Performance of serum apolipoprotein-A1 as a sentinel of Covid-19

Findings During the 30 weeks of 2020 we observed in two large cohorts of patients at risk of liver fibrosis a highly significant decrease of serum apolipoprotein-A1, not observed in previous years. This decrease was highly correlated to and in parallel with the daily increase in confirmed Covid-19 cases, including the recovery period. Meaning Apolipoprotein-A1 could be used as a sentinel of the pandemic in existing routine surveillance of the general population. Additional Information: Question Response Financial Disclosure Enter a financial disclosure statement that describes the sources of funding for the work included in this submission. Review the submission guidelines for detailed requirements. View published research articles from PLOS Medicine for specific examples. This statement is required for submission and will appear in the published article if the submission is accepted. Please make sure it is accurate. Unfunded studies Enter: The author(s) received no specific funding for this work. Funded studies Enter a statement with the following details: Initials of the authors who received each award • Grant numbers awarded to each author• The full name of each funder• URL of each funder website• Did the sponsors or funders play any role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript? • NO - Include this sentence at the end of your statement: The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. • YES - Specify the role(s) played.• European Grant EIT health ProCoP 20879 Patrice Cacoub Powered by Editorial Manager® and ProduXion Manager® from Aries Systems Corporation All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted September 3, 2020. ; https://doi.org/10.1101/2020.09.01.20186213doi: medRxiv preprint * typeset Competing Interests Use the instructions below to enter a competing interest statement for this submission. On behalf of all authors, disclose any competing interests that could be perceived to bias this work—acknowledging all financial support and any other relevant financial or non- financial competing interests. This statement will appear in the published article if the submission is accepted. Please make sure it is accurate. View published research articles from PLOS Medicine for specific examples. NO authors have competing interests Enter: The authors have declared that no competing interests exist. Authors with competing interests Enter competing interest details beginning with this statement: I have read the journal's policy and the authors of this manuscript have the following competing interests: [insert competing interests here] * typeset I have read the journal's policy and the authors of this manuscript have the following competing interests: Thierry Poynard is the inventor of FibroTest, founder of BioPredictive, the patents belong to the public organization Assistance Publique-Hôpitaux de Paris. Olivier Deckmyn, Valentina Peta, Yen Ngo, Jean Marie Castille, Fabienne Drane and Clemence Franc are full employees of BioPredictive. The other authors have nothing to declare. Powered by Editorial Manager® and ProduXion Manager® from Aries Systems Corporation All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted September 3, 2020. ; https://doi.org/10.1101/2020.09.01.20186213doi: medRxiv preprint Cover letter Dear Editor We sub mit to PlosMedicine an article entitled “Performance of serum apolipoprotein-A1 as a sentinel of Covid-19”. I share my professional life with apolipoprotein-A1 (ApoA1) since 1982, with a prospective cohort of 975 heavy French drinkers, followed by several other cohorts looking for fibrosis biomarkers. In February 2020, serendipity scenario. Our Hepatology department of Pitié-Salpêtrière, was transformed in a Covid-19 center and I rapidly observed in these severe patients a specific profile of Apo-A1 decrease and Haptoglobin increase, never observed before. The night after, I realize that we have online more than 1 million of ApoA1 and Haptoglobin in the cohorts of patients followed by FibroTest for a risk of liver fibrosis. Therefore, we compared daily the ongoing year 2020 to previous years “Covid -Free”. Indeed, there was a highly significant decrease in ApoA1 parallel (2 weeks before) to the confirmed Covid- 19 cases. The most original result was that this decrease started few weeks before the incidence of confirmed cases, suggesting that ApoA1 detected infected cases. Furthermore, the liver function biomarkers as well as the Haptoglobin did not change so early. We observed the normalization of apolipoprotein-A1 in the sera of the French cohort in June, in parallel with the decrease of daily confirmed Covid-19 cases, as well as the plateau in the US cohort. Apolipoprotein-A1 could be a very simple early marker of the Covid-19 pandemic, as suggested in two ongoing surveillance databases, in France and USA with high sensitivity and specificity. It can be done immediately and with low cost in routine surveillance already dosing apolipoprotein-A1, as well as in patients with a clinical suspicion of Covid-19 but with a negative virologic test. Cover Letter All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted September 3, 2020. ; https://doi.org/10.1101/2020.09.01.20186213doi: medRxiv preprint We do think that this article will help clinicians immediately, in the phase of a possible second wave, and epidemiologists for estimating the prevalence of infected subjects in general population, using other surveillance cohorts. Furthermore, our results suggest a potential intestinal transmission without liver signals. This manuscript has not been presented or is not submitted elsewhere. None of the material concerning the large surveillance databases and the prospective observational cohort has been published or is under consideration for publication elsewhere. According to the context of the pandemic, for assessing the specificity we used controls previously published in large studies of liver diseases. This permitted to have a wide spectrum of false positive risk. T hese measurements have been all performed on fresh serum collected prospectively and analyzed in the biochemistry unit of Pitié-Salpêtrière Hospital, Paris, France, with the same biochemical methods as for the prospective Covid-19 cases. The prospective observational study in Covid-19 patients was approved by CER-Sorbonne University IRB, N° 2020– CER-2020-14, with a signed informed consent. Clinical investigation was conducted according to the principles of the Declaration of Helsinki. All authors had access to the study data and reviewed and approved the final manuscript. The components of Fibrotest and Nash-FibroTest were measured in routine in all patients with a risk of liver disease in France. The different databases integrated were used on sera anonymously. All the retrospective database analyses previously published either non-interventional study, without supplementary blood sample, were exempt from institutional review board (IRB) review (ethical committee of ‘Comité de Protection des Personnes of Paris, Ile -de-France’ FIBROFRANCE project. CPP - IDF-VI, 10– 1996-DR-964, DR-2012-222 and USA-NCT01927133). This project was funded by the European Grant EIT health ProCoP 20879, Patrice Cacoub