Impact of Angiotensin-converting Enzyme Inhibitors and Angiotensin Receptor Blockers on Inflammatory Responses and Viral Clearance in COVID-19 Patients: A Multicenter Retrospective Cohort Study

Objectives: To evaluate the impact of angiotensin-converting enzyme inhibitors (ACEIs) or angiotensin receptor blockers (ARBs) on coronavirus disease 2019 (COVID-19) patients. Methods: : We included 229 patients with confirmed COVID-19 in a multicenter, retrospective cohort study. Propensity score matching at a ratio of 1:4 was introduced to eliminate the potential confounders. Patients were assigned to the ACEI/ARB group (n=40) or control group (n=160) according to whether they were current users of medication. Results: : Compared to the control group, patients in the ACEI/ARB group had lower levels of plasma IL-1β [(6.27±0.50) vs. (8.23±0.39) pg/ml, P =0.028], IL-8 [(35.74±4.00) vs. (45.88±2.06) pg/ml, P =0.037] and TNF-α [(8.79±0.40) vs. (10.91±0.21) pg/ml, P <0.01]. Patients with the current use of ACEIs/ARBs had a higher risk of shock (23% vs. 8%, P <0.01). Decreased lymphocyte counts [(0.85±0.45) vs. (1.02±0.52)*10


Introduction
Up to March 31, 2020, the total number of patients with coronavirus disease 2019 has risen sharply to nearly 700,000 globally, with a mortality of nearly 5%. Meanwhile, this epidemic seems to be spreading at an exponential rate and has become an urgent public health emergency of international concern.
Several large retrospective studies have revealed that pre-existing cardiovascular disease and diabetes were the most frequent comorbidities of coronavirus disease 2019 (COVID-19) patients [1][2][3] ; these patients even had a higher risk of mortality [4,5] than those with underlying respiratory disease.
As the molecular biology of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is well established, it appears to bind to its target cells through ACE2, which is expressed by epithelial cells of the lung to enable it to infect host cells [10,11] . The expression of ACE2 is substantially increased in patients who are treated with ACE inhibitors and ARBs [12] , which promotes SARS-CoV-2 entry into the body, increasing the risk of developing COVID-19 [13,14] .
The controversial pathogenesis as well as the mixed results of several clinical studies [15,16] of pneumonia with other pathogens made it difficult for physicians to determine whether the use of ACE inhibitors or ARBs should be terminated in patients with COVID-19.
To date, the actual impact of ACE inhibitor and ARB prescriptions on COVID-19 patients has not been assessed in current studies. Therefore, we aimed to evaluate the clinical manifestations, inflammatory responses, viral clearance and outcomes by a multicenter, retrospective cohort study.

Study design and population
We retrospectively included patients with confirmed cases of COVID-19 according to the World Health Organization (WHO) [17] and Chinese official guidelines [18] in a multicenter retrospective cohort study performed at three tertiary hospitals in Wuhan, Hubei Province, China (Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology; Zhongnan Hospital of Wuhan University; and the Central Hospital of Wuhan) from February 15, 2020 to March 25, 2020.

Exclusion criteria
1) Patients younger than 18 years old.
2) Noncurrent users of ACE inhibitors or ARBs prior to hospitalization.
3) Patients still hospitalized at the end of the study.
All patients were treated according to the standard protocols for antiviral, antibiotic, glucocorticoid, and Chinese medicine treatments.
The ethics committee of China-Japan Friendship Hospital approved this study (2020-21-K16). Written informed consent was waived due to the rapid emergence of this infectious disease. Written informed consent was waived due to the rapid emergence of this infectious disease.

Group division
We divided the patients into two groups. The ACEI/ARB group included patients who were current users of ACE inhibitor or ARB medication, while the control group included noncurrent users. Patients in the ACEI/ARB group were further divided into subgroups of a continued medication group and a terminated medication group according to the application of ACE inhibitors or ARBs during hospitalization.

