key: cord-0974176-dtx3rpvh
authors: Chouchana, L.; Beeker, N.; Garcelon, N.; Rance, B.; Paris, N.; Salamanca, E.; Polard, E.; Burgun, A.; Treluyer, J.-M.; Neuraz, A.; collaboration, AP-HP Universities Inserm COVID-19 research
title: Association of antihypertensive agents with the risk of in-hospital death in patients with Covid-19
date: 2020-11-24
journal: nan
DOI: 10.1101/2020.11.23.20237362
sha: d736f0491902fc31dc8ddb1a66fd2e3da87388be
doc_id: 974176
cord_uid: dtx3rpvh

In this retrospective multicenter cohort study, we aimed to investigate the association between antihypertensive agent exposure and in-hospital mortality in patients with Covid-19. Of 8,078 hospitalized patients for Covid-19, 3,686 (45.6%) had hypertension including 2043 (55.4%) patients exposed to a renin-angiotensin-aldosterone inhibitors (RAASi), 1624 (44.1%) to calcium channel blockers (CCB) and 1154 (37.7%) to beta-blockers. Overall in-hospital 30-day mortality was 23.1%. Compared to non-users, the risk of mortality was lower in CCB (aOR, 0.83 [0.70-0.99]) and beta-blockers users (aOR, 0.80 [0.67-0.95]), and not different in RAASi users. These findings support the continuation of antihypertensive agents in patients with Covid-19.

32 . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) preprint

The copyright holder for this this version posted November 24, 2020. ; https://doi.org/10.1101 https://doi.org/10. /2020 

The coronavirus disease-2019 , due to the novel severe acute respiratory syndrome 49 coronavirus-2 (SARS-CoV-2) is a global pandemic representing a major treat to global health. 50 Data from several cohorts have shown that age and cardiovascular comorbidities, including 51 hypertension, were among the main determinants of severe disease. 1, 2 As SARS-CoV-2 uses 52 the angiotensin-converting enzyme 2 (ACE2) receptor as an entrance door into cells, 53

questions have been raised about the role of the renin-angiotensin-aldosterone system (RAAS) 54 activity on Covid-19 pathophysiology. 3 RAAS inhibitors directly acts on RAAS in two 55 different ways, both leading to ACE2 upregulation. 3 On the contrary, some authors have 56 hypothesized that the use of RAAS inhibitors may protect against acute lung injury caused by 57 angiotensin II accumulation. 4 Therefore, since these drugs such as ACE inhibitors (ACEi) or 58 angiotensin receptor blockers (ARB) are common treatments for hypertension, controversies 59 about their role on Covid-19 severity have been raised. To date, scientific societies have 60 advised against their discontinuation in patients with hypertension, until robust clinical 61 evidence is available. 5 Previous reports showed that RAAS inhibitors do not increase the risk 62 of severe or fatal outcome in Covid-19 patients. 1, 6, 7 The role of other antihypertensive agents 63 is unclear. However, these analyses are retrospective and additional data are needed to 64 ascertain the beneficial or harmful effect of antihypertensive agents on Covid-19 mortality. 65

In order to provide additional information, we performed a multicenter retrospective cohort 66 study in patients hospitalized for Covid-19 in Greater Paris area, France. 67 68 69 . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) preprint

The copyright holder for this this version posted November 24, 2020. Hypertension as well as other comorbidities have been retrieved in EHRs using specific codes 84 from the International Classification of Diseases 10 th version, combined with natural language 85 processing (regular expressions) as previously described, within the six months previous to 86 the hospital stay for Covid-19 (Suppl Table S1 ). 8 Antihypertensive agent exposure have been 87 identified in EHRs at admission of the hospital stay for Covid-19 and within the six previous 88 months, using both specific Anatomical Therapeutic Chemical classification (ATC) codes and 89 deep learning algorithm on clinical narratives as previously described (Suppl Table S2 ). 8 As 90 described in Neuraz et al. 8 , we extracted drug mentions and their attributes (dose, frequency, 91 duration, route, condition) using a deep-learning pipeline based on fine-tuned multilingual 92 embeddings and bidirectional long-short term memory units combined with conditional 93 random fields. 9 The drug mentions were then normalized to ATC using exact string matching. 94 . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) preprint