being the awarded author. All the coauthors have reviewed this manuscript. The data that support the findings of this study are fully available from the corresponding author, upon request. I have a potential conflict of interest as the inventor of FibroTest, and founder of BioPredictive a spinoff company of Sorbonne University. The patent belongs to the French public organization Assistance Publique Hôpitaux de Paris. Kind regards Thierry Poynard All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted September 3, 2020. ; https://doi.org/10.1101/2020.09.01.20186213doi: medRxiv preprint 1 Performance of serum apolipoprotein-A1 as a sentinel of Covid-19 1 2 3 Short title: Apolipoprotein-A1 decrease in Covid-19 4 5 6 7 Thierry Poynard1*, Olivier Deckmyn2*, Marika Rudler 3, Valentina Peta 2, Yen Ngo2, Mathieu 8 Vautier4, Sepideh Akhavan5, Vincent Calvez6, Clemence Franc2, Jean Marie Castille2, Fabienne 9 Drane2, Mehdi Sakka 6, Dominique Bonnefont-Rousselot6, Jean Marc Lacorte7, David 10 Saadoun4, Yves Allenbach4, Olivier Benveniste 4, Frederique Gandjbakhch9, Julien Mayaux10 11 Olivier Lucidarme 11, Bruno Fautrel 9, Vlad Ratziu 3,12, Chantal Housset 1, Dominique 12 Thabut3 and Patrice Cacoub4 13 14 Authors’ affiliations 15 1. Assistance Publique-Hôpitaux de Paris (AP-HP), Sorbonne University, INSERM, Centre de 16 Recherche Saint-Antoine (CRSA), Institute of Cardiometabolism and Nutrition (ICAN) 17 2. BioPredictive, Research, Paris, France 18 3. AP-HP Pitié-Salpêtrière, Sorbonne University, Department of Hepatology, Paris, France 19 4. AP-HP Pitié-Salpêtrière, Sorbonne University, Department of Internal Medicine and Clinical 20 Immunology, Paris, France 21 5. AP-HP Pitié-Salpêtrière, Sorbonne University, Department of Virology, Paris, France 22 6. AP-HP Pitié-Salpêtrière, Sorbonne University, Department of Metabolic Biochemistry, 23 Paris, France 24 7. AP-HP Pitié-Salpêtrière, Sorbonne University, Department of Biochemistry, Endocrinology 25 and Oncology, Paris, France 26 8. AP-HP Pitié-Salpêtrière, Sorbonne University, Institut of Cardiometabolism and Nutrition 27 ICAN, Paris, France 28 9. AP-HP Pitié-Salpêtrière, Sorbonne University, Department of Rhumatology, Paris, France 29 10. AP-HP Pitié-Salpêtrière, Sorbonne University, Department of Intensive Care, Paris, France 30 11. AP-HP Pitié-Salpêtrière, Sorbonne University, Department of Radiology, Paris, France 31 12. INSERM Sorbonne University UMRS 1269 Nutriomique, service de Nutrition, APHP *These authors contributed equally to this work. 32 33 Corresponding author: Thierry Poynard, Hepatology Groupe Hospitalier Pitie-Salpêtrière, 57 34 Bd Hôpital 75013 Paris, France. [email protected] . +(33) Tel: 142161022 Fax: 35 142161427 36 37 Funding source: Grant EIT health ProCoP 20879 Pr Patrice Cacoub APHP France 38 39 Manuscript Click here to access/download;Manuscript;Sentinel of Covid19 TP V22 noFig.docx All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted September 3, 2020. ; https://doi.org/10.1101/2020.09.01.20186213doi: medRxiv preprint 2 Electronic word count: 3,997 words (Including legends) 40 41 Number of figures and tables: 3 figures. 2 tables. 42 43 Conflict of interest statement: a statement to declare any conflict of interest. 44 Thierry Poynard is the inventor of FibroTest, founder of BioPredictive, the patents belong to 45 the public organization Assistance Publique-Hôpitaux de Paris. Olivier Deckmyn, Valentina 46 Peta, Yen Ngo, Jean Marie Castille, Fabienne Drane and Clemence Franc are full employees of 47 BioPredictive. The other authors have nothing to declare. 48 49 50 Authors’ contribution 51 Thierry Poynard, M.D. PhD (Conceptualization: Lead; Data curation: Equal; Formal analysis: Equal; Funding 52 acquisition: Equal; Investigation: Equal; Methodology: Lead; Project administration: Lead; Resources: Equal; 53 Supervision: Lead; Validation: Lead; Visualization: Lead; Writing – original draft: Lead; Writing – review & 54 editing: Lead) 55 Olivier Deckmyn, PhD (Conceptualization: Equal; Data curation: Equal; Methodology: Equal; Software: Lead; 56 Validation: Equal; Visualization: Equal; Writing – original draft: Equal; Writing – review & editing: Equal) 57 Marika Rudler, MD, PhD (Data curation: Equal; Investigation: Equal; Resources: Equal) 58 Valentina Peta, PhD (Data curation: Equal; Investigation: Equal; Resources: Equal; Writing original draft: Equal; 59 Writing review & editing: Equal) 60 Yen Ngo, MD PhD (Data curation: Equal; Investigation: Equal; Methodology: Equal; Validation: Equal) 61 Mathieu Vautier, MD (Data curation: Equal; Investigation: Equal; Resources: Equal) 62 Sepideh Akhavan, MD PhD (Data curation: Equal; Investigation: Equal; Resource s: Equal; Validation: Equal) 63 Vincent Calvez, MD PhD (Resources: Equal; Supervision: Equal; Validation: Equal) 64 Clemence Franc, MS (Formal analysis: Equal; Investigation: Equal; Writing – original draft: Equal) 65 Jean Marie Castille, PhD (Funding acquisition: Equal; Project administration: Equal; Validation: Equal) 66 Fabienne Drane, Ms (Investigation: Equal; Resources: Equal; Validation: Equal) 67 Mehdi Sakka, MD PhD (Investigation: Equal; Resources: Equal; Validation: Equal) 68 Dominique Bonnefont-Rousselot, Md PhD (Resources: Equal; Supervision: Equal; Validation: Equal) 69 Jean Marc Lacorte, MD PhD (Resources: Equal; Supervision: Equal; Validation: Equal) 70 David Saadoun, MD PhD (Data curation: Equal; Investigation: Equal; Resources: Equal) 71 Yves Allenbach, MD PhD (Data curation: Equal; Investigation: Equal; Resources: Equal) 72 Olivier Benveniste, MD PhD (Investigation: Equal; Resources: Equal; Supervision: Equal) 73 Frederique Gandjbakhch, MD PhD (Data curation: Equal; Investigation: Equal; Resources: Equal) 74 Julien Mayaux, MD PhD (Data curation: Equal; Investigation: Equal; Resources: Equal) 75 Olivier Lucidarme, MD PhD (Investigation: Equal; Resources: Equal; Validation: Equal) 76 Bruno Fautrel, MD PhD (Investigation: Equal; Resources: Equal; Supervision: Equal) 77 Vlad Ratziu, MD PhD (Formal analysis: Equal; Investigation: Equal; Resources: Equal) 78 Chantal Housset, MD PhD (Formal analysis: Equal; Supervision: Equal; Validation: Equal; Writing – original 79 draft: Equal) 80 Dominique Thabut, MD PhD (Project administration: Equal; Resources: Equal; Supervision: Equal; Validation: 81 Equal) 82 Patrice Cacoub, MD PhD (Conceptualization: Lead; Formal analysis: Lead; Funding acquisition: Equal; 83 Investigation: Equal; Project administration: Lead; Resources: Equal; Supervision: Lead; Validation: Lead; 84 Visualization: Lead; Writing – original draft: Equal; Writing review & editing: Equal) 85 86 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted September 3, 2020. ; https://doi.org/10.1101/2020.09.01.20186213doi: medRxiv preprint 3 Key Points 87 Question: Does serum apolipoprotein-A1 decrease could be a very early biomarker of SARS-88 CoV-2 pandemic? 89 Findings: During the 30 weeks of 2020 we observed in two large cohorts of patients at risk of 90 liver fibrosis a highly significant decrease of serum apolipoprotein-A1, not observed in previous 91 years. This decrease was highly correlated to and in parallel with the daily increase in confirmed 92 Covid-19 cases, including the recovery period. 93 Meaning: Apolipoprotein-A1 could be used as a sentinel of the pandemic in existing routine 94 surveillance of the general population. 95 96 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted September 3, 2020. ; https://doi.org/10.1101/2020.09.01.20186213doi: medRxiv preprint 4

