key: cord-0764880-fz8do7qy
authors: Mitacchione, Gianfranco; Schiavone, Marco; Curnis, Antonio; Arca, Marcello; Antinori, Spinello; Gasperetti, Alessio; Mascioli, Giosuè; Severino, Paolo; Sabato, Federica; Caracciolo, Maria M.; Arabia, Gianmarco; D'Erasmo, Laura; Viecca, Maurizio; Mancone, Massimo; Galli, Massimo; Forleo, Giovanni B.
title: Impact of prior statin use on clinical outcomes in COVID-19 patients: data from tertiary referral hospitals during COVID-19 pandemic in Italy
date: 2020-12-29
journal: J Clin Lipidol
DOI: 10.1016/j.jacl.2020.12.008
sha: 7d9d8ae61d70431234445750eaa62a05af429cfe
doc_id: 764880
cord_uid: fz8do7qy

BACKGROUND: Epidemiological evidence suggests that anti-inflammatory and immunomodulatory properties of statins may reduce the risk of infections and infection-related complications. OBJECTIVE: We aimed to assess the impact of prior statin use on coronavirus disease (COVID-19) severity and mortality. METHODS: In this observational multicenter study, consecutive patients hospitalized for COVID-19 were enrolled. In-hospital mortality and severity of COVID-19 assessed with National Early Warning Score (NEWS) were deemed primary and secondary outcomes, respectively. Propensity score (PS) matching was used to obtain balanced cohorts. RESULTS: Among 842 patients enrolled, 179 (21%) were treated with statins before admission. Statin patients showed more comorbidities and more severe COVID-19 (NEWS 4 [IQR 2–6] vs 3 [IQR 2–5], p < 0.001). Despite having similar rates of intensive care unit admission, noninvasive ventilation, and mechanical ventilation, statin users appeared to show higher mortality rates. After balancing pre-existing relevant clinical conditions that could affect COVID-19 prognosis with PS matching, statin therapy confirmed its association with a more severe disease (NEWS ≥5 61% vs. 48%, p = 0.025) but not with in-hospital mortality (26% vs. 28%, p = 0.185). At univariate logistic regression analysis, statin use was confirmed not to be associated with mortality (OR 0.901; 95% CI: 0.537 to 1.51; p = 0.692) and to be associated with a more severe disease (NEWS≥5 OR 1.7; 95% CI 1.067–2.71; p = 0.026). CONCLUSIONS: Our results did not confirm the supposed favorable effects of statin therapy on COVID-19 outcomes. Conversely, they suggest that statin use should be considered as a proxy of underlying comorbidities, which indeed expose to increased risks of more severe COVID-19.

Coronavirus disease 2019 (COVID-19) is a viral infection caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), potentially leading to acute respiratory distress syndrome (ARDS) requiring invasive ventilation and intensive care unit (ICU) admission. It has to be underlined though, that some patients may be asymptomatic or have mildly symptomatic COVID-19. [1] [2] [3] Despite the exact pathophysiologic mechanisms of COVID-19 remain largely unknown, it has been suggested that, during the response to SARS-CoV-2 infection, the immune dysregulation and the high levels of proinflammatory cytokines might represent pivotal causes of tissue injury. [4] [5] [6] The severity of COVID-19 seems to be related with older age and comorbidities. In particular, patients with pre-existing cardiovascular disease (CVD) have an elevated risk of adverse outcomes, and, on the other hand, SARS-CoV-2 infection is associated with vascular and arrhythmic complications. [7] [8] [9] [10] [11] [12] Traditional cardiovascular (CV) risk factors, such as dyslipidemia, are recognized to have an impact on the immune system, [13] [14] [15] and therefore it has been postulated that pre-existing CV risk factors may be related with the immune response to SARS-CoV-2 infection and with COVID-19 severity. 7 Hydroxymethyl-glutaryl Coenzyme A (HMG-CoA) reductase inhibitors, or statins, which are widely used to prevent and treat CVD because of their lipid-lowering ability, 16 have been proposed to have a role as an add-on treatment for COVID-19 patients. 17, 18 This role has been suggested not only in light of their proven lipid-lowering ability, but mostly because of their known pleiotropic effects on vascular balance, thrombogenic response, 19 oxidative stress, and inflammation, at different levels. 20 Indeed, some observational studies conducted among outpatients and hospitalized patients with sepsis and/or Pneumonia, have demonstrated that statin therapy is associated with a reduced risk of pneumonia, a milder disease, and a reduced mortality, [21] [22] [23] [24] while others 25, 26 did not show a beneficial relationship between statin use and clinical outcomes in the same settings. Despite mechanisms of actions and advantages of statins that have been proposed, 17 little is known about their real world clinical impact in COVID-19 patients. Hence, we performed a retrospective study to assess the association of statin therapy at hospital admission with clinical outcomes in COVID-19 patients.

