key: cord-0795327-2g0r72hv authors: Kaur, U.; Chakrabarti, S. S.; Patel, T. K. title: RAAS blockers and region-specific variations in COVID-19 outcomes: findings from a systematic review and meta-analysis date: 2020-09-13 journal: nan DOI: 10.1101/2020.09.09.20191445 sha: 2dd58da83d68264ca70ae7a58451bafb6bd1befd doc_id: 795327 cord_uid: 2g0r72hv Background: Coronavirus disease 2019 (COVID-19) has evolved as a global crisis with high mortality seen in elderly and people with cardiometabolic diseases. The use of renin angiotensin aldosterone system (RAAS) blockers in these patients is known to enhance the expression of ACE-2, the chief binding receptor of SARS-CoV-2 and may potentially enhance infectivity. Objective: To provide a pooled estimate of the effect of RAAS blocker usage on COVID-19 outcomes. Data Sources: An electronic literature search was performed for published (using MEDLINE/PubMed and Google Scholar) and preprint (using bioRxiv and medRxiv) studies of interest. The last search was conducted on 9th July 2020. Study Selection: Studies reporting data on RAAS blocker use and COVID-19 mortality and severity were included in the review. Data Extraction and Synthesis: Mortality data and severity data including hospitalization, intensive care unit (ICU) admission, invasive ventilation, steroid use and acute kidney injury (AKI) were recorded. Pooled Odds ratio (OR) estimates were reported with 95% CIs and level of heterogeneity (I2). Main Outcomes and Measures: Odds of mortality in users of RAAS blockers with respect to non-users was the primary outcome. Odds of severity, hospitalization, ICU admission, mechanical ventilation, steroid use, and AKI in users with respect to non-users of RAAS blockers were the secondary outcomes. Results: Of 1348 articles identified, 48 published studies were included in the final analysis, with a total of 26432 patients from 31 studies included in mortality analysis and 20127 patients from 23 studies included in severity analysis. Majority of the studies (41.6%) were from China. No increased risk of mortality (Pooled OR 0.91 (0.65-1.26), I2=89%) or severity (Pooled OR 1.08 (0.79-1.46), I2=88%) was seen with RAAS blockers. The drug class was protective in hypertension (pooled OR 0.63 (0.46-0.86), I2=58%). Severity of COVID-19 outcomes was found to be high for Europeans (Pooled OR 2.08 (1.52-2.85), I2=77%) and US patients (Pooled OR 1.87 (1.62-2.17) in users of RAAS-blockers. A nearly 4 times higher risk of hospitalization, two times higher risk of ICU admission and mechanical ventilation was observed in US patients on RAAS blockers. No net effect on mortality and severity outcomes was seen in Chinese patients. RAAS blocker usage did not have any effect on corticosteroid use and AKI in Chinese patients. Conclusions and Relevance: Use of RAAS blockers is not associated with increased risk of mortality in COVID-19 patients. Reduced mortality is seen in hypertensive patients with COVID-19 and therefore the drugs should be continued in this subset. US and European patients are at higher risk of severe outcomes. Pharmacogenomic differences may explain the ethnicity related variations. Corona Following this, multiple observational studies were conducted to assess the relationship between the use of RAAS blockers and severity of COVID-19. This systematic review and meta-analysis aim to compile the information obtained • All types of study setting (outpatient, inpatient, nursing homes, home care approach) were included. • All age groups of study population were included • Studies focusing on individual RAAS blockers only. • Studies focusing only on outcomes based on laboratory parameters (e.g., serum or urinary ACE2 expression). • Non-comparative studies, review articles, in-vitro studies, animal studies, viewpoints. All relevant abstracts were scrutinized, and full text was searched for those found useful. In case of lack of clarity in the abstracts, full text was analysed. This was done by author UK assisted by author SSC and confirmed by author TKP assisted by author SSC. From the included studies, data was extracted in a Microsoft Excel Sheet. Data included author name, publication year, country, study design, total duration of study, mean or median follow up, characteristics of patients or specific population of COVID-19 patients in whom the particular study was conducted, age, gender, sample size, use of RAAS blockers, mortality outcomes, severity outcomes, need of hospitalization, care in intensive care unit (ICU), need of mechanical ventilation, corticosteroid use and occurrence of acute kidney injury (AKI). All rights reserved. No reuse allowed without permission. preprint (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 this version posted September 13, 2020. . https://doi.org/10.1101/2020.09.09.20191445 doi: medRxiv preprint the included studies as per the Newcastle -Ottawa quality assessment scale (NOS) adapted for the cross-sectional design. The criteria considered were representativeness of the study sample, sample size, non-respondents, ascertainment of the exposure, comparability of study groups for the confounders (age and major co-morbidities), assessment of outcome and statistical tests. The maximum possible score was 10.