97

98 Since 1920, a decrease in serum cholesterol has been identified as a marker of severe 99 pneumonia. We have assessed the performance of serum apolipoprotein -A1, the main 100 transporter of HDL-cholesterol, to identify the early spread of coronavirus disease 2019 (Covid-101 19) in the general population and its diagnostic performance for the Covid-19. 102

103 We compared the daily mean serum apolipoprotein-A1 during the first 30 weeks of 2020 in a 104 population that is routinely followed for a risk of liver fibrosis risk in the USA (183,112 sera) 105 and in France (18,316 sera) in relation to a local increase in confirmed cases, and in comparison 106 to the same period in 2019 (respectively 234,881 and 26,056 sera). We prospectively assessed 107 the sensitivity of this marker in an observational study of 136 consecutive hospitalized cases 108 and retrospectively evaluated its specificity in 7,48 1 controls represent ing the general 109 population. 110

111 The mean serum apolipoprotein-A1 levels in these populations began decreasing in January 112 2020, compared to the same 30 weeks in 2019. This decrease was highly correlated to and in 113 parallel with the daily increase in confirmed Covid-19 cases in the following 30 weeks, in both 114 France and USA, including the June and mid-July recovery periods in France. Apolipoprotein-115 A1 at the 1.25 g/L cutoff had a sensitivity of 90.6% (95%CI84.2 -95.1) and a specificity of 116 96.1% (95.7-96.6%) for the diagnosis of Covid-19. The area under the characteristics curve was 117 0.978 (0.957-0.988), and outperformed haptoglobin and liver function tests. The adjusted risk 118 ratio for survival without transfer to intensive care unit was 5.61 (95%CI 1.02-31.0;P=0.04). 119

120 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted September 3, 2020. ; https://doi.org/10.1101/2020.09.01.20186213doi: medRxiv preprint 5 Apolipoprotein-A1 could be both a sentinel of the pandemic in existing routine surveillance of 121 the general population with no new blood sample, as well as a candidate predictor of suspected 122 Covid-19 in multivariate analysis in cases with a negative virologic al test. NCT01927133, 123 CER-2020-14. 124 125

126 SARS-CoV-2; COVID-19; ApoA1; HDL-cholesterol; A2M; haptoglobin; ALT, GGT; 127 pandemic early biomarker; FibroFrance cohort 128 129 130 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted September 3, 2020. ; https://doi.org/10.1101/2020.09.01.20186213doi: medRxiv preprint 6

131 There is an urgent need to detect people in the general population who are at risk of 132 being admitted to the hospital for Covid-19. Although v iral nucleic acid testing and chest 133 computed tomography are standard methods for diagnosing Covid-19 in patients with 134 symptoms, these are time consuming. A review reported that there are only three models for 135 predicting hospital admission in the healthy general population, and these are limited by a high 136 risk of bias, and by using proxy outcomes.1 137 In the meantime, several early detection tests of cytokine burst could be immediately 138 available. In 1920, Harold A. Kipp found that a decrease in serum cholesterol was a marker of 139 severe pneumonia.2 One hundred years later, a meta -analysis confirmed that levels of high-140 density lipoprotein cholesterol (HDL) and a level below the median of its transporter , 141 apolipoprotein-A1, was associated with a two-fold increase in mortality in patients with severe 142 sepsis (supplementaryFig1,supplementaryTable1).3 Its decrease seems to occur earlier than 143 the increase in haptoglobin, another marker of sepsis.4,5 Furthermore, unlike haptoglobin, which 144 is synthesized mainly by the liver, apolipoprotein-A1 is synthesized by both the liver and the 145 intestine.5 This hypolipidemia begins in Covid-19 patients with mild symptoms.6,7 Therefore, 146 in the general population, apolipoprotein -A1 might be a sensitive marker of infection in 147 asymptomatic subjects, in those without the pulmonary symptoms of standard Covid-19, or in 148 patients with an intestinal route of infection (supplementaryFile1).8,9 149 150 Methods. 151 Ethics 152 The prospective observational study in Covid-19 patients was approved by CER-Sorbonne 153 University IRB,CER-2020-14, with a signed informed consent. All of the previously published 154 patient analyses from retrospective databases were non -interventional studies, without 155 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted September 3, 2020. ; https://doi.org/10.1101/2020.09.01.20186213doi: medRxiv preprint 7 supplementary blood samples, and were exempt from a review of the IRB 156 (NCT01927133,supplementaryFile2). The investigation was performed according to the 157 principles of the Declaration of Helsinki. All authors had access to the study data and reviewed 158 and approved the final manuscript. 159 Decrease of apolipoprotein-A1 160 We used three cohorts of sera from subjects at risk of liver fibrosis followed by FibroTest 161 (FibroSure in USA ,supplementaryFile2),10 a large private-laboratories US cohort (“US-162 cohort”), a cohort with an intermediate risk of Covid-19 “(French-cohort”) were patients being 163 followed in academic and private-laboratories, and a high-risk cohort which included patients 164 at the Pitié-Salpêtrière hospital Paris, France (“APHP-PSL”). The c ore temporal analysis 165 compared the first 30 weeks of consecutive anonymous sera 2020 vs. the sera 2019, from these 166 three routinely followed cohorts (supplementaryFile2, supplementaryFig2C). 167 Confounding factors 168 Decrease in apolipoprotein-A1 may be due to direct liver toxicity from SARS-CoV-2,11 but also 169 to drug-induced liver disease (DILI) caused by medications (supplementaryFile2, 170 supplementaryTable2). We analyzed the kinetics of alpha2-macroglobulin (A2M) a specific 171 marker of liver fibrosis ,10 and of haptoglobin, a sensitive biomarker of severe acute phase, as 172 well in the US-cohort of all the other components, according to non-alcoholic fatty liver disease 173 (NAFLD) sera, or chronic hepatitis C (HCV) sera(supplementaryFile2). 174 Daily apolipoprotein-A1 and spread of Covid-19 175 The number of confirmed Covid-19 cases in France and in the USA was assessed according to 176 published data from the European Centre for Disease Prevention and Control 177 (https://ourworldindata.org/coronavirus-data) (supplementaryFig2A ,supplementaryFig2B). 178 Sensitivity and prognostic value 179 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted September 3, 2020. ; https://doi.org/10.1101/2020.09.01.20186213doi: medRxiv preprint 8 Sensitivity was assessed in a prospective study of Covid-19 patients hospitalized in APHP-PSL, 180 described in supplementaryFile2. The primary endpoint was the survival without transfer in 181 ICU at 28 days, adjusted on age, gender, haptoglobin, and liver tests. (supplementaryTable1, 182 supplementaryTable2,supplementaryFile1. 