The study was approved by the institutional ethics board of the Italian hospitals involved (Luigi Sacco Hospital, Milan; Policlinico Umberto I Hospital, Rome; Spedali Civili Hospital, Brescia; Humanitas Gavazzeni Hospital; Bergamo). From February 23, 2020 to March 31, 2020, 987 consecutive patients presenting to the emergency department (ED) of the hospitals involved were screened. Confirmed cases of COVID-19 were defined with a positive result on a reverse-transcriptase-polymerase-chain-reaction assay performed on a nasopharyngeal swab, in accordance with World Health Organization (WHO) guidelines. 27 Inclusion criteria for study enrollment were: age $ 18 years; an in-hospital stay of at least 24 h, defined as ED management and subsequent early discharge or hospital admission; all patients included in the study had completed their hospital course (discharge or death) at the time of data analysis.

Patients were divided into 2 groups (statin vs. no-statin group): patients were included in the statin group if they were taking statins for at least 1 month before hospital admission for COVID-19. Data recorded at the time of ED admission/hospitalization included age, sex, medical history, symptoms reported to ED presentation, time from symptom onset, vital signs, laboratory values, arterial blood gas analysis, home medical therapy. Data regarding inhospital management, medical treatment and clinical outcomes were evaluated during data collection. Chronic kidney disease (CKD) was defined by the presence of an estimated glomerular filtration rate (eGFR) # 60 mL/ min/m 2 , calculated with the CKD-EPI equation. Fever was defined as axillary temperature of at least 37.5 C. Sepsis and septic shock were defined according to the 2016 Third International Consensus Definition. 28 Myocardial injury was defined in accordance with the fourth universal definition of myocardial infarction (MI) 29 (high sensitivity cardiac troponin [hs-cTn] above the 99th percentile of the upper reference limit). National Early Warning Score (NEWS) 30 was retrospectively assessed to determine the severity of illness and clinical deterioration at the moment of hospital admission. The six parameters included in this scoring system are respiratory rate, oxygen saturation, temperature, systolic blood pressure, heart rate and level of consciousness (Table 1 -data supplement). Patients with moderate to severe disease were defined by a NEWS $ 5 while patients with mild disease were defined by a NEWS , 5, in order to establish a group of patients that in real-world multicenter registry may exhibit a more severe disease. The primary outcome was in-hospital mortality. The secondary outcome of interest was the COVID-19 severity at presentation, defined by a NEWS $ 5.

Continuous variables are expressed as mean 6 standard deviation (SD) or median and interquartile range (IQR) if not normally distributed according to D'Agostino-Pearson test. Categorical data are expressed as absolute values and proportions. Student's t-test for independent samples with a confidence interval of 95% was used for continuous variables; otherwise, either logarithmic transformation of variables or nonparametric tests (Mann-Whitney U) were used. A chi-squared test or a Fisher exact test were performed to examine whether there was an association between categorical variables. Univariate linear regression was used to assess correlation between variables. Odds ratio (OR) and 95% confidence interval (CI) were reported. A two tailed P values , 0.05 were considered significant. All statistical analyses were performed using IBM SPSS Statistics (version 26.0).

We used the propensity score (PS) method to minimize the effect of potential confounders on selection bias. The PS for an individual was defined as the conditional probability of taking statin given the individual's clinical pre-existing conditions (covariates). To estimate the scores, we created a logistic regression model including twelve relevant covariates potentially related to COVID-19 clinical presentation and outcome. Variables considered for the PS matching were age $ 65 years, male gender, obesity, hypertension, diabetes, smoke, coronary artery disease (CAD), atrial fibrillation (AF), heart failure (HF), chronic kidney disease (CKD), chronic obstructive pulmonary disease (COPD), cerebrovascular disease and malignancy. Patients with prior statin therapy use were matched 1:1 to no-statin therapy patients by PS, using the nearest neighbor method with a caliper of 0.1 and no replacement. Absolute standard difference and t-test were applied for balancing check of covariates both before and after matching. For good balancing, t-test for equality of means should be non-significant after matching, and the absolute standard difference should be , 0.1 after matching (Table 1) .