(6) The primary outcome was odds of mortality in the users of RAAS blockers with respect to non-users among confirmed cases of COVID-19. The secondary outcomes were odds of severity, hospitalization, ICU admission, mechanical ventilation, steroid use, and AKI in users of RAAS blockers with respect to non-users. In the absence of universally accepted definitions, severity was considered as defined by the authors in the included studies. When outcomes were reported both under 'critical' and 'severe' headings, we considered the more serious outcome under severity analysis. In case of multiple time-point for the outcome estimation, we considered data at the end of study period. A subgroup analysis of all outcomes was performed based on the geographical locations (country or continent of origin) of the included studies. The mortality outcome was further analysed as per study sub-populations (e.g., patients with hypertension). The severity outcome was stratified based on definitions considered by individual authors. A sensitivity analysis was performed for each outcome after excluding studies with high risk of bias. The studies with score 8 All outcomes were dichotomous variables. They are reported as odds ratio (OR) with 95% confidence intervals (CI). The meta-analysis was weighted with inverse variance method. An I 2 test was used to assess the heterogeneity between the studies. Fixed-effect model was used if heterogeneity was less than 50% and random-effect was applied in case heterogeneity exceeded 50%. Funnel plot method was used for reporting publication bias. The meta-analysis was performed using Review Manager Software version 5.4. As shown in Figure 1 , a total of 1348 articles were retrieved. Out of 70 full text articles assessed, 48 studies satisfying the selection criteria were included for detailed qualitative and quantitative analysis in this review. A total of 26432 patients from 31 studies (6030 users of RAAS blockers and 20402 non-users) were included in the mortality analysis. The use of RAAS blockers was not associated with increased risk of mortality (pooled OR 0.91 (0.65-1.26), I 2 = 89%) (Figure 2) . A similar trend was observed in the sensitivity analysis after excluding studies with high risk of bias (pooled OR 1.09 (0.71-1.67, I 2 = 91%). Funnel plot was asymmetrical on visual inspection (Supplementary Figure 1) . The subgroup analysis of mortality outcome based on geographical locations showed use of RAAS blockers conferred a protection from mortality in the Chinese population (OR 0.71 (0.52-0.97)) ( Figure 2) . However, in sensitivity analysis, no difference in mortality was observed in studies with low risk of bias (pooled OR 0.85 (0.48-1.50), I 2 =25%). Neither benefit nor risk was observed with the use of RAAS blockers in patients in the US (pooled OR 0.96 (0.59-1.56), I 2 = 81%), Europe (pooled OR 1.19 (0.74-1.91), I 2 =89%), and South Korea (pooled OR 1.12 (0.18-7.01), I 2 =97%) (Figure 2) . The results were consistent in sensitivity analysis ( Table 2) . All rights reserved. No reuse allowed without permission. preprint (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 this version posted September 13, 2020. . https://doi.org/10.1101/2020.09.09.20191445 doi: medRxiv preprint On indication or disease-wise comparison, use of RAAS blockers was found to reduce the overall risk of mortality when prescribed for hypertension (pooled OR 0.63 (0.46-0.86), I 2 =58%). Similar trend was observed in sensitivity analysis (pooled OR 0.48 (0.36-0.63), I 2 =0%). Ten out of fifteen studies reporting mortality in hypertensive patients were from China. (Figure 3) . The results were consistent in sensitivity analysis (pooled OR 1.81 (1.28-2.58), I 2 =81%). Region/country specific analysis also showed an increased risk of poor health outcomes in European patients (pooled OR 2.08 (1.52-2.85), I 2 =77%) and US All rights reserved. No reuse allowed without permission. preprint (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 this version posted September 13, 2020. . patients (OR 1.87 (1.62-2.17)) ( Figure 5) . A similar trend was observed in sensitivity analysis. In contrast, no effect on severity with the use of RAAS blockers was evident in the Chinese population in overall (pooled OR 0.69 (0.45-1.06), I 2 =51%), and sensitivity analysis (pooled OR 0.68 (0.3-1.53), I 2 =58%) ( Figure 5 ). A total of 16441 patients (4060 RAAS blocker users and 12381 non-users) from 13 studies were analysed for the assessment of risk of ICU admission. No increased risk of ICU admission was observed with the use of RAAS blockers in the overall (pooled OR 1.37 (0.86-2.19), I 2 =91%) and sensitivity analyses (pooled OR 1.55 (0.79-3.02), I 2 =93%). Country-specific analysis showed an increased risk of ICU admission in the US population in overall (pooled OR 1.47 (1.15-1.87), I 2 = 37%) and sensitivity analyses (pooled OR 1.82 (1.29-2.58), I 2 =0%). No effect on ICU admission was observed in Chinese patients All rights reserved. No reuse allowed without permission. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (Supplementary Figure 4) . Seven studies (n=1854) commented on corticosteroid use in relation to RAAS blocker use. All of these were from China. Use of RAAS blockers did not affect the requirement for corticosteroid use in the overall analysis (pooled OR 0.82 (0.65-1.04), I 2 =38%) and also in the sensitivity analysis (pooled OR 1.01 (0.64-1.6), I 2 = 35%) (Supplementary Figure 6) . All rights reserved. No reuse allowed without permission. preprint (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 this version posted September 13, 2020. . Five studies (n=2143) reporting on AKI were analysed. Use of RAAS blockers was not associated with increased or decreased risk of AKI in overall analysis (pooled OR 0.94 ( 0.76-1.16), I 2 =0%) and also in the sensitivity analysis (pooled OR 1.23 (0.52-2.89), I 2 =0%). The latter was based on two studies, both from China (Supplementary Figure 7) .(16,53) Hypertension, diabetes mellitus, cerebrovascular disease, and ischemic heart disease are co-morbidities which are commonly prevalent and found to be responsible for adverse prognosis in patients with COVID-19.(54) RAAS blockers are used in majority of these diseases and are known for their disease modifying roles in ischemic heart disease, congestive heart failure, and diabetic nephropathy. With the observation that SARS-CoV-2 binds preferentially to ACE2 as its receptor and that ACE2 is prone to upregulation by RAAS blockers, speculations were made that the continuation of RAAS blockers would increase binding of the virus to host cells and enhance its infectivity. On the contrary, ACE2 is known to be protective against lung injury via the Ang armamentarium. (60) The confusion surrounding the use of RAAS blockers led All rights reserved. No reuse allowed without permission. preprint (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 this version posted September 13, 2020. . to a spurt of observational studies with a focus on establishing relationship if any, between the use of RAAS blockers and COVID-19 outcomes. In this systematic review, we have tried to compile information from all such studies and provide insights on association between the use of RAAS blockers and COVID-19 morbidity and mortality outcomes. In our meta-analysis, use of RAAS blockers was not associated with an increased risk of mortality. A reduced risk of mortality was seen in the Chinese population, but the effect was nullified in sensitivity analysis. RAAS blockers were found to reduce mortality in hypertensive patients. On the other hand, an increased risk of composite outcome of ICU admission/death was seen with the use of RAAS blockers and this effect persisted in sensitivity analysis. With respect to severity of COVID-19 disease, although no overall effect of RAAS blockers was evident, a reduced risk of 'critical' form of the disease (defined as per NHC China) was observed. The same protection was not validated, however, in sensitivity analysis. Further, while RAAS blockers did not produce any adverse effect on disease severity when analysed in the entire population, the outcomes differed considerably between the countries. RAAS blockers were found not to affect disease severity in Chinese patients but the use of such agents was associated with nearly a two times higher risk of severe disease in US patients and Europeans. Nearly a four times increased risk of hospitalisation was seen with the use of RAAS blockers in US patients. Similarly, no increase in the risk of ICU admission and invasive ventilation was seen with RAAS blockers in Chinese patients, whereas the US patients on RAAS blockers had an approximately two times higher risk of getting admitted in the ICU or receiving mechanical ventilation. Further, with respect to requirement of corticosteroids and causation of renal injury, no risk could be attributed to RAAS blockers. This interpretation is primarily based on the sensitivity analysis involving Chinese studies. All rights reserved. No reuse allowed without permission. preprint (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 this version posted September 13, 2020. . preprint (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 