key: cord-0755008-ub718jpv
authors: Diaz-Arocutipa, C.; Saucedo-Chinchay, J.; Hernandez, A. V.
title: Association Between ACEIs or ARBs Use and Clinical Outcomes in COVID-19 Patients: A Systematic Review and Meta-analysis
date: 2020-06-08
journal: nan
DOI: 10.1101/2020.06.03.20120261
sha: ade7c144c0b2fa29b07022ab08aaabdd8bff23da
doc_id: 755008
cord_uid: ub718jpv

Importance: There is a controversy regarding whether or not to continue angiotensin-converting enzyme inhibitors (ACEIs) and angiotensin receptor blockers (ARBs) in patients with coronavirus disease 2019 (COVID-19). Objective: To evaluate the association between ACEIs or ARBs use and clinical outcomes in COVID-19 patients. Data Sources: Systematic search of the PubMed, Embase, Scopus, Web of Science, and Cochrane Central Register of Controlled Trials from database inception to May 31, 2020. We also searched the preprint servers medRxiv and SSNR for additional studies. Study Selection: Observational studies and randomized controlled trials reporting the effect of ACEIs or ARBs use on clinical outcomes of adult patients with COVID-19. Data Extraction and Synthesis: Risk of bias of observational studies were evaluated using the Newcastle-Ottawa Scale. Meta-analyses were performed using a random-effects models and effects expressed as Odds ratios (OR) and mean differences with their 95% confidence interval (95%CI). If available, adjusted effects were pooled. Main Outcomes and Measures: The primary outcome was all-cause mortality and secondary outcomes were COVID-19 severity, hospital discharge, hospitalization, intensive care unit admission, mechanical ventilation, length of hospital stay, and troponin, creatinine, procalcitonin, C-reactive protein (CRP), interleukin-6 (IL-6), and D-dimer levels. Results: 40 studies (21 cross-sectional, two case-control, and 17 cohorts) involving 50615 patients were included. ACEIs or ARBs use was not associated with all-cause mortality overall (OR 1.11, 95%CI 0.77-1.60, p=0.56), in subgroups by study design and using adjusted effects. ACEI or ARB use was independently associated with lower COVID-19 severity (aOR 0.56, 95%CI 0.37-0.87, p<0.01). No significant associations were found between ACEIs or ARBs use and hospital discharge, hospitalization, mechanical ventilation, length of hospital stay, and biomarkers. Conclusions and Relevance: ACEIs or ARBs use was not associated with higher all-cause mortality in COVID-19. However, ACEI or ARB use was independently associated with lower COVID-19 severity. Our results support the current international guidelines to continue the use of ACEIs and ARBs in COVID-19 patients with hypertension.

Introduction 1 1

Main characteristics of the 40 included studies (n=50615) were summarized in Table 1 . 3

The exposure variable in almost all studies was the chronic use of ACEIs and ARBs (i.e. 4 before hospital admission) as registered in medical records, although in two studies 20,21 it 5 was defined as in-hospital use. Also, only six studies [22] [23] [24] [25] [26] [27] reported that ACEIs and ARBs 6

were not discontinued during hospitalization. The definition of outcomes was the same in 7 almost all included studies. In contrast, the definition of COVID-19 severity was very 8 heterogeneous across the studies due to different clinical guidelines used for management 9

of COVID-19 in each country. The most common criteria for COVID-19 severity was 10 critical/severe vs mild/moderate which was used in five studies 24,28-31 (Table 1) . 11

Our search in clinicaltrials.gov identified 16 registered RCTs (eTable 3), of which 12 six evaluate the impact of continuation or discontinuation of ACEIs and ARBs on COVID-13 19 outcomes and four placebo-controlled trials assess the efficacy of ARB (losartan and 14 valsartan) and ACEI (ramipril) in COVID-19 patients who are not previously taking a RAS 15 inhibitor. 16

In 22 studies (11 cross-sectional and 11 cohorts, n=23059), the use of ACEIs or ARBs was 24 not associated with higher odds of all-cause mortality (OR 1.11, 95%CI 0.77-1.60, p=0.56) 25 and heterogeneity was high among studies ( Figure 2 ). In the subgroup analysis by study 26 . CC-BY-NC-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 preprint this version posted June 8, 2020. . https://doi.org/10.1101/2020.06.03.20120261 doi: medRxiv preprint design, ACEIs or ARBs use were not associated with all-cause mortality in 11 cross-1 sectional studies (OR 0.86, 95%CI 0.65-1.15, p=0.32) and 11 cohort studies (OR 1.30, 2 95%CI 0.71-2.38, p=0.40) (Figure 2 ). The funnel plot did not show asymmetry and the 3 Egger's test was not significant (p=0.39) (eFigure 1). ACEI use was not associated with all-4 cause mortality (OR 1.18, 95%CI 0.83-1.66, p=0.35) (eFigure 2), but ARB use was 5 associated with increased odds of all-cause mortality (OR 1.79, 95%CI 1.07-3.00, p=0.03) 6 (eFigure 3). 7