183 Specificity 184 We collected prospective data on apolipoprotein-A1 from and since the first cohort of alcoholic 185 liver disease in 1982, and from the FibroFrance cohort (supplementaryFile2). More recently, 186 we collected six cohorts, which allowed us to retrospectively validate the specificity in a large 187 group of subjects (supplementaryFile2).12-16 The measurements were all performed on fresh 188 prospectively collected serum and analyzed in the biochemistry unit of the APHP-PSL hospital, 189 with the same methods as the Covid-19 cases ( supplementaryTable3). The core control 190 population was a group of healthy volunteers that was representative of the French population.16 191 Forty-three patients without suspected Covid-19 were excluded (supplementaryTable4). In 192 order to identify a profile of patients with a possible intestinal route of infection, we compared 193 the subsets of patients with or without diarrhea. 194 Apolipoprotein-A1, haptoglobin, A2M, gammaglutamyl transpeptidase (GGT), alanine 195 aminotransferase (ALT) were assessed (supplementaryTable2,supplementaryFile1), 196 following BioPredictive (Paris, France) analytical recommendations.17 The virological methods 197 for the diagnosis used to diagnose SARS -CoV2 in respiratory samples , were detailed in 198 supplementaryFile2. 199 Statistical methods 200 The first primary endpoint was to demonstrate a significant decrease of the mean daily value of 201 apolipoprotein-A1 in the 30 weeks of the year 2020, vs. those of the year 2019 . We defined a 202 significant decrease of apolipoprotein-A1 as below 1.25 g/L, the optimal cutoff defined by the 203 highest Youden index (sensitivity + specificity -1) in the severe patients. The daily proportion 204 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted September 3, 2020. ; https://doi.org/10.1101/2020.09.01.20186213doi: medRxiv preprint 9 of sera below this cutoff defined a significant risk of Covid-19 . The second analysis was to 205 exclude co nfounding factors detailed in supplementaryFile1. We compared the l inear 206 regression curves with 95% confidence intervals (95%CI) between the mean daily levels of all 207 markers, and the daily proportion of low apolipoprotein -A1, between the same 30 weeks 208 periods per year, by the F-test. A significant difference was defined as P<0.001. The third 209 primary endpoint was to assess the association between the spread of Covid-19 and the decrease 210 in apolipoprotein -A1. The diagnostic performance of biomarkers was assessed using non -211 parametric AUROCs. The prognostic values were assessed by survival curves, using Kaplan-212 Meir method, compared by Logrank test (cutoff being the median of this context of use) and 213 adjusted by Cox model. The repeated sera assess ments were compared by repeated ANOVA 214 and Tukey-Kramer multiple -comparison test . R and NCSS -2020 were used as statistical 215 software. 216 217 218

219 Decrease of apolipoprotein-A1 220 The mean daily levels of apolipoprotein-A1 decreased globally (fig1A) and in the three cohorts 221 (fig1B; all P<0.001), during the first 30 weeks of 2020. This was already highly significant in 222 January 2020 in the US-cohort (lower panel), compared to first 30 weeks of 2019 and 2018. 223 The mean daily proportion of sera with low apolipo protein-A1 was 32.4%(95%CI 32.4-32.5) 224 on 30 weeks, that is 5.1% higher vs. 26.8%(27.2-27.4;P<0.001) in 2019 (fig1C). There was no 225 significant difference between the years 2019 and 2018 for the daily mean globally (fig1A), or 226 by cohort ( fig1B). The proportion of low apolipoprotein -A1, in 2018 was 26.5%(26.3-26.6), 227 that is a small 0.8% difference vs. the year 2019 (fig1C). 228 Confounding factors 229 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted September 3, 2020. ; https://doi.org/10.1101/2020.09.01.20186213doi: medRxiv preprint 10 The same significant kinetics in apolipoprotein-A1 levels were observed after stratification of 230 the regression curves for gender and age in the three cohorts (supplementaryFig3). The US-231 cohort was the only sample that had the necessary power to compare these two factors together 232 between 2020 and 2019 and 2018 (fig1D), and in the subset of patients with HCV 233 (supplementaryFig3C). Apolipoprotein-A1 decrease was similar during Covid-19 spread 234 versus 2019 and 2018 in the US-cohort(supplementaryFile2, supplementaryFig3). 235 The kinetics of apolipoprotein-A1 were not associated with those of haptoglobin 236 (supplementaryFig4). 237 For A2M, there was in the US-cohort only a significant lower mean serum value, when 238 compared to 2019 and even more compared to 2018 (supplementaryFig5A), and detailed in 239 supplementaryFig2, supplementaryFig5), and highest in HCV vs. NAFLD sera . After 240 adjustment on age and gender there was no significant decrease of A2M between 2020 and 241 2019 (supplementaryFig5). 242 The other significant differences were, in the French-cohort s, GGT increased during the 243 pandemic peak and returned to previous years’ value thereafter (supplementaryFig6), and an 244 increase in ALT in April 2020, in the US-cohort (supplementaryFig7). 245 No significant changes were observed for total cholesterol (supplementaryFig8), triglycerides 246 (supplementaryFig9), fasting glucose (supplementaryFig10), weight (supplementaryFig11) 247 and height (supplementaryFig12). 248 Temporal associations between Covid-19 cases and apolipoprotein-A1 249 The daily mean number of confirmed Covid-19 cases paralleled about 10 days after the daily 250 proportion of low apolipoprotein -A1 in sera. In USA, the first ten Covid-cases were declared 251 mid-January 2020, when the proportion of low apolipoprotein-A1 had already increase by 252 several percent (fig1E). The first peak of cases (n=50,000) was reached mid-April 253 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted September 3, 2020. ; https://doi.org/10.1101/2020.09.01.20186213doi: medRxiv preprint 11 (supplementaryFig2B), and the first peak of low apolipoprotein-A1 (37 .0%) on April 7 th 254 (fig1E, supplementaryFile2). 255 In France, the first ten Covid-cases were declared first week of March 2020 256 (supplementaryFig2A), when the proportion of low apolipoprotein -A1 started to increase by 257 several percent (fi g1C). The first peak of cases (n= 7,000) was reached mid -April 258 (supplementaryFig2A), as well as the first peak of low apolipoprotein -A1 (40.0%) on April 259 14th (fig1C, supplementaryFile2). 260 Diagnostic performance of apolipoprotein-A1 261 A total of 136 consecutive patients with severe Covid-19 , but who did not require ICU were 262 include. Their characteristics and were similar to those published in such severity profiles 263 (Table 2,supplementaryFile3). 264 The characteristics of patients included for the assessment of specificity are presented in 265 supplementaryTable3, the differences between characteristics of the subsets in 266 supplementaryTable3, in fig2A for the median value of apolipoprotein-A1 and in fig2B for 267 haptoglobin at inclusion. 268 The area under the characteristics curve (AUROC;95%CI) in 136 Covid-19 cases and 7,481 269 controls was 0.978(0.957 -0.988), which outperformed haptoglobin and liver function tests 270 (fig2C). Apolipoprotein-A1 at a cutoff of 1.25 g/L, had the best Youden index (86.7%) with a 271 sensitivity of 90.6%(84.2 -95.1) and a specificity of 96.1%(95.7 -96.6) for the diagnos is of 272 Covid-19. 