A total of 842 patients met the inclusion criteria and were therefore enrolled as study cohort. Propensity score matching (PS) identified 145 patients in each group according to prior statin use.

Baseline and characteristic on admission before and after propensity score matching Patient demographic and clinical characteristics for the unadjusted and PS matched patients are shown in Table 2 . As many as thirteen covariates, identifying demographic and pre-existing clinical conditions (age, elderly patients [$65 years], gender, obesity, hypertension, diabetes, dyslipidemia, smoke, CAD, AF, HF, CKD and cerebrovascular disease), were significantly different (p , 0.05) between the two groups before matching. However, all covariates were comparable and well balanced (p . 0.05) between both groups after PS matching. Before PS matching of demographic and pre-existing clinical variables, patients in statin therapy showed lower oxygen saturation, PaO2, PaO2/FiO2 ratio, and a higher NEWS (4 [IQR 2-6) vs. 3 [IQR 2-5] p , 0.001) with a higher incidence of patients on statin therapy presenting moderate/severe disease (as meant as a NEWS$5) at the moment of hospital admission (58% vs. 44%, p , 0.001). After PS matching, only the presence of moderate/severe disease remained significantly more represented in the group of patients on statin therapy respect to patients in non-statin group (61% vs. 48%, p 5 0.025). This secondary outcome remained significantly associated with statin use also to univariate logistic regression (OR 1.7; 95% CI 1.067-2.71; p 5 0.026).

In-hospital management and outcomes before and after propensity score matching Data regarding in-hospital management and clinical outcomes are summarized in Table 3 . In the overall population, patients in the statin group were more likely to receive oxygen support (86% vs. 75% p 5 0.002) and to develop myocardial injury (20% vs. 9%, p , 0.001) than patients in non-statin group, as well as showing higher all-cause mortality rate (29% vs. 20%, p 5 0.006) and a longer in-hospital stay. With regard to laboratory and radiological findings (Table 4 ), in the non-matched population, patients in the statin group showed a higher white blood cell count, lactate dehydrogenase, creatine phosphokinase, creatinine, and hs-cTn, with lower levels of hemoglobin compared with patients in non-statin group patients. Based on radiological findings, pneumonia was diagnosed in the majority of patients (89%) on admission, with a higher incidence among patients in statin therapy (94 vs. 87%; p 5 0.013) and also showed worse radiological features. After PS matching, only hemoglobin level remained significantly lower in the statin group respect to nostatin group (13.2; IQR 12-14.4 vs. 13.6; IQR 12.4-15; p 5 0.037). Of note, in the statin patient cohort, a significantly lower level of C-reactive protein was noted with respect to the control group (20.2; IQR 8.13-99.7 vs. 36; IQR 11.6-99.7; p 5 0.021). Moreover, both groups showed a similar prevalence of pneumonia on admission, while worse radiological features were confirmed after PS matching in the statin group. While mortality rates significantly differed among the two unmatched groups (29% vs. 20%, p 5 0.006), after PS matching no significant differences were found (26% vs. 28%, p 5 0.185). At univariate logistic regression analysis for mortality, statin use was confirmed not to be associated with the primary outcome (OR 0.901; 95% CI: 0.537 to 1.51; p 5 0.692).

In-hospital mortality and severity of disease within the balanced cohort of statin users, stratified by statin intensity, 31 are shown in Table 5 . The majority of patients were treated with high intensity (48%) and moderateintensity statins (39%) respect to low-intensity (13%) (p , 0.05) (Fig. 1) . At univariate logistic regression, patients using high intensity and moderate intensity statins showed a worse clinical presentation of COVID-19 (NEWS $ 5) at hospital presentation respect to patients 

