Advertisement

The effect of anticoagulation on clinical outcomes in novel Coronavirus (COVID-19) pneumonia in a U.S. cohort

  • Lei Lynn
    Correspondence
    Corresponding authors at: 900 23rd St NW, Washington, DC 20037, United States of America.
    Affiliations
    Division of Hospital Medicine, The George Washington University School of Medicine and Health Sciences, The GW Medical Faculty Associates, Washington, DC, United States of America
    Search for articles by this author
  • Juan A. Reyes
    Correspondence
    Corresponding authors at: 900 23rd St NW, Washington, DC 20037, United States of America.
    Affiliations
    Division of Hospital Medicine, The George Washington University School of Medicine and Health Sciences, The GW Medical Faculty Associates, Washington, DC, United States of America
    Search for articles by this author
  • Katrina Hawkins
    Affiliations
    Department of Anesthesiology and Critical Care Medicine, The George Washington University School of Medicine and Health Sciences, The GW Medical Faculty Associates, Washington, DC, United States of America
    Search for articles by this author
  • Arjun Panda
    Affiliations
    Department of Medicine, The George Washington University School of Medicine and Health Sciences, The GW Medical Faculty Associates, Washington, DC, United States of America
    Search for articles by this author
  • Laura Linville
    Affiliations
    Department of Medicine, The George Washington University School of Medicine and Health Sciences, The GW Medical Faculty Associates, Washington, DC, United States of America
    Search for articles by this author
  • Walaa Aldhahri
    Affiliations
    Department of Medicine, The George Washington University School of Medicine and Health Sciences, The GW Medical Faculty Associates, Washington, DC, United States of America
    Search for articles by this author
  • Ghazal Kango
    Affiliations
    Division of Hospital Medicine, The George Washington University School of Medicine and Health Sciences, The GW Medical Faculty Associates, Washington, DC, United States of America
    Search for articles by this author
  • Sneha Shah
    Affiliations
    Department of Medicine, The George Washington University School of Medicine and Health Sciences, The GW Medical Faculty Associates, Washington, DC, United States of America
    Search for articles by this author
  • Shant Ayanian
    Affiliations
    Division of Hospital Medicine, The George Washington University School of Medicine and Health Sciences, The GW Medical Faculty Associates, Washington, DC, United States of America
    Search for articles by this author
  • Karolyn Teufel
    Affiliations
    Division of Hospital Medicine, The George Washington University School of Medicine and Health Sciences, The GW Medical Faculty Associates, Washington, DC, United States of America
    Search for articles by this author
Published:November 05, 2020DOI:https://doi.org/10.1016/j.thromres.2020.10.031

      Highlights

      • Anticoagulation strategies for patients with COVID-19 infections remain unclear.
      • Outcomes of prophylactic versus therapeutic anticoagulation were compared.
      • In our cohort, therapeutic anticoagulation provided no mortality benefit over thromboprophylaxis.
      • Therapeutic anticoagulation was associated with increased adverse events.

      Abstract

      Background

      COVID-19 infection is associated with D-dimer elevations, high rates of thrombus formation, and poor clinical outcomes. We sought to determine if empiric therapeutic anticoagulation (AC) affected survival in COVID-19 patients compared to standard prophylactic AC.

      Methods

      Retrospective analysis of 402 COVID-19 patients hospitalized between March 15 and May 31, 2020 was performed. Clinical outcomes were compared between 152 patients treated with therapeutic AC to 250 patients on prophylactic AC. An elastic net logistic regression was designed to first identify the important variables affecting mortality. These variables were then included as covariates to AC in standard multivariate logistic regression models studying the effect of AC on death. Nonparametric survival analysis was conducted, and Kaplan Meier curves were constructed.

      Results

      Increased mortality was associated with therapeutic AC [OR 3.42 (2.06, 5.67)]. The log-rank test was statistically significant at p = 0.001 showing higher mortality for patients treated with therapeutic AC compared to prophylactic AC. Subset analysis of critically ill and intubated patients had similar survival curves regardless of AC dose. The log-rank test was not significant even with Prentice modification. For non-ICU patients, the log rank test favoring prophylactic AC disappeared when the analysis was stratified by D-dimer level less or greater than 3 μg/mL. Approximately 9% of patients receiving therapeutic AC experienced clinically significant bleeding or thrombocytopenia, versus 3% in those receiving prophylactic AC.