Data collection and analysis
We collected data on the following parameters from the hospital electronic medical record systems, nursing records, laboratory examination systems, and radiological examinations and obtained standardized data collection forms: demographic characteristics, comorbidities, medication history within one month, symptoms at admission, laboratory finding changes from day 1 to day 14, radiological manifestations, treatment during hospitalization and outcome data that contained the rate of in-hospital death and progression, the duration of viral shedding, the length of hospital stay and the time from onset to death or discharge. Patients with cardiovascular disease and diabetes often combined the medication with statins [19] and oral hypoglycemic agents, especially thiazolidinediones, which have been reported to have an impact on the level of ACE2 by several studies [14,20] . To further control for potential confounders, data on the use of statins, thiazolidinediones and other antihypertensive agents (β receptor blocking agent and diuretics) prior to admission in each group were calculated within 90 days [6] .
Two researchers also independently reviewed the data collection forms to double check the data collected. Any missing or uncertain records of the epidemiological, medication and symptom data were collected and clarified through direct communication with patients and their families.
We compared the two groups in terms of the above aspects to identify the differences between ACE inhibitor or ARB users with nonusers prior to admission. Then, a subgroup analysis was conducted by comparing the dynamic changes of indicators involved in immune status and inflammatory reactions, as well as the outcomes between patients who continued and terminated medication during hospitalization. As hypertension itself could activate the RAS, patients with hypertension were excluded to avoid potential confounders. A comparison of the immune status, inflammatory reactions and outcomes between the ACEI/ARB and control groups in patients without hypertension was conducted.

Cytokine and chemokine measurement
To evaluate the impact of coronavirus and additional ACE inhibitors or ARBs on the production of cytokines or chemokines in the acute phase of the illness, plasma cytokines and chemokines [interleukin 1β (IL-1β), IL-2R, IL-6, IL-8, IL-10 and tumor necrosis factor α (TNF-α)] were measured using chemiluminescent immunoassay (CLIA) (CFDA approved) by Siemens IMMULITE 1000 for patients according to the manufacturer's instructions.

Definitions
Medications classified as ACE inhibitors were benazepril, perindopril and fosinopril, while the ARBs of the included patients were candesartan, irbesartan, valsartan, olmesartan, telmisartan and losartan.
The effective half-life period of the above agents varied from 6 to 50 hours; therefore, it took no more 10 days to achieve a stable blood drug concentration.
Patients were considered a current user of medication if they had a supply of medication to last until the date of hospitalization assuming an 80% compliance rate [6,21] . The patients who did not meet the definition were regarded as noncurrent users. ACE inhibitors or ARBs were considered to be continued if they were given more than 50% of the days during hospitalization [8] ; otherwise, they were considered to be terminated.
In-hospital progression was defined as a decline in PaO 2 /FiO 2 of more than 100 mmHg or the need for invasive positive pressure ventilation (IPPV) and/or extracorporeal membrane oxygenation (ECMO) during hospitalization.
Shock was defined according to the interim guidance of the WHO for novel coronavirus [22,23] . Acute kidney injury (AKI) was identified and classified on the basis of the highest serum creatinine level or urine output criteria according to the Kidney Disease Improving Global Outcomes Classification (KDIGO) [23,24]. Respiratory failure, coagulation and liver failure were defined as a Sequential Organ Failure Assessment (SOFA) score greater than or equal to two points.

Statistical analysis
Descriptive statistics included proportions for categorical variables and the mean (standard deviation) or median (interquartile range) for continuous variables. Data were unadjusted unless specifically stated otherwise.

Processing of missing data
When the missing rate of vital variables involved in our study was less than 15%, we used SAS predictive mean matching imputation to replace missing values within each variable, while the variables were abandoned when the missing rate reached 20%.

Processing of the unbalanced sample size: propensity score matching
The propensity score matching (PSM) method was applied at a ratio of 1:4 between the ACEI/ARB group and the control group. Age, sex and medication history prior admission which might be potential confounders associated with ACEI/ARB were matched variables in PSM to derive the cohort.
Proportions were compared using χ2 or Fisher's exact tests, and continuous variables were compared using the t test or Wilcoxon rank sum test, as appropriate. Statistical significance was defined as a 2tailed P value of ≤.05. SAS software, version 9.4 (SAS Institute Inc.) was used for all analyses.

