Low molecular weight heparin is useful in adult COVID-19 inpatients. Experience during the first Spanish wave: observational study

ABSTRACT BACKGROUND: The intensity of the thromboprophylaxis needed as a potential factor for preventing inpatient mortality due to coronavirus disease-19 (COVID-19) remains unclear. OBJECTIVE: To explore the association between anticoagulation intensity and COVID-19 survival. DESIGN AND SETTING: Retrospective observational study in a tertiary-level hospital in Spain. METHODS: Low-molecular-weight heparin (LMWH) status was ascertained based on prescription at admission. To control for immortal time bias, anticoagulant use was analyzed as a time-dependent variable. RESULTS: 690 patients were included (median age, 72 years). LMWH was administered to 615 patients, starting from hospital admission (89.1%). 410 (66.7%) received prophylactic-dose LMWH; 120 (19.5%), therapeutic-dose LMWH; and another 85 (13.8%) who presented respiratory failure, high D-dimer levels (> 3 mg/l) and non-worsening of inflammation markers received prophylaxis of intermediate-dose LMWH. The overall inpatient-mortality rate was 38.5%. The anticoagulant nonuser group presented higher mortality risk than each of the following groups: any LMWH users (HR 2.1; 95% CI: 1.40-3.15); the prophylactic-dose heparin group (HR 2.39; 95% CI, 1.57-3.64); and the users of heparin dose according to biomarkers (HR 6.52; 95% CI, 2.95-14.41). 3.4% of the patients experienced major hemorrhage. 2.8% of the patients developed an episode of thromboembolism. CONCLUSIONS: This observational study showed that LMWH administered at the time of admission was associated with lower mortality among unselected adult COVID-19 inpatients. The magnitude of the benefit may have been greatest for the intermediate-dose subgroup. Randomized controlled trials to assess the benefit of heparin within different therapeutic regimes for COVID-19 patients are required.


INTRODUCTION
The disease caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), known as coronavirus disease , is a new pandemic that appeared in the city of Wuhan, China, in December 2019. 1,2 Over the past four months, COVID-19 has become a worldwide pandemic, such that 32,150,495 cases and 982,680 deaths globally have been reported. In Spain, 682,267 cases and 45,252 deaths were reported up to September 24, 2020. Although the majority of COVID-19 cases have resolved spontaneously, some have developed various fatal complications, including organ failure, septic shock, pulmonary edema, severe pneumonia and SARS. 3 Current data support the concept that disseminated intravascular coagulation (DIC) in sepsis is a coagulation disorder induced by infection, and that it also represents an acute systemic inflammatory response that leads to endothelial dysfunction. 4,5 Recent data in the literature show that severe COVID-19 is commonly complicated with coagulopathy and that DIC might exist in the majority of deaths. 6 Moreover, a remarkably high incidence of venous thromboembolism (VTE) has been reported in patients hospitalized with  Heparin may have positive effects on COVID-19 patients. 8 The American College of Chest Physicians (ACCP) recommends use of the Padua prediction score, which is a validated risk assessment model, in order to identify hospitalized medical patients who are at high risk of VTE and who should therefore I MD, PhD. Physician, Department of Hematology, Complejo Asistencial Universitario de Salamanca (CAUSA), Instituto de Investigación Biomédica de Salamanca (IBSAL), Universidad de Salamanca (USAL), Salamanca, Spain.
Based on these findings, it seems that prophylactic doses of heparin for patients with severe COVID-19 and coagulopathy could be useful, and this has been recommended by some expert consensuses. 8,[10][11][12] Nonetheless, the high incidence of coagulopathy and thrombotic complications that is seen among COVID-19 patients despite use of antithrombotic prophylaxis could be important for decision-making with regard to the intensity of thromboprophylaxis to be applied. Therefore, the benefits of high doses of antithrombotic drugs in COVID-19 cases need to be clarified.

OBJECTIVE
The present study was designed to explore the intensity of the thromboprophylaxis needed as a potential factor for preventing in-hospital mortality associated to COVID-19.