The copyright holder for this this version posted November 24, 2020. ; https://doi.org/10.1101/2020.11.23.20237362 doi: medRxiv preprint

The primary outcome was all-cause 30-day in-hospital mortality. We assessed the association 97 between antihypertensive agent class exposure and primary outcome using a multivariate 98 logistic regression. Analyses were adjusted on age, sex and main chronic diseases to take into 99 account confounding factors known to be associated with Covid-19 mortality. Data are 100 presented as median (interquartile range [IQR] ) and numbers (%). The results of the logistic 101 regression model are presented as crude odds-ratio (OR) or adjusted OR (aOR) and their 95% 102 confidence interval ([95% CI]). All analyses were performed using R statistical software. 103 104 105 . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) preprint

The copyright holder for this this version posted November 24, 2020. In total, there were 853 (23.1%) all-cause in-hospital 30-day deaths. Patients who died were 117 older and there were more likely to be men, compared to survivors. They also had chronic 118 diseases more frequently, except for respiratory diseases and obesity. The risk of mortality 119 Table S3) . 122 123 124 . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) preprint

The copyright holder for this this version posted November 24, 2020. ; https://doi.org/10.1101/2020.11.23.20237362 doi: medRxiv preprint

This study, one of the largest multicenter retrospective to date on more than 3600 hospitalized 126

Covid-19 patients with hypertension, provides two main findings. First, based on more than 127 2,000 patients exposed to RAAS inhibitors, there is no association with the use of ACEi or 128 ARB and the risk of in-hospital death. Second, there is a significant protective effect of 129 calcium-channel blockers (CCB) or beta-blocker use on the risk of death in hypertensive 130 patients with Hypertension and cardiovascular comorbidities have been previously reported as risk factors 132 for severe disease and fatal outcome. 1, 10, 11 The underlying pathogenic mechanisms of these 133 comorbidities remain unclear and may involve the RAAS as being a double-edged sword. On 134 one hand, RAAS inhibitors increase the expression of ACE2, which could promote virus 135 entry. 3 On the other hand, RAAS inhibitors, especially ARB, could reverse deleterious effects 136 of unopposed angiotensin II accumulation resulting from ACE2 down-regulation associated to 137 SARS-CoV-2 entry in cells. 4, 12 Clinical evidence based on a large study including 12,594 138 tested patients for SARS-CoV-2 showed no increase of likelihood to have a positive test 139 among RAAS users. 13 Furthermore, in line with our findings, a meta-analysis on four 140 retrospective studies that include 921 ACEi or ARB users found no difference in risk of death 141 compared to non-users (OR, 0.88 [0.68-1.14]). 14 142

In our study, we found that CCB and beta-blockers were more frequently used in survivors 143 (OR 0.83 [0.70-0.99] and 0.80 [0.67-0.95], respectively) than in non-survivors. These findings 144 were statistically significant. Previous Covid-19 cohort studies did not found any beneficial or 145 deleterious effect of CCB use on Covid-19 severity or mortality. 7,15 However, these studies 146 included few CCB users and were probably lacking power. Recently, an in vitro study on an 147 emerging porcine coronavirus showed that CCB could prevent from infection and intracellular 148 calcium homeostasis dysregulation. 16 Regarding beta-blockers, one could hypothesizes that 149 . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this this version posted November 24, 2020. ; https://doi.org/10.1101/2020.11.23.20237362 doi: medRxiv preprint they can counteract deleterious sympathetic activation during the SARS-CoV-2 induced 150 cytokine storm and severe disease. Further clinical data are needed to explore these early 151