      Conclusions

      In our cohort, therapeutic anticoagulation provided no mortality benefit over thromboprophylaxis, independent of co-morbidities or disease severity. More adverse events were observed with therapeutic AC.
      The clinical presentation of the novel Coronavirus infection (COVID-19) is highly variable. One of the features described in the literature is the thrombogenic nature of a coagulopathy attributed to COVID-19 infection [
      • Gorlinger K.
      • Dirkmann D.
      • Gandhi A.
      • Simioni P.
      COVID-19 associated coagulopathy and inflammatory response: what do we know already and what are the knowledge gaps?.
      ]. Autopsy studies on patients who died of COVID-19 have revealed high rates of thromboses [
      • Wichmann D.
      • Sperhake J.P.
      • Lutgehetmann M.
      • et al.
      Autopsy findings and venous thromboembolism in patients with COVID-19: A prospective cohort study.
      ], and retrospective studies suggest a higher frequency of thromboses significantly associated with elevated D-dimer levels [
      • Cui S.
      • Chen S.
      • Li X.
      • Liu S.
      • Wang F.
      Prevalence of venous thromboembolism in patients with severe novel coronavirus pneumonia.
      ]. Consequently, some clinicians have implemented protocols for the use of intermediate or therapeutic AC as part of routine care for severe COVID-19 patients [
      • Bikdeli B.
      • Madhavan M.V.
      • Jimenez D.
      • et al.
      COVID-19 and thrombotic or thromboembolic disease: implications for prevention, antithrombotic therapy, and follow-up: JACC State-of-the-Art Review.
      ]. To date, data remains scarce as to the benefit of empiric use of therapeutic AC [
      • Turshudzhyan A.
      Anticoagulation options for coronavirus disease 2019 (COVID-19)-induced coagulopathy.
      ].
      At the beginning of the pandemic, all patients admitted with COVID-19 to the George Washington University Hospital (GWUH) in Washington D.C. were treated with prophylactic AC, unless contraindicated. As awareness of the elevated thrombotic risk developed, most COVID-19 patients admitted to our ICU were empirically started on therapeutic AC per provider discretion unless contraindicated. The medicine wards were advised to consider initiation of empiric therapeutic AC when the D-dimer level (checked daily) exceeded 3 μg/mL. We have previously reported this level as having a statistically significant increase in mortality (OR 7.5 [3.8, 14.6]) [
      • Ayanian S.
      • Reyes J.
      • Lynn L.
      • Teufel K.
      The association between biomarkers and clinical outcomes in novel coronavirus pneumonia in a US cohort.
      ], and it is estimated as having the highest positive predictive value for VTE per Cui, et al. [
      • Cui S.
      • Chen S.
      • Li X.
      • Liu S.
      • Wang F.
      Prevalence of venous thromboembolism in patients with severe novel coronavirus pneumonia.
      ].
      In this retrospective study, we compared the clinical outcomes of COVID-19 patients on therapeutic AC to those on prophylactic AC. Between March 15, 2020 and May 31, 2020, 431 patients with confirmed COVID-19 infection were admitted to our hospital. Four hundred and two (402) patients were included in the analysis, while 29 patients who received neither prophylactic nor therapeutic AC were excluded. Patients were followed for their entire hospital course up until discharge or death. This study was approved by IRB with full exemption (NCR202580).
      Baseline characteristics of the 402 patients included in this study are summarized in Table 1. Notably, our patient population was 53.7% male, 57.2% age >60; 26.9% required ICU admission, and 15.7% were intubated. Patients who received therapeutic AC were statistically significantly more likely to be older, be admitted to the ICU, or have heart disease.
      Table 1Baseline characteristics by anticoagulation status.
      Prophylactic AC (N = 250)Therapeutic AC (N = 152)Total (N = 402)p value
      Sex0.54
      Pearson's Chi-squared test.
       Female120 (47.4%)66 (44.3%)186 (46.3%)
       Male133 (52.6%)83 (55.7%)216 (53.7%)
      Age0.01
      Trend test for ordinal variables.
       <3012 (4.7%)4 (2.7%)16 (4.0%)
       30–3925 (9.9%)7 (4.7%)32 (8.0%)
       40–4932 (12.6%)15 (10.1%)47 (11.7%)
       50–5948 (19.0%)29 (19.5%)77 (19.2%)
       60–6958 (22.9%)35 (23.5%)93 (23.1%)
       70–7943 (17.0%)31 (20.8%)74 (18.4%)
       >8035 (13.8%)28 (18.8%)63 (15.7%)
      African American0.96