Study design and population
We performed a retrospective observational study in Spain on all patients with a diagnosis of COVID -19 who

Laboratory procedures and data collection
The baseline characteristics of the patients were retrospectively collected from the electronic medical record system and from the concomitant therapies. We started a registry of patients hospi-

Baseline characteristics, treatment received, UCI admission and mortality among the patients (comparison between heparin subgroups)
Out of the 615 patients who received heparin, 410 (66.7%) received a prophylactic dose, 120 (19.5%) received a therapeutic dose and 85 other patients (13.8%) undergoing LMWH prophylaxis presented respiratory failure, high D-dimer levels (> 3 mg/l) and non-worsening of inflammation markers, and thus received an intermediate heparin dose (heparin-dose group according to biomarkers).      to biomarkers. The gender, body mass index (BMI), presence of pneumonia and levels of fibrinogen and D-dimer at diagnosis were similar in all the heparin groups. The percentage of intensive care unit (ICU) admission was higher in the heparin-dose group according to biomarkers. The inpatient mortality was lower in the heparin-dose group according to biomarkers (28.2%) and the prophylactic-dose group (32.7%). Figure 1 shows the overall survival based on type of heparin use.

COVID-19 survival in relation to use of heparin
The anticoagulant nonuser group presented higher mortality risk than any LMWH users (HR 2.1; 95% CI: 1.40-3.15). Three other variables retained their independent prognostic value for predicting higher inpatient mortality: age, DIC-ISTH score and LDH levels.

Bleeding and thromboembolic complications
Among the 690 patients, 24 patients (3.4%) experienced major hemorrhage, but only one case was fatal ( Table 4). Two cases of major bleeding complications occurred in patients without heparin (2.6%), eight cases of major hemorrhage occurred in the lowheparin-dose group (1.9%), six cases of major bleeding complications occurred among the patients with therapeutic-dose heparin (5%) and eight cases of major bleeding occurred in the heparindose group according to biomarkers (9.4%) (P = 0.007, between heparin groups). Nineteen patients (2.8%) developed an episode of thromboembolism, which was fatal in three cases ( Table 5).