Results
From February 15, 2020 to March 25, 2020, a total of 229 patients with confirmed cases of COVID-19 were admitted, and 175 patients not taking an ACE inhibitor or ARB medication. Among the 54 patients with a medication history, three of them were regarded as noncurrent users according to the definition and were excluded ultimately. The propensity score matching (PSM) method was applied at a ratio of 1:4 between the ACEI/ARB group (n=40) and the control group (n=160). Age, sex and medication history 90 days prior admission were matched variables in PSM to derive the cohort.
Among the patients with the ACEI/ARB medication, 17 continued medication during hospitalization, while the other 34 terminated medication (Figure 1). The mean age was 58 ± 17 years, male patients accounted for 58% (n=115), and the disease severity status [18] at admission was general for most of the study cohort (n=144, 72%). (Table 1) The ACEI/ARB group included more patients with hypertension (75% vs. 23%, P<0.01) than the control group. The demographic characteristics, other comorbidities, severity of the condition and possible medication histories might have influenced the ACE2 level but did not differ significantly between the two groups. No significant difference was found between the two groups in time from onset to hospitalization and to COVID-19 diagnosis. (Table 2) The symptoms, including fever, cough, hemoptysis, dyspnea, fatigue/myalgia and diarrhea, as well as vital signs, were not significantly different between the ACEI/ARB group and the control group. For

Comparison of organ function, treatment and outcomes during hospitalization between the ACEI/ARB and control groups (Table 3)
Patients with current use of ACE inhibitors or ARBs seemed to have a higher risk of developing shock (23% vs. 8%, P<0.01) and the application of vasoconstrictive agents (20% vs. 6%, P<0.01) than the control group; however, the percentages of respiratory failure, AKI, coagulation failure and liver failure were not different from those of the control group. In addition, the necessities for invasive IPPV and ECMO were not decreased in the ACEI/ARB group. were longer in the ACEI/ARB group than in the control group, while no difference was found in the rate of in-hospital progression or death.

Subgroup analyses: comparison between patients who continued and terminated medication during hospitalization
Among the patients with ACEI/ARB medication, 17 continued medication during hospitalization, while the other 34 terminated medication for several reasons. The dynamic changes in indicators involved in immune status and inflammatory reactions at the first, seventh, and fourteenth days of hospitalization as well as the outcomes were compared between the two groups. The missing rates of IL-2R and IL-8 at seven days and fourteen days after admission were extremely high and were not included in the analysis. Patients with continued use of ACE inhibitor or ARB had a consistently lower level of IL-1β and TNF-α at seventh day; while maintained higher level of the IL-10 from the seventh day to the fourteenth day than patients terminated the medication. However, the patients who terminated the medication had a trend of elevated lymphocyte counts [day 1, day 7, day 14: (  Figure 2). The rate of in-hospital progression and death were not significantly different between the two groups ( Table 5).

Subgroup analyses: a comparison of the immune status, inflammatory reactions and outcomes between the ACEI/ARB and control groups in patients with hypertension
Among 66 patients with hypertension, 30 patients were divided into the study group (ACEIs/ARBs group) and the other 36 patients were in the control group. in the study group than in the control group; however, no difference was detected in the rate of inhospital progression and death between the two groups.

Discussion
To our knowledge, this is the first study to evaluate the clinical manifestations, inflammatory responses, and viral clearence of COVID-19 patients treated with ACE inhibitor and ARB medication by a multicenter, retrospective cohort control study and to allow a dynamic observation of inflammatory responses by continuous monitoring from the first to fourteenth day.
The major findings of our study were that ACE inhibitor or ARB medication had an impact on inhibiting the proinflammatory response but promoted the anti-inflammatory response persistently and extended the duration of viral shedding, had an impact on hemodynamics as well. The inflammatory factors of COVID-19 patients should be monitored as route examination. For patients with excessive inflammatory responses and stable hemodynamics, ACEIs or ARBs might be tried to relieve the inflammatory storm, but the antiviral treatment should be enforced and the hemodynamics should be monitored closely; for patients with low levels of proinflammatory factors or instability hemodynamics, the agents might not be used to avoid a delay in viral clearance or increase the risk of shock.
Inflammation is mediated by proinflammatory cytokines and anti-inflammatory cytokines.
Inappropriate elevated expression of proinflammatory cytokines can result in sepsis, tissue destruction, or death [21,25] . Our study revealed that the plasma levels of IL-1β, IL-8 and TNF-α in patients taking ACE inhibitor or ARB were lower than those in patients not taking medication; in addition, persistently lower levels of proinflammatory factors were maintained in patients who continued medication during hospitalization, which was consistent with the previous experimental results by Gullestad L and his colleagues [26] with the conclusion that high-dose enalapril was associated with a significant decrease in IL-6 activity in patients with severe chronic heart failure. The specific organ and systemic inflammatory responses were postulated to attenuate through a reduction in the level of cytokines, which might be explained by the attenuating effects of ACE inhibitors through the deactivation of the ACE-AngII-AT1 axis but the stimulation of the ACE2-Ang-(1-7)-Mas axis in a feedback mechanism [9,27,28] as a negative regulator with attenuated cytokines and thus protecting the patients from organ injury. Consequently, some authors [29,30] have speculated that the use of ACEIs/ARBs might actually be a potentially beneficial intervention in those with COVID-