      Pearson's Chi-squared test.
       No77 (30.4%)45 (30.2%)122 (30.3%)
       Yes176 (69.6%)104 (69.8%)280 (69.7%)
      White0.68
      Pearson's Chi-squared test.
       No235 (92.9%)140 (94.0%)375 (93.3%)
       Yes18 (7.1%)9 (6.0%)27 (6.7%)
      Hispanic0.88
      Pearson's Chi-squared test.
       No216 (85.4%)128 (85.9%)344 (85.6%)
       Yes37 (14.6%)21 (14.1%)58 (14.4%)
      BMI over 300.45
      Pearson's Chi-squared test.
       No11 (4.3%)9 (6.0%)20 (5.0%)
       Yes242 (95.7%)140 (94.0%)382 (95.0%)
      ICU admission<0.01
      Pearson's Chi-squared test.
       No214 (84.6%)80 (53.7%)294 (73.1%)
       Yes39 (15.4%)69 (46.3%)108 (26.9%)
      Intubation<0.01
      Pearson's Chi-squared test.
       No230 (90.9%)109 (73.2%)339 (84.3%)
       Yes23 (9.1%)40 (26.8%)63 (15.7%)
      Chronic pulmonary disease0.69
      Pearson's Chi-squared test.
       Missing808
       No182 (74.3%)108 (72.5%)290 (73.6%)
       Yes63 (25.7%)41 (27.5%)104 (26.4%)
      Cancer0.37
      Pearson's Chi-squared test.
       Missing808
       No232 (94.7%)144 (96.6%)376 (95.4%)
       Yes13 (5.3%)5 (3.4%)18 (4.6%)
      Cardiovascular disease<0.01
      Pearson's Chi-squared test.
       No157 (62.1%)65 (43.6%)222 (55.2%)
       Yes96 (37.9%)84 (56.4%)180 (44.8%)
      Hypertension0.71
      Pearson's Chi-squared test.
       No77 (30.4%)48 (32.2%)125 (31.1%)
       Yes176 (69.6%)101 (67.8%)277 (68.9%)
      Diabetes0.53
      Pearson's Chi-squared test.
       No144 (56.9%)80 (53.7%)224 (55.7%)
       Yes109 (43.1%)69 (46.3%)178 (44.3%)
      a Pearson's Chi-squared test.
      b Trend test for ordinal variables.
      Unadjusted hospital mortality for patients treated with therapeutic AC due to severe COVID-19 was 34.8% compared to 15.2% in patients receiving prophylactic AC [OR 3.42 (2.06, 5.67)]. Patients treated with therapeutic AC empirically for COVID coagulopathy had a higher mortality (38.3%) than patients treated for other indications (26.4%) such as Atrial Fibrillation/Atrial flutter/Prior VTE. However, this was not statistically significant.
      We used an elastic net logistic regression model to identify the important variables affecting mortality of the cohort. The non-zero variables were used in a multivariate logistic regression model to compare the effect of prophylactic AC to therapeutic AC (heparin drip, 1 mg/kg twice a day or 1.5 mg/kg daily subcutaneous enoxaparin, and direct oral anticoagulants) on in-hospital mortality. The R squared for the model with mortality as the outcome was 0.75. The OR for therapeutic anticoagulation was not significant. Although there was a statistically significant difference in mortality between the two groups, this was not related to anticoagulation; rather, the difference was due to factors associated with severity of illness.
      Nonparametric survival analysis was conducted as well, and Kaplan Meier curves were constructed and stratified for ICU status, intubation, and elevated D-dimer. A Cox Proportional Hazards Model was generated for non-ICU patients to evaluate the effect of an elevated D-dimer and conduct power analysis. Power analysis conducted on the multivariate Cox model indicated a beta error of 13%.
      While the initial survival curve showed significant difference between patients receiving therapeutic and prophylactic AC, the survival benefit disappeared once adjusted for disease severity, by performing subset analysis on critically ill requiring ICU admission (further stratified by intubation status) and non-critically ill patients. No statistical difference was found in either of the survival curves (Fig. 1). While our cohort of patients who were intubated and treated with therapeutic AC appeared to have a slightly higher survival probability as compared to patients who were on prophylactic AC, it was only seen during the first few days of AC treatment. This benefit disappeared after 4 days. Our mean duration for therapeutic AC (7.2 days) was longer than the Mt. Sinai cohort (median duration of therapeutic AC = 3 days). Our results suggest that the survival benefit highlighted by Paranjbpe, et al. may represent a timing bias for effect of AC on survival for intubated patients [
      • Paranjpe I.
      • Fuster V.
      • Lala A.
      • et al.
      Association of treatment dose anticoagulation with in-hospital survival among hospitalized patients with COVID-19.
      ].
      Fig. 1
      Fig. 1Survival curves for critically ill and non-critically ill patients.