273 For a prevalence of 1.8% (136/7617;1.5-2.1) of Covid-19 cases, the positive predictive value 274 was 30.0%(25.6-34.7) and the negative predictive value was 99.8%(99.7 -99.9). The adjusted 275 predictive values according to prevalence predicted in the French population, 16 were detailed 276 in supplementaryFile2 as well as the specificity -sensitivity including blood donors 277 (supplementaryFig12), and patients with rheumatologic al disease ( supplementaryFig13), 278 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted September 3, 2020. ; https://doi.org/10.1101/2020.09.01.20186213doi: medRxiv preprint 12 and the integrated six databases ( supplementaryFig14), and the sensitivity in the 19 patients 279 with negative viral nucleic acid testing. 280 The prognostic value of apolipoprotein-A1 at inclusion for predicting the primary outcome was 281 significant, r isk-ratio (RR;95%CI) =5.61 (1.02-31.0;P=0.04), adjusted on age (1.04;1.01-282 1.07;P=0.04), GGT (2.88;1.01-8.19;P=0.04). The 71 patients with apolipoprotein-A1 283 value>=0.84 g/L, the median value at inclusion, had a significant higher survival without ICU 284 (93.0%;87.0-98.9) than the 65 patients with lower value (75.8%; 65.1-86.5; P=0.02) (fig2D). 285 Repeated assessments of 305 sera (supplementaryFig15) showed that in patients who survived 286 without ICU, the mean apolipoprotein-A1 raised significantly already at 10 days, as well as the 287 16 patients who survived after their transfer to ICU. 288 The prevalence of diarrhea on initial presentation was 29 out of 131 cases (22.1%;95%CI 15.4-289 30.2). In this subset, the only significant difference was a lower median number of polynuclear 290 leucocytes (supplementaryTable5, supplementaryFile2). 291 292 293

294 Our study shows in three cohorts of patients at risk of liver fibrosis, that apolipoprotein-A1 had 295 a highly significant decrease in 2020 vs previous years, and a highly significant negative daily 296 time-related association with the number of Covid-19 cases. Apolipoprotein-A1 decrease had 297 a high sensitivity in prospective hospitalized patients, with a high specificity in retrospective 298 controls, and an independent prognostic value for the survival without transfer to ICU . These 299

have certain strengths and limitations. 300 Decrease of apolipoprotein-A1 in 2020 301 The originality of these results was not the decrease in apolipoprotein-A1 during the peak of 302 the pandemic in April, as very low levels of HDL -cholesterol in sera collected in Covid -19 303 were known in severe pneumonia since 1920 (supplementaryTable1).2 More intriguing was 304 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted September 3, 2020. ; https://doi.org/10.1101/2020.09.01.20186213doi: medRxiv preprint 13 the very early decrease observed since January 2020 in the USA when the number of Covid-19 305 cases was unknown. The first known Covid-19 patient was detected on 27/12/2019 and 306 19/01/2020 in France and the USA, respectively(supplementaryFile3). The larger sample size 307 of the US surveillance population , compared to the French, allowed detection of a significant 308 1% increase in the proportion of subjects possibly infected using the 1.25 g/L cutoff in January 309 (fig1C), without any inflammatory signal using haptoglobin. We hypothesized that the SARS-310 CoV2 virus influenced the liver or intestinal synthesis of apolipoprotein-A1, in asymptomatic 311 patients or in those with unusual mild symptoms. 312 Confounding factors 313 The decrease of apo lipoprotein-A1 in 2020 vs. previous years, as well as the time -related 314 association of apolipoprotein-A1 in 2020 and Covid-19 might be due to numerous confounding 315 factors. In the context of the pandemic we used cohorts of subjects requiring surveillance of 316 liver fibrosis biomarkers which represent at least 30% of the general adult population in the 317 USA and in France. In these cohorts 70% of the subjects had no or minimal fibrosis. The 318 decrease in apolipoprotein-A1 in 2020 compared to 2019, and 2018 cannot be explained by bias 319 due to gender, age, the cause of liver disease (table1, supplementaryFile3), and the severity 320 of liver diseases (supplementaryFile4). The prevalence of severe cases cannot explain the 321 significant decrease in apolipoprotein -A1 already observed in January 2020 . GGT a very 322 sensitive liver biomarkers did not change during the first 3 months (supplementaryFig6). 323 Finally, in these severe liver diseases, haptoglobin should be also significantly decreased 324 (fig2A), which was not observed. 325 In the US-cohort the proportion of sera with NAFLD was increased by 1.8% in 2020 vs 2019 326 (table1). However, after stratification for age and gender, no significant changes were observed 327 for all other biomarkers (supplementaryFig8,supplementaryFig9,supplementaryFig10, 328 supplementaryFig11,supplementaryFile12). ALT was the only biomarker of the liver tests 329 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted September 3, 2020. ; https://doi.org/10.1101/2020.09.01.20186213doi: medRxiv preprint 14 which increased significantly at the 13th week of 2020, above the usual mean value observed in 330 2019 (supplementaryFile7). This increase in ALT was not associated with any other changes 331 (supplementaryFig10, supplementaryFig11,supplementaryFig12). We have no clear 332 explanation. We hypothesize that another confounding factor could be a DILI, including oral 333 acetaminophen or hydroxychloroquine misuse during the pandemic. Such factor could explain 334 an increase in ALT, without increase in haptoglobin, in subjects with mild symptoms. 335 The absence of haptoglobin change ( supplementaryFig4) associated with the linear 336 apolipoprotein-A1 decrease (fig1B), has never been described before. It was known that in 337 patients with severe fibrosis these two proteins decrease.10 It was also known that in severe 338 pneumonia the haptoglobin increase was associated to the apolipoprotein-A1 decrease 339 (supplementaryTable1), as we observed in the cohorts with high prevalence of severe Covid-340 19 (fig2B)(supplementaryFile3). The 30 weeks followup permitted to see the return to normal 341 values of haptoglobin in the se cohorts, associated to the decrease of severe Covid -19 cases 342 admissions (supplementaryFig4A). Furthermore, the recovering patients followed by repeated 343 sera in the prospec tive study, had both a significant increase of apolipoprotein-A1 344 (supplementaryFig15B) and a significant decrease in haptoglobin ( supplementaryFig15D), 345 10 days after inclusion. 346 Temporal associations between Covid-19 confirmed cases and apolipoprotein-A1 decrease 347 Our results (fig1C) suggest that the spread of the pandemic in the US-cohort would be at least 348 around 5.6%, and at least of 4.3% in the French-cohort. In France, this estimate does not differ 349 from the recent French model that predicted a rate of infection of between 2.8% to 7.2% in the 350 general population.18 351 This temporal association of low apolipoprotein-A1 with the number of confirmed cases, 352 persisted in the two different pandemic changes. In France, after the national lockdown the 353 proportion of low apolipo protein-A1 returned to the cohort usual 2018-2019 values in June 354 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted September 3, 2020. ; https://doi.org/10.1101/2020.09.01.20186213doi: medRxiv preprint 15 together with the dramatic regression of confirmed cases, still maintained in July. In the US, 355 the apolipoprotein A1 and confirmed cases had the same kinetics with the dramatic increase in 356 April and a plateau in June, still ongoing in July (fig1E). 