Our study has some limitations. First, in this retrospective observational design on 153 hospitalized Covid-19 patients, hypertension or antihypertensive exposure could have affected 154 the chance of hospitalization, which may limit the generalizability of our results. Second, 155 although analyses were adjusted on major comorbidities, some unmeasured or unknown 156 confounders may have not been ruled out, including indication bias. Finally, ascertainment of 157 medication use from EHRs may not capture all drug prescriptions and not reflect actual drug 158 exposure. However, considering antihypertensive agents are chronic treatments, we consider 159 exposure to these drugs up to six months prior to hospitalization for 161 CONCLUSION 162

In summary, we did not find any association between the use of RAAS inhibitors and the risk 163 of in-hospital death in Covid-19 patients with hypertension. Interestingly we found a reduced 164 mortality among CCB and beta-blockers users. Further studies are needed to confirm this 165 protective role. Our findings, such as previous reports, support the continuation of 166 antihypertensive agents in Covid-19 patients, in line with the current guidelines. 167 168 169 . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this this version posted November 24, 2020. ; https://doi.org/10.1101 https://doi.org/10. /2020 is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) preprint

The copyright holder for this this version posted November 24, 2020. ; https://doi.org/10. 1101 /2020 Data available on request. The data underlying this article will be shared on reasonable 194 request to the corresponding author. CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) preprint

The copyright holder for this this version posted November 24, 2020. ; https://doi.org/10.1101 https://doi.org/10. /2020 CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) preprint

The copyright holder for this this version posted November 24, 2020. ; https://doi.org/10.1101 https://doi.org/10. /2020 These groups are not exclusive and one patient can be exposed to more than one 287 pharmacological class. Non-survivors are considered in relation with in-hospital 30-day 288 mortality. 289 ARB: angiotensin II receptor blockers; ACEi: angiotensin-converting enzyme inhibitors; 290 CCB: calcium channel blockers. 291 292 . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) preprint

The copyright holder for this this version posted November 24, 2020. ; https://doi.org/10.1101/2020.11.23.20237362 doi: medRxiv preprint These groups are not exclusive and one patient can be exposed to more than one 298 pharmacological class. Analyses have been adjusted on age, sex and main chronic diseases 299 (i.e. hypertension, chronic kidney disease, cerebrovascular disease, cardiovascular disease, 300 cardiac failure, diabetes, respiratory disease, obesity and malignancies). 301 ARB: angiotensin II receptor blockers; ACEi: angiotensin-converting enzyme inhibitors; 302 CCB: calcium channel blockers 303 304 . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) preprint

The copyright holder for this this version posted November 24, 2020. 

Association of hypertension and 219 antihypertensive treatment with COVID-19 mortality: a retrospective observational 220 study

Clinical Characteristics of Coronavirus Disease 2019 in 225 China

Clinical Implications of SARS-CoV-2 227

Interaction With Renin Angiotensin System: JACC Review Topic of the Week

Angiotensin receptor blockers as tentative SARS-CoV-2 therapeutics. Drug 230

Patients taking 233 ACE-i and ARBs who contract COVID-19 should continue treatment, unless otherwise 234 advised by their physician

Use of 241 renin-angiotensin-aldosterone system inhibitors and risk of COVID-19 requiring 242 admission to hospital: a case-population study

Association of Renin-Angiotensin System 244

Inhibitors With Severity or Risk of Death in Patients With Hypertension Hospitalized for 245

COVID-19) Infection in Wuhan, China

Natural Language Processing for Rapid 250 Response to Emergent Diseases: Case Study of Calcium Channel Blockers and 251 Hypertension in the COVID-19 Pandemic

Neural Architectures 253 for Named Entity Recognition

Clinical characteristics of 113 deceased patients with coronavirus disease 2019: 260 retrospective study

Clinical course and risk 263 factors for mortality of adult inpatients with COVID-19 in Wuhan, China: a 264 retrospective cohort study

Renin-Angiotensin-Aldosterone System Inhibitors in Patients with Covid-19

Renin-Angiotensin-Aldosterone System Inhibitors and Risk 271 of Covid-19

Treatment with ACE inhibitors or ARBs and 274 risk of severe/lethal COVID-19: a meta-analysis