DISCUSSION
We report in this retrospective observational study how the administration of LMWH at the time of admission was associated with a reduced mortality rate among unselected adult COVID-19 patients. The magnitude of the benefit may have been greatest for the group of patients who received a heparin dose according to biomarkers. It should be noted that overall, although major bleeding was more frequently reported in the higher dose groups, only one fatal event was reported. In addition, young patients Non-heparin Prophylactic-dose heparin Therapeutic-dose heparin Heparin dose according to biomarkers P-value < 0.001 between non-heparin group and prophylactic-dose group; P-value < 0.001 between non-heparin group and therapeutic-dose group; P-value < 0.001 between non-heparin group and heparin-dose group according to biomarkers. There is no strong evidence to support the idea that routine anticoagulation therapy would be effective for preventing sepsis. 19 A meta-analysis on randomized controlled trials comparing LMWH versus placebo in sepsis suggested that LMWH might reduce mortality among septic patients. 20 Another recent meta-analysis suggested that anticoagulation therapy would be beneficial only for patients with sepsis-induced DIC and not for the entire population of patients with sepsis. 21 Moreover, the guidance for diagnosis and treatment of DIC provided by the ISTH states that use of therapeutic doses of heparin should be considered in cases of DIC in which thrombosis is predominant. 22 A multicenter cohort study conducted by Japanese institutions reported that use of high-intensity anticoagulation therapy was associated with better outcomes among patients with sepsis-induced DIC. 23 Currently, there is little information on the use of LMWH in relation to COVID-19. Anticoagulant therapy that was implemented mainly using LMWH at a prophylactic dose was associated with a better prognosis in a series of COVID-19 patients in China, but the infection level was severe in all patients. 8 11 However, in addition, we suggest that the intensity of the thromboprophylaxis used may be a potential factor for preventing in-hospital mortality associated with COVID-19.
Besides its use as an anticoagulant, heparin has demonstrated excellent anti-inflammatory properties in animal models and clinical trials. 24 Use of LMWH was found to reduce serum IL-6 levels, which are a key factor in patients with severe COVID-19, and to reduce TNF-α levels. 25 Heparin has been seen to exert an inhibitory effect on replication activity and against attachment and entry of enveloped viruses, in relation to several viruses: human herpes simplex virus (HSV), human immunodeficiency virus (HIV), SARS coronavirus and influenza virus (H5N1). 26 Moreover, heparin prevents Zika virus-induced cell death of human neural progenitor cells. 27 Therefore, the potential anti-inflammatory and antiviral properties of LMWH might partly explain its beneficial mechanism.
Throm boprophylaxis using high doses of LMWH may lead to bleeding, which can be fatal. In our series, major bleeding was presented in 3.4% of the patients and the bleeding rate was significantly higher in the high-heparin-dose group (7.3%). The incidence of major bleeding in critically ill patients who received LMWH prophylaxis was reported to range from 1.2% to 5.4% in three trials.
The rate of thrombosis in our series seemed very low (2.8%).
The exact prevalence or incidence of venous thromboembolism in COVID-19 patients is unknown. Different reports have indicated VTE rates ranging from 11% to 31%, and the highest incidence of VTE has been found among patients admitted to intensive care units. 28 Thrombotic complications have only rarely been described in    presented D-dimer levels that were 2.5 to 5-fold higher than those in patients without this. 6,[30][31][32][33][34][35] found an association between higher D-dimer levels (9-fold higher) and mortality among patients with severe COVID-19. 34 The risk of severe illness was more frequent in patients with D-dimer levels above 0.5 mg/l. 33 A pooled analysis on four retrospective observational studies found that D-dimer levels were considerably higher in COVID-19 patients with severe disease than in those without this (weighted mean difference: 2.97 mg/l; 95% CI: 2.47-3.46 mg/l), but the heterogeneity across the four studies was relatively high (i.e. I 2 = 94%; P < 0.001). 36 Petrilli et al. showed that there was a relationship between D-dimer level and its trajectory and the frequency of adverse clinical events. 37 In our unselected cohort of COVID-19 patients, the median D-dimer level on admission was significantly higher in non-survivors (2.1 g/l) than in survivors (0.9 g/l), in the univariate analysis, but the prognostic impact of this finding was not maintained in the multivariate analysis. According to our model, the DIC-ISTH score, which includes D-dimer data, is a more valuable criterion with independent prognostic value for predicting inpatient mortality risk.
In addition, the mortality risk index also included age, LDH, underlying diseases, DIC-ISTH score and use of LMWH, which would facilitate identification of patients with high mortality risk, among unselected adult COVID-19 cases at the time of hospital admission. In fact, the reported area under the receiver operating characteristic curve for this category is of great value (COVID-19 mortality index 0.869).
The benefit of heparin doses needs to be balanced against the risk of bleeding. We observed an excess of bleeding complications in patients who received the highest heparin dose.
Along the same lines, bleeding events were observed in another study in 7.8% of the patients hospitalized with COVID-19 and were sensitive to use of escalated doses of anticoagulants and to the nature of data collection. 29 The limitations of our study are those that are inherent to an observational retrospective single-center study. Potential selection and immortal time bias do exist in this kind of study. Through assessing the potential role and magnitude of this confounding, the inherent differences between the heparin groups can be understood. We had detailed information on patient characteristics among the heparin groups. The analyses were adjusted for multiple background variables to minimize bias. The outcome was survival at the time of the analysis: at that time, only one patient was still hospitalized. On the other hand, to control for immortal time bias, the anticoagulant dose was analyzed as a time-dependent variable.
Although our study focused on coagulation parameters, other variables could also impact on mortality. The concomitant therapy, including LMWH, was not assessed in relation to a control.
The true rate of VTE was also perhaps underestimated due to the impossibility of carrying out imaging studies on some patients with clinically suspected VTE. Nonetheless, our report describes the experience of a single center with a large patient population that was homogeneously managed in accordance with the local guidelines, which were regularly updated with the emerging information. If a multicenter study had been conducted, this might have given rise to introduction of additional confounding factors, due to the heterogeneity of management protocols across the centers.

CONCLUSIONS
Our results suggest that application of LMWH at the time of admission significantly reduced the mortality rate among these unselected adult COVID-19 inpatients. The LMWH dose could have prognostic impact, although overall, major bleeding was more frequently reported in the high-dose group.
Further research is needed to tailor heparin prophylaxis and ascertain the correct dose for adults COVID-19 patients.