      Non-critically ill patients were examined as a whole and stratified based on D-dimer level. Fig. 1 illustrates all non-critically ill patients and shows a p value of 0.006 for the log-rank test, favoring prophylactic dose AC. This trend disappeared when the analysis was stratified by D-dimer level less or greater than 3 μg/mL. To our knowledge, ours is the first study that utilizes D-dimer for disease severity to stratify non-critically ill patients and compare the effect of therapeutic AC to prophylactic AC on in-hospital mortality.
      Clinically significant adverse outcomes were defined as hemorrhage resulting in a decrease in hemoglobin greater than 2 g/dL with transfusion requirements, or a clinically significant decrease in platelet count (based on judgement of treating provider) resulting in discontinuation of any form of AC. Approximately 9% of our patients receiving therapeutic AC experienced clinically significant adverse events resulting in discontinuation of treatment; 11 patients had clinically significant bleeding while 3 patients developed clinically significant thrombocytopenia. This contrasts with the 3% risk seen in those patients receiving prophylactic AC. Our overall bleeding rate of 7.2% is higher than previously reported [
      • Paranjpe I.
      • Fuster V.
      • Lala A.
      • et al.
      Association of treatment dose anticoagulation with in-hospital survival among hospitalized patients with COVID-19.
      ,
      • Nadkarni G.N.
      • Lala A.
      • Bagiella E.
      • Chang H.L.
      • Moreno P.
      • Pujadas E.
      • Arvind V.
      • Bose S.
      • Charney A.W.
      • Chen M.D.
      • Cordon-Cardo C.
      • Dunn A.S.
      • Farkouh M.E.
      • Glicksberg B.
      • Kia A.
      • Kohli-Seth R.
      • Levin M.A.
      • Timsina P.
      • Zhao S.
      • Fayad Z.A.
      • Fuster V.
      Anticoagulation, mortality, bleeding and pathology among patients hospitalized with COVID-19: a single health system study.
      ]. Out of the 11 patients in our therapeutic AC cohort who had clinically significant bleeding, 9 were critically ill and 3 of these patients had significant bleeding associated with ECMO. We suspect our higher bleeding rates are driven by this observation, when contrasting critically ill versus non-critically ill COVID-19 patients [
      • Al-Samkari H.
      • Karp Leaf R.S.
      • Dzik W.H.
      • et al.
      COVID-19 and coagulation: bleeding and thrombotic manifestations of SARS-CoV-2 infection.
      ].
      We recognize that our study has some limitations. We are unable to derive causality given the observational nature of the study. Further, our patients were not uniformly selected for therapeutic AC based on D-dimer or critical illness. Therapeutic AC was empirically initiated for ICU patients regardless the level of specific biomarker and critical illness was determined by bed location (ICU vs Medicine floor) without additional metrics to measure concurrently for disease severity.
      Our analysis, however, identified no evidence that therapeutic AC empirically prescribed to patients with severe COVID-19 infection provides any mortality benefit over standard thromboprophylaxis, even after controlling for confounders including disease severity and comorbidities. We identified the degree of severity of infection as the primary driver of mortality, which was unaffected by the initiation of therapeutic AC.
      Further, empiric initiation of therapeutic anticoagulation is not without serious consequences as there were significant hematologic adverse events in COVID-19 patients treated with therapeutic AC compared to usual thromboprophylaxis. Randomized prospective clinical trials are needed to ascertain the appropriate indications, patient selection, and dosing of AC in COVID-19 infection [
      • Marietta M.
      • Vandelli P.
      • Mighali P.
      • et al.
      Randomised controlled trial comparing efficacy and safety of high versus low Low-Molecular Weight Heparin dosages in hospitalized patients with severe COVID-19 pneumonia and coagulopathy not requiring invasive mechanical ventilation (COVID-19 HD): a structured summary of a study protocol.
      ]. Future directions of study should consider the role of biomarkers of inflammation and coagulopathy in guiding therapeutic decisions. Despite the limitations, our study provides important insight regarding the use of anticoagulation for COVID-19 patients. These findings may help inform clinical practice in this ongoing global crisis.