To the best of our knowledge, this is the first study specifically focused on the role of statins in COVID-19 patients. Hereby, we report data demonstrating that patients on prior statin therapy had a worse disease severity at the time of hospital admission for COVID-19 with respect to non-statin users, but nevertheless, statin therapy was not associated with an increased risk of in-hospital death. The proportion of our patients receiving statins (21%) was similar to that reported in other studies. 32 In the statin group, 46% of patients took statins for secondary prevention, whereas 54% took statins for dyslipidemia and primary prevention. The overall mortality rate reported in our cohort is higher than reported in similar series from China 3, 9, 33, 34 but is consistent with other cohorts from Italy and USA. [35] [36] [37] Our findings in the overall population confirmed that patients with CVD, and therefore receiving statins, have a higher mortality rate (29%), similar to what reported in other cohorts (Inciardi et al., 37 mortality in cardiac patients: 35.8%). Indeed, statin use underlies the presence of CV risk factors as dyslipidemia or CAD, taking into account that patients with CVD are supposed to exhibit worse outcomes. 34, 38 This finding can explain why patients with COVID-19 taking domiciliary statin therapy showed an increased in-hospital mortality. This negative relation is further confirmed by the reduction in this mortality rate after balancing pre-existing CV risk factors with PS matching.

The higher COVID-19 severity in the statin group, assessed with NEWS, is not matched with a higher mortality rate in these patients. In our series, COVID-19 severity was evaluated with NEWS to avoid confounding factors related with in-hospital treatment and ICU admission, also because, due to the Italian critical care increasing demand during the COVID-19 outbreak, 39,40 we could not rely on ICU admission. Indeed, mechanical ventilation and ICU admission were sometimes reserved to younger patients with less comorbidities that were more likely to recover from severe COVID-19. Indeed, while SOFA score showed its ability to predict the in-hospital mortality in sepsis patients already admitted in ICU, NEWS showed a greater ability to predict the risk of ICU admission among patients evaluated in the ED. 41 For these reasons, in contrast with other similar studies, 2,37 we preferred NEWS over SOFA score or other scores useful to track patients' status during the stay in ICU.

Several studies have evaluated the role of statins as an add-on treatment during infections. In a metanalysis, Tleyeh et al. 42 examined nine cohorts addressing the role 26 showed that statin use was not associated with decreased risk of pneumonia among immunocompetent community-dwelling older people. It has been postulated that statins may have beneficial effects in influencing the clinical course of COVID-19 by mitigating the impact of viral infection through their immunomodulatory and anti-inflammatory properties. 43 In an animal model, statins have shown to stabilize MYD88 levels and attenuate myocardial remodeling after a proinflammatory trigger such as hypoxia 44 that is a crucial feature of SARS-CoV-2 infection. This supposed beneficial effect was only partially noticed in our cohort, as patients receiving domiciliary statin therapy showed a lower C-reactive protein level, although a worse PaO2 and PaO2/ FiO2 ratio on admission. Moreover, COVID-19 patients with CV comorbidities exhibit direct cardiac involvement and worse outcomes. 34, 38 In this setting, COVID-19 patients may theoretically take advantage of statin therapy also for their cardioprotective actions and for the upregulation of Angiotensin-Converting Enzyme (ACE) 2 via epigenetic modifications, 45 that SARS-CoV-2 employs for cellular entry. 46 Our results did not confirm these supposed favorable effects of statin therapy, even after propensity score matching, so that our different results can mainly be explained by a ''healthy user effect'' and with indication bias noticed in other observational studies on statin use in similar clinical settings. 42 The higher severity of COVID-19, exhibited in patients on high-intensity statin treatment, could be partially explained with the enhanced severity of underlying CV risk factors and CVD in those patients.

Because of the retrospective nature of our study, whether our patients received statins was extracted from our inhospital medical records, and specific data on the inhospital drug regimen of statins (continuation vs. discontinuation) are lacking, though generally these drugs were continued during hospitalization, at least in the non-ICU setting. Indeed, precise data on drug administration in the ICU setting are lacking. Another potential limitation is represented by the struggle to assess compliance with statin therapy up to the patient's admission to the hospital. Given the logistical limitations during the COVID-19 outbreak, some laboratory data (i.e. lipid profile) were not collected in all patients. Moreover, our in-hospital based study cannot be considered representative of the overall ''statin population'' affected by COVID-19. Thus, since this was not a population-based study, it is still possible that the use of statin therapy could be associated with improved outcome in mild disease where patients do not seek hospital care. Moreover, patients who came to the ED but were discharged early (,24 h from admission) were not included. Therefore, the potential beneficial effects of statin therapy in COVID-19 patients treated outside of the hospital setting and its impact on mild disease need to be further investigated.

