Data and code availability

Anonymised data and code used in conducting the analyses will be made available upon request directed to the corresponding author. Proposals will be reviewed and approved by the authors with scientific merit and feasibility as the criteria. After approval of a proposal, data can be shared via a secure online platform after signing a data access and confidentiality agreement. Data will be made available for a maximum of 5 years after a data sharing agreement has been signed.

## Acknowledgments

All authors are grateful to all patients who contributed samples.

This work was supported in part by funding through the European Union’s Horizon Europe Marie Sklodowska-Curie Actions Doctoral Networks - Industrial Doctorates Programme (HORIZON – MSCA – 2021 – DN-ID, “DisCo-I”, grant agreement No 101072828) to AL, AV, JPS and HM.

*   Received April 28, 2023.
*   Revision received April 28, 2023.
*   Accepted May 1, 2023.


*   © 2023, Posted by Cold Spring Harbor Laboratory

This pre-print is available under a Creative Commons License (Attribution-NonCommercial-NoDerivs 4.0 International), CC BY-NC-ND 4.0, as described at [http://creativecommons.org/licenses/by-nc-nd/4.0/](http://creativecommons.org/licenses/by-nc-nd/4.0/)

## Reference List

1.  1.Staessen, J. A., Wendt, R., Yu, Y. L., Kalbitz, S., Thijs, L., Siwy, J., Raad, J., Metzger, J., Neuhaus, B., Papkalla, A. et al. (2022) Predictive performance and clinical application of COV50, a urinary proteomic biomarker in early COVID-19 infection: a prospective multicentre cohort study. Lancet Digit.Health 4, e727–e737.
    
    

2.  2.Wendt, R., Thijs, L., Kalbitz, S., Mischak, H., Siwy, J., Raad, J., Metzger, J., Neuhaus, B., Leyen, H. V., Dudoignon, E. et al. (2021) A urinary peptidomic profile predicts outcome in SARS-CoV-2-infected patients. EClinicalMedicine. 36, 100883.
    
    

3.  3.Siwy, J., Wendt, R., Albalat, A., He, T., Mischak, H., Mullen, W., Latosinska, A., Lubbert, C., Kalbitz, S., Mebazaa, A. et al. (2021) CD99 and polymeric immunoglobulin receptor peptides deregulation in critical COVID-19: A potential link to molecular pathophysiology? 21, e2100133.
    
    

4.  4.Mebazaa, A., Casadio, M. C., Azoulay, E., Guidet, B., Jaber, S., Levy, B., Payen, D., Vicaut, E., Resche-Rigon, M., and Gayat, E. (2015) Post-ICU discharge and outcome: rationale and methods of the The French and euRopean Outcome reGistry in Intensive Care Units (FROG-ICU) observational study. BMC.Anesthesiol. 15, 143.
    
    

5.  5.Nkuipou-Kenfack, E., Latosinska, A., Yang, W. Y., Fournier, M. C., Blet, A., Mujaj, B., Thijs, L., Feliot, E., Gayat, E., Mischak, H. et al. (2020) A novel urinary biomarker predicts 1-year mortality after discharge from intensive care. Crit Care 24, 10.
    
    

6.  6.Latosinska, A., Siwy, J., Mischak, H., and Frantzi, M. (2019) Peptidomics and proteomics based on CE-MS as a robust tool in clinical application: The past, the present, and the future. Electrophoresis 40, 2294–2308.
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1002/elps.201900091&link_type=DOI) 

7.  7.Rodriguez-Suarez, E., Siwy, J., Zurbig, P., and Mischak, H. (2013) Urine as a source for clinical proteome analysis: From discovery to clinical application. Biochim.Biophys.Acta 1844, 884–898.
    
    

8.  8.Gomez-Cabrero, D., Walter, S., Abugessaisa, I., Minambres-Herraiz, R., Palomares, L. B., Butcher, L., Erusalimsky, J. D., Garcia-Garcia, F. J., Carnicero, J., Hardman, T. C. et al. (2021) A robust machine learning framework to identify signatures for frailty: a nested case-control study in four aging European cohorts. Geroscience. 43, 1317–1329.
    