357 Specificity of apolipoprotein-A1 358 There is a high risk of overestimating the specificity of a test in Covid-19 when the participants 359 enrolled in the studies might not be representative of targeted populations. 1 However, our 360 previous cohorts of patients with severe liver diseases allowed us to identify the major risks of 361 a significant decrease in apolipoprotein-A1 (fal se positives), mainly due to severe hepatic 362 insufficiency and severe fibrosis. An impact of malnutrition on apolipoprotein-A1 values, was 363 not possible as no changes in weight and height were observed. 364 Sensitivity of apolipoprotein-A1 365 There is also a high risk of overestimating the sensitivity. 1 It is difficult to confirm the 366 sensitivity for asymptomatic infection, due to the absence of validations of SARS -CoV2-367 antibodies. To validate the sensitivity of apolipoprotein -A1, a large number of asymptomatic 368 “apparently healthy” subjects who are positive for SARS -CoV-2 viral nucleic acid testing is 369 needed. Our patients included with Covid-19 symptoms had a median of 7 0 years of age and 370 were severe enough to justify admission to hospital , but none of them required mechanical 371 ventilation at admission, 86% survived, only 6% were transferred to the intensive care unit, and 372 9% died ( table1). The sensitivity in the prospective part of our study was similar in the 19 373 patients who were negative for viral nucleic acid testing (94.7%) to that in positive patients 374 (89.9%). Thus, a simple measurement of apolipoprotein-A1 could be useful for clinicians due 375 to the high percentage of false negatives in available viral nucleic acid testing.18Apolipoprotein-376 A1 or HDL -cholesterol are already being assessed in many large ongoing studies in patients 377 with chronic diseases, which could rapidly validate our results. 378 Prognostic performance 379 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted September 3, 2020. ; https://doi.org/10.1101/2020.09.01.20186213doi: medRxiv preprint 16 The univariate prognostic value of apolipoprotein- A1 was confirmed ( fig2D) but to our 380 knowledge, it was the first time that its prognostic value persisted after adjustment by age, a 381 marker of acute inflammation (haptoglobin), a sensi tive marker of liver injury (GGT), and a 382 marker of liver fibrosis (A2M). Among patients with recovery, there was a significant increase 383 of apolipoprotein -A1 10 days a fter admission (supplementaryFig15). Therefore, 384 apolipoprotein-A1 measurement could help for the decision to transfer to ICU. 385 Mechanisms of the early decrease in apolipoprotein-A1 before recognition of the pandemic 386 We never observed such profile of biomarkers in our experience since 2001 with more than 387 three million of FibroTest assessed in liver diseases (supplementaryFile1).10,12-17 388 Although the mechanisms explaining the decrease in apolipoprotein-A1 in late severe Covid-389 19 pneumonia are known (supplementaryTable1), the reason for the early decrease before the 390 acute phase, when haptoglobin remained normal is unclear. In patients with severe pneumonia, 391 apolipoprotein-A1 decrease was associated with acute inflammation and the “cytokine storm” 392 with an increase in IL6 and acute phase proteins such as CRP and haptoglobin. This dissociation 393 suggests that different mechanisms play a role in the early influence of the SARS-CoV2 virus 394 on the synthesis of apolipoprotein-A1, and the intestine could be more involved than the liver. 395 The SARS-CoV2 virus could impact several pathways leading to a decrease in the intestinal 396 synthesis and absorption of apolipoprotein-A1 in the small intestine resulting in the decrease in 397 serum.3-9 These mechanisms are discussed in supplementaryFile1. The first could the 398 inhibition of lysophosphatidylcholine-acyltransferase-3 activity. There is evidence of direct 399 SARS-CoV2 infection of the endothelial cell and diffuse endothelial inflammation in the 400 intestine.7,8,9 SARS-CoV2 uses angiotensin-converting enzyme-2 receptor (ACE-2) expressed 401 on endothelial cells, to infect the host, widely expressed in the lung, intestine and liver.9,20 The 402 second mechanism could be an impact of the virus through the intestinal mucus .21 403 Apolipoprotein-A1 is released as a free apolipoprotein from the apical side of enterocytes into 404 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted September 3, 2020. ; https://doi.org/10.1101/2020.09.01.20186213doi: medRxiv preprint 17 the lumen in the fasting state. Apolipoprotein-A1 had faster turnover in mucus, which could be 405 a target for SARS-CoV2.7,8,21 406

407 Despite the limitations of this study, these results suggest that apolipoprotein-A1 could 408 be a component of multi-analyte Covid-19 diagnostic and prognostic tests. It could also help to 409 manage patients with a clinical suspicion of Covid-19 and a negative virologic al test. These 410

must be v alidated in independent large cohorts, ideally with virologic and efficiency 411 endpoints. The role of SARS-CoV2 in the possible “asymptomatic” decrease in apoliprotein-412 A1 could be related to intestin al infection without or before overt pulmonary disease. Finally, 413 one hundred years after surrogate sentinel HDL-cholesterol for pneumonia,2 apolipoprotein-A1 414 for the second time in a century, could be “one of the early warning systems that alert the world 415 to potential outbreaks”.22 416 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted September 3, 2020. ; https://doi.org/10.1101/2020.09.01.20186213doi: medRxiv preprint 18

417 1. Wynants L, Van Calster B, Bonten MMJ, et al. Prediction models for diagnosis and prognosis 418 of Covid-19 infection: systematic review and critical appraisal. BMJ 2020;369:m1328. 419 2. Kipp HA. Variation in the cholesterol content of the serum in pneumonia. J Biol Chem 1920; 420 44:215-237. 421 3. Liu SH, Liang HY, Li HY, et al. Effect of low high-density lipoprotein levels on mortality 422 of septic patients: A systematic review and meta -analysis of cohort studies. World J Emerg 423 Med 2020;11:109–116. 424 4. Delanghe JR, Langlois MR, Boelaert JR, et al. Haptoglobin polymorphism, iron metabolism 425 and mortality in HIV infection. AIDS 1998;12:1027-32. 426 5. Yang H, Wang H, Wang Y, et al. The haptoglobin beta subunit sequesters HMGB1 toxicity 427 in sterile and infectious inflammation. J Intern Med 2017;282:76-93. 428 6. Wei X, Zeng W, Su J, et al. Hypolipidemia is associated with the severity of COVID -19. J 429 Clin Lipidol. 2020;14:297‐ 304. 430 7. Cao X, Yin R, Albrecht H,et al. Cholesterol: A new game player accelerating endothelial 431 injuries caused by SARS-CoV-2? Am J Physiol Endocrinol Metab. 2020;319:E197-E202. 