      CRediT authorship contribution statement

      L. Lynn, MD: Collected the data, contributed to the interpretation of the results, took the lead in writing the manuscript, provided critical feedback, helped shape the analysis and manuscript.
      J. A. Reyes, MD MPH: Helped conceive the original idea and research plan, Collected the data, contributed to the interpretation of the results, provided critical feedback, helped shape the analysis and manuscript.
      K. Hawkins, MD: Contributed to the interpretation of the results, provided critical feedback, helped shape the analysis and manuscript.
      A. Panda MS: Verified the analytical methods, contributed to the interpretation of the results, provided critical feedback.
      L. Linville, MD: Collected the data. Discussed the results and contributed to the final manuscript.
      W. Aldhahri, MD: Collected the data. Discussed the results and contributed to the final manuscript.
      G. Kango, MD: Collected the data. Discussed the results and contributed to the final manuscript.
      S. Shah, MD: Collected the data. Discussed the results and contributed to the final manuscript.
      S. Ayanian, MD MS: Helped conceive the original idea and research plan, processed the experimental data, performed the analysis, contributed to the interpretation of the results, provided critical feedback, helped shape the analysis and manuscript.
      K. Teufel, MD: Helped conceive the original idea and research plan, Collected the data, contributed to the interpretation of the results, provided critical feedback, helped shape the analysis and manuscript.

      Declaration of competing interest

      The authors have no conflict of interest to disclose.

      Acknowledgements

      We are grateful to receive the valuable input from Dr. Linda Lesky regarding our study design and data interpretation, as well as Mr. Timothy P. Shields for his review of the data process. The R project for Statistical Computing (https://www.r-project.org/).

      Funding source

      None.

      References

        • Gorlinger K.
        • Dirkmann D.
        • Gandhi A.
        • Simioni P.
        COVID-19 associated coagulopathy and inflammatory response: what do we know already and what are the knowledge gaps?.
        Anesth Analg. 2020 Nov; 131: 1324-1333
        • Wichmann D.
        • Sperhake J.P.
        • Lutgehetmann M.
        • et al.
        Autopsy findings and venous thromboembolism in patients with COVID-19: A prospective cohort study.
        Ann. Intern. Med. 2020 Aug 18; 173: 268-277
        • Cui S.
        • Chen S.
        • Li X.
        • Liu S.
        • Wang F.
        Prevalence of venous thromboembolism in patients with severe novel coronavirus pneumonia.
        J. Thromb. Haemost. 2020 Jun; 18: 1421-1424
        • Bikdeli B.
        • Madhavan M.V.
        • Jimenez D.
        • et al.
        COVID-19 and thrombotic or thromboembolic disease: implications for prevention, antithrombotic therapy, and follow-up: JACC State-of-the-Art Review.
        J. Am. Coll. Cardiol. 2020 Jun 16; 75: 2950-2973
        • Turshudzhyan A.
        Anticoagulation options for coronavirus disease 2019 (COVID-19)-induced coagulopathy.
        Cureus. 2020; 12e8150
        • Ayanian S.
        • Reyes J.
        • Lynn L.
        • Teufel K.
        The association between biomarkers and clinical outcomes in novel coronavirus pneumonia in a US cohort.
        Biomark. Med. 2020 Aug; 14: 1091-1097
        • Paranjpe I.
        • Fuster V.
        • Lala A.
        • et al.
        Association of treatment dose anticoagulation with in-hospital survival among hospitalized patients with COVID-19.
        J. Am. Coll. Cardiol. 2020 Jul 7; 76: 122-124
        • Nadkarni G.N.
        • Lala A.
        • Bagiella E.
        • Chang H.L.
        • Moreno P.
        • Pujadas E.
        • Arvind V.
        • Bose S.
        • Charney A.W.
        • Chen M.D.
        • Cordon-Cardo C.
        • Dunn A.S.
        • Farkouh M.E.
        • Glicksberg B.
        • Kia A.
        • Kohli-Seth R.
        • Levin M.A.
        • Timsina P.
        • Zhao S.
        • Fayad Z.A.
        • Fuster V.
        Anticoagulation, mortality, bleeding and pathology among patients hospitalized with COVID-19: a single health system study.
        J. Am. Coll. Cardiol. 2020; https://doi.org/10.1016/j.jacc.2020.08.041
        • Al-Samkari H.
        • Karp Leaf R.S.
        • Dzik W.H.
        • et al.
        COVID-19 and coagulation: bleeding and thrombotic manifestations of SARS-CoV-2 infection.
        Blood. 2020; 136: 489-500
        • Marietta M.
        • Vandelli P.
        • Mighali P.
        • et al.
        Randomised controlled trial comparing efficacy and safety of high versus low Low-Molecular Weight Heparin dosages in hospitalized patients with severe COVID-19 pneumonia and coagulopathy not requiring invasive mechanical ventilation (COVID-19 HD): a structured summary of a study protocol.
        Trials. 2020; 21: 574