    

9.  9.Pape, H. C., Moore, E. E., McKinley, T., and Sauaia, A. (2022) Pathophysiology in patients with polytrauma. Injury 53, 2400–2412.
    
    

10. 10.Wendt, R., Thijs, L., Kalbitz, S., Mischak, H., Siwy, J., Raad, J., Metzger, J., Neuhaus, B., Leyen, H. V., Dudoignon, E. et al. (2021) A urinary peptidomic profile predicts outcome in SARS-CoV-2-infected patients. EClinicalMedicine 36, 100883.
    
    

11. 11.Currie, G. E., von Scholten, B. J., Mary, S., Flores Guerrero, J. L., Lindhardt, M., Reinhard, H., Jacobsen, P. K., Mullen, W., Parving, H. H., Mischak, H. et al. (2018) Urinary proteomics for prediction of mortality in patients with type 2 diabetes and microalbuminuria. Cardiovasc.Diabetol. 17, 50.
    
    

12. 12.Verbeke, F., Siwy, J., Van Biesen W., Mischak, H., Pletinck, A., Schepers, E., Neirynck, N., Magalhaes, P., Pejchinovski, M., Pontillo, C. et al. (2021) The urinary proteomics classifier chronic kidney disease 273 predicts cardiovascular outcome in patients with chronic kidney disease. Nephrol Dial.Transplant 36, 811–818.
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1093/ndt/gfz242&link_type=DOI) 

13. 13.Martens, D. S., Thijs, L., Latosinska, A., Trenson, S., Siwy, J., Zhang, Z. Y., Wang, C., Beige, J., Vlahou, A., Janssens, S. et al. (2021) Urinary peptidomic profiles to address age-related disabilities: a prospective population study. Lancet Healthy.Longev. 2, e690–e703.
    
    

14. 14.Batra, R., Uni, R., Akchurin, O. M., Alvarez-Mulett, S., Gomez-Escobar, L. G., Patino, E., Hoffman, K. L., Simmons, W., Whalen, W., Chetnik, K. et al. (2023) Urine-based multi-omic comparative analysis of COVID-19 and bacterial sepsis-induced ARDS. Mol Med. 29, 13.
    
    

15. 15.Bannaga, A., Metzger, J., Voigtlander, T., Pejchinovski, M., Frantzi, M., Book, T., James, S., Gopalakrishnan, K., Mischak, H., Manns, M. P. et al. (2021) Pathophysiological Implications of Urinary Peptides in Hepatocellular Carcinoma. Cancers (Basel) 13, 3786.
    
    

16. 16.Wei, D., Melgarejo, J. D., Thijs, L., Temmerman, X., Vanassche, T., Van Aelst L., Janssens, S., Staessen, J. A., Verhamme, P., and Zhang, Z. Y. (2022) Urinary Proteomic Profile of Arterial Stiffness Is Associated With Mortality and Cardiovascular Outcomes. J Am Heart Assoc. 11, e024769.
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1161/JAHA.121.024769&link_type=DOI) 

17. 17.Piedrafita, A., Siwy, J., Klein, J., Akkari, A., Amaya-Garrido, A., Mebazaa, A., Sanz, A. B., Breuil, B., Montero Herrero L., Marcheix, B. et al. (2022) A universal predictive and mechanistic urinary peptide signature in acute kidney injury. Crit Care 26, 344.
    
    

18. 18.He, T., Melgarejo, J. D., Clark, A. L., Yu, Y. L., Thijs, L., Diez, J., Lopez, B., Gonzalez, A., Cleland, J. G., Schanstra, J. P. et al. (2021) Serum and urinary biomarkers of collagen type-I turnover predict prognosis in patients with heart failure. Clin Transl.Med. 11, e267.
    
    

19. 19.Gayat, E., Cariou, A., Deye, N., Vieillard-Baron, A., Jaber, S., Damoisel, C., Lu, Q., Monnet, X., Rennuit, I., Azoulay, E. et al. (2018) Determinants of long-term outcome in ICU survivors: results from the FROG-ICU study. Crit Care 22, 8.
    