Our data suggest that statin therapy does not affect inhospital mortality in COVID-19 patients. Therefore, caution should be warranted in attributing benefits to statin therapy as an adjuvant or independent therapy and a part of the standard care for COVID-19 patients. These effects may instead reflect unmeasured baseline differences between users and non-users, that in our study were limited by using propensity score matching.

Pathological findings of COVID-19 associated with acute respiratory distress syndrome

Clinical characteristics of 138 hospitalized patients with 2019 novel coronavirus-infected pneumonia in Wuhan, China

Clinical characteristics of coronavirus disease 2019 in China

Characteristics of and important lessons from the coronavirus disease 2019 (COVID-19) outbreak in China: summary of a report of 72314 cases from the Chinese center for disease control and prevention

Characteristics of lymphocyte subsets and cytokines in peripheral blood of 123 hospitalized patients with 2019 novel coronavirus pneumonia (NCP). medRxiv

Acute coronary syndromes and covid-19: exploring the uncertainties

Prevalence and impact of cardiovascular metabolic diseases on COVID-19 in China

COVID-19 and the cardiovascular system

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

Ventricular tachycardia storm management in a COVID-19 patient: a case report

Oral anticoagulation and clinical outcomes in COVID-19: an Italian multicenter experience

Redefining the prognostic value of high-sensitivity troponin in COVID-19 patients: the importance of concomitant coronary artery disease

Inflammation in atherosclerosis. From pathophysiology to practice

inflammation and innate immunity

Inflammatory mechanisms linking obesity and metabolic disease

ESC/EAS Guidelines for the management of dyslipidaemias: lipid modification to reduce cardiovascular risk

Statin therapy in COVID-19 infection

The role of lipid metabolism in COVID-19 virus infection and as a drug target

Pleiotropic effects of statins

The potential role of statins in pneumonia

Statins and the risk of pneumonia: a population-based, nested case-control study

Statin treatment and reduced risk of pneumonia in patients with diabetes

Risk of community-acquired pneumonia and the use of statins, ace inhibitors and gastric acid suppressants: a population-based case-control study

The effect of statins on mortality in patients with bacteremia

Understanding the potential role of statins in pneumonia and sepsis

Statin use and risk of community acquired pneumonia in older people: population based case-control study

2019-nCoV) in suspected human cases. WHO -Interim Guid

Assessment of clinical criteria for sepsis for the third international consensus definitions for sepsis and septic shock (sepsis-3)

Fourth universal definition of myocardial infarction

The ability of the National Early Warning Score (NEWS) to discriminate patients at risk of early cardiac arrest, unanticipated intensive care unit admission, and death

PCNA guideline on the management of blood cholesterol: a report of the American college of cardiology/American heart association task force on clinical practice guidelines

Prior statin therapy is associated with a decreased rate of severe sepsis

Clinical features of patients infected with 2019 novel coronavirus in Wuhan

Cardiovascular implications of fatal outcomes of patients with coronavirus disease 2019 (COVID-19)

Characteristics and outcomes of 21 critically ill patients with COVID-19 in Washington state

Covid-19 in critically ill patients in the seattle region -case series

At the heart of COVID-19

Prevalence and impact of myocardial injury in patients hospitalized with COVID-19 infection

Critical care utilization for the COVID-19 outbreak in lombardy, Italy: early experience and forecast during an emergency response

Quis custodiet ipsos custodes: are we taking care of healthcare workers in the Italian COVID-19 outbreak?

Comparison of the accuracy of three early warning scores with SOFA score for predicting mortality in adult sepsis and septic shock patients admitted to intensive care unit

Statins for the prevention and treatment of infections: a systematic review and meta-analysis

Immune modulatory effects of statins

Atorvastatin attenuates myocardial remodeling induced by chronic intermittent hypoxia in rats: partly involvement of TLR-4/MYD88 pathway

Hiding in plain sight: an approach to treating patients with severe covid-19 infection

Renin-angiotensin-aldosterone system inhibitors in patients with covid-19

Supplementary data related to this article can be found online at https://doi.org/10.1016/j.jacl.2020.12.008.