Six studies 21,32-36 reported adjusted effect measures of association between 8

ACEIs/ARBs use and all-cause mortality (eTable 7). The pooled estimate of three 9 studies 21,35,36 with similar adjusted variables (age, sex, and cardiovascular comorbidities) 10

found that ACEIs or ARBs use was not associated with all-cause mortality (aHR 0.83, 11 95%CI 0.49-1.38, p=0.47) (eFigure 4). Furthermore, the pooled estimate of ACEI 12 studies 32,34,35 (aHR 0.97, 95%CI 0.83-1.13, p=0.67) and ARB studies 32,34,35 (aHR 1.14, 13 95%CI 0.98-1.34, p=0.08) showed no association either (eFigure 5 and 6). 14 15 Secondary Outcomes 16

In 18 studies (11 cross-sectional, two case-control, and five cohorts, n=11870), the use of 18

ACEIs or ARBs was not associated with COVID-19 severity (OR 0.79, 95%CI 0. 59-1.07, 19 p=0.13) and showed high heterogeneity among studies (Figure 3 ). The funnel plot showed 20 asymmetry, suggesting publication bias which was confirmed by the Egger's test (p<0.01) 21 (eFigure 7). Subgroup analysis by study design showed that ACEIs or ARBs use was only 22 associated with lower COVID-19 severity in five cohort studies (OR 0.61, 95%CI 0.39-23 0.95, p=0.03) (Figure 3 ). In contrast, the use of ACEI (OR 1.10, 95%CI 0.55-2.18, p=0.79) 24

and ARB (OR 1.00, 95%CI 0.77-1.29, p=0.98) separately were not associated with 25 COVID-19 severity (eFigure 8 and 9). 26 . CC-BY-NC-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)

The copyright holder for this preprint this version posted June 8, 2020.

The pooled adjusted estimate of four studies 23,24,37,38 showed that the ACEIs or 1

ARBs use (aOR 0.56, 95%CI 0.37-0.87, p<0.01) was independently associated with lower 2 COVID-19 severity (eFigure 10). However, adjusted estimates of ACEI (aOR 0.66, 95%CI 3 0.37-1.18, p=0.15) and ARB (aOR 0.97, 95%CI 0.79-1.20, p=0.81) use separately were 4 not associated with COVID-19 severity (eFigure 11 and 12). In three studies 29,31,39 (two cross-sectional and one cohort, n=301), the use of ACEIs or 8 CC-BY-NC-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)

The copyright holder for this preprint this version posted June 8, 2020. .

In seven studies 21,28,29,33,40,44,45 (three cross-sectional and four cohorts, n=6533), the use of 1 In five studies 26,29,31,39,46 (four cross-sectional and one cohort, n=699), the use of ACEIs or 6

ARBs was not associated with length of hospital stay (MD -0.96 days, 95%CI -2.50 to 7 0.57, p=0.22) (eFigure 21). 8 9

In four studies 26-28,31 (two cross-sectional and two cohorts, n=580), the use of ACEIs or 11

ARBs was not associated with troponin level (MD -0.01 μg/L, 95%CI -0.04 to 0.02, p=0.37) 12 (eFigure 22). 13 14

In six studies 24,26-28,31,47 (two cross-sectional and four cohorts, n=716), the use of ACEIs or 16

ARBs was not associated with creatinine level (MD -0.58 μmol/L, 95%CI -8.72 to 7.56, 17 p=0.89) (eFigure 23). 18

In five studies 26-28,31,47 (two cross-sectional and three cohorts, n=651), the use of ACEIs or 21

ARBs was not associated with procalcitonin level (MD -0.02 ng/mL, 95%CI -0.05 to 0.01, 22

p=0.21) (eFigure 24). 23 24 CRP level 25 . CC-BY-NC-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)

The copyright holder for this preprint this version posted June 8, 2020. .