this version posted September 13, 2020. . This systematic review has some limitations. The pooled analysis is mainly based on observational studies, which are more likely to have study populations with difference in baseline characteristics and co-interventions than randomized controlled trials. The country specific subgroup analysis was based on only a small number of studies. Further, the current meta-analysis aimed to generate data related to RAAS blockers and therefore excluded those studies (n=11) which focussed on ACEI and ARB class in isolation and did not provide information about the outcomes in combined RAAS blocker class. We did not All rights reserved. No reuse allowed without permission. preprint (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 this version posted September 13, 2020. . compare the outcomes between users of ACEIs and ARBs also. However, such analyses can be done in the future to deduce any class specific differences that can potentially influence COVID-19 outcomes. There is a need to investigate racial or region/country specific differences in the clinical outcomes of COVID-19. Genetic polymorphisms may govern the pharmacodynamic response to RAAS blockers in different population groups, as seen in our meta-analysis and should be explored actively in future. There is a need to explore excess risk of ICU admission and mechanical ventilation in the US and increased severity of COVID-19 disease in Europeans, both of which were found to be associated with RAAS blocker usage. Overall, the use of RAAS blockers does not seem to have any impact on COVID-19 mortality and severity. In the presence of a protective effect in patients with hypertension, it may be advisable to continue these drugs in those patients with pre-COVID indication for the same. Funding Support: None Acknowledgement: Dr Patel and Dr Kaur wish to thank the administration of the All India Institute of Medical Sciences, Gorakhpur for providing research support. 1. Coronavirus [Internet] . [cited 2020 Aug 28]. Available from: https://www.worldometers.info/coronavirus All rights reserved. No reuse allowed without permission. preprint (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 this version posted September 13, 2020. . Xie M, Chen Q. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. preprint (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 this version posted September 13, 2020. . in New York: retrospective case series. medRxiv [Internet] . 2020 Jan All rights reserved. No reuse allowed without permission. preprint (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 this version posted September 13, 2020. . preprint (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 this version posted September 13, 2020. All rights reserved. No reuse allowed without permission. preprint (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 this version posted September 13, 2020. . preprint (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 this version posted September 13, 2020. . All rights reserved. No reuse allowed without permission. preprint (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 this version posted September 13, 2020. . preprint (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 this version posted September 13, 2020. . preprint (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 this version posted September 13, 2020. . preprint (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 this version posted September 13, 2020. . patients All rights reserved. No reuse allowed without permission. preprint (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 this version posted September 13, 2020. . preprint (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 this version posted September 13, 2020. . Non-users gp- All rights reserved. No reuse allowed without permission. preprint (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 this version posted September 13, 2020. . preprint (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 this version posted September 13, 2020. . preprint (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 this version posted September 13, 2020. . preprint (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 this version posted September 13, 2020. . Risk of pneumonia associated with use of angiotensin converting enzyme inhibitors and angiotensin receptor blockers: systematic review and meta-analysis Angiotensin-converting enzyme inhibitor use and protection against pneumonia in patients with diabetes The pivotal link between ACE2 deficiency and SARS-CoV-2 infection Polymorphisms of ACE2 Gene are Associated With Essential Hypertension and Antihypertensive Effects of Captopril in Women ACE2 polymorphisms and individual susceptibility to SARS-CoV-2 infection: insights from an