    

20. 20.Mischak, H., Kolch, W., Aivalotis. M, Bouyssie, D, Court, M., Dihazi, H., Dihazi G.H., Franke, J., Garin, J., Gonzales de Peredo, A. et al. (2010) Comprehensive human urine standards for comparability and standardization in clinical proteome analysis. Proteomics Clin Appl. 4, 464–478.
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1002/prca.200900189&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=21137064&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2023%2F05%2F01%2F2023.04.28.23289257.atom) 
    
    [Web of Science](http://medrxiv.org/lookup/external-ref?access_num=000277221700011&link_type=ISI) 

21. 21.Mavrogeorgis, E., Mischak, H., Latosinska, A., Siwy, J., Jankowski, V., and Jankowski, J. (2021) Reproducibility Evaluation of Urinary Peptide Detection Using CE-MS. Molecules. 26.
    
    

22. 22.Mischak, H., Vlahou, A., and Ioannidis, J. P. (2013) Technical aspects and inter-laboratory variability in native peptide profiling: The CE-MS experience. Clin.Biochem. 46, 432–443.
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/j.clinbiochem.2012.09.025&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=23041249&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2023%2F05%2F01%2F2023.04.28.23289257.atom) 

23. 23.Rudnicki, M., Siwy, J., Wendt, R., Lipphardt, M., Koziolek, M. J., Maixnerova, D., Peters, B., Kerschbaum, J., Leierer, J., Neprasova, M. et al. (2021) Urine proteomics for prediction of disease progression in patients with IgA nephropathy. 37, 42–52.
    
    

24. 24.Frantzi, M., van Kessel, K. E., Zwarthoff, E. C., Marquez, M., Rava, M., Malats, N., Merseburger, A. S., Katafigiotis, I., Stravodimos, K., Mullen, W. et al. (2016) Development and Validation of Urine-based Peptide Biomarker Panels for Detecting Bladder Cancer in a Multi-center Study. Clin Cancer Res. 22, 4077–4086.
    
    [Abstract/FREE Full Text](http://medrxiv.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6MTA6ImNsaW5jYW5yZXMiO3M6NToicmVzaWQiO3M6MTA6IjIyLzE2LzQwNzciO3M6NDoiYXRvbSI7czo1MDoiL21lZHJ4aXYvZWFybHkvMjAyMy8wNS8wMS8yMDIzLjA0LjI4LjIzMjg5MjU3LmF0b20iO31zOjg6ImZyYWdtZW50IjtzOjA6IiI7fQ==) 

25. 25.Packham, D. K., Wolfe, R., Reutens, A. T., Berl, T., Heerspink, H. L., Rohde, R., Ivory, S., Lewis, J., Raz, I., Wiegmann, T. B. et al. (2012) Sulodexide fails to demonstrate renoprotection in overt type 2 diabetic nephropathy. J.Am.Soc.Nephrol. 23, 123–130.
    
    [Abstract/FREE Full Text](http://medrxiv.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6ODoiam5lcGhyb2wiO3M6NToicmVzaWQiO3M6ODoiMjMvMS8xMjMiO3M6NDoiYXRvbSI7czo1MDoiL21lZHJ4aXYvZWFybHkvMjAyMy8wNS8wMS8yMDIzLjA0LjI4LjIzMjg5MjU3LmF0b20iO31zOjg6ImZyYWdtZW50IjtzOjA6IiI7fQ==) 

26. 26.Huang, Q. F., Trenson, S., Zhang, Z. Y., Yang, W. Y., Van Aelst L., Nkuipou-Kenfack, E., Wei, F. F., Mujaj, B., Thijs, L., Ciarka, A. et al. (2017) Urinary Proteomics in Predicting Heart Transplantation Outcomes (uPROPHET)-Rationale and database description. PLoS.ONE. 12, e0184443.
    