In five studies 26,28,29,31,47 (two cross-sectional and three cohorts, n=709), the use of ACEIs 1 or ARBs was not associated with CRP level (MD -6.39 mg/L, 95%CI -16.19 to 3.41, 2 p=0.20) (eFigure 25). 3 4 IL-6 level 5

In four studies 26,27,31,47 (two cross-sectional and two cohorts, n=601), the use of ACEIs or 6

ARBs was not associated with IL-6 level (MD -4.41 pg/mL, 95%CI -13.24 to 4.42, p=0.33) 7 (eFigure 26). In six studies 26-29,31,47 (two cross-sectional and one cohort, n=751), the use of ACEIs or 11

ARBs was not associated with D-dimer level (MD -0.91 nmol/L, 95%CI -2.77 to 0.94, 12 p=0.33) (eFigure 27). 13 14

The results of the sensitivity analyses are reported in the eTable 8. Overall, the results 16 showed that ACEIs or ARBs use was independently with lower COVID-19 severity, ARB 17 use was independently associated with higher mortality, and ACEIs or ARBs use was 18 associated with higher ICU admission. 19 20

We found that the use of ACEIs or ARBs was not significantly associated with all-cause 22 mortality in COVID-19 patients, and when analyzed by study design or when using 23 adjusted effects. In contrast, ACEIs or ARBs use was independently associated with lower 24 COVID-19 severity. Although ACEIs or ARBs use was associated with an increased odds 25 of ICU admission, this effect disappeared when ACEIs and ARBs were analyzed 26 . CC-BY-NC-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)

The copyright holder for this preprint this version posted June 8, 2020. . https://doi.org/10.1101/2020.06.03.20120261 doi: medRxiv preprint individually. No significant associations were found between ACEIs or ARBs use and other 1 clinical outcomes or biomarkers. Risk of bias was low or moderate across studies. 2

It has been proposed that RAS play a crucial role in the pathogenesis of infection 3 by SARS-CoV-2, since it uses the ACE2 receptor to enter into cells, with the subsequent 4 downregulation of this surface protein. 5 The reduction of ACE2 expression in infected cells 5

can lead to a tissue and systemic RAS imbalance with a predominance of the dangerous 6 ACE/Ang II/AT1R axis. 5 This phenomenon can be particularly harmful in the elderly 7 population since they have already a lower level of ACE2 expression compared to young 8 people. 48 This could partly explain the higher mortality observed in older patients with 9

COVID-19. 49 Recent evidence from a Chinese cohort of 12 COVID-19 patients showed 10 that circulating Ang II levels were markedly elevated compared to healthy controls and 11 linearly associated with viral load and lung injury. 50 Moreover, in an animal experiment of 12 acute lung injury induced by acid, the SARS-CoV spike protein enhances the pulmonary 13

Ang II levels and lung injury severity. 51 Altogether, these data suggest that SARS-CoV-2 14 can mediate the damage to lungs and possibly to other organs through the absence of 15 degradation of Ang II. Therefore, RAS modulators such as ACEIs and ARBs can be used 16 as potential therapeutic agents in COVID-19 patients. This is currently under investigation 17 in several ongoing clinical trials. 18

In general, we found that the use of ACEIs or ARBs had a neutral effect on all-19 cause mortality and other clinical outcomes in COVID-19 patients. Two studies 21,52 with 20 larger samples and adjustment for confounders reported a significant reduction of all-21 cause mortality and severity in these patients. Nowadays, there is controversy regarding 22 the use of ACEIs and ARBs in patients with COVID-19 and hypertension. Initially it was 23 suggested that the use of these drugs could increase ACE2 expression; however, there is 24 conflicting evidence about the effect of ACEIs and ARBs on ACE2 tissue expression in 25 animal models. 53 Besides, studies in humans showed no effect of ACEIs and ARBs 26 . CC-BY-NC-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 preprint this version posted June 8, 2020. . https://doi.org/10.1101/2020.06.03.20120261 doi: medRxiv preprint administration on ACE2 protein levels in urine and plasma. 54,55 Furthermore, a recent 1

Mendelian randomization study revealed a lack of association between genetically proxy 2 ACE inhibition and lung ACE2 expression or circulating ACE2 levels. 56 Likewise, in a study 3 on human myocardial samples, there was no significant difference in ACE2 expression in 4 tissue samples with and without exposure to ACEI. 57 Overall, these findings suggest that 5

ACEIs and ARBs are unlikely to raise ACE2 in humans. Thus, it seems reasonable that 6

ACEIs and ARBs could exert its effect on COVID-19 mainly through inhibition of the 7 ACE/Ang II/AT1R axis. 8