432 433 8. Lamers MM, Beumer J, van der Vaart J, et al. SARS-CoV-2 productively infects human 434 gut enterocytes. Science. 2020; eabc1669. 435 436 9. Ding S, Liang TJ. Is SARS-CoV-2 Also an enteric pathogen with potential fecal-oral 437 Transmission? A COVID-19 virological and clinical review. Gastroenterology 438 2020;159:53-61. 439 10. Imbert-Bismut F, Ratziu V, Pieroni L, et al. Biochemical markers of liver fibrosis in patients 440 with hepatitis C virus infection: a prospective study. Lancet 2001; 357:1069-75. 441 11. Zhang C, Shi L, Wang FS. Liver injury in COVID-19: management and challenges. Lancet 442 Gastroenterol Hepatol 2020; 5:428-430. 443 12. Peta V, Tse C, Perazzo H, et al. Serum apolipoprotein A1 and haptoglobin, in patients with 444 suspected drug -induced liver injury (DILI) as biomarkers of recovery. PloS One 2017; 445 12:e0189436. 446 13. Rudler M, Mouri S, Charlotte F, et al. Validation of AshTest as a non-invasive alternative 447 to transjugular liver biopsy in patients with suspected severe acute alcoholic hepatitis. PLoS 448 One 2015; 10:e0134302. 449 14. Perazzo H, Pais R, Munteanu M, et al. Variability in definitions of transaminase upper limit 450 of the normal impacts the APRI performance as a biomarker of fibrosis in patients with chronic 451 hepatitis C: "APRI c'est fini ?". Clin Res Hepatol Gastroenterol 2014; 38:432‐ 439. 452 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted September 3, 2020. ; https://doi.org/10.1101/2020.09.01.20186213doi: medRxiv preprint 19 15. Church RJ, Kullak-Ublick GA, Aubrecht J, et al. Candidate biomarkers for the diagnosis 453 and prognosis of drug -induced liver injury: An international collaborative effort. Hepatology 454 2019; 69:760-773. 455 16. Poynard T, Lebray P, Ingiliz P, et al. Prevalence of liver fibrosis and risk factors in a general 456 population using non-invasive biomarkers (FibroTest). BMC Gastroenterol 2010; 10:40. 457 17. Poynard T, Munteanu M, Deckmyn O, et al. Applicability and precautions of use of liver 458 injury biomarker FibroTest. A reappraisal at 7 years of age. BMC Gastroenterol. 2011;11:39. 459 18. Salje H, Tran Kiem C, Lefrancq N, et al. Estimating the burden of SARS-CoV-2 in France 460 2020;368:6498. Science. 2020;eabc3517. 461 19. Li Y, Yao L, Li J, et al. Stability issues of RT-PCR testing of SARS-CoV-2 for 462 hospitalized patients clinically diagnosed with COVID-19. J Med Virol 2020; 92:903-908. 463 464 20. Xu H, Zhong L, Deng J, et al. High expression of ACE2 receptor of 2019-nCoV on the 465 epithelial cells of oral mucosa. Int J Oral Sci 2020;12:8. 466 467 21. Danielsen EM, Hansen GH, Rasmussen K, et al. Apolipoprotein A-1 deposition in, and 468 release from, the enterocyte brush border: a possible role in transintestinal cholesterol efflux 469 (TICE)? Biochim Biophys Acta 2012; 1818:530‐ 536. 470 22. Gates B. Responding to Covid-19.A Once-in-a-Century Pandemic? New Engl J Med 2020; 471 382:1677-1679. 472 473 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted September 3, 2020. ; https://doi.org/10.1101/2020.09.01.20186213doi: medRxiv preprint 20 Table1. Characteristics of the sera included in US and French-cohorts 474 475 Characteristics 2020 2019 P-Value US-cohort N=183,112 N=234,881 Median age [IQR) 54.1 [40.4-63.3] 55.0 [41.7-63.3] <0.001 Age category n (%) <0.001 < 50 year 75,782(41.4) 90,853 (38.5) 50 to = 70 year 18,733 (10.2) 22,771 (9.7) Male sex (%) 103,672 (56.6) 131,056 (55.8) <0.001 Apolipoprotein-A1<1.25 g/L (%) 59,419 (32.4) 62,898 (26.8) <0.001 Median laboratory (IQR) Apolipoprotein-A1 g/L 1.37 [1.19-1.58] 1.41 [1.23-1.68] <0.001 Haptoglobin g/L 1.23 [0.82-1.68] 1.23 [0.82-1.69] 0.72 Alpha-2-macroglobulin g/L 2.16 [1.65-2.93] 3.07 [1.70-3.07] <0.001 GGT IU/L 36 [21-77] 38 [21-79] 0.87 ALT IU/L1 38 [22-70] 38 [23-68] <0.001 Total bilirubin micromol/L 6.84 [5.13-10.26] 6.84 [5.13-10.26] 0.01 Fibrosis stage by FibroTest (%) <0.001 F0 94,170 (51.4) 116,666 (49.7) F1 36,738 (20.1) 47,123 (20.1) F2 13,330 (7.3) 17,834 (7.6) F3 17,914 (9.8) 24,514 (10.4) F4 19,648 (10.7) 27,030 (11.5) Non interpretable 1,312 (0.7) 1,714 (0.7) NAFLD (surveillance by NASH-FibroTest)2 N=37,298 (20.2) N=42,271 (18.0) <0.001 Weight (Kg) 88.9 (74.8-105.2) 88.5 (74.8-104.3) 0.003 Aspartate amino transferase (AST) 34 (25-52) 34 (25-52) 0.55 Total cholesterol 4.53 (3.80.5.33) 4.53 (3.80-5.28) 0.001 Triglycerides 1.57 (1.13-2.24) 1.55 (1.11-2.21) 0.02 Fasting glucose 5.67 (5.06-6.89) 5.61 (5.06-6.78) <0.001 Table 3 continued French-cohort N= 18,316 N=26,056 Median age [IQR) 51.5 (41.2-64.4) 52.6 (39.7-63.1) <0.001 Age category n (%) <0.001 < 50 year 7,514 (40.9) 11,517 (44.2) 50 to = 70 year 2,615 (14.3) 3,199 (14.3) Male sex (%) 10,619 (58.0) 14,908 (57.2) <0.001 Apolipoprotein-A1<1.25 g/L (%) 5,694 (31.1) 7,057 (27.1) <0.001 Median laboratory (IQR) Apolipoprotein-A1 g/L 1.38 (1.20-1.59) 1.40 (1.23-1.61) <0.001 Haptoglobin g/L 1.18 (0.79-1.62) 1.15 (0.78-1.57) <0.001 Alpha-2-macroglobulin g/L 1.96 (1.52-2.61) 2.01 (1.56-2.69) <0.001 GGT IU/L 36 (21-79) 34 (20-73) <0.001 ALT IU/L 31 (21-50) 31 (21-50) 0.42 Total bilirubin micromol/L 8.7 (6-12) 8.7 (6-12) 0.20 Fibrosis stage by FibroTest (%) F0 9,437 (51.5) 13,671 (52.5) <0.001 F1 4,080 (22.3) 5,708 (21.9) F2 1,328 (7.3) 1,877 (7.2) F3 1,664 (9.1) 2,263 (8.7) F4 1,655 (9.0) 2,336 (9.0) Non interpretable 190 (0.8) 201 (0.8) NAFLD (surveillance by NASH-FibroTest)2 4152 (19.2) 6,110 (20.4) <0.001 Weight 83 (71-97) 83 (71-97) 0.41 Aspartate amino transferase (AST) 30 (23-42) 30 (23-41) 0.47 Total cholesterol 4.79 (4.00-5.55) 4.75 (4.00-5.62) 0.30 Triglycerides 1.37 (1.00-1.96) 1.40 (1.00-2.00) 0.02 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted September 3, 2020. ; https://doi.org/10.1101/2020.09.01.20186213doi: medRxiv preprint 21 Fasting glucose 5.76 (5.15-6.98) 5.72 (5.17-6.90) 0.87 1 Not normal distribution, the means of ALT were in 2020 43.4 IU/L vs 42.7 IU/L in 2019 476 2 NASH-FibroTest included more components than FibroTest: weight, aspartate aminotransferase, total 477 cholesterol, triglycerides, and fasting glucose 478 479 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted September 3, 2020. ; https://doi.org/10.1101/2020.09.01.20186213doi: medRxiv preprint 22 Table2. Characteristics of Covid-19 patients of the prospective study. 