19.
Apart from organ protection by attenuating the inflammatory response, basic investigation has shown that bradykinin and substance P produced by ACE inhibitors sensitize the sensory nerves of the airways and enhance the cough reflex [31,32] , which plays a protective role against pathogens. These two mechanics made it possible to improve the outcome in patients with pneumonia. Mortensen et al [6] found a significant decrease in mortality, the length of hospital stay, and mechanical ventilation in patients taking ACE inhibitors or ARBs who were hospitalized with pneumonia compared to a matched cohort. A meta-analysis [33] that included 19 studies noted that patients taking ACE inhibitors were associated with a significant approximately one-third reduction in the risk of pneumonia compared with controls. In addition, a recent study [8] by Christopher Henry also observed lower rates of death and intubation with continued use of ACE inhibitors than with terminated use (OR=0.25; 95% CI, 0.09-0.64) throughout the hospital stay in cases of viral pneumonia not due to coronavirus.
Unfortunately, our study did not find decreased mortality in patients with current use of ACE inhibitors or ARBs, even though we analyzed patients with continued medication during hospitalization and without hypertension to avoid potential confounding factors. The most likely explanation was that our study included a small number of patients, while most of their patients had mild cases as determined by diseases severity status and without excessive inflammatory reactions, which was the target for ACE inhibitors or ARBs.
What is noteworthy is that ACE inhibitors or ARBs increased the risk of shock and the necessity of vasoconstrictive agents. It could be explained by the nature of the antihypertensive agents and came as a revelation to us that it should be avoided in patients with instability hemodynamics.
Our research also revealed that ACE inhibitors or ARBs led to prolonged viral shedding and extended the length of hospitalization. SARS-CoV-2 appears to bind to its target cells through angiotensinconverting enzyme 2 (ACE2). ACE inhibitors or ARBs upregulate ACE2 receptor expression in humans [34] by blocking the classic ACE pathway; thus, it is theoretically possible that the pre-existing use of these drugs might predispose a person to infection with a greater viral load of SARS-CoV-2 [13] . This hypothesis was supported by the evidence of Ferrario that there was a 4.7-fold increase in cardiac ACE2 mRNA by an ACE inhibitor [35] . Decreased lymphocyte counts and elevated plasma levels of IL-10 were also important discoveries in patients with ACE inhibitors or ARBs. Moreover, the lymphocyte counts in patients with continued use of an ACE inhibitor or ARB during hospitalization recovered slowly, as observed by successive monitoring on the first to fourteenth days. The immune status was weakened by lymphocytopenia and elevated anti-inflammatory cytokines in patients taking ACEIs, which might be another reason for the slow viral clearance. As the important criterion for discharge was the negative conversion of the SARS-CoV-2, prolonged viral shedding led to an extended length of hospitalization. This might be the defect of the ACE inhibitor or ARB and could explain the mixed results and controversy about their prescription in COVID-19 patients. For this reason, antiviral therapy in patients taking ACE inhibitors or ARBs for the purpose of relieving the inflammatory response should be reinforced, and their viral load should be monitored closely; however, in patients with low levels of proinflammatory factors, ACE inhibitors or ARBs should be not be used to avoid a delay in viral clearance.
An autopsy report revealed that mononuclear inflammatory infiltration dominated by lymphocytes was observed in the lungs, but no virus inclusion bodies were found [36] . We could then propose a hypothesis that cytokines released by inflammatory storms secondary to viral infection might be more important in the death of critically ill patients with COVID-19 than the viral infection itself in a certain period. From this perspective, it is possible that ACE inhibitors or ARBs might improve the outcome in critically ill patients with excessive inflammatory responses or severe multiple organ failure; when the inflammatory storm gradually diminishes, the focus of therapy should be on clearance of the virus and the enhancement of the immune system; and ACE inhibitors or ARBs with the adverse effect of prolonged viral shedding and weakening the defense of the immune system to some degree should be terminated. Prospective cohort and randomized controlled trials are needed to confirm this hypothesis and examine potential mechanisms of action.
Our study was limited by the small number of patients included and by not strictly excluding confounding factors, such as other medications, diabetes, coronary heart disease and chronic heart failure that might influence the RAS [37] . However, the PSM method was applied to eliminate bias. We also noticed that the number of patients with hypertension was much higher in the ACEI/ARB group, which might be an important confounding factor. By subgroup analyze, we found a similar result. We encourage further prospective randomized controlled studies designed by increasing the sample size and strictly excluding potential confounders to explore the impact of ACE inhibitors or ARBs on inflammatory responses, viral clearance and mortality in COVID-19 patients.