The lung is the target organ in COVID-19; however, other organs may potentially 9 be involved. A recent study reported that acute cardiac injury, manifested as elevated 10 troponin levels, was present in 20% of COVID-19 patients and was independently 11 associated with worse outcomes. 58 Although the pathophysiological basis of this finding is 12 not fully understood, it has been proposed that SARS-CoV-2 can cause cardiac injury 13 through several mechanisms: direct viral damage, systemic inflammatory response, 14 microangiopathy, and myocardial infarction. 59 Similarly, acute kidney injury was observed 15 in up to 27% of COVID-19 patients, 60 this is probably related to alterations in renal 16 microvasculature, kidney cell viral infection, and systemic inflammation. 61 Dysregulation of 17 contribute to the tissue and systemic damage caused by SARS-CoV-2. Thus, considering 1 that RAS inhibitors are capable of regulating both tissue and systemic RAS, it has been 2 suggested that could have a beneficial effect in COVID-19. However, our study did not find 3 a significant association of ACEIs or ARBs use on troponin, inflammatory markers 4 (procalcitonin, CRP, and IL-6), creatinine, and D-dimer levels in COVID-19 patients. 5

Further research is needed to clarify the potential therapeutic role of RAS inhibitors on 6 multiorgan dysfunction associated with COVID-19. concluding that there is no evidence of association between ACEI/ARB use with more 18 severe COVID-19 disease. Compared to these reviews, our study included 40 studies and 19 evaluated 13 outcomes. Additionally, to our knowledge, our review is the first that pooled 20 adjusted effect estimates for mortality and COVID-19 severity. 21

Our study has some limitations. First, given most of the studies did not use 22 adjusted effects, there is an increased risk of bias in their pooled effect measures. Thus, 23 these results should be considered with caution. However, we also reported meta-24 analyses of adjusted estimates of a few available studies. Second, the majority of the 25 included studies were of cross-sectional design, thus causality cannot be concluded due to 26 . CC-BY-NC-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 preprint this version posted June 8, 2020. . https://doi.org/10.1101/2020.06.03.20120261 doi: medRxiv preprint the methodological limitations of this design. Third, heterogeneity was high in most of the 1 evaluated outcomes. Possible reasons for heterogeneity include sample size, differences 2 in outcome definitions, heterogeneous population, among others. Fourth, given that 3 discontinuation of ACEIs or ARBs during hospitalization was not reported consistently 4 across studies, this could influence the significance of pooled estimates. Finally, we could 5 not adequately evaluate the effects of ACEIs and ARBs by separate, since were mainly 6 reported as aggregate due to scarcity of studies. 7 8

In conclusion, the use of ACEIs or ARBs was not associated with higher all-cause mortality 10 in COVID-19 patients, based on the meta-analysis of cross-sectional and cohort studies 11 and also using adjusted effects. ACEI or ARB use was independently associated with 12 lower COVID-19 severity. Also, there was no evidence of association between ACEIs or 13 is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 8, 2020. . is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 8, 2020. .

Wuhan, China: a single-center retrospective observational study. medRxiv [Preprint] . CC-BY-NC-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 preprint this version posted June 8, 2020. . CC-BY-NC-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 preprint this version posted June 8, 2020. . https://doi.org/10.1101/2020.06.03.20120261 doi: medRxiv preprint . CC-BY-NC-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 preprint this version posted June 8, 2020. Figure 1 . CC-BY-NC-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)

The copyright holder for this preprint this version posted June 8, 2020. . https://doi.org/10.1101/2020.06.03.20120261 doi: medRxiv preprint Figure 2 . CC-BY-NC-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)

The copyright holder for this preprint this version posted June 8, 2020. . https://doi.org/10.1101/2020.06.03.20120261 doi: medRxiv preprint . CC-BY-NC-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)

The copyright holder for this preprint this version posted June 8, 2020. . https://doi.org/10.1101/2020.06.03.20120261 doi: medRxiv preprint 

Elevated 11 troponin in patients with Coronavirus Disease 2019: possible mechanisms

Survival rate in acute kidney injury superimposed 14 COVID-19 patients: a systematic review and meta-analysis

Therapy, and Mortality in Covid-19. NEJM. 2020

Hydroxychloroquine or chloroquine with 3 or without a macrolide for treatment of COVID-19: a multinational registry analysis

A mysterious company's coronavirus papers in top medical journals may be

top-medical-journals-may-be-unraveling

Expression of Concern: Mehra MR et al. Cardiovascular Disease, Drug

Therapy, and Mortality in Covid-19. NEJM. 2020

Expression of concern: Hydroxychloroquine or chloroquine with or without a macrolide 11 for treatment of COVID-19: a multinational registry analysis

ACEI/ARB use and risk of infection or severity or 14 mortality of COVID-19: A systematic review and meta-analysis

Decreased mortality of COVID-19 with renin-angiotensin-17

aldosterone system inhibitors therapy in patients with hypertension: a meta-analysis