480 481 Characteristics COVID-19 diagnostic PCR positive PCR negative Adjudicated P value Total Number n (%) 117 (100%) 19 (100%) 136 Median age (IQR) year 72 (57-82) 65.1 (49-74) 0.15 72 (59-83) Age category 0.16 <50 year 14 (12.0) 5 (26.4) 19 (14.0) 50 to =70 year 66 (56.4) 7 (36.8) 73 (53.7) Male sex 66 (56.4) 12 (63.2) 0.58 78 (57.4) Geographic origin 0.53 Caucasian 67 (57.3) 14 (73.7) 81 (59.6) Subsaharan 17 (14.5) 1 (5.3) 18 (13.2) North African, Middle East 20 (17.1) 2 (10.5) 22 (16.9) Asian 13 (11.1) 2 (10.5) 15 (11.0) Oxygen-support category 0.75 Invasive oxygen support 0 0 0 (0-0) Noninvasive oxygen support 82 (70.1) 14 (73.7) 96 (70.6) None 35 (29.9) 5 (26.3) 40 (29.4) Coexisting conditions Hypertension 65 (55.6) 9 (47.4) 0.51 74 (54.4) Diabetes 32 (27.4) 3 (15.8) 0.29 35 (25.7) Hyperlipidemia 32 (27.4) 7 (36.8) 0.40 39 (28.7) Liver disease 11 (9.4) 4 (21.1) 0.13 15 (11.03) Pulmonary disease 13 (11.6) 1 (5.6) 0.44 14 (10.8) Severe disease associated 64 (55.2) 11(57.9) 0.82 75 (55.6) Tobacco (ongoing or stopped) 35 (29.9) 8 (42.1.3) 0.29 43 (31.6) Alcohol consumption 15 (12.9) 5 (26.3) 0.13 20 (14.8) Initial presentation Anosmia (13 missing) 12 (11.1) 3 (20.0) 0.32 15 (12.2) Ageusia (13 missing) 16 (14.8) 3 (20.0) 0.60 19 (15.5) Headache (13 missing) 10 (9.3) 4 (26.7) 0.05 14 (11.4) Dyspnea (12 missing) 59 (54.1) 8 (53.3) 0.95 67 (54.0) Wheezing (12 missing) 3 (2.75) 0 (0.00) 0.52 3 (2.42) Cough without spitting (12 missing) 47 (43.1) 6 (40.0) 0.82 53 (42.7) Cough with spitting (12 missing) 18 (16.5) 2 (13.3) 0.75 20 (16.1) Fatigue 61 (56.5) 8 (53.3) 0.82 69 (59.1) Time clinic-inclusion (days) 8 (5-13.8) 11 (9-16.5) 0.13 9 (5-14) Followup (days) 38 (27-53) 28 (26-36) 0.05 32 (27-49) BMI (Kg/m2) 25 (23-29) 24 (22-28) 0.56 25 (23-28) Apolipoprotein <= 1.25g/L 104 (88.9) 19 (100) 0.13 123 (90.4) Median laboratory (IQR) Apolipoprotein-A1 g/liter 0.85 (0.73-1.06) 0.74 (0.61-0.87) 0.06 0.84 (0.70-1.03) Haptoglobin g/liter 3.16 (2.20-4.19) 3.18 (1.59-3.86) 0.59 3.16 (2.22-4.08) Alpha-2 macroglobulin g/liter 1.49 (1.24-2.06) 1.46 (1.12-2.01) 0.41 1.49 (1.22-2.05) GGT IU per liter 47 (28-114) 58 (32-102) 0.27 49 (30-114) ALT IU per liter 29 (20-47) 40 (15-57) 0.80 31 (20-51) Total bilirubin micromol/L 7 (6-8) 9 (8-13) 0.01 8 (5-12) Platelets 106 per ml 211 (162-278) 279 (194-330) 0.06 221 (164-287) Creatinine µmol per liter 78 (62-103) 78 (70-120) 0.97 78 (63-103) Fasting glucose 6 (5.4-8.0) 6.6 (5.2-7.9) 0.38 6.1 (5.4-7.9) Albumin g/L (18 missing) 31(27-33) 29 (27-39) 0.89 30 (27-33) Procalcitonin (17 missing) 0.13 (0.08-0.27) 0.14 (0.09-0.23) 0.98 0.13 (0.09-0.27) C-reactive protein (13 missing) 56 (20.3-95.6) 101 (15.4-252.9) 0.21 59 (20.3-102.3) InterLeukin-6 (38 missing) 31 (13.5-55.2) 29 (8.05-119) 0.96 31 (12.3-55.5) CPK (18 missing) 101 (51.5-243.3) 47.5 (37.5-127) 0.05 94 (45.8-233) AST (11 missing) 44 (30-60) 34 (25-42) 0.04 41 (30-59) LDH (16 missing) 347 (278-419) 353 (289-440) 0.68 347 (278-420) D-dimer (24 missing) 1050 (550-2030) 1570 (1140-3980) 0.09 1125 (570-2158) Troponin (14 missing) 16.4 (9.85-34.63) 17.2 (9.56-36.5) 0.99 16.4 (9.85-34.63) Prothrombin time (16 missing) 92 (83-100) 82 (39-89) 0.005 90 (81-100) White cells Neutrophil missing (14 missing) 4185 (2773-6145) 5945 (3963-7550) 0.05 4420 (2928-6200) Eosinophil (12 missing) 0 (0-30) 10 (0-150) 0.14 0.05 (0-30) All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted September 3, 2020. ; https://doi.org/10.1101/2020.09.01.20186213doi: medRxiv preprint 23 Lymphocyte (14 missing) 965 (670-1235) 1150 (798-1700) 0.25 975 (685-1295) Hemoglobin g/L (12 missing) 12 (11.2-13.6) 12 (10.5-14.7) 0.64 12 (11.1-13.8) Treatment at risk of DILI Paracetamol oral (2-4 g/day) 43 (36.8) 5 (26.3) 0.38 48 (35.3) Antibiotics 91 (77.8) 15 (78.9) 0.91 106 (77.9) None 26 (22.2) 4 (21.1) 30 (22.1) Without clavulinate 49 (41.9) 5 (26.3) 54 (39.7) Whit clavulinate 68 (58.1) 14 (73.7) 82 (60.3) Steroids 12 (10.3) 1 (5.3) 0.49 13 (9.56) Hydroxy chloroquine 47 (40.1) 6 (31.6) 0.48 53 (39.0) No Intensive care no death 99 (84.6) 17 (89.5) 0.23 116 (85.3) Transfer intensive care unit 18 (15.4) 2 (10.5) 0.58 20 (14.7%) Death (4 in intensive care unit) 15 (12.8) 1(5.3) 0.34 16 (11.8) 482 483 484 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted September 3, 2020. ; https://doi.org/10.1101/2020.09.01.20186213doi: medRxiv preprint 24 Figures legends 485 486 Fig1. Decrease of apolipoprotein-A1 value in the first 30 weeks of 2020. 487 1A Global decrease of apolipoprotein-A1 among the sera of patients with risk of liver 488 fibrosis in the first 30 weeks of 2020 in USA and France compared to the two previous 489 years (P<0.001). 490 1B Quantitative decrease of apolipoprotein-A1 by cohort. Apolipoprotein-A1 decreased 491 (P<0.001) in the three cohorts, starting early in January 2020 (red line with 95% confidence 492 interval) in the US cohort (lower panel). 493 1C Proportion of serum with low apolipoprotein-A1 by cohorts. Low apolipoprotein-A1 494 was defined as below 1.25 g/L. Details in supplementaryFile2. 495 1D Decrease of apolipoprotein- A1 by gender and age in US cohort. The same significant 496 kinetics were observed, P<0.001 between 2020 and previous years. 497 1E Number of confirmed Covid-19 cases per day and proportion of low (<1.25g/L) 498 apolipoprotein-A1 in the US cohort during the first 30 weeks of 2020. 499 The red graph is the number of confirmed cases per day in logarithmic scale. The black line is 500 the daily mean proportion of low apolipoprotein-A1 (<1.25g/L; blue line;95%CI in grey). 501 502 Fig 2. Performances of apolipoprotein-A1 and haptoglobin for the diagnostic and 503 prognostic of Covid-19 patients. Details in supplementaryFile2. 504 Fig2A. Apolipoprotein-A1 median with IQR between the 6 populations 505 Fig2B. Haptoglobin CPAM: general population ASH: severe acute alcoholic hepatitis, DILI: 506 drug induced liver disease, RHE: rheumatologic disease, BD: blood donors. 507 Fig2C. AUROCs of each FibroTest components. 508 Fig2D. Survival without transfer to intensive care unit (ICU). 509 The 71 patients with apolipoprotein-A1 value >= 0.84 g/L, the median value at inclusion, had 510 a significant higher survival without ICU (93.0%;87.0-98.9) than the 65 patients with lower 511 value (75.8%;65.1-86.5;P=0.02). 512 513 All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted September 3, 2020. ; https://doi.org/10.1101/2020.09.01.20186213doi: medRxiv preprint Figure Click here to access/download;Figure;Fig1A.tiff All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted September 3, 2020. ; https://doi.org/10.1101/2020.09.01.20186213doi: medRxiv preprint Figure Click here to access/download;Figure;Fig1B.tiff All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted September 3, 2020. ; https://doi.org/10.1101/2020.09.01.20186213doi: medRxiv preprint Figure Click here to access/download;Figure;Fig1C.tiff All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted September 3, 2020. ; https://doi.org/10.1101/2020.09.01.20186213doi: medRxiv preprint Figure Click here to access/download;Figure;Fig1D.tiff All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted September 3, 2020. ; https://doi.org/10.1101/2020.09.01.20186213doi: medRxiv preprint Figure Click here to access/download;Figure;Fig2A.tiff All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted September 3, 2020. ; https://doi.org/10.1101/2020.09.01.20186213doi: medRxiv preprint Figure Click here to access/download;Figure;Fig2B.tiff All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted September 3, 2020. ; https://doi.org/10.1101/2020.09.01.20186213doi: medRxiv preprint Figure Click here to access/download;Figure;Fig2C.tiff All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted September 3, 2020. ; https://doi.org/10.1101/2020.09.01.20186213doi: medRxiv preprint Figure Click here to access/download;Figure;Fig2D.tiff All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted September 3, 2020. ; https://doi.org/10.1101/2020.09.01.20186213doi: medRxiv preprint Supporting Information - Compressed/ZIP File Archive Click here to access/download Supporting Information - Compressed/ZIP File Archive Supplementary material V22 TP.docx.zip All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted September 3, 2020. ; https://doi.org/10.1101/2020.09.01.20186213doi: medRxiv preprint