Consent for publication
Not applicable.

Ethics approval and consent to participate
The ethics committee of China-Japan Friendship Hospital approved this study (2020-21-K16). Written informed consent was waived due to the rapid emergence of this infectious disease.

Availability of data and materials
The datasets used and/or analysed during the current study are available from the corresponding author upon reasonable request.

Competing Interests
The authors declare that they have no competing interests.

Funding
The study was supported by the grants from the Zhejiang University special scientific research fund for COVID-19 prevention and control (2020XGZX008). The funder had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Authors' contributions
All authors made substantial contributions to the conception and design of the study or to the data acquisition, analysis or interpretation; reviewed and approved the final manuscript; and significantly contributed to this study. Dr. Zhan took full responsibility for the integrity of the submission and publication and was involved in the study design. Drs. Linna Huang and Ziying Chen involved in data collection, had full access to all of the data in the study, took responsibility for the integrity of the data and were responsible for data verification, as well as the drafting of the manuscript. Pro. Lin Hua took the responsibility for statistical analysis and the accuracy of the data analysis. Others involved in in data collection, had full access to all of the data in the study, and took responsibility for the integrity of the data. sensitization of airway sensory nerves: a mechanism for ACE-inhibitor cough. Nat Med 1996, 2(7):814-817.  Tables   Table 1. Baseline variables in the two groups prior to admission a : P<0.05; b : P<0.01; *defined according to the Chinese management guideline for COVID-19 (version 7.0) [18] .   [22,23] . AKI was identified and classified on the basis of the highest serum creatinine level or urine output criteria according to kidney disease, improving global outcome classification [23,24] . Respiratory failure, coagulation and liver failure were defined as a SOFA score greater than or equal to two points. # , defined as a decline in PaO 2 /FiO 2 greater than 100 mmHg or the need for IPPV and/or ECMO during hospitalization. AKI: acute kidney injury; HFNO: high flow nasal oxygenation; NPPV: noninvasive positive pressure ventilation; IPPV: invasive positive pressure ventilation; ECMO: extracorporeal membrane oxygenation b : P<0.01; IL-1β, interleukin 1β; TNF-α, tumor necrosis factor α; D1, the first day after admission; D7, the seventh day after admission; D14, the fourteenth day after admission Table 5. Outcomes in patients who continued and those who terminated ACEIs/ARBs during hospitalization a : P<0.05; # , defined as a decline in PaO 2 /FiO 2 greater than 100 mmHg or the need for IPPV and/or ECMO during hospitalization.   The dynamic changes in the lymphocyte counts and inflammatory factors between patients who continued and those who terminated ACEIs/ARBs during hospitalization. Patients with continued use of ACE inhibitor or ARB had a consistently lower level of IL-1β and TNF-α at seventh day; while maintains higher level of the IL-10 from the seventh day to the fourteenth day than patients terminated the medication. The patients who terminated the medication had a trend of elevated lymphocyte counts from the first day to the fourteenth day, while they had a trend of higher levels of IL-1β and TNF-α at the seventh day and returned to the baseline level on the fourteenth day.