    

27. 27.Alkhalaf, A., Zürbig, P., Bakker, S. J., Bilo, H. J., Cerna, M., Fischer, C., Fuchs, S., Janssen, B., Medek, K., Mischak, H. et al. (2010) Multicentric validation of proteomic biomarkers in urine specific for diabetic nephropathy. PLoS One. 5, e13421.
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1371/journal.pone.0013421&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=20975990&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2023%2F05%2F01%2F2023.04.28.23289257.atom) 

28. 28.Rossing, K., Bosselmann, H. S., Gustafsson, F., Zhang, Z. Y., Gu, Y. M., Kuznetsova, T., Nkuipou-Kenfack, E., Mischak, H., Staessen, J. A., Koeck, T. et al. (2016) Urinary Proteomics Pilot Study for Biomarker Discovery and Diagnosis in Heart Failure with Reduced Ejection Fraction. PLoS.ONE. 11, e0157167.
    
    

29. 29.Htun, N. M., Magliano, D. J., Zhang, Z. Y., Lyons, J., Petit, T., Nkuipou-Kenfack, E., Ramirez-Torres, A., von Zur Muhlen C., Maahs, D., Schanstra, J. P. et al. (2017) Prediction of acute coronary syndromes by urinary proteome analysis. PLoS.ONE. 12, e0172036.
    
    

30. 30.Kuznetsova, T., Mischak, H., Mullen, W., and Staessen, J. A. (2012) Urinary proteome analysis in hypertensive patients with left ventricular diastolic dysfunction. Eur.Heart J 33, 2342–2350.
    
    

31. 31.He, T., Mischak, M., Clark, A. L., Campbell, R. T., Delles, C., Diez, J., Filippatos, G., Mebazaa, A., McMurray, J. J. V., Gonzalez, A. et al. (2021) Urinary peptides in heart failure: a link to molecular pathophysiology. Eur.J Heart Fail. 23, 1875–1887.
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1002/ejhf.2195&link_type=DOI) 

32. 32.Rotbain Curovic, V, Magalhaes, P., He, T., Hansen, T. W., Mischak, H., and Rossing, P. (2021) Urinary peptidome and diabetic retinopathy in the DIRECT-Protect 1 and 2 trials. Diabet.Med. 38, e14634.
    
    

33. 33.Tofte, N., Lindhardt, M., Adamova, K., Bakker, S. J. L., Beige, J., Beulens, J. W. J., Birkenfeld, A. L., Currie, G., Delles, C., Dimos, I. et al. (2020) Early detection of diabetic kidney disease by urinary proteomics and subsequent intervention with spironolactone to delay progression (PRIORITY): a prospective observational study and embedded randomised placebo-controlled trial. Lancet Diabetes Endocrinol. 8, 301–312.
    
    

34. 34.Lindhardt, Morten, Persson, Frederik, Zuerbig, Petra, Stalmach, Angelique, Mischak, Harald, de Zeeuw, Dick, Heerspink Hiddo Lambers, Klein, Ronald, Orchard, Trevor, Porta, Massimo et al. (2017) Urinary proteomics predict onset of microalbuminuria in normoalbuminuric type 2 diabetic patients, a sub-study of the DIRECT-Protect 2 study. 32, 1866–1873.
    
    

35. 35.Zhang, Z., Staessen, J. A., Thijs, L., Gu, Y., Liu, Y., Jacobs, L., Koeck, T., Zurbig, P., Mischak, H., and Kuznetsova, T. (2014) Left ventricular diastolic function in relation to the urinary proteome: a proof-of-concept study in a general population. Int.J Cardiol. 176, 158–165.
    
    

36. 36.Delles, C., Schiffer, E., von Zur Muhlen C., Peter, K., Rossing, P., Parving, H. H., Dymott, J. A., Neisius, U., Zimmerli, L. U., Snell-Bergeon, J. K. et al. (2010) Urinary proteomic diagnosis of coronary artery disease: identification and clinical validation in 623 individuals. J.Hypertens. 28, 2316–2322.
    
    [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1097/HJH.0b013e32833d81b7&link_type=DOI) 
    
    [PubMed](http://medrxiv.org/lookup/external-ref?access_num=20811296&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2023%2F05%2F01%2F2023.04.28.23289257.atom) 
    
    [Web of Science](http://medrxiv.org/lookup/external-ref?access_num=